Subject: nontarget effect Date: Wed, 13 Jan 1999 11:44:18 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "workshop" Dear colleagues It is true that we need data on ecological effect of biological agents released in an area. Did you ever think that we need also studies to answer the question what would be the ecological effects of not releasing any biological control agents ? In case of biological control of weeds we need to evaluate not only the benefits of released agents, in term of $, but also replacement of native vegetation, eradication of rare species, and reduction of biodiversity are important matters for which we have hardly any data. Subject: Predicting impacts on non-target species Date: Wed, 13 Jan 1999 10:29:10 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Two months have elapsed since this list-server was established and 155 people have registered. Today, the first comment was posted, calling for assessment of the potential impacts of alternatives to biocontrol introductions. I think everyone would agree that such assessments would be worthwhile. However they would also be extremely difficult, perhaps as difficult as predicting non-target impacts of introductions. Our inability to predict effects of perturbations in ecosystems lies at the heart of the controversies surrounding biological control introductions. For the sake of discussion, let us assume that most candidates for introduction are not species-specific in the hosts or prey they attack in their region of origin. Let us also assume that we can reliably predict host/prey range of candidates on the basis of pre-introduction studies and that non-target species in the region of introduction may be attacked. In this case, we must either forgo most introductions or find a way to assess potential impacts on non-target species. Predicting the impact of natural enemy attack on the population dynamics of non-target species (let alone indirect impacts) appears beyond the power of ecological theory and is likely to remain so for a long time. However, analyses of the current impacts of previous and ongoing introductions would allow assessment of the frequency and magnitude of non-target impacts and the conditions under which they are occur. Thus, the most important work that could be done to assess the risks of impacts on non-target species would be to evaluate what the actual impacts have been in a variety of systems. Do you agree or disagree and if so why? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: RE: impact of biological control Date: Wed, 13 Jan 1999 11:09:43 -0500 From: Kevin Floate Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear All: I agree entirely with the comment made by Rouhollah SOBHIAN, who indicated the need, not just to discuss the risks of releasing biological control agents, but also to discuss the risk associated with NOT implementing some type of control strategy (e.g., biocontrol, chemical, cultural) where such measures are warranted. I suggest that, ideally, any biological control programme would include a realistic economic assessment using 'hard' data (i.e., no unsubstantiated, anectodal-type observations) of the problem as part of the justification for the research. Time to 'stir the pot'... (I note that broad discussion of biocontrol issues was encouraged by the listowners.) I would be interested to know what readers define as 'biological control' strategies. Specifically, would they include transgenic plants in their definitions? Why or why not? I don't ask this question lightly. I would not, but some of my colleagues would. Organic growers in the US reject transgenic plants as 'organic'. Hence, unless the definition of 'biological control' excluded transgenic plants, biocontrol research would not necessarily be supported by organic growers, who should be among our strongest advocates. I think it incumbent upon biocontrol practioners and researchers to clearly define 'biological control', and to resist efforts to broaden its definition until it becomes almost meaningless. Now to stir things up a little more, and to return to the central theme of the list (i.e., risks associated with the release of biocontrol agents) If we accept transgenic plants as a form of biological control, then arguably, the dispersal of pollen or seeds from these plants is equivalent to the dispersal of biocontrol agents from a point of release. Hence, should the companies that sell transgenic crops be held accountable for damages ***IF*** these crops cause problems in later years? There have been recent reports in Canada, of farmers who planted herbicide resistant 'transgenic' canola (i.e., rapeseed, Brassica sp.) last year. Although canola was not planted in these field this spring, 'volunteer' herbicide-resistant canola appeared. Similarly, one can easily envisage seeds from these resistant plants blowing off the back of farm trucks and ending up along road margins. Trying to stimulate discussion on a sluggish list, Sincerely, Dr. Kevin D. Floate, Research Scientist Biological Control of Livestock Insect Pests Agriculture and Agri-Food Canada Lethbridge Research Centre Lethbridge, AB, CANADA T1J 4B1 ph: (403) 317-2242, FAX (403) 382-3156 http://res.agr.ca/leth/scitech/kdf/intro.htm ! ! Subject: Re: nontarget effect Date: Wed, 13 Jan 1999 08:35:31 -1000 From: "Marshall W. Johnson" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu A very good point. Dr. Marshall W. Johnson, Professor & Entomologist Department of Entomology Vice Chair, Department of Plant & Environmental Protection Sciences (In transition) University of Hawaii at Manoa 3050 Maile Way, 310 Gilmore Hall Honolulu, Hawaii 96822 U.S.A. Phone: 808-956-8432 FAX: 808-956-2428 E-mail: mjohnson@Hawaii.edu Subject: Re: Predicting impacts on non-target species Date: Wed, 13 Jan 1999 15:26:59 -0500 From: Howard Frank Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I agree that it would be useful to evaluate effects of imported biological control agents on non-target species. This leaves a lot of things unsaid, and these things should be part of the documentation. 1. It should be made clear whether the example is of a classical biological control agent which has become established or a biopesticide (read also inundative biological control agent) which may or may not have become established by the application. 2. You wrote of generalizing that the biological control agents are not species-specific in their region of origin (and presumably not bothering to document those that ARE species-specific). But perhaps species-specific agents should be documented too as part of the record - with any indication that any of these has switched its diet after importation. How about those that are genus-specific or tribe- specific in their region of origin - and by this level of specialization avoid affecting non-target organisms in their new territories. 3. What about documentation of the positive effects of the biological control agents? For example, importation of U reduced populations of V (its target) but this had a positive effect on native plants W and X (which are attacked by V), and a negative effect on native birds Y and Z (which use V as part of their diet). 4. And, if U successfully reduces populations of V, then fewer pesticidal applications are used against V, and there are positive effects on a range of other non-target organisms. Examples such as the imaginary example above can get very complicated - expensive and time- consuming to document, and difficult to present simply. I agree that it would be useful to document them, but let us not jump in expecting to get simple measures that can be reduced easily to frequency and magnitude. Howard Frank At 10:29 AM 1/13/99 -0500, you wrote: >Two months have elapsed since this list-server was established and 155 >people have registered. Today, the first comment was posted, calling for >assessment of the potential impacts of alternatives to biocontrol >introductions. I think everyone would agree that such assessments >would be worthwhile. However they would also be extremely difficult, >perhaps as difficult as predicting non-target impacts of introductions. >Our inability to predict effects of perturbations in ecosystems lies at >the heart of the controversies surrounding biological control >introductions. > >For the sake of discussion, let us assume that most candidates for >introduction are not species-specific in the hosts or prey they attack >in their region of origin. Let us also assume that we can reliably >predict host/prey range of candidates on the basis of pre-introduction >studies and that non-target species in the region of introduction may be >attacked. In this case, we must either forgo most introductions or find >a way to assess potential impacts on non-target species. Predicting the >impact of natural enemy attack on the population dynamics of non-target >species (let alone indirect impacts) appears beyond the power of >ecological theory and is likely to remain so for a long time. However, >analyses of the current impacts of previous and ongoing introductions >would allow assessment of the frequency and magnitude of non-target >impacts and the conditions under which they are occur. Thus, the most >important work that could be done to assess the risks of impacts on >non-target species would be to evaluate what the actual impacts have >been in a variety of systems. Do you agree or disagree and if so why? > >Yours, Keith Hopper > >****************************************************************** >Keith R. Hopper, Research Entomologist >USDA, ARS, Beneficial Insect Introduction Research Unit >University of Delaware, 501 South Chapel Street, Newark, Delaware >Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 >****************************************************************** > Subject: Re: impact of biological control Date: Wed, 13 Jan 1999 16:04:54 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu In response to Kevin Floate's message, I would like to note that the focus of this list is actually rather narrow. It is to discuss research needs for assessing and reducing non-target impacts of biological control introductions. The emphasis in our discussions will be introductions of arthropods to control arthropod pests and weeds, but we do not exclude discussion of other sorts of introductions. The purpose of our mail list is to promote discussion among researchers working on this problem. The sort of introduction meant is from one region of the globe to another. Trans-genic plants in particular and trans-genic organisms in general are beyond the scope of this workshop. This is not because the issues involved with trans-genic organisms are unimportant, but because the issues are very different from those involved with transferring insects from one region to another. Furthermore, the goal of the workshop is to delineate research needs. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: 'Biological control' strategies Date: Wed, 13 Jan 1999 16:35:14 -0500 From: "E. Alan Cameron" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Folks; A comment from an historical perspective on the question of inclusion of transgenic plants as being included in 'biological control' (question posed by Kevin Floate): Several decades ago, when 'pheromones' were in their infancy in terms of being identified and exploited for pest management, there was a camp that argued that these chemicals are really a part of 'biological control' because they originate in a biological organism. Further, they generate a behavioral (='biological') response in the receiving organism. I recall some passionate discussions on both sides of the fence, often in the halls or cocktail lounges at various regional and national meetings. I see a broad parallel between 'pheromones' and 'transgenic plants' in this kind of an argument. I am quite comfortable NOT including these behavior-modifying chemicals under the rubric of 'biological control.' And for many of the same kinds of reasons (recognizing, however, that transgenic plants are often capable of propogation of another generation, something that pheromones cannot do alone), I would not include them in 'biological control.' It might help to gain short-term political acceptance for transgenics, largely among the public at large (biocontrol still wears white - or at worst, grey - hats) if they could be included in BC, but I do not believe it is for the long-term good for BC that this adoption be made. Floate gave one very good specific example of a liability - and from the political arena as well ('organic' growers). Interesting question. Cheers! Alan ********************** E. Alan Cameron, B.C.E. Phone: (814) 863-2867 Professor of Entomology Fax: (814) 865-3048 501 A.S.I. Building Penn State University mailto:acameron@psu.edu University Park, PA 16802 Subject: Re: Predicting impacts Date: Wed, 13 Jan 1999 14:48:32 -0700 From: "Alec McClay" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I'm glad to see some discussion starting up on this list! >Predicting the >impact of natural enemy attack on the population dynamics of non-target >species (let alone indirect impacts) appears beyond the power of >ecological theory and is likely to remain so for a long time. Even predicting the impact on _target_ species is beyond our power! >Thus, the most >important work that could be done to assess the risks of impacts on >non-target species would be to evaluate what the actual impacts have >been in a variety of systems. Do you agree or disagree and if so why? I agree. However, in addition to evaluating the actual impacts I think it is important to compare them with any predictions that may have been made before release of the agent. In weed biocontrol, there have been many agents approved for release despite the fact that testing showed that some level of attack might occur on particular non-target species. These releases may have been approved because (a) attack on non-targets was considered to be an artefact of test procedures, (b) because impact was expected to be much lower than on the target, (c) because the non-target species were not considered important, or (d) because of the urgent need for control of the target species. [E.g. Hasan & Delfosse (1995) for attack by a rust on native Australian Heliotropium sp., by Hill & Hulley (1995) and Olckers et al. (1995) for attack by biocontrol agents for introduced Solanum spp. on eggplant (Solanum melongena L.) in South Africa, by Kok et al. (1992) and Blossey et al. (1994) for attack by biocontrol agents for Lythrum salicaria L. on native Lythraceae in North America, and by Willis et al. (1996) for attack by an eriophyid mite on native Hypericum species in Australia.] Cases such as these give a clear pointer as to where we should be looking for non-target impacts. It should be relatively straightforward to do the necessary monitoring, at least at the level of comparing rates of attack on targets and non-targets. Hopefully by comparing outcomes of these releases with the predictions we can at least assess how well we have been doing at predicting non-target impact. Alec McClay Research Scientist, Biological Control of Weeds alec@aec.arc.ab.ca Alberta Research Council Phone (403) 632-8207 Bag 4000, Vegreville Fax (403) 632-8612 Alberta T9C 1T4, Canada Please note: On Jan. 25 1999 the area code will change from (403) to (780) http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML Subject: RE: impact of biological control Date: Wed, 13 Jan 1999 15:52:39 -0600 From: "Julio S. Bernal" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Kevin Floate revived a question that was put forward in the late 1980's (NAS, Garcia et al.), and which is as valid today as it was then. Hopefully, a majority of researchers (in general, and not only BC practitioners) would favor a narrower (i.e. parasitoids, predators, and diseases) rather than a broader definition. However, because the question has returned 10 years after it was first posed indicates that probably the broader definition is currently accepted more than (personally) presumed. In my case, the favor for the narrower definition is due to reasons similar to those exposed by Garcia et al.; i.e. the broader definition will tend to dilute already scarce research funds, and BC's (narrow sense here) proven track record will be "used" by technologies that are still being tested. Perhaps, as Kevin Floate indicates, it is time to seriously revive the question concerning the accepted definition of BC. Julio Bernal ps: Inclusion of transgenic plants as BC technologies is a a very long a stretch of the definition of BC; because of Terminator gene technology and other restrictions, they do not persist (as in classical BC), they can not be reproduced (as in augmentation BC), nor can they be conserved (as in conservation BC). ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julio S. Bernal, Ph.D. Biological Control Laboratory Texas A&M University College Station, TX 77843-2475 Tel. (409) 862-8378 Fax (409) 845-7977 e-mail jsb7473@unix.tamu.edu ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Subject: Re: nontarget effect Date: Wed, 13 Jan 1999 16:57:03 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Marshall, To which point are you referring? Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: nontarget effect Date: Wed, 13 Jan 1999 14:49:14 -1000 From: "Marshall W. Johnson" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Keith: I was originally referring to the comments of Rouhollah SOBHIAN who said that we should also be addressing what are the environmental effects if we do not introduce natural enemies. A common response of some individuals is just use pesticides (intead of natural enemies) which is somewhat terrifying to one who has strived to reduce pesticide usage. Marshall Dr. Marshall W. Johnson, Professor & Entomologist Department of Entomology Vice Chair, Department of Plant & Environmental Protection Sciences (In transition) University of Hawaii at Manoa 3050 Maile Way, 310 Gilmore Hall Honolulu, Hawaii 96822 U.S.A. Phone: 808-956-8432 FAX: 808-956-2428 E-mail: mjohnson@Hawaii.edu Subject: Re: impact of biological control Date: Wed, 13 Jan 1999 17:14:01 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu This post is in response to Rouholah Sobian's comment. Of course he is correct that we must balance in the cost of not doing classical biological control. Unfortunately, we must also add in the cost of making classical introductions that are ineffective against the target pest but may have nontarget effects. In other words we have an elaborate distribution of potential outcomes at the outset of a classical bc program. These include 9 possible outcomes if we do a classical program 1-3)a (1) highly ([2]moderately, or [3]completely un-) successful program with no nontarget effects 4-6) as above but with modest nontarget rearing records etc. 7-9) as above but with important nontarget effects. Then we have possible scenarios (4 presented) if we choose not to make classical BC intros: Pest continues unabaited either destroying lots of habitat, injuring commodities or causing a pesticide management program to persist Biorational controls for th pest are developed and implemented (but at the expense they engender) An IPM management program that promotes activity of endemic natural enemies is developed that allows conservation bc to significantly manage the pest. In unmanaged settings or crops where few insecticides are used, generalist natural enemies build up and the exotic pest slowly subsides because of the action of these beneficials. Given these various alternatives, clearly identifying a general direction of research becaomes problematic. Clearly it makes no sense to insist on proving that the introduction will be successful before hand. That has been a goal of biocontrol/ecologists for many, many years and I do not think it has been demonstrated we can do this. However, I do believe it is clear that we can do research to characterize host range and its correlates and applying this infromation minimize nontarget effects. In other words we may be alot better at identifying what won't work (in this case on nontargets) than what will. Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Re: nontarget effect Date: Wed, 13 Jan 1999 17:15:41 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Yeah Marshal! Tom Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Apologies Date: Thu, 14 Jan 1999 11:17:39 +0000 From: "E. Alan Cameron" To: Keith Keith; My apologies for sending my note, in light of your reiteration of the scope of the listserver. Your reminder came after I had sent my comments. Will try to remember not to step out of bounds in future. Ironically, that is one of the few comments I have made to lists. I monitor several largely for teaching and background research purposes, but tend to try not to engage actively in what, as you know, too often becomes endless and repetitive debate. Cheers! Alan ********************** E. Alan Cameron, B.C.E. Phone: (814) 863-2867 Professor of Entomology Fax: (814) 865-3048 501 A.S.I. Building Penn State University mailto:acameron@psu.edu University Park, PA 16802 Subject: Trans-genic organism Date: Thu, 14 Jan 1999 17:45:18 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "Workshop" Hi everybody I agree with Keith that the subject matter does not fit into this workshop. However, because it is a very important issue, it may be handled in a separate workshop. Subject: Subject of mail-list Date: Thu, 14 Jan 1999 10:47:59 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Again, I would like to reiterate that the subject of this workshop is to delineate research needs for assessing and reducing impacts of biological control agents introduced from one region to another. Although risk-benefit analysis of all strategies for pest management is a worthwhile goal, increasing the scope of the workshop to this degree is likely to impede progress on this primary goal. Although weighing the various strategies for pest control will ultimately be necessary, do you think it possible to assess impact of introduced biological control agents on non-target species and to develop ways of reducing the likelihood of such impacts, independently of evaluating risk-benefits of all strategies? Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: impact on target species Date: Thu, 14 Jan 1999 11:45:16 -0600 From: Manuel Vázquez Navarro Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact I find a very interesting discussion: Alec Mcley wrote: Even predicting the impact on _target_ species is beyond our power! It´s a good point. ¿Is realy beyond our power? ¿Do we have a good advanced in assesing the impact of our releases on target species? Maybe a starting point for research needs. Manuel Vazquez Navarro Division de estudios de Posgrado Fac. de Agricultura y Zootecnia, Univ. Juarez del Edo. de Durango Jalisco 141 Col. Las Rosas Gomez Palacio, Durango. 35090 Mexico mavazna@coah1.telmex.net.mx Subject: Another Viewpoint Date: Thu, 14 Jan 99 11:16:04 PST From: "Allan Deutsch" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu ----------------------------------------------------------------- As a starting point, would it be useful to review, analyze, and catalog the BC introductions that, while well-intentioned, went very wrong? Maybe this has been done, but it would seem that there might be some hindsights gained that would help avoid future problems. It would also seem that extracting an organism from the native ecology where it flourishes and plunking it down in what appears to be, but in fact may not be, a similar set of eco conditions is a gamble. Maybe that's the gist of the reason for this newsgroup: to try to devise methodologies that minimize (ideally) or eliminate (not realistic) the odds for error. My own 2 cents, A.E. Deutsch Oregon, US Subject: Keep risk-benefits linked Date: Thu, 14 Jan 1999 16:01:13 -0600 From: Tim Kring Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Hopper wrote: "do you think it possible to assess impact of introduced biological control agents on non-target species and to develop ways of reducing the likelihood of such impacts, independently of evaluating risk-benefits of all strategies? " Yes....BUT. I believe to focus on the impact on non-target species without concurrent evaluation of the natural enemies impact in a risk-benefit context does a disservice to biological control. As an aside, as biological control workers we do a poor job of evaluating the impact of our introductions on the TARGET...so I have VERY low expectations for us providing valuable studies on often poorly known (or unknown) species sometimes trophic levels away. -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Tim Kring, Univ. of Arkansas-Entomology ** (501) 575-3186 Virology-Biocontrol Laboratory ** (501) 575-3348 (FAX) 213 Young Avenue ** Fayetteville AR 72704 ** tkring@comp.uark.edu -=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- At 10:47 AM 01/14/1999 -0500, you wrote: >Again, I would like to reiterate that the subject of this workshop is to >delineate research needs for assessing and reducing impacts of >biological control agents introduced from one region to another. >Although risk-benefit analysis of all strategies for pest management is >a worthwhile goal, increasing the scope of the workshop to this degree >is likely to impede progress on this primary goal. Although weighing >the various strategies for pest control will ultimately be necessary, do >you think it possible to assess impact of introduced biological control >agents on non-target species and to develop ways of reducing the >likelihood of such impacts, independently of evaluating risk-benefits of >all strategies? > >Yours, Keith Hopper >-- >****************************************************************** >Keith R. Hopper, Research Entomologist >USDA, ARS, Beneficial Insect Introduction Research Unit >University of Delaware, 501 South Chapel Street, Newark, Delaware >Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 >****************************************************************** > > Subject: Re: Trans-genic organism Date: Thu, 14 Jan 1999 15:15:53 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I would point out to Dr. Hopper that running a listserver for scientists is much like herding cats. Best to let them have their way and then they may settle down. If we are to confine our discussion to risk rating arthropod introductions, a workably narrow topic, and ignore the reasons for undertaking the biological control introduction in the first place then I offer the following questions laced with opinion. DOes the literature on host range provide anything useful? Some like Shaw and Hawkins would argue that it is almost worthless. I disagree. I believe it is especially useful for identifying species that have very narrow host ranges. For sake of further simplicity lets say that we can divide the world of candidate species into those with narrow host ranges and those with fairly broad host ranges (say within a couple of families of hosts in the case of insects, and a genus or two in hte case of weed agents). Given that, do the readers believe that there is some difference in the standards of evidence required for each type of species in order to risk rate their possible nontarget problems in each of these type of species? What do we accept to be a negative nontarget effect? It must devolve to some effect on population dynamics of the nontarget but when do we consider it biologically meaningful? I do not think we can do this with any rigor at all because we don't know enough about the ecology at the outset and because the potential nontargets are a new spectrum of species in new habitats. I believe it has more to do with the taxonomic distribution of the existing host range. In other words are their taxa of concern in the nominal host range distribution? Are we worried about nontarget attacks of aphids or scale insects? Of course from a biodiversity viewpoint we all probably are if we think extinction of nontargets is going to be possible. Barring that, then what? How does one make host range data meaningful? I believe it must be interpreted with the host utilization behavior in mind. For example a parasitoid that attacks a fairly small group of hosts, but ones which are hidden away in refugia such as bark, are likely to make some oviposition mistakes which will expand the host range. Also the researcher is very likely to make some rearing/identification mistakes further expanding the host range. SO how do we cull spurious rearing records and misidentifications from the global literature? Can we somehow rank the strength of rearing data based on percent parasitism or percent utilization? Must we redo rearing studies and create large data bases for all species prior to introduction? Or can we confirm field host range data from laboratory studies for those species where we are in a grey zone? How do we identify the grey zone? Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Re: Another Viewpoint Date: Thu, 14 Jan 1999 18:58:36 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu In response to the message from Allan Deutsch, what are the introductions of insects to control insects or weeds that have gone very wrong? I've looked for cases where impact on non-target species has been documented and have found no cases of intentional insect introductions with solid evidence of population-level impacts on non-target species. Indeed, this is why I suggested that retrospective analysis of a variety of introductions would be perhaps the most useful project for assessing the risks of impacts on non-target species. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Nontarget effect Date: Fri, 15 Jan 1999 11:50:46 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "Workshop" Dear colleagues Here are a few comments on the subject : 1- I response to Allan Deutsch, I agree with Keith Hopper. My colleagues in the laboratory and I do not know of any case where well-intentioned releases went very wrong. If an insect like Rhynocyllus controlled Carduus thistles in large areas and attacked also a few native Cirsium species, it is not a release that went very wrong. Because without the insect the thistles would not only negatively effect on the Cirsium species, but the thistles would replace or eradicate other native species also. 2- Keith, Yes any step towards reducing damage to the nontarget organisms is desirable. But equally or even more important is the post release evaluation, both in term of $ value and of ecological aspects. The studies made by a student in Beltsville demonstrated that the rate of cost benefit in biological control was 1:100. If we do not have this kind of data, we will not have justification for what we are doing. If somebody could demonstrate how many species of native plants were reduced or eradicated due to alien weeds or how many of the native species mowed into an area after an alien weed was controlled by an insect and how much was the benefits of these introductions, then biological control would get more support. Often the alien weeds are the worst enemies of the native vegetation. 3- This is a comment on Tim Kring message: Yes it is very difficult to predict or study the nontarget effects of released organisms, but this is not a good reason not to look for solutions and not to try to find a way to do it. 4- This is in response to To Unruh: You are right we need to consult the world literature. BUT, often there are wrong or misleading information in the old literature, which is mainly due to the misidentification of organisms. For example while I was working on Bangasternus orientalis, a seed feeder on yellow starthistle, in Greece, I found a citation in the Russian literature that said the species also attacks safflower. I could not believe this and tested the insect. My field and laboratory experiments showed that the insect does not oviposits on safflower. The insect has been released in the USA in 1985 and it never attacked anything except yellowstarthistle. Another species of the genus, Bangasternus planifrons, attacks wild and cultivated safflower and probably this was wrongly identified as B. orientalis. In a second case, I found a weevil that attacks Tamarix and its larvae makes galls on the roots of its host plant. This is Liocleonus clathratus, a about 1 cm long weevil, which has been reported in the Russian literature on camel thorn, Alhagi. Alhagi is a Leguminosae and has no relation with Tamarix, which is a very separate group of plants. I discussed this matter with Kovalev and other entomologists at the Zoological Museum in St. Petersburg and E. Colonnelli, in Rome, and everybody agrees with me that this must be due to a misidentification or mislabeling of the specimens. We are going to test the insect on Alhagi. Sorry, I am not an English speaker, but I hope that my comments will be understandable. Subject: Re: Trans-genic organism Date: Fri, 15 Jan 1999 09:08:22 GMT From: "Nick Birch" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I would welcome a workshop (separate?) on non-target effects of transgenic organisms. The IOBC has initiated a steering committee (under Angelika Hilbeck) to discuss these important issues and standardise and good experimental protocols for testing target and non-target (eg tri-trophic impacts of pest-resistant transgenic crops on pests and their natural enemies). Good luck with your BC workshop! Nick Birch. Dr Nick Birch Snail mail: Soft Fruit & Perennial Crops Dept., Scottish Crop Research Institute Invergowrie, DUNDEE DD2 5DA SCOTLAND, U.K. Email address: N.Birch@scri.sari.ac.uk Phone number: +44 (0)1382 562731 Fax number: +44 (0)1382 562426 Subject: Re: Subject of mail-list Date: Fri, 15 Jan 1999 10:50:45 +0000 From: Lisa Danko Reply-To: bc-ntimpact@UDel.Edu Organization: Mercyhurst College To: bc-ntimpact@UDel.Edu Keith, Im not sure whether we can ever be 100% sure we have evaluated all the risk involved in introducing a new insect/ organism as a new bio control agent. There should be a specific procedure to follow which tries to simulate all habitats, weather conditions, variables which might influence the success or demise of the event. Currently I know that the Galerucella sp. beetles have been quite freely released, after trials were done, but yet in the lab they will eat other plant materials. Can we be sure that all the natural stressor in the environment have been tested? Probably not. Risk vs Need. I believe we must always proceed with educated caution. But also be careful about what we call a weed, Id hate to see the milkweed decline any further. -- Lisa Danko M.S. Laboratory Supervisor/Lecturer Department of Biology Mercyhurst College 501 East 38th St. Erie, PA 16546 Voice: 814-824-2373 FAX: 814-824-2188 email : ldanko@mercyhurst.edu Subject: Re: Another Viewpoint Date: Fri, 15 Jan 1999 07:07:10 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: bc-ntimpact@UDel.Edu Our theme song is the Old Woman Who Swallowed the Fly; a classic example of biocontrol gone wrong. She swallowed the fly, you will recall, to catch the spider ... Asian ladybugs, Asian praying mantids, all those little parasitic wasps that invade any caterpillar they find ... well, until recently, we thought anything that ate insects was wonderful. Here in South Florida the pace of official introductions increases, as the ecology reels under the impact of unofficial introductions, such as the cane toad. Anyone who wants to can release ladybugs by the gazillion, in happy disregard for their impact on the local bug-eating multitude. Success crown your efforts. Anne Kilmer South Florida Allan Deutsch wrote: > > ----------------------------------------------------------------- > > As a starting point, would it be useful to review, analyze, and catalog > the BC introductions that, while well-intentioned, went very wrong? Maybe > this has been done, but it would seem that there might be some hindsights > gained that would help avoid future problems. > > It would also seem that extracting an organism from the native ecology > where it flourishes and plunking it down in what appears to be, but in fact > may not be, a similar set of eco conditions is a gamble. Maybe that's the > gist of the reason for this newsgroup: to try to devise methodologies that > minimize (ideally) or eliminate (not realistic) the odds for error. > > My own 2 cents, A.E. Deutsch > Oregon, US Subject: Re: and Another Viewpoint Date: Fri, 15 Jan 1999 10:26:52 -0500 From: Tom McAvoy Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Colleagues, Here's one article that did what Allen Deutsch suggested, catalog the success & failures of weed Bio Control: Crawley, M. J. 1989. The successes and failures of weed biocontrol using insects. Biocontrol. News Info. 10:213-223. This is 10 years old so it would be useful to update. Some examples of weed BC agents feeding on nontargets: 1. Cactoblastis sp. has been found feeding on threatened Opuntia sp. in south Florida (Kass 1990, Robertson 1990), 2. C. quadrigemina was found feeding on Hypericum calycenum, an exotic plant extensively used as a ground cover in northern California (Ehler 1991), 3. R. conicus (one species L. T. Kok & I are working with) has been found feeding on nontarget Cirsium sp. thistles (Strong 1997)& Louda (1998) But overall, weed bio control has had a very good track record very few if any agents have 'host shifted' - switched to another host that was not predicted in host feeding studies. Host feeding studies on R. conicus showed that other Carduus and Cirsium speices are hosts of R. conicus. But the benefit of reducing C. nutans populations has resulted in a very large reduction in herbicide application. More long term studies on changes in plant diversity after release and the economic and ecological benefits need to be documented. One problem with weed BC is often not having control data after release ie no plots or areas where the agents are not released, but this could be solved with some good planning. One result I'd like to see come out of this is more communication as is happening now with this list serve among entomologist PLUS botanist &/or plant ecologist, with regards to what weeds are being studied for exotic introductions. Espcially if weed biocontrol moves into non-agricultural habitats and used as a tool in controlling exotics to restore and conserve native habitats. Plus, seeds and plant material for host study feeding could be obtained more readily through cooperation between groups, which has been a problem for many researchers in the past. Tom McAvoy At 11:16 AM 1/14/99 PST, you wrote: >----------------------------------------------------------------- > > As a starting point, would it be useful to review, analyze, and catalog >the BC introductions that, while well-intentioned, went very wrong? Maybe >this has been done, but it would seem that there might be some hindsights >gained that would help avoid future problems. > > It would also seem that extracting an organism from the native ecology >where it flourishes and plunking it down in what appears to be, but in fact >may not be, a similar set of eco conditions is a gamble. Maybe that's the >gist of the reason for this newsgroup: to try to devise methodologies that >minimize (ideally) or eliminate (not realistic) the odds for error. >Allan > My own 2 cents, A.E. Deutsch > Oregon, US > Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: Re: Subject of mail-list Galerucella & regulations Date: Fri, 15 Jan 1999 15:34:27 -0500 From: Tom McAvoy Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu In response to the note on Galerucella spp. released for the control of purple loosestrife (L. salicaria) it was found to be a host of only one native species, Lythrum alatum (Kok et al. 1992) (L. virgatum also supported larval development to the adult stage but this species is considered a subspecies of L. salicaria). Blossey also reported that Galerucella reared on L. alatum had reduced life spans, low fecundity and very low egg hatch. For a species to qualify as a host females must oviposite on the plant and support larval development to the adult stage. Oviposition and adult & larval feeding did occur, at a very low rate, on several nontarget species tested but this does not qualify them to be hosts, since they did not support larval development and in the absence of L. salicaria would not survive. As far as testing goes much can and was done in Germany (coutry of origin) by Blossey testing loosestrife BC agents under more natural settings in the field. Most of you probally know this but just in case: in 1958, the Working Group on Biological Control of Weeds (WGBCW) was formed as an advisory committee for the USDI and USDA-APHIS regarding host specificity testing of potential biological control agents of weeds. Subsequently, the Technical Advisory Group (TAG) on the introduction of Biological Control Agents of Weeds was established in 1987. This group, made up of specialists in many fields, recommends whether petitions for release of new biological control agents should be approved. Its purpose is to provide advice involving importation testing and field release of exotic organisms for the classical biological control of weeds. Check out APHIS'S web page at: http://www.aphis.usda.gov/nbci/nbci.html They have soon new regulations posted, quite long as you'd excpet from the government. I copied the following from their web page on regulations: "With input from investigators, quarantine facility personnel, and identifiers, APHIS conducts a first-tier risk assessment that asks the following questions to determine plant pest risk: Does the organism feed on, infect, or parasitize living plant tissues? Does the organism feed on, infect, or contaminate plant products such as stored grain, stored fruit, or lumber? Does the organism transmit plant pathogens? Does the organism develop as a secondary parasite, pathogen, or predator of a primary natural enemy of a herbivore or plant pathogen? Does the organism adversely affect commercially important pollinators or important herbivores or plant pathogens that control weeds? An affirmative answer to any one of these questions will give APHIS reason to believe that the organism presents a plant pest risk and is, therefore, subject to the Agency's authority and regulations. " Tom McAvoy Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: Weed bc non-target examples Date: Fri, 15 Jan 1999 13:46:17 -0800 (PST) From: John Herr Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu As was pointed out by Tom McAvoy, there are a few examples of non-target impact caused by weed biocontrol agents. Another is the Cinnabar moth, which feeds on a native Senecio (Diehl & McEvoy 1990), and the Klamath weed beetle, which feeds on a native Hypericum in addition to the previously mentioned non-native ground cover species (Andres 1985). However, the critical issue is whether this impact is biologically significant. There are at least two additional examples where this does indeed appear to be the case: 1) The Cactoblastis moth has caused significant damage to a rare endemic Opuntia cactus in Florida (Pemberton 1996 ?). However, in this case the agent was not intentionally introduced to Florida, and may have come in with a contaminated nursery shipment or on its own from release sites in the Caribbean. (If it turns out to be the latter, this would be a good argument to include dispersal potential in pre-release assessments). 2) The introduction of the thistle head weevil Rhinocyllus conicus to the US is definitely a case where biologically significant impact has occurred in non-target populations. Louda (1998) has documented the expansion of this weevil's host range into two native Cirsium thistles in Nebraska, and she has shown through long term studies that seedling recruitment in populations of these non-targets is limited by seed predation (Louda & Potvin 1995). I am currently working on this issue with rare California Cirsium species, and have found additional examples to add to the R. conicus non-target host range list since this problem was first reported by Turner et al. in 1987. Fortunately, it appears that a lack of synchrony between the R. conicus oviposition period and the native Cirsium flowering phenology limits the overall impact in the species I have studied. However, since distinct host races of R. conicus are known from Europe, is possible that strains more adapted to the novel California hosts will develop over time, and if this occurs I would expect the non-target impact to be much greater. In contrast to the earlier comments of Rouhollah Sobhian, I would have difficulty justifying the introduction of R. conicus today, given what we know now. Musk thistle (the main target weed), was primarily a pest of pasture land, and was not likely to be a problem in undisturbed native plant communities. In defense of those involved with this program, it should be noted that the broad host range of R. conicus was well known at the time of introduction (1969), but that in those days, impact to native plants (particularly native thistles!) was not considered to be a problem. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= John Herr jherr@nature.berkeley.edu UC Berkeley, Division of Insect Biology (510)642-8414 USDA-ARS-WRRC, Plant Protection Research (510)559-5974 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Subject: solid evidence Date: Fri, 15 Jan 1999 16:08:33 -1000 From: messing@hawaii.edu (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu In response to Kieth's statement (below)..... we've published data which show that braconids imported to Hawaii for fruit fly control clearly have "population-level impacts" on non-target tephritids imported for biocontrol of weeds (i.e., lantana). Of course, the impact varies from site to site, season to season, etc. (maximum parasitism was 28%), and the gall fly is not a very effective biocontrol agent to begin with..... There are also a few other cases in Hawaii I could cite. "Population-level impacts" are hard to define, but we'd do better to admit the impacts that do occur, and discuss them in terms of cost/benefit or risk/benefit.... rather than to deny them entirely and lower our credibility. Russell Messing _________________________________________ "I've looked for cases where impact on non-target species has >been documented and have found no cases of intentional insect >introductions with solid evidence of population-level impacts on >non-target species......." > >Yours, Keith Hopper Russell Messing University of Hawaii Dept. of Entomology 7370 Kuamoo Rd. Kapaa, Hawaii, 96746 tel (808) 822-4984 x 223 fax (808) 822-2190 email: messing@hawaii.edu Subject: WGBCW and TAG Date: Fri, 15 Jan 1999 23:04:02 -0500 From: Gary Buckingham Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I would like to add to the comments of Tom McAvoy that the Working Group was formed in 1958, as I understand it, at the request of the ARS biocontrol researchers who wanted the opinions of other agencies. GAry Buckingham Subject: A New Zealand perspective Date: Mon, 18 Jan 1999 13:10:15 +1200 From: Pauline Syrett Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear colleagues Even though I am reasonably confident that the non-target impacts of nearly all weed biological control agents are minimal, I am supportive of more research to measure them. I am looking forward to seeing more contributions to this workshop on non-target impacts of biological control agents of insects to find out whether my prejudices as an entomologist (that this is the REAL area of concern!) are supported. I agree with Alec McClay that a fruitful area of research is to examine some earlier introductions to see whether predictions from laboratory host specificity determinations have been borne out in the field. We have been asked to do this for one of our recent introductions, broom seed beetle (Bruchidius villosus), by the New Zealand Department of Conservation. Although I could conduct a survey for broom seed beetle on non-target plants closely related to introduced broom, I would have little confidence in the consequent negative results at this stage (already more than 10 years after the first release) because 1) it is not easy to find natural populations of closely related non-target plants in the vicinity of our release sites, and 2) populations of beetles are still increasing, even at sites where they are doing well, and there are many areas where beetles are not yet established. So examining populations of biological control agents that have been established for a long time seems like a good idea. We have plans to examine the field host-range in New Zealand of three insect species feeding on ragwort (Senecio jacobaea): ragwort flea beetle (Longitarsus jacobaeae), cinnabar moth (Tyria jacobaeae) and magpie moth (Nyctemera annulata). Host tests indicated that L. jacobaeae has a narrow host range, and we are unlikely to find it on plants other than a couple of very close relatives, also introductions from Europe. T. jacobaeae, in New Zealand as in Oregon, is sometimes found on native Senecios. Its impact on populations has not been studied here, but it is likely that if results from host specificity tests conducted in the 1920's were presented now, permission to release this insect would not be obtained. Less value was placed on native Senecios then than now. N. annulata is a native oligophage that will be included to make the survey less boring! And from comparisons between field records, and laboratory host-range test results, for this insect, we hope to gain some insights into the predictability of host-range from such experiments. One other observation that people may find interesting relates to the recent accidental introduction of the Echium leaf miner, Dialectica scalariella, to New Zealand. When this moth was first released in Australia, it was recognised that host tests indicated it might attack some Myosotis spp. So when, a couple of years ago, the moth was discovered in the northern part of the South Island of New Zealand by a ¡retired¢ lepidopterist, entomologists and botanists throughout the country were alerted, and are actively checking both Echium and Myosotis spp. for signs of the moth. The point I want to make is that if we don¢t find anything on Myosotis, no-one will know we have looked, and Louda et al. will continue to assume that it is only in areas like Florida Keys that observant ecologists would have found Cactoblastis on a rare, native Opuntia. And if someone says, ¡then why don¢t you publish it¢, I¢ll need some suggestions on how to quantify lots of people looking at lots of plants over a longish period of time! My contention is that in most countries where biological control of weeds is practised widely, there are botanists around who keep a pretty close watch on their rare natives. I¢d certainly be pretty worried about being caught out if I released something that looked like it would attack a native. We are regularly presented with mis-identifications: for example, a close relative of gorse spider mite on native Coprosma spp., native seed weevils (Peristoreus spp.) on native brooms (Carmichaelia spp.). I am not advocating complacency. I do believe we should seriously look for non-target impacts. I am just not convinced that for weeds, anyway, the problem is anything like as bad as the scientific community is currently being led to believe, based on the undisputed impact of two agents that were known to have a broader host range than the target species at the time of introduction. Pauline Syrett Subject: Comments to various contributions by our colleagues Date: Mon, 18 Jan 1999 12:29:46 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "workshop" 1- Dear J. Herr, nobody has anything against studies on nontarget effects, which would lead us to more caution and more detailed studies on the natural enemies before their release, in order to minimize their risk. Fortunately, the examples that you mentioned as having negative impact on native vegetation, were from the old days, when the nontarget effect was not considered to be important and the agents would not be released today (As you said correctly). In addition, they are using some of the native plants as marginal hosts. As Russell Messing indicated correctly, we should consider also the cost/benefit of our releases and as Pauline Syrette said, the problem of nontarget effect is not as bad as the scientific community is currently being led to believe. 2- The comments by Pauline Syrette, Russell messing and Tom McAvoy are excellent contributions. Subject: Re: A New Zealand perspective Date: Mon, 18 Jan 1999 09:45:27 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: bc-ntimpact@UDel.Edu Pauline Syrett wrote: > > The point I want to make is that if we don¢t find anything on Myosotis, no-one will know we have looked, and Louda et al. will continue to assume that it is only in areas like Florida Keys that observant ecologists would have found Cactoblastis on a rare, native Opuntia. > I am not advocating complacency. I do believe we should seriously look for non-target impacts. I am just not convinced that for weeds, anyway, the problem is anything like as bad as the scientific community is currently being led to believe, based on the undisputed impact of two agents that were known to have a broader host range than the target species at the time of introduction. > > Pauline Syrett Well, about that cactoblastis ... it is also wiping out common varieties of Opuntia in South Florida. Our nice prickly pear melts and dies. Bad enough when a rare plant vanishes ... to render scarce a common plant is an appalling act. I understand that it was introduced here accidentally, though probably from someplace where it was deliberately introduced. Ron Boender at Butterfly World (in Pompano Beach) complains long and loud about parasitic wasps released in his area to benefit the farmers. That might be a interesting place to do studies, especially at this time of year. I, too, think our real problem is going to be the insects that eat other insects ... there seems to have been an agreement until recently that anything that eats insects is perfectly wonderful. Glad you're on the job. Anne Kilmer south Florida Subject: Re: A New Zealand perspective Date: Tue, 19 Jan 1999 14:01:26 GMT+1200 From: "Karyn Froud" Reply-To: bc-ntimpact@UDel.Edu Organization: HortResearch, Mt Albert, N.Z. To: bc-ntimpact@UDel.Edu Anne Kilmer wrote: Ron Boender at Butterfly World (in Pompano Beach) complains long and loud about parasitic wasps released in his area to benefit the farmers. That might be a interesting place to do studies, especially at this time of year. I, too, think our real problem is going to be the insects that eat other insects ... there seems to have been an agreement until recently that anything that eats insects is perfectly wonderful. Glad you're on the job. In New Zealand we are equally careful when introducing a new biological control agent for insects as we are for weeds, and have very tight regulations regarding non-target impacts. This has not always been the case however and many of the common examples of non-target effects that are used as arguments against new introductions come from a time when there were few restrictions on bc agents. Almost all the non-target impacts reported on native or beneficial insects in NZ were actually predicted at the time of introduction (mostly the 1930's). So I think that at least in New Zealand we have learnt from this and no longer allow potential benefits to outweigh likely risks. The problem still remains however, on how to determine host- specificity and prove negliable risk to an organism within the confines of an insect quarantine facility, with all the inherent difficulties of caging and confinement altering both parasitoid and host behaviour. And how much non-target parasitism is too much? Any? >5%? What about choice versus non-choice host testing, are these valuable tools or irrelevent tests? I believe that the major difficulty is that no two cases are, or can be treated the same and what an experimental design may show for one introduction may be completely impractical for another. So are there any basic standard procedures that can be implemented as a starting point for all introductions? I think that the weed bc people are a long way ahead of us insect bc'ers in this respect. Cheers from sunny NZ, Karyn Froud ------------------------------------------------------------------------ Karyn J Froud Entomologist HortResearch, Mt. Albert Research Centre, Private Bag 92169, Auckland New Zealand Phone:+64-9-8493660, Mobile:025 2931080 Fax:+64-9-8154202,E-mail:kfroud@hort.cri.nz Subject: Re: Predicting impacts on non-target species Date: Tue, 19 Jan 1999 18:48:12 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting from Howard Frank Wed, 13 Jan 1999 15:26:59 -0500: Howard Frank wrote: > 2. You wrote of generalizing that the biological > control agents are not species-specific in their > region of origin (and presumably not bothering to > document those that ARE species-specific). But > perhaps species-specific agents should be documented > too as part of the record - with any indication that > any of these has switched its diet after importation. > How about those that are genus-specific or tribe- > specific in their region of origin - and by this > level of specialization avoid affecting non-target > organisms in their new territories. Although I would argue that there are few insects that will have no potential non-target hosts in the area of origin, these few should be considered first in any introduction program. This may be my bias from work with natural enemies of insects. For natural enemies of weeds, it appears possible to find tribes and even families that occur in one region and not in another. However, my point was actually about how to go about assessing risks of impacts and the difficulty of doing this before introduction and thus the desirability of a solid body of evidence concerning post-introduction non-target impacts for a variety of systems. > 3. What about documentation of the positive effects > of the biological control agents? For example, > importation of U reduced populations of V (its > target) but this had a positive effect on native > plants W and X (which are attacked by V), and a > negative effect on native birds Y and Z (which use > V as part of their diet). > 4. And, if U successfully reduces populations of > V, then fewer pesticidal applications are used > against V, and there are positive effects on a > range of other non-target organisms. Of course documentation of positive effects it worthwhile, but the point of this workshop is to develop a research agenda for assessing and reducing negative impacts. > Examples such as the imaginary example above > can get very complicated - expensive and time- > consuming to document, and difficult to present > simply. I agree that it would be useful to > document them, but let us not jump in expecting > to get simple measures that can be reduced > easily to frequency and magnitude. If it were easy, it would already have been done. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Predicting impacts Date: Tue, 19 Jan 1999 18:55:19 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Responses to posting from Alan McClay Wed, 13 Jan 1999 14:48:32 -0700: Alec McClay wrote: > > Even predicting the impact on _target_ species is beyond our power! At present, but perhaps not forever. However, we can measure impact after introduction. It would seem worthwile doing this for non-targets as well ast targets. > I agree. However, in addition to evaluating the actual impacts I think it > is important to compare them with any predictions that may have been made > before release of the agent. In weed biocontrol, there have been many > agents approved for release despite the fact that testing showed that some > level of attack might occur on particular non-target species. These > releases may have been approved because (a) attack on non-targets was > considered to be an artefact of test procedures, (b) because impact was > expected to be much lower than on the target, (c) because the non-target > species were not considered important, or (d) because of the urgent need > for control of the target species. [E.g. Hasan & Delfosse (1995) for attack > by a rust on native Australian Heliotropium sp., by Hill & Hulley (1995) > and Olckers et al. (1995) for attack by biocontrol agents for introduced > Solanum spp. on eggplant (Solanum melongena L.) in South Africa, by Kok et > al. (1992) and Blossey et al. (1994) for attack by biocontrol agents for > Lythrum salicaria L. on native Lythraceae in North America, and by Willis > et al. (1996) for attack by an eriophyid mite on native Hypericum species > in Australia.] Cases such as these give a clear pointer as to where we > should be looking for non-target impacts. It should be relatively > straightforward to do the necessary monitoring, at least at the level of > comparing rates of attack on targets and non-targets. I quite agree that one should look at cases where predictions were made before introduction to see whether they have been born out after introduction. However, if it is straightforward to do this, why hasn't it been done? How would one go about measuring impact on population sizes of non-target species post-introduction. What would one use for controls? I'm pursuing some ideas about this, but am interested in seeing how others are approaching this problem. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Keep risk-benefits linked Date: Tue, 19 Jan 1999 19:07:24 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Tim Kring Thu, 14 Jan 1999 16:01:13 -0600: Tim Kring wrote: > > Hopper wrote: > "do you think it possible to assess impact of introduced biological control > agents on non-target species and to develop ways of reducing the likelihood > of such impacts, independently of evaluating risk-benefits of all > strategies? " > > Yes....BUT. I believe to focus on the impact on non-target species > without concurrent evaluation of the natural enemies impact in a > risk-benefit context does a disservice to biological control. > As an aside, as biological control workers we do a poor job of evaluating > the impact of our introductions on the TARGET...so I have VERY low > expectations for us providing valuable studies on often poorly known (or > unknown) species sometimes trophic levels away. I've thought some about quantitative risk-benefit analyses for biological control introductions and for other strategies. Here is what I concluded for a meeting in Australia last September: Economic analyses have been done of the benefits to agriculture of arthropod introductions for biological control (Tisdell 1990; Greathead 1995). However, although the need for weighing the risks versus benefits of biological control introductions has been emphasized by critics and practioners alike (Howarth 1991; Carruthers and Onsager 1993; McEvoy 1996; Simberloff and Stiling 1996), quantitative risk analysis is extremely difficult because there is so little evidence on which to base estimates of the risks of biological control or comparisons of risks among various alternatives in managing pests. Such analyses are complicated for any form of pest management and especially complicated for biological control introductions. A major problem is defining criteria for what are good versus bad impacts. For example, biological control of a pest may reduce competitive suppression of other plants or insects which then become pests. Thus, although the target may no longer be a pest, the agroecosystem or nature reserve has gained nothing by such an introduction. On the other hand, competition from introduced natural enemies which reduces populations of native natural enemies may in some cases be useful when the native natural enemies themselves threaten rare species, especially if the native natural enemy densities have been inflated by previous human activity. Another problem is that risks and benefits are often unequally distributed among geographical regions and segments of society. Furthermore, decisions are made for political units, yet because of dispersal, introductions affect biogeographical units. Probably the greatest barrier to risk analysis is the lack of adequately predictive theory in population, community, and evolutionary ecology. This lack explains to a great extent the emphasis on host range screening as the major tool for determining whether introductions are advisable: at least one can make some attempt at measuring host range. What do others think? Are we likely to actually be able to evaluate risks/benefits of various strategies in a common currency in such a way that we can actually compare strategies quantitatively? Does anyone know of examples of introductions where this has been done? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Trans-genic organism Date: Tue, 19 Jan 1999 19:36:51 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Tom Unruh Thu, 14 Jan 1999 15:15:53 -0800: Tom Unruh wrote: > > I would point out to Dr. Hopper that running a listserver for scientists is > much like herding cats. Best to let them have their way and then they may > settle down. I have 5 cats. They don't herd at all, but they will accompany one, if one goes to interesting places. I hope this workshop will go to sufficiently interesting places. > If we are to confine our discussion to risk rating arthropod introductions, > a workably narrow topic, and ignore the reasons for undertaking the > biological control introduction in the first place then I offer the > following questions laced with opinion. > > DOes the literature on host range provide anything useful? > Some like Shaw and Hawkins would argue that it is almost worthless. I > disagree. I believe it is especially useful for identifying species that > have very narrow host ranges. Doubt this question has a simple answer. In several recent projects, I've found complexes of sibling species (e.g., Aphelinus spp., Aphidius spp.) that greatly complicated interpretation of the literature on their host ranges. Other groups seem to be much better understood. Would it be possible to catagorize taxonomic groups for which the literature is or is not likely to prove very useful? > For sake of further simplicity lets say that we can divide the world of > candidate species into those with narrow host ranges and those with fairly > broad host ranges (say within a couple of families of hosts in the case of > insects, and a genus or two in hte case of weed agents). Given that, do the > readers believe that there is some difference in the standards of evidence > required for each type of species in order to risk rate their possible > nontarget problems in each of these type of species? It would appear that risk will always depend on the target environment as well as the candidate. An herbivore that attacks species in a family that does not have native species in North America would not present much risk here, but might be a problem in South Africa if the plant family does occur there. When you say standards of evidence, do you mean for host range tests actually to be conducted or for analysis of the implications of the host range, however host range was determined? > What do we accept to be a negative nontarget effect? It must devolve to > some effect on population dynamics of the nontarget but when do we consider > it biologically meaningful? I do not think we can do this with any rigor at > all because we don't know enough about the ecology at the outset and because > the potential nontargets are a new spectrum of species in new habitats. I > believe it has more to do with the taxonomic distribution of the existing > host range. In other words are their taxa of concern in the nominal host > range distribution? Are we worried about nontarget attacks of aphids or > scale insects? Of course from a biodiversity viewpoint we all probably are > if we think extinction of nontargets is going to be possible. Barring that, > then what? I think most would agree that the non-target effects of concern are changes in distribution and abundance of non-target species (although one might also changes in gene frequencies, this may be far too problematic). However, I would argue that judging what are acceptable or unacceptable changes in distribution and abundance are societal decisions, and thus not really up to biocontrol practioners or their critics. Rather this will be decided by our legislatures, courts, and ultimately by the people as a whole. Admittedly, the decisions will be slow in coming and often contradictory. Our role is to provide the best quantitative estimates of impacts possible. Of course, there are thresholds below which changes in distribution and abundance will be empirically undetectable. What are these thresholds likely to be? > How does one make host range data meaningful? I believe it must be > interpreted with the host utilization behavior in mind. For example a > parasitoid that attacks a fairly small group of hosts, but ones which are > hidden away in refugia such as bark, are likely to make some oviposition > mistakes which will expand the host range. Also the researcher is very > likely to make some rearing/identification mistakes further expanding the > host range. SO how do we cull spurious rearing records and > misidentifications from the global literature? Can we somehow rank the > strength of rearing data based on percent parasitism or percent utilization? > Must we redo rearing studies and create large data bases for all species > prior to introduction? Or can we confirm field host range data from > laboratory studies for those species where we are in a grey zone? How do we > identify the grey zone? The answers to most of these questions are going to be system specific and no doubt subject to debate. However, in the process of making these determinations and then testing how well they worked, we should build up the data needed to generalize. Again, retrospective analyses seem a very good shortcut to getting some of these questions answered relatively quickly. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Predicting impacts Date: Wed, 20 Jan 1999 15:59:35 GMT+1200 From: "Karyn Froud" Reply-To: bc-ntimpact@UDel.Edu Organization: HortResearch, Mt Albert, N.Z. To: bc-ntimpact@UDel.Edu Keith Hooper wrote: I quite agree that one should look at cases where predictions were made before introduction to see whether they have been born out after introduction. However, if it is straightforward to do this, why hasn't it been done? ** In New Zealand this type of post-introduction research has largely been ignored due to funding constraints, it is hard enough to find the money for research on the success (or not) of the bc agent on the target host. Increasing the profile of bc agents going wrong may actually help in increasing the funding avaliable for these types of studies (while also probably destroying the chances of any new introductions). And... How would one go about measuring impact on population sizes of non-target species post-introduction. What would one use for controls? I'm pursuing some ideas about this, but am interested in seeing how others are approaching this problem. **We are currently going through the motions of applying for the release of a new bc agent for a thrips species and through a post graduate research project we managed to get population dynamics sampling of both the host and the potential non-target thrips (only one thrips is likely to be impacted) before introduction so that we could assess the post-release impact on the populations (money and release acceptance allowing). In this case it was quite easy as the host records show that there is only one thrips species to worry about however, this will all fall down if there were unpredicted host shifts onto other thrips species (I'm very confident that will not be the case). I believe our biggest restraint is the funding to pursue this kind of work. Yours Karyn ------------------------------------------------------------------------ Karyn J Froud Entomologist HortResearch, Mt. Albert Research Centre, Private Bag 92169, Auckland New Zealand Phone:+64-9-8493660, Mobile:025 2931080 Fax:+64-9-8154202,E-mail:kfroud@hort.cri.nz Subject: Date: Wed, 20 Jan 1999 08:15:39 +0200 From: Dan Gerling Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear Colleagues I am intrigues how excellent a discussion group this has turned out to be. It were very beneficial if, at the outset, the different ideas and agruments would be collated into a volume (not too thick), becuase it is rare that so many practical examples and so many ideas from pepole who are familiar with the work, are exposed. Dan Gerling Subject: Risk-benefit context Date: Wed, 20 Jan 1999 15:08:24 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "Workshop" Keith Hopper commented on the message of 13 Jan. 99, by Howard Frank with the following statement on Jan. 20: "Of course documentation of positive effects is worthwhile (Only worthwhile ? I would say it very essential), but the point of this workshop is to develop a research agenda for assessing and reducing non-target impacts". I would like to make the following comments: I suggest to change the subject of the workshop. It is true that the non-target effect was the subject of the workshop, but as you see the majority of colleagues favor research on risk/benefit issues in bc. It is not fair to leave the field to the minority that is seeking only the negative effects of bc. It is like leaving the country to a few percent of the people of a country (Europe and elsewhere) to do what they want, bomb a train station, burn an opera house, kill people etc. I fully agree with colleagues such as Howard Frank, Tim Kring, etc. Doing bc is never 100% free of risk. just like bombing the military installations in Iraq. You may kill a few civilian people in the operation, but sometime you have to make a decision. I would like the silent majority to break their silence and talk. Who is in favor of changing the subject of the workshop to risk/benefits of bc. Actually we need people that evaluate the ecological effects of bc, both positive and negative effects, which could be done at the same time. When you reduce the population of an alien weed, it is expected that the native vegetation would be restored, which would be a positive effect etc. etc. Subject: RE: Predicting impacts Date: Wed, 20 Jan 1999 08:13:27 -0500 From: Jack Coulson Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I submit that it is also often difficult to present meaningful information on the impact of the target itself, a starting point for cost:benefit studies. Jack Coulson -----Original Message----- From: Keith R. Hopper [mailto:khopper@UDel.Edu] Sent: Tuesday, January 19, 1999 6:55 PM To: bc-ntimpact@UDel.Edu Subject: Re: Predicting impacts Responses to posting from Alan McClay Wed, 13 Jan 1999 14:48:32 -0700: Alec McClay wrote: > > Even predicting the impact on _target_ species is beyond our power! At present, but perhaps not forever. However, we can measure impact after introduction. It would seem worthwile doing this for non-targets as well ast targets. > I agree. However, in addition to evaluating the actual impacts I think it > is important to compare them with any predictions that may have been made > before release of the agent. In weed biocontrol, there have been many > agents approved for release despite the fact that testing showed that some > level of attack might occur on particular non-target species. These > releases may have been approved because (a) attack on non-targets was > considered to be an artefact of test procedures, (b) because impact was > expected to be much lower than on the target, (c) because the non-target > species were not considered important, or (d) because of the urgent need > for control of the target species. [E.g. Hasan & Delfosse (1995) for attack > by a rust on native Australian Heliotropium sp., by Hill & Hulley (1995) > and Olckers et al. (1995) for attack by biocontrol agents for introduced > Solanum spp. on eggplant (Solanum melongena L.) in South Africa, by Kok et > al. (1992) and Blossey et al. (1994) for attack by biocontrol agents for > Lythrum salicaria L. on native Lythraceae in North America, and by Willis > et al. (1996) for attack by an eriophyid mite on native Hypericum species > in Australia.] Cases such as these give a clear pointer as to where we > should be looking for non-target impacts. It should be relatively > straightforward to do the necessary monitoring, at least at the level of > comparing rates of attack on targets and non-targets. I quite agree that one should look at cases where predictions were made before introduction to see whether they have been born out after introduction. However, if it is straightforward to do this, why hasn't it been done? How would one go about measuring impact on population sizes of non-target species post-introduction. What would one use for controls? I'm pursuing some ideas about this, but am interested in seeing how others are approaching this problem. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Date: Wed, 20 Jan 1999 08:14:57 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Dan Gerling Wed, 20 Jan 1999 08:15:39 +0200 Dan Gerling wrote: > > Dear Colleagues > I am intrigues how excellent a discussion group this has turned out to be. > It were very beneficial if, at the outset, the different ideas and > agruments would be collated into a volume (not too thick), becuase it is > rare that so many practical examples and so many ideas from pepole who are > familiar with the work, are exposed. This is what we plan to do: assuming that there is sufficient material, it will be compiled, summarized and distributed to the participants and anyone else interested. Everyone who contributes will be listed, and where major contributions come from a single source, this will also be indicated. My thinking now is to distribute a hard-copy version and also make it accessible on the web. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: trophic levels Date: Wed, 20 Jan 1999 16:12:10 +0000 From: John LaSalle Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I have enjoyed the discussions. I would like to make a couple of comments of my own for people to think about. My main interest is biocontrol of insect pests, rather than weeds. One of my concerns is that most of the discussion concerning non-target effects of introduced natural enemies concentrates on the effect of these parasitoids on native phytophages. Although The following is an excerpt from a chapter on insect agrobiodiversity which I currently have in press: Firstly, the introduction of non-native species in biological control programs is not a problem, it is one possible solution. Such action is stimulated by the actual problem, generally in the form of an introduced pest species, a potential threat both to agriculture and biodiversity. Possible solutions include: chemical treatments, doing nothing, introducing exotic natural enemies, and control strategies which minimize chemicals and employ other methodologies such as cultural control. There are potential problems with all possible solutions, which could include environmental damage, loss of food to hungry populations, or simply that they do not work effectively. Indeed, the failure to introduce a biological control agent might have more devastating effects on biodiversity if it means that a starving population is forced to overexploit natural resources in search of food. So, despite the potential for non-target effect, it may be that introducing a natural enemy is still the most environmentally sound method of responding to a pest problem. Secondly, much of the focus of non-target organisms has been placed on the wrong trophic level. Although Howarth (1991) mentioned upper trophic levels, the predominance of the literature is directlly concerned with the threat of introduced natural enemies of herbivores to non-target herbivores. This may not be the most serious threat. The upper trophic levels, and particularly parasitoids, are more extinction-prone than their herbivorous hosts (LaSalle and Gauld, 1992; LaSalle, 1993; Unruh and Messing, 1993). When introduced parasitoids move over onto non-target hosts, there may be a greater likelihood that they endanger other native parasitoids through competition than they endanger non-target herbivores through regulatory pressure. Having shown that these native parasitoids represent a pool of potential biological control agents, we should take care not to expose them to undue threats. A good example of the decision making dilemma in species associated with non-target organisms again comes from the citrus leafminer. An undescribed species of Quadrastichus (Hymenoptera: Eulophidae) is known from Asia, where it has been considered to be a specific parasitoid of CLM. It was released, and is now apparently established, as an introduced natural enemy of CLM in Israel (Argov and Rossler, 1996). It was recently discovered that this species also attacks Liriomyza (Diptera: Agromyzidae) in Asia (LaSalle, unpublished), and it is being reared on Liriomyza sp. in quarantine in Spain (J. Jacas, Valencia, April 1998, pers. comm.). This presents an interesting dilemma. The fact that it can use Liriomyza as an alternative host makes it a very attractive parasitoid from the point of view of biological control. If weeds are allowed to persist in the vicinity of citrus groves, they will surely harbour populations of Liriomyza spp. which can serve as an alternate host to Quadrastichus sp., and this will benefit any attempts to control CLM using this parasitoid. This very attribute that makes Quadrastichus sp. so good from a pest management point of view makes it very bad from a conservation biology point of view. It is known that this is a polyphagous species, it is known that its introduction might threaten native parasitoids, and it is known that these native parasitoids are potentially very valuable as biological control agents. Balancing this type of conflicting information when making choices about which species of parasitoids should be introduced should become an important part of the decision making process in biological control. **NOTE NEW MAILING ADDRESS** As of the end of December, any post going to my old address (International Institute of Entomology, 56 Queen's Gate, London) will be returned to the sender. Please use one of the following addresses if you wish to send me post. ********************************************** John LaSalle Unit of Parasitoid Systematics CABI Bioscience UK Centre (Ascot) Department of Biology Imperial College at Silwood Park Ascot, Berks SL5 7PY UK or John LaSalle CABI Bioscience c/o Department of Entomology The Natural History Museum Cromwell Road, South Kensington London, SW7 5BD, UK ********************************************** Subject: Date: Wed, 20 Jan 1999 16:24:05 +0000 From: John LaSalle Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Sorry that last message went off before I really wanted it to (I hit the send button by accident). The point I wanted to make was that indigenous parasitoids might be more at risk from introductions of parasitoids then their hosts. Russell Messing stated In response to Kieth's statement (below)..... we've published data which show that braconids imported to Hawaii for fruit fly control clearly have "population-level impacts" on non-target tephritids imported for biocontrol of weeds (i.e., lantana). Of course, the impact varies from site to site, season to season, etc. (maximum parasitism was 28%), and the gall fly is not a very effective biocontrol agent to begin with..... But what he didn't discuss, and what would be interesting, is if these parasitoids displaced any natives from those non-target tephritids. Of course, this is an area where we really do not have enough baseline data - we can't go surveying parasitoid faunas of every native pests in the case an introduced parasitoid moves over onto one of them. But we have seen that rare (previously unknown) indigenous parasitoids have become very important in providing fortuitous biocontrol of citrus leafminer (i.e. preserved biodiversity represents among other things a pool of potential biological control agents). We assume that rare parasitoids are more at threat than their hosts. We know that species introductions are a serious threat to survival of native species (Reid and Miller (1989: 51) stated that "Introduced species threaten 19 percent of all endangered, vulnerable and rare species of vertebrates", making them the third worst threat behind habitat loss and fragmentation (67%) and overexploitation (37%)). So, I have to wonder if the more important threat with bc introductions is to other potential biocontrol agents rather than their host species. **NOTE NEW MAILING ADDRESS** As of the end of December, any post going to my old address (International Institute of Entomology, 56 Queen's Gate, London) will be returned to the sender. Please use one of the following addresses if you wish to send me post. ********************************************** John LaSalle Unit of Parasitoid Systematics CABI Bioscience UK Centre (Ascot) Department of Biology Imperial College at Silwood Park Ascot, Berks SL5 7PY UK or John LaSalle CABI Bioscience c/o Department of Entomology The Natural History Museum Cromwell Road, South Kensington London, SW7 5BD, UK ********************************************** Subject: Re: Risk-benefit context Date: Wed, 20 Jan 1999 12:25:42 -0500 From: sdp@nersp.nerdc.ufl.edu (Sanford D. Porter) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu > Who is in >favor of changing the subject of the workshop to risk/benefits of bc. >Actually we need people that evaluate the ecological effects of bc, both >positive and negative effects, which could be done at the same time. When >you reduce the population of an alien weed, it is expected that the native >vegetation would be restored, which would be a positive effect etc. etc. I think that it is essential to consider risks along with potential benefits when considering ecological and economic impacts. That is the only way that risks can be considered in the proper context. Sanford D. Porter, PhD Research Entomologist ------------------------------------ USDA-ARS, CMAVE P.O. Box 14565 Gainesville, FL 32604 USA Office: 352 374-5914 Secretary: 374-5903 FAX: 374-5818 Internet: SDP@NERSP.NERDC.UFL.EDU ********Please note that my e-mail address has changed slightly************ ------------------------------------ Subject: Re: Risk-benefit context Date: Wed, 20 Jan 1999 11:37:10 -0700 From: "Alec McClay" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu In fairness to Keith, who set up this discussion group to address a fairly specific issue, I do not think we should change the topic. Clearly it is important to consider non-target impacts in a broader risk/benefit context, but if we broaden the discussion too much we risk not making any progress on the specific questions of how to conceptualize, measure and predict non-target effects. There are other lists such as BIOCONTROL-L available for discussion on the broader issues. At 15:08 99.01.20 +0300, you wrote: >I would like the silent majority to break their silence and talk. Who is in >favor of changing the subject of the workshop to risk/benefits of bc. Alec McClay Research Scientist, Biological Control of Weeds alec@aec.arc.ab.ca Alberta Research Council Phone (403) 632-8207 Bag 4000, Vegreville Fax (403) 632-8612 Alberta T9C 1T4, Canada Please note: On Jan. 25 1999 the area code will change from (403) to (780) http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML Subject: Re: Nontarget effect Date: Wed, 20 Jan 1999 14:12:18 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by R. Sobhian Fri, 15 Jan 1999 11:50:46 +0300: Rouhollah SOBHIAN wrote: > > Dear colleagues > > Here are a few comments on the subject : > 1- I response to Allan Deutsch, I agree with Keith Hopper. My colleagues in > the laboratory and I do not know of any case where well-intentioned > releases went very wrong. If an insect like Rhynocyllus controlled Carduus > thistles in large areas and attacked also a few native Cirsium species, it > is not a release that went very wrong. Because without the insect the > thistles would not only negatively effect on the Cirsium species, but the > thistles would replace or eradicate other native species also. > > 2- Keith, Yes any step towards reducing damage to the nontarget organisms > is desirable. But equally or even more important is the post release > evaluation, both in term of $ value and of ecological aspects. The studies > made by a student in Beltsville demonstrated that the rate of cost benefit > in biological control was 1:100. If we do not have this kind of data, we > will not have justification for what we are doing. If somebody could > demonstrate how many species of native plants were reduced or eradicated > due to alien weeds or how many of the native species mowed into an area > after an alien weed was controlled by an insect and how much was the > benefits of these introductions, then biological control would get more > support. Often the alien weeds are the worst enemies of the native > vegetation. Although economic analyses of the benefits of biological control have been published several times, it is true that these have concentrated on direct benefits to agricultural and have not usually addressed indirect or non-target benefits. However, these are extremely difficult to measure given that they involve 2nd or 3rd order interactions. I would be happy if we could in this workshop develop relatively simple measures of 1st order impacts on non-target species. > 3- This is a comment on Tim Kring message: > Yes it is very difficult to predict or study the nontarget effects of > released organisms, but this is not a good reason not to look for solutions > and not to try to find a way to do it. > > 4- This is in response to To Unruh: > You are right we need to consult the world literature. BUT, often there are > wrong or misleading information in the old literature, which is mainly due > to the misidentification of organisms. For example while I was working on > Bangasternus orientalis, a seed feeder on yellow starthistle, in Greece, I > found a citation in the Russian literature that said the species also > attacks safflower. I could not believe this and tested the insect. My field > and laboratory experiments showed that the insect does not oviposits on > safflower. The insect has been released in the USA in 1985 and it never > attacked anything except yellowstarthistle. Another species of the genus, > Bangasternus planifrons, attacks wild and cultivated safflower and probably > this was wrongly identified as B. orientalis. > In a second case, I found a weevil that attacks Tamarix and its larvae > makes galls on the roots of its host plant. This is Liocleonus clathratus, > a about 1 cm long weevil, which has been reported in the Russian literature > on camel thorn, Alhagi. Alhagi is a Leguminosae and has no relation with > Tamarix, which is a very separate group of plants. I discussed this matter > with Kovalev and other entomologists at the Zoological Museum in St. > Petersburg and E. Colonnelli, in Rome, and everybody agrees with me that > this must be due to a misidentification or mislabeling of the specimens. We > are going to test the insect on Alhagi. It appears that in some cases the literature on host ranges of candidates for introduction is flawed. Are there ways of telling this without extensive additional host range testing? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: and Another Viewpoint Date: Wed, 20 Jan 1999 14:23:03 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Tom McAvoy Fri, 15 Jan 1999 10:26:52 -0500: Tom McAvoy wrote: > > Colleagues, > Here's one article that did what Allen Deutsch suggested, catalog the > success & failures of weed Bio Control: > Crawley, M. J. 1989. The successes and failures of weed biocontrol using > insects. Biocontrol. News Info. 10:213-223. > This is 10 years old so it would be useful to update. > > Some examples of weed BC agents feeding on nontargets: > 1. Cactoblastis sp. has been found feeding on threatened Opuntia sp. in > south Florida (Kass 1990, Robertson 1990), > 2. C. quadrigemina was found feeding on Hypericum calycenum, an exotic > plant extensively used as a ground cover in northern California (Ehler 1991), > 3. R. conicus (one species L. T. Kok & I are working with) has been found > feeding on nontarget Cirsium sp. thistles (Strong 1997)& Louda (1998) These are examples of feeding, not of changes in distribution/abundance of the non-target species. Attack on non-target species has been well-documented, what is lacking are studies showing population level consequences of such attack. > But overall, weed bio control has had a very good track record very few > if any agents have 'host shifted' - switched to another host that was not > predicted in host feeding studies. Host feeding studies on R. conicus > showed that other Carduus and Cirsium speices are hosts of R. conicus. But > the benefit of reducing C. nutans populations has resulted in a very large > reduction in herbicide application. What are the cases where the host range after establishment was not predicted well prior to introduction? > More long term studies on changes in plant diversity after release and the > economic and ecological benefits need to be documented. One problem with > weed BC is often not having control data after release ie no plots or > areas where the agents are not released, but this could be solved with some > good planning. Do you have suggestions about how would to conduct a study of non-target impacts after release when there are no before-release data on the non-target species? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Weed bc non-target examples Date: Wed, 20 Jan 1999 15:04:18 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by John Herr Fri, 15 Jan 1999 13:46:17 -0800 (PST): John Herr wrote: > > As was pointed out by Tom McAvoy, there are a few examples of non-target > impact caused by weed biocontrol agents. Another is the Cinnabar moth, > which feeds on a native Senecio (Diehl & McEvoy 1990), and the Klamath > weed beetle, which feeds on a native Hypericum in addition to the > previously mentioned non-native ground cover species (Andres 1985). > > However, the critical issue is whether this impact is biologically > significant. There are at least two additional examples where this does > indeed appear to be the case: > > 1) The Cactoblastis moth has caused significant damage to a rare endemic > Opuntia cactus in Florida (Pemberton 1996 ?). However, in this case the > agent was not intentionally introduced to Florida, and may have come in > with a contaminated nursery shipment or on its own from release sites in > the Caribbean. (If it turns out to be the latter, this would be a good > argument to include dispersal potential in pre-release assessments). Why is Opuntia spinossima rare? It is not because of attack by C. cactorum because the cactus was rare long before C. cactorum was accidentally introduced. By the way, given the small number of plants in the population, it would seem fairly simple to prevent impact of C. cactorum on O. spinossima by using insecticides, physical barriers, or by hand-picking the insects from the plants. > 2) The introduction of the thistle head weevil Rhinocyllus conicus to the > US is definitely a case where biologically significant impact has occurred > in non-target populations. Louda (1998) has documented the expansion of > this weevil's host range into two native Cirsium thistles in Nebraska, and > she has shown through long term studies that seedling recruitment in > populations of these non-targets is limited by seed predation (Louda & > Potvin 1995). I am currently working on this issue with rare California > Cirsium species, and have found additional examples to add to the R. > conicus non-target host range list since this problem was first reported > by Turner et al. in 1987. Fortunately, it appears that a lack of > synchrony between the R. conicus oviposition period and the native Cirsium > flowering phenology limits the overall impact in the species I have > studied. However, since distinct host races of R. conicus are known from > Europe, is possible that strains more adapted to the novel California > hosts will develop over time, and if this occurs I would expect the > non-target impact to be much greater. First, my understanding is that R. conicus did not expand its host range in an evolutionary sense. Attack on native Cirsium species was understood to be likely before introduction. What I would like to see for R. conicus impact on Cirsium spp. is evidence that the weevil has or is changing the distribution/abundance of the non-target host plants and the magnitude of such changes. This is not a system I know well, but the evidence that Turner et al. (1987) Louda et al. (1997) and Louda & Potvin (1995) are not as strong as what would be required to demonstrate conclusively that the beetle is affecting distribution/abundance of non-target plant species. This is not to say that it does not have such impacts, I would just like to see stronger evidence. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: solid evidence Date: Wed, 20 Jan 1999 15:45:53 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Russell Messing Fri, 15 Jan 1999 16:08:33 -1000: Russell Messing wrote: > > In response to Kieth's statement (below)..... we've published data which > show that braconids imported to Hawaii for fruit fly control clearly have > "population-level impacts" on non-target tephritids imported for biocontrol > of weeds (i.e., lantana). Of course, the impact varies from site to site, > season to season, etc. (maximum parasitism was 28%), and the gall fly is > not a very effective biocontrol agent to begin with..... Certainly you and your colleagues have done excellent work on attack by introduced natural enemies on non-target species. However, as you yourself pointed out, it is not clear from your work whether Diachasmimorpha tryoni plays a major role in regulating Eutreta xanthochaeta (Duan et al. 1998 Env. Entomol. 27: 419-426). A maximum parasitism of 28% by itself does not demonstrate any impact on the abundance of the species parasitized, as researchers studying dynamics of target pests must woefully attest. > There are also a few other cases in Hawaii I could cite. "Population-level > impacts" are hard to define, but we'd do better to admit the impacts that > do occur, and discuss them in terms of cost/benefit or risk/benefit.... > rather than to deny them entirely and lower our credibility. I disagree that population-level impacts are hard to define. What I mean is long-term effects on the distribution and/or abundance of non-target species. However, I would agree that such effects are difficult to measure; indeed that is what this workshop is in part about. How can we measure effects of introduced natural enemies on the distribution/abundance of non-target species? I'm interested in any cases where such effects have been shown so I would like to have the cites on the other cases in Hawaii you mention. I should say that I did not deny that such impacts occur, I just failed to find strong evidence for them in the literature. Pointing out the weakness of evidence concerning impacts on distribution/abundance of non-target species and calling for more research does not lower our credibility at all. However, I do agree that we would be best served by more solid research and less polemics. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: A New Zealand perspective Date: Wed, 20 Jan 1999 15:54:48 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Pauline Syrett Mon, 18 Jan 1999 13:10:15 +1200: Pauline Syrett wrote: > One other observation that people may find interesting relates to the recent accidental introduction of the Echium leaf miner, Dialectica scalariella, to New Zealand. When this moth was first released in Australia, it was recognised that host tests indicated it might attack some Myosotis spp. So when, a couple of years ago, the moth was discovered in the northern part of the South Island of New Zealand by a ¡retired¢ lepidopterist, entomologists and botanists throughout the country were alerted, and are actively checking both Echium and Myosotis spp. for signs of the moth. The point I want to make is that if we don¢t find anything on Myosotis, no-one will know we have looked, and Louda et al. will continue to assume that it is only in areas like Florida Keys that observant ecologists would have found Cactoblastis on a rare, native Opuntia. And if someone says, ¡then why don¢t you publish it¢, I¢ll need some suggestions on how to quantify lots of people looking at lots of plants over a longish period of time! My contention is that in most countries where biological control of weeds is practised widely, there are botanists around who keep a pretty close watch on their rare natives. I¢d certainly be pretty worried about being caught out if I released something that looked like it would attack a native. We are regularly presented with mis-identifications: for example, a close relative of gorse spider mite on native Coprosma spp., native seed weevils (Peristoreus spp.) on native brooms (Carmichaelia spp.). Your plans for evaluating host range and non-target impacts of past introductions sound excellent. For negative evidence concerning infestation of Myosotis spp., is there no way to get a rough estimate of the number of plants examined in total? Because if there were, you could calculate a threshold below which frequency of infested plant would have to fall to give the observed results. If this value is very low, most would agree that Dialectica scalariella is unlikely to have much effect on Myosotis spp. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: A New Zealand perspective Date: Wed, 20 Jan 1999 15:58:15 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Anne Kilmer Mon, 18 Jan 1999 09:45:27 -0500: Anne Kilmer wrote: > Well, about that cactoblastis ... it is also wiping out common varieties > of Opuntia in South Florida. Our nice prickly pear melts and dies. > Bad enough when a rare plant vanishes ... to render scarce a common > plant is an appalling act. Do you have evidence (i.e., experiments, life-history studies, etc.) or can you cite evidence for reductions in abundance of common Opuntia species being caused by C. cactorum? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: A New Zealand perspective Date: Wed, 20 Jan 1999 16:15:47 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Karyn Froud Tue, 19 Jan 1999 14:01:26 GMT+1200: Karyn Froud wrote: > The problem still remains however, on how to determine host- > specificity and prove negliable risk to an organism within the > confines of an insect quarantine facility, with all the inherent > difficulties of caging and confinement altering both parasitoid and > host behaviour. And how much non-target parasitism is too > much? Any? >5%? What about choice versus non-choice host > testing, are these valuable tools or irrelevent tests? I believe that > the major difficulty is that no two cases are, or can be treated the > same and what an experimental design may show for one > introduction may be completely impractical for another. So are > there any basic standard procedures that can be implemented as > a starting point for all introductions? I think that the weed bc > people are a long way ahead of us insect bc'ers in this respect. If the threshold for parasitism of potential non-target species is 0, then we will be doing few introductions in the future. Because of the complexity of factors affecting population dynamics, the amount of parasitism beyond 0 does not translate readily into level of reduction in host abundance. Thus, we cannot use a fixed threshold. Even if we knew a great deal about factors regulating the abundance of non-target species, it would still be difficult to predict impacts. Also, whether no-choice or choice tests will be most useful will depend at least on the spatial and temporal distributions of target and non-target species and on the foraging behavior and dispersal capacities of the candidates for introduction. For these reasons protocols for evaluating risks before introduction will have to be system specific. However, if we had a good set of case histories concerning impact after introduction for a variety of systems, we would be much better able to design protocols for evaluating risks before introduction. This is one reason I would really like to see some solid studies of impacts of introduced natural enemies on the distribution/abundance of non-target species and why I find the literature to date so frustrating. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: trophic levels Date: Wed, 20 Jan 1999 17:37:57 -0500 (EST) From: Roy G Vandriesche Reply-To: bc-ntimpact@UDel.Edu To: John LaSalle CC: bc-ntimpact@UDel.Edu John LaSalle mentioned that parasitoids are at risk of extinction when polyphagous parasitoids preempt their hosts. We have a research project underway in MA in the lab of Joe Elkinton (by Jeff Boettner) that suggest that highly polyphagous tachinid Compsilura conncinata (introduced against gypsy moth about 1905) has affected various silkmoths (Cercropia, Polyphemus etc) and that recoveries of some parasitoids associated with these species have not been made in the decades following the action of this tachinid (exact time frames vary by region). This may be a case of what LaSalle was discussing. Roy Van Drieche, University of Massachusetts Subject: Pre-release research Date: Wed, 20 Jan 1999 14:38:00 -0800 (PST) From: "Andre Gassmann (BPM-Weeds)" Reply-To: bc-ntimpact@UDel.Edu To: "bc-ntimpact@udel.edu" Dear colleagues, At this point of the discussion, I would like to make a few comments: 1)Everybody agrees that we need, a) post-introduction studies on the target pest and non target organisms on current programmes; b) retrospective analysis on previous programmes. 2) John Herr very rightly pointed that the critical issue is whether the non-target impacts are biologically significant. I would like to add that we need also to know whether these biologically significant non-target impacts are long-lasting, or whether the system will return to equilibrium after a relatively "short period of time". Any example ? Most people mentionned that we need to assess what the actual impacts have been in a variety of systems, and I do agree. For future introductions in particular, it is essential to compare these assessments with the outcome of pre-release researches. My main concern regarding this discussion is that I haven't yet seen many contributions on PRE-RELEASE research. To the participants of this discussion, I would like to ask questions like: 1) Do you feel there is a need to improve the methodology and design of host range testing, whether these are run in lab or in field conditions, in confined or open environments (especially in weed classical biological control)? 2) Do you feel there a need to improve the analysis and interpretation of the results of host range studies ? For example, how to interprete a 2% larval development on a non target host as compared to 30% on the target host ? The result will be statistically significantly different, but what is the biological mean of such results? Is there a realistic way to solve such a question ? 3) How essential are field studies and observations in the native area of the potential biocontrol agents studied ? Andre Gassmann ----------------------------------------------- Dr. Andre Gassmann CABI Bioscience Centre Switzerland (International Institute of Biological Control) Chemin des Grillons 1 CH - 2800 Delemont Switzerland E-Mail: a.gassmann@cabi.org Subject: Re: Risk-benefit context Date: Wed, 20 Jan 1999 17:46:58 -0500 From: Tom McAvoy Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I second Alec's motion. Stay with original intent of this workshop, measure and predict non-target effects. Too broad a topic will diminish any concrete conclusions. Tom McAvoy At 11:37 AM 1/20/99 -0700, you wrote: >In fairness to Keith, who set up this discussion group to address a fairly >specific issue, I do not think we should change the topic. Clearly it is >important to consider non-target impacts in a broader risk/benefit >context, but if we broaden the discussion too much we risk not making any >progress on the specific questions of how to conceptualize, measure and >predict non-target effects. There are other lists such as BIOCONTROL-L >available for discussion on the broader issues. > >At 15:08 99.01.20 +0300, you wrote: >>I would like the silent majority to break their silence and talk. Who is in >>favor of changing the subject of the workshop to risk/benefits of bc. > >Alec McClay >Research Scientist, Biological Control of Weeds > >alec@aec.arc.ab.ca Alberta Research Council >Phone (403) 632-8207 Bag 4000, Vegreville >Fax (403) 632-8612 Alberta T9C 1T4, Canada > >Please note: >On Jan. 25 1999 the area code will change from (403) to (780) > >http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML > > > > Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: Date: Wed, 20 Jan 1999 16:39:34 -0800 From: "Peter B. McEvoy" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Ketch Hopper wrote: I quite agree that one should look at cases where predictions were made before introduction to see whether they have been born out after introduction. However, if it is straightforward to do this, why hasn't it been done? How would one go about measuring impact on population sizes of non-target species post-introduction. What would one use for controls? I'm pursuing some ideas about this, but am interested in seeing how others are approaching this problem. My reply: 1. Defining expectations. Predictions made prior to release are often not published, so we go round and round on the issue of whether the outcomes observed post release were "expected." To counter this problem, we are forming a team of auditors to review and evaluate biocontrol decisions related to selecting target weeds, screening control organisms for safety and effectiveness, and importing and releasing weed control organisms. We will review proposals made to a federal regulatory authority (USDA-APHIS) over a 40-year span, and we will draft a "report card" on how effectively science is applied in making these decisions, where improvements might be made, etc. This "sociological research" is one form of learning from experience. 2.Predicting biocontrol. A set of explicit predictions were made and published prior to releasing the control organisms for purple loosestrife (see reference below) -- and a number of us are busy testing these predictions following release in North America. If pre-release predictions fail, post-release evaluation may tell us why. Malecki, R. A., B. Blossey, S. D. Hight, D. Schroeder, L. T. Kok, and J. R. Coulson. 1993. Biological control of purple loosestrife. BioScience 43:680-686. Cheers from rainy Oregon, Peter Dr. Peter B. McEvoy Professor of Ecology and Biological Control Rm 2046 Cordley Hall Department of Entomology Oregon State University Corvallis, OR 97331-2907 Phone (541) 737-5507 Fax (541) 737 3643 Email mcevoyp@bcc.orst.edu Subject: Re: Risk-benefit context Date: Wed, 20 Jan 1999 17:26:55 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu I concur with Sobhian. This discussion group should consider both positive and negative impacts of introductions and how we predict them beforehand because the standards of evidence for positive impact (control of a pestinsect or weed) and negative impact (control of a nontarget) are the same. They are ecological evaluations that must be (as pointed out by Hopper) interpretted economically and politically (because esthetic, moral, and other aspects may apply). One man's weed is another man's smoke or other trite phrases come to mind. There are lots of books and review chapters written on the topic of arthropod introductions from various perspective including: predicting pest status of newly colonized pests or predicting their pest potential before they colonize, predicting geographic range based on climatic adaptation, predicting impact based on population models (including work by Godfray and others for BC agents), relationship of establishment rates to inoculum size etc. etc. It is clear that we cannot address all aspects of the factors associated with colonization success and biological control success. The subject remains too broad. SO I pose the following question: If we can only predict colonization success very poorly how can we weight the predicted biocontrol benefits against the risks? Aren't we just participating in wichful thinking when we say "if it establishes, and if it controls the pest, then the benefits will be such and such?" I believe we must assume the same rule used by many doctors, first do no harm. The biological control benefits remain nebulous. SO imagine this scenario, we develop a model using life history and ecological data that indicates that species x will establish and produce some measurable ecological effect on species a. Now we transform that model to consider to consider a suite of notargets in turn and ask if it will produce a measurable effect on each of these. I wonder just how many of the pest systems we work in are known well enough to take this approach. Very few I would warrant (fruitflies in Hawaii and a few other discrete examples are possible). I would suggest that we address certain components of a risk rating cascade in order. I would suggest also that we come to some consensus on the value of these and identify research areas that are adressable in them. I apologize for being repetitive but I would suggest that host range is critical and with parasitoids and predators it (literature data) may be sufficient in many cases. What standards of evidence are required to make it sufficient before we must go into host range testing in the lab or in the native home? I believe we must take another approach. Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: RE: Nontarget effect Date: Thu, 21 Jan 1999 13:36:20 +1000 From: Day Michael Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu The problem facing many biocontrol workers is that we are often working on agents where initially little is known. Its not just that the literature is flawed but non-existent. This has occurred for many agents where the only reference(s) available is in the form of a checklist or when the agent was first described. Keith you may recall at the host specificity workshop in Brisbane last year we discussed many of the issues currently being aired including the questions that Andre raised. From the workshop we concluded that we could identify/predict reasonably well which non-target species are likely to be attacked using a combination of field observations/studies in the country of origin (which addresses the lack of information in the literature) as well as some rather well planned detailed host testing experiments in quarantine. As a result of the discussions, there was greater awareness in the design of such experiments so that we are in a better position to predict what will happen in the field if the agent is released. The process as I recall became rather extensive if we really wanted to be sure of the agent's potential host range. One area discussed in the workshop was the quality of the adults reared on the non-target and whether they were viable. Some adult development may be acceptable on non-targets if the resulting adults cannot develop eggs. Therefore non-targets may only to subject to occasional attack rather than supporting an increase in population of the agent. Andre your comment on the significance of 2% larval development is quite valid. It may appear that 2% is insignificant in the lab where only a few adults may emerge, but in the field when populations are much higher, this could translate to many more insects. Whether the damage to a non-target is significant is another question. For post release evaluation, the work being suggested by a number of contributors is very worthwhile. It is these studies which will then confirm how well we are conducting host testing and therefore how accurate are our predictions. If need be, we may then have to modify our host testing or pre-release studies. Regards Mike Michael Day Alan Fletcher Research Station PO Box 36 Sherwood Qld 4075 AUSTRALIA Ph: (07) 3375 0725 International: 61 7 3375 0725 Fax: (07) 3379 6815 International: 61 7 3379 6815 Email: Michael.Day@dnr.qld.qov.au Subject: Re: Nontarget effect Date: Thu, 21 Jan 1999 09:47:04 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: ---------- > From: Keith R. Hopper > To: bc-ntimpact@UDel.Edu > Subject: Re: Nontarget effect > Date: Wednesday, January 20, 1999 10:12 > > Response to posting by R. Sobhian Fri, 15 Jan 1999 11:50:46 +0300: > > Rouhollah SOBHIAN wrote: > > > > Dear colleagues > > > > Here are a few comments on the subject : > > 1- I response to Allan Deutsch, I agree with Keith Hopper. My colleagues in > > the laboratory and I do not know of any case where well-intentioned > > releases went very wrong. If an insect like Rhynocyllus controlled Carduus > > thistles in large areas and attacked also a few native Cirsium species, it > > is not a release that went very wrong. Because without the insect the > > thistles would not only negatively effect on the Cirsium species, but the > > thistles would replace or eradicate other native species also. > > > > 2- Keith, Yes any step towards reducing damage to the nontarget organisms > > is desirable. But equally or even more important is the post release > > evaluation, both in term of $ value and of ecological aspects. The studies > > made by a student in Beltsville demonstrated that the rate of cost benefit > > in biological control was 1:100. If we do not have this kind of data, we > > will not have justification for what we are doing. If somebody could > > demonstrate how many species of native plants were reduced or eradicated > > due to alien weeds or how many of the native species mowed into an area > > after an alien weed was controlled by an insect and how much was the > > benefits of these introductions, then biological control would get more > > support. Often the alien weeds are the worst enemies of the native > > vegetation. > > Although economic analyses of the benefits of biological control have > been published several times, it is true that these have concentrated on > direct benefits to agricultural and have not usually addressed indirect > or non-target benefits. However, these are extremely difficult to > measure given that they involve 2nd or 3rd order interactions. I would > be happy if we could in this workshop develop relatively simple measures > of 1st order impacts on non-target species. > > > 3- This is a comment on Tim Kring message: > > Yes it is very difficult to predict or study the nontarget effects of > > released organisms, but this is not a good reason not to look for solutions > > and not to try to find a way to do it. > > > > 4- This is in response to To Unruh: > > You are right we need to consult the world literature. BUT, often there are > > wrong or misleading information in the old literature, which is mainly due > > to the misidentification of organisms. For example while I was working on > > Bangasternus orientalis, a seed feeder on yellow starthistle, in Greece, I > > found a citation in the Russian literature that said the species also > > attacks safflower. I could not believe this and tested the insect. My field > > and laboratory experiments showed that the insect does not oviposits on > > safflower. The insect has been released in the USA in 1985 and it never > > attacked anything except yellowstarthistle. Another species of the genus, > > Bangasternus planifrons, attacks wild and cultivated safflower and probably > > this was wrongly identified as B. orientalis. > > In a second case, I found a weevil that attacks Tamarix and its larvae > > makes galls on the roots of its host plant. This is Liocleonus clathratus, > > a about 1 cm long weevil, which has been reported in the Russian literature > > on camel thorn, Alhagi. Alhagi is a Leguminosae and has no relation with > > Tamarix, which is a very separate group of plants. I discussed this matter > > with Kovalev and other entomologists at the Zoological Museum in St. > > Petersburg and E. Colonnelli, in Rome, and everybody agrees with me that > > this must be due to a misidentification or mislabeling of the specimens. We > > are going to test the insect on Alhagi. > > It appears that in some cases the literature on host ranges of > candidates for introduction is flawed. Are there ways of telling this > without extensive additional host range testing? I would say NO! If the wrong information is based on a wrong identification or on mislabeling there is probably no other way to verify it. > > Yours, Keith Hopper > > ****************************************************************** > Keith R. Hopper, Research Entomologist > USDA, ARS, Beneficial Insect Introduction Research Unit > University of Delaware, 501 South Chapel Street, Newark, Delaware > Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 > ****************************************************************** Subject: safety of bc agents Date: Thu, 21 Jan 1999 12:10:58 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "Workshop" A. Gassmann was asking what means 2% attack on nontarget versus 30% on target, which is significantly different. I would say anything over 0% is unacceptable, because these few organisms that developed on the nontarget may be genetically different and have broader host range. Thus, they may build up a population that would be well adapted to the nontarget. Subject: Re: Predicting impacts Date: Thu, 21 Jan 1999 10:52:19 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Karyn Froud Wed, 20 Jan 1999 15:59:35 GMT+1200: Karyn Froud wrote: > > Keith Hooper wrote: > I quite agree that one should look at cases where predictions were made > before introduction to see whether they have been born out after > introduction. However, if it is straightforward to do this, why hasn't > it been done? > > ** In New Zealand this type of post-introduction research has > largely been ignored due to funding constraints, it is hard enough > to find the money for research on the success (or not) of the bc > agent on the target host. Increasing the profile of bc agents > going wrong may actually help in increasing the funding avaliable > for these types of studies (while also probably destroying the > chances of any new introductions). I think that the current concerns about exotic species make a favorable climate for funding of retrospective and non-target studies. I disagree that increased funding for this will necessarily destroy the chances of new introductions; it depends on what we find out. > And... > How would one go about measuring impact on population > sizes of non-target species post-introduction. What would one use for > controls? I'm pursuing some ideas about this, but am interested in > seeing how others are approaching this problem. > > **We are currently going through the motions of applying for the > release of a new bc agent for a thrips species and through a post > graduate research project we managed to get population > dynamics sampling of both the host and the potential non-target > thrips (only one thrips is likely to be impacted) before introduction > so that we could assess the post-release impact on the > populations (money and release acceptance allowing). In this > case it was quite easy as the host records show that there is only > one thrips species to worry about however, this will all fall down if > there were unpredicted host shifts onto other thrips species (I'm > very confident that will not be the case). What I meant by phrased poorly, was how does one do retrospective analyses of past introductions where data concerning the status of non-targets was not measured before introduction so that one cannot make pre- vs post-introduction comparisons. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: RE: Predicting impacts Date: Thu, 21 Jan 1999 11:10:38 -0500 From: Tom McAvoy Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu In response to Keith Hopper: A few possiblities on how to predict impact on non-targets Although every weed and every insect offers unique and challenging possibilities no two cases are the same but some guidelines for all are possible: PRE-RELEASE: 1. Collect data on target & nontarget density (sq. m plots), diversity, biomass, seed production etc. Density of insect spp. in same feeding guild as intended BC agents. Threats to native insects need to be documented. Louda et al. (1997) reported a significant decrease in native picture-winged flies [Tephritidae: Paracantha culta (Weid.) and Orellia occidentalis (Snow)] found in Cirsium canescens Nutt. infested with R. conicus larvae. Many native insect spp. have been found on exotic weed spp. is this a benefit? Sounds like it. Oh what a tangled web nature is! 2. Economic impact of weed needs to be documented. This is one of the biggest holes in BC research. If we could document economic impact of the target weed we could better document benefits & I feel they are very high of BC. Resulting in better acceptance of BC & increased funding. HOST FEEDING STUDIES 1. Field studies in country of origin done with as many taxonomically related nontargets as possible cooperation needed here from botanist/plant ecologist to obtain seeds/ roots etc. 2. Quaratine studies in US. Continue host feeding tests. A host plant is one that allows insect to completely develop & lay viable eggs. Do you release an insect that has mult-hosts? Some feeding has been found to stimulate seed production ex.: soybeans. POST-RELEASE 1. Continue surveying plant diversity & density as described above along with introduced BC agent density Has the target species declined to the point of allowing non-targets to repopulate the habitat? Document economic impacts. But maybe the plant species that invades the habitat left vacant by the exotic is worse than the exotic!? Often as is the case with R. conicus the many potential hosts are not attacked due to poor host synchronization with the BC agent as was pointed out in an earlier email. ie Cirsium flower buds are present before or after the egg laying period thus avoiding attack by R. conicus. Most non-target species have no economic value associated with them except for their ecological value as part of a native community. How to asses that I'm not sure. But significant decline in density & reproduction due to BC agents would be cause for concern. Lauda has documented decline in native cirsium seed production but I have not seen any data on changes in plant density which is the bottomline. We're working with Lythrum salicaria & Carduus nutans. For L. salicaria we're slowly planting taxonomically related nontarget species in several loosestrife stands where the three Bio control agents are established Galerucella pusilla, G. calmariensis & H. transversovitattus. & sampling m2 plots for plant density & diversity we hope to do this for the next 10 yrs at least. Often the importment nontargets are not growing in the immediate area where the agents are established. We will be surveying thistle sites looking for nontargets in areas where the BC agents have been established for over 20 yrs. Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: Re: safety of bc agents Date: Thu, 21 Jan 1999 09:21:02 -0700 From: "Alec McClay" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu At 12:10 99.01.21 +0300, you wrote: >A. Gassmann was asking what means 2% attack on nontarget versus 30% on >target, which is significantly different. I would say anything over 0% is >unacceptable, because these few organisms that developed on the nontarget >may be genetically different and have broader host range. Thus, they may >build up a population that would be well adapted to the nontarget. If we eliminate anything which shows over 0% development on non-targets in testing, we will reduce the number of agents released drastically. Even in weed biocontrol, it is exceptional for pre-release tests to show that agents are absolutely monophagous on a single well-defined target weed species. I would expect that this is even rarer in the case of parasitoids, and almost unheard-of for predators. Normally there is at least some level of development on a few closely-related species. If any of these occur in the proposed area of introduction, it is then a judgement call as to whether they are likely to be attacked in the field, and if so how significant the consequences of such attack will be. In general we seem to have made those judgement calls quite well. The point about genetic variability is a valid one, however. What tests could one do to distinguish the cases where (a) all individuals in the population being tested have a 2% success rate on the non-target species, and (b) 2% of the population have a genetic difference which allows them to be 100% successful on the non-target? I think the risk of selecting a population adapted to the non-target would be much higher in case (b). Alec McClay Research Scientist, Biological Control of Weeds alec@aec.arc.ab.ca Alberta Research Council Phone (403) 632-8207 Bag 4000, Vegreville Fax (403) 632-8612 Alberta T9C 1T4, Canada Please note: On Jan. 25 1999 the area code will change from (403) to (780) http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML Subject: Re: Risk-benefit context Date: Thu, 21 Jan 1999 12:06:25 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Rouhollah Sobhian Wed, 20 Jan 1999 15:08:24 +0300: Rouhollah SOBHIAN wrote: > > Keith Hopper commented on the message of 13 Jan. 99, by Howard Frank with > the > following statement on Jan. 20: "Of course documentation of positive > effects is worthwhile (Only worthwhile ? I would say it very essential), > but the point of this workshop is to develop a research agenda for > assessing and reducing non-target impacts". > I would like to make the following comments: > I suggest to change the subject of the workshop. It is true that the > non-target effect was the subject of the workshop, but as you see the > majority of colleagues favor research on risk/benefit issues in bc. It is > not > fair to leave the field to the minority that is seeking only the negative > effects of bc. It is like leaving the country to a few percent of the > people of a country (Europe and elsewhere) to do what they want, bomb a > train station, burn an opera house, kill people etc. I fully agree with > colleagues such as Howard Frank, Tim Kring, etc. Doing bc is never 100% > free of risk. just like bombing the military installations in Iraq. You may > kill a few civilian people in the operation, but sometime you have to make > a decision. > I would like the silent majority to break their silence and talk. Who is in > favor of changing the subject of the workshop to risk/benefits of bc. > Actually we need people that evaluate the ecological effects of bc, both > positive and negative effects, which could be done at the same time. When > you reduce the population of an alien weed, it is expected that the native > vegetation would be restored, which would be a positive effect etc. etc. Rouhollah's proposal is somewhat ambiguous. Does he mean (1) that we should discuss yet again the pros and cons of biological control introductions based on the information currently available in the literature or (2) that we should discuss methods for evaluating benefits as well as risks of introductions? I strongly oppose number 1 because there have been many papers and workshops on this subject which have often cited the same shaky examples and covered the same reasoning. What we need is to delineate good methods and collect solid evidence concerning both risks and benefits. Because benefits have been much more studied than risks and risks are harder to evaluate, risks need more work and new and innovative methods. On the other hand, many of the same techniques can be used to evaluate benefits and well as risks so naturally both sorts of evaluation will profit from a discussion of research methods for retrospective studies on measuring impacts on abundances of target and non-target species, predicting such impacts before introduction, and reducing non-target impacts while increasing target impacts. Thus number 2 fits very well with the goal of this workshop. The workshop will remain a forum for discussing methods for evaluating impacts (positive or negative) of introductions, and not become yet another debate on the pros and cons of introductions. Those who do not like this constraint are free to start their own workshops. I trust that it will not be necessary to monitor every message before posting to ensure that we do not stray too far from the subject at hand. It is certainly a large enough subject without any expansion. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Date: Thu, 21 Jan 1999 12:21:48 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by John LaSalle Wed, 20 Jan 1999 16:24:05 +0000: John LaSalle wrote: > The point I wanted to make was that indigenous parasitoids might be more at > risk from introductions of parasitoids then their hosts. What is the evidence that parasitoids are more extinction prone than their hosts? Although it may be that the native parasitoids are more at risk, it would seem easier to measure impacts on non-target host species. Unless interactions between introduced and native parasitoids are behavioral, changes in non-target host species (e.g., abundances, stage structures, or phenologies) would be required to cause changes in abundances of native parasitoids and changes in host species would be easier to measure. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: safety of bc agents Date: Thu, 21 Jan 1999 09:57:16 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu The recent comment by SObhian highlights the very real difference in addressing this problem for arthropod versus weed bc. With weed agents, especially when the nontarget is a food crop or other desirable plant his statement is probably true. This may not apply at all when the nontarget is an arthropod of little importance. I posit that 30% may not even be important when the nontarget is a relatively undesirable microlep but that may just be my entomological bias against microleps. >A. Gassmann was asking what means 2% attack on nontarget versus 30% on >target, which is significantly different. I would say anything over 0% is >unacceptable, because these few organisms that developed on the nontarget >may be genetically different and have broader host range. Thus, they may >build up a population that would be well adapted to the nontarget. > > Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Re: Nontarget effect Date: Thu, 21 Jan 1999 10:19:27 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu >> > Tamarix, which is a very separate group of plants. I discussed this >matter >> > with Kovalev and other entomologists at the Zoological Museum in St. >> > Petersburg and E. Colonnelli, in Rome, and everybody agrees with me >that >> > this must be due to a misidentification or mislabeling of the >specimens. We >> > are going to test the insect on Alhagi. >> >> It appears that in some cases the literature on host ranges of >> candidates for introduction is flawed. Are there ways of telling this >> without extensive additional host range testing? > >I would say NO! If the wrong information is based on a wrong identification >or on mislabeling there is probably no other way to verify it. > > >> >> Yours, Keith Hopper It is clear from the literature on host range data that it is full of holes, that misidentifications of either the bc agent or their hosts are common, and that rearings are often not careful enough. Still, I suggest that iif one does not look at the details but instead thinks of the pattern we can identify whether a species is stenophagous, oligophagous, or polyphagous. In this form host range data is useful. In fact it should be conservative by representing species as more polyphagous than they really are. The latter can happen by having a complex of bc agents, or a misidentification of the bc agent in some studies. Data quality can also be judged by how many different studies of the system have occurred. In other words, the host range data should be our first cut, and subsequently should guide us in how to conduct quarantine-based testing. Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Date: Thu, 21 Jan 1999 10:44:00 -0800 (PST) From: "Andre Gassmann (BPM-Weeds)" Reply-To: bc-ntimpact@UDel.Edu To: "bc-ntimpact@udel.edu" I agree with Alec's and Tom's motions.I do not like to change the topic. In addition, we want to minimize the risks also when there will be no benefit from biocontrol. This happens sometimes unfortunately...... This consideration leads me to ask another question: Are the biologically significant non-target impacts which have been observed so far (and to a lesser extend all new associations recorded) usually associated with some level of success ?. This seems to be the case at least in weed biological control. Is this because successes are more attractive to researches than failures ? Or is this because of the demographic response (or other biological traits which would need to be study in depth in the context of understanding non-target impacts) of successful biocontrol agents ? That's all for today. Andre Gassmann Subject: Re: Date: Thu, 21 Jan 1999 10:26:58 -1000 From: messing@hawaii.edu (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu John LaSalle wrote: > >"But what he didn't discuss, and what would be interesting, is if these >parasitoids displaced any natives from those non-target tephritids. Of >course, this is an area where we really do not have enough baseline data - >we can't go surveying parasitoid faunas of every native pests in the case an >introduced parasitoid moves over onto one of them". actually, we have conducted fairly extensive surveys of several endemic tephritids as well as beneficial tephritids introduced for biocontrol of weeds. in both cases we found that there is a robust parasitoid fauna already attacking the non-targets, but it is composed primarily of "junk" species of chalcidoids that were accidentally introduced to the islands by unknown means. no endemic parasitoids were ever recovered. i agree with John that effects on the upper trophic levels can be important, but i disagree that it is impractical to do surveys and otherwise get reliable baseline information. these surveys can provide high quality data that may serve multiple purposes: (1) conducting specific non-target species risk analyses; (2) providing information on potential biocontrol agents for other areas/ cropping systems; and (3) generating a broader empirical database of parasitoid guild formation and host associations which can help solidify the theoretical underpinnings of biological control and pest management in general. Russell Messing Subject: Cost/Benefit Studies Date: Thu, 21 Jan 1999 15:34:13 -0500 From: Jack Coulson Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Tom McAvoy wrote: "2. Economic impact of weed needs to be documented. This is one of the biggest holes in BC research. If we could document economic impact of the target weed we could better document benefits & I feel they are very high of BC. Resulting in better acceptance of BC & increased funding." This was the point of my earlier comment about the difficulty of documenting economic impacts of BC targets. But from comments received re: cost/benefit studies, I admit that this expands the reason for this discussion group beyond the "simple" risk of BC introductions. But how does one arrive at weighing potential benefits versus risks of an introduction without some information on the "cost" of the target? Jack Coulson Subject: Re: A New Zealand perspective Date: Thu, 21 Jan 1999 15:47:54 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: bc-ntimpact@UDel.Edu, coxanne@ibm.net Keith R. Hopper wrote: > > Response to posting by Anne Kilmer Mon, 18 Jan 1999 09:45:27 -0500: > > Anne Kilmer wrote: > > > Well, about that cactoblastis ... it is also wiping out common varieties > > of Opuntia in South Florida. Our nice prickly pear melts and dies. > > Bad enough when a rare plant vanishes ... to render scarce a common > > plant is an appalling act. > > Do you have evidence (i.e., experiments, life-history studies, etc.) or > can you cite evidence for reductions in abundance of common Opuntia > species being caused by C. cactorum? > > Yours, Keith Hopper > > ****************************************************************** > Keith R. Hopper, Research Entomologist > USDA, ARS, Beneficial Insect Introduction Research Unit > University of Delaware, 501 South Chapel Street, Newark, Delaware > Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 > ****************************************************************** Subject: Cactoblastis Date: Thu, 21 Jan 1999 09:36:29 -0500 From: Anne Cox To: Anne Kilmer Anne, I have one paper on Cactoblastis in Florida: Habeck, D.H. and F.D. Berg. Cactoblastis cactorum Berg (LepidopteraL Pyralidae), a Phycitine new to Florida. FL Dept. Agric. & Consumer Serv. Division of Plant Industry. Entomology Circular # 333, Aug 1990. I am not certain whether a quantitative study has been conducted since then, but I can make some calls. I seem to recall that someone did a survey, whether it was published, I don't know. Let me know. Anne Cox Ph.D. Dear Anne, You are most kind. I believe the people in this news group would like further information. They're discussing non-target impacts of biocontrol introductions. I have noticed that the cactus exhibit in the botanical garden is pretty pitiful looking ... I remembered your pointing out the cacti at Jonathan Dickinson Park and attributing the damage to Cactoblastis. Thanks Anne Kilmer Subject: Measuring bc impacts Date: Thu, 21 Jan 1999 14:28:24 -0700 From: larbaker@wyoming.com To: owner-bc-ntimpact@UDel.Edu Kieth and All, I have enjoyed the debate. I am a practitioner of bc in Wyoming and a benefactor of the research many of you do. I have been trying to document the impact of the weed bioagents we have released over the last twenty years and am generally frustrated by the lack of statistically significant findings. I have made some observations which mostly lead to questions. I offer these observations to the discusion. 1)I have file drawers full of photos taken of release sites at the time of release. Few capture enough detail to actually show population level trends. I always seem to focus on the wrong thing. Thus, when impacts are observable, I do not have base line photos for comparison. I have run many transects. In a few cases they have been useful because the bc impacts have fallen on the transect, but usually, I can see some impacts nearby the area where data was collected, but to significant change has taken place along the transect. Now time may take care of this. If I live long enough, the bioagent may actually make a change. Thus, I am concerned that the time required to actually watch agent establishment and target/non-target impacts in the field will be so long as to be of little use from a management standpoint. 2) I released Aphthona nigriscutis on Leafy Spurge in 1990 at 18 sites in Fremont County. By 1992, I was able to confirm establishment at 7 sites and decided the rest had been poor locations. That may still be true, but last year, 1998, I confirmed establishment at 2 more sites for a total of 10 out of the original 18. After two years I was ready to delacle releases as usccess or failure and draw conclusions as to what habitats the insects preferred. Today, I do not know where they will go or what they like. I actually think, eventually, A. n. will populate all 18 sites. The difference in impact is dramatic. I have removed over a million insects for redistribution from 6 sites over the last few years, but at others I only get a few adults for 10 sweeps. Thus, any measurements of bc impact will be questionable for many years, until some degree of stability is reached. At least today, I hope that the impacts I see in a few locations will become more universal. 3) I have also released many other agents that have not established, or having established, have no effect. Yet, I have collected the data and photos. So one reason that there is little documentation of bc success and failure appears to me that it is a lot of work and most of what has been done fails to meet the tests for scientific significance and experimental design. And most importantly, it takes a lot of time, thus exceeding the life span of most graduate programs, the tenure of researchers at institutions, assignments within agencies, and even careers. Weed control is especially so as the spread of some species is too slow to measure from year to year and only produces heart burn when viewed decade to decade. 4) In Fremont County, there is a native, Euphorbia robusta, that was identified early on for host plant testing. I was able to collect enough of this plant to supply testing material for several investigators. In the lab, there is some feeding on robusta by Aphthona. I have watched this plant every year to see if there is any impact and have never seen any adults on the plants or evidence of feeding, until late summer when I took pictures of adults and feeding damage on robusta plants growing among leafy spurge that was being attacked by Aphthona nigriscutis. I still looked at 50 plants in other locations where there was no feeding and no nearby spurge. But, how many plants would I need to observe, and for how many years, before I could determine what the actual impact was? For critics of biocontrol, there would always be the chance that if we waited long enough, there would be an impact, and thus justification to block introduction. Leafy spurge will continue to spread, and I see no reason to believe that it will not eventually dominate all of the E. robusta habitat. Whether dispaced by leafy spurge or eaten by Aphthona, robusta is history. The case of death was the introduction of leafy spurge, not the control activity. Its only chance in Fremont County is the control of spurge. We do not have the option of erasing the introduction, only choices for management. 5) In 1978, Musk thistle was scattered across 60,000 acres of my county and totally dominated 5,000 acres to the exclusion of agricultural use of the land. We release R. conicus that year. In 1984, we released T. horridus. Both are wide spread. Today, I still have Musk thistle, but if you define a weed as a plant the interfers with management objectives, then Musk Thistle is not a weed. I have so few plants that it is hard work to collect enough insects for redistribution. I have watched other thistles for impact. R. conicus lightly attacks Canada thistle. I have observed hundreds of Cirsium undulatum, a native in the sagebrush community and even where it grows along side musk thistle, I have never seen R. conicus on the plants, feeding damage, or larvae in the heads. But, is that enough observation? How would I document my observations so I could say R. conicus does not attact C. undulatum? Cirsium foliosum, a native of high meadows is heavily attcked. In June the plants are black with adult R. conicus. The nearest Musk Thistle is miles away although Canada thistle may act as a bridge. Disection of mature flowers shows few or no developing larvae. There is seed production and this plant is a perennial. How many years would I need to monitor these plants, and at how many sites, before a population level trend could be measured? I am interested in the work. I want to do a better job. So, the out come of this discusion will be of use to me. But, while I watch the development of the bioagents and collect data with little usefulness, I also see some 20 species of exotic plants expand exponentially across the land and am simply not comfortable waiting to introduce bc agents until every possible impact has been quantified. Thus, I think we need to devise a test protocol that will try to reduce non-target impacts to the minimum without delaying introduction for decades. In most cases, political necessity will drive risk/benefit analysis to the point where bc releases will be made long before anyone can build the knowledge base needed to make scientifically based decisions. So how do you balance the equation? This is not an academic exercise. Biological control is a really important tool for weed control on rangeland. Making the choice to "take no action" is really choosing exotics over native species and monoculture over diversity. John L. Baker, Supervisor Fremont County Weed and Pest 450 N. 2nd St., Room 315 Lander, WY 82520 307-332-1052 Phone 307-332-1132 Fax larbaker@wyoming.com Subject: Re: Predicting impacts Date: Thu, 21 Jan 1999 16:56:23 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: bc-ntimpact@UDel.Edu Tom McAvoy wrote: > snip > Many native insect spp. have been found on exotic weed spp. is this a > benefit? Sounds like it. > Oh what a tangled web nature is! Probably it's not a benefit. Eumaeus atala, the atala butterfly, nectars freely on melaleuca and Brazilian pepper, both noxious exotics. Therefore it maintains a large population year-round, and is able to overwhelm and kill its host plant, a "nearly extinct" native cycad which is much propagated and used in landscapes. Causing a native insect to become "visible" to gardeners does not necessarily do it a favor. > > 2. Economic impact of weed needs to be documented. This is one of the > biggest holes in BC research. If we could document economic impact of the > target weed we could better document benefits & I feel they are very high > of BC. Resulting in better acceptance of BC & increased funding. > > HOST FEEDING STUDIES > > 1. Field studies in country of origin done with as many taxonomically > related nontargets as possible cooperation needed here from botanist/plant > ecologist to obtain seeds/ roots etc. > 2. Quaratine studies in US. Continue host feeding tests. > A host plant is one that allows insect to completely develop & lay viable > eggs. Do you release an insect that has mult-hosts? Some feeding has been > found to stimulate seed production ex.: soybeans. > > POST-RELEASE > 1. Continue surveying plant diversity & density as described above along > with introduced BC agent density > Has the target species declined to the point of allowing non-targets to > repopulate the habitat? Document economic impacts. But maybe the plant > species that invades the habitat left vacant by the exotic is worse than > the exotic!? > You got that right. Kill the water hyacinth, you get the elodea (well fertilized by the decomposition); kill the elodea you go to the yellow algae. The desired natives don't come back; you have to fetch them back. But nobody notices that our native parasitic wasps are being crowded out by the ones they imported to kill citrus leaf miner (for instance) because we, the ordinary gardeners, wouldn't recognize one if you handed it to us on a plate with parsley round it. We take them on faith. Why, when a local insect is attacking a new pest, do "they" feel it necessary to add an exotic wasp? Why, because the public cannot tolerate the signs of insect damage in the garden. The Department of Symbolism, in other words, has not yet been adequately clued in. This is why most of you are having to work with one foot in a bucket. > Often as is the case with R. conicus the many potential hosts are not > attacked due to poor host synchronization with the BC agent as was pointed > out in an earlier email. ie Cirsium flower buds are present before or after > the egg laying period thus avoiding attack by R. conicus. > > Most non-target species have no economic value associated with them except > for their ecological value as part of a native community. How to asses that > I'm not sure. But significant decline in density & reproduction due to BC > agents would be cause for concern. Lauda has documented decline in native > cirsium seed production but I have not seen any data on changes in plant > density which is the bottomline. > > We're working with Lythrum salicaria & Carduus nutans. > For L. salicaria we're slowly planting taxonomically related nontarget > species in several loosestrife stands where the three Bio control agents > are established Galerucella pusilla, G. calmariensis & H. > transversovitattus. & sampling m2 plots for plant density & diversity we > hope to do this for the next 10 yrs at least. > Often the importment nontargets are not growing in the immediate area where > the agents are established. > We will be surveying thistle sites looking for nontargets in areas where > the BC agents have been established for over 20 yrs. > > Tom McAvoy > Dept. of Entomology (0319) > Price Hall > Virginia Tech > Blacksburg, VA 24061 > Phone: (540)231-6320 > Fax: (540)231-9131 Good stuff. And the USDA is being more careful now, but the rush of accidental introductions is so great that what they're doing is overwhelmed. The whitefooted ant Technomyrmex albipes is going to be a major player in this battle; plants where it grows its herds usually seem to thrive. Interesting times, my friends. Anne Kilmer South Florida Subject: Re: Pre-release research Date: Thu, 21 Jan 1999 17:18:49 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Andre Gassmann Wed, 20 Jan 1999 14:38:00 0800 (PST) Andre Gassmann (BPM-Weeds) wrote: > 1)Everybody agrees that we need, a) post-introduction studies on the target > pest and non target organisms on current programmes; b) retrospective > analysis on previous programmes. > 2) John Herr very rightly pointed that the critical issue is whether the > non-target impacts are biologically significant. I would like to add that we > need also to know whether these biologically significant non-target impacts > are long-lasting, or whether the system will return to equilibrium after a > relatively "short period of time". Any example ? I'm inclined to argue that (1b) is likely to produce more information in the short term than (1a). In part this is because of point 2: we want to know what the long-term impacts are and such impacts may take a long time to unfold. Because we have introductions that were done a relatively long time ago, it makes sense to look at them in detail to see what has happened. The problem with this the lack of data before introduction for most systems. > Most people mentionned that we need to assess what the actual impacts have > been in a variety of systems, and I do agree. For future introductions in > particular, it is essential to compare these assessments with the outcome of > pre-release researches. My main concern regarding this discussion is that I > haven't yet seen many contributions on PRE-RELEASE research. Again, our ability to predict impact before release is hampered by our lack of strongly predictive ecological theory. This is why I emphasize analysis of past introductions: if the impacts of these can be fathomed, it should give us our best predictors. The best predictor of invasiveness of introduced plants is whether they have been invasive elsewhere. > To the participants of this discussion, I would like to ask questions like: > 1) Do you feel there is a need to improve the methodology and design of host > range testing, whether these are run in lab or in field conditions, in > confined or open environments (especially in weed classical biological > control)? Certainly this is so for insects introduced for control of insects and from what I've read it may also be true for insects introduced to control insects. > 2) Do you feel there a need to improve the analysis and interpretation of > the results of host range studies ? For example, how to interprete a 2% > larval development on a non target host as compared to 30% on the target > host ? The result will be statistically significantly different, but what is > the biological mean of such results? Is there a realistic way to solve such > a question ? Whether survival of 2% of the larvae on a non-target host is likely to reduce the abundance of the non-target will depends on a variety of factors, for example: (1) whether such survival would support a population of the candidate in the absence of the target, (2) how many populations of the target and non-target species are likely to persist in proximity for a long time, so that even without persistence on the non-target, immigration could maintain high levels of larvae on non-targets at high levels, (3) whether the population of larvae is sufficiently numerous to affect individual plants, (4) whether effects on individual plants translate into effects on populations of plants, (5) whether the 2% surviving are the result of selection of heritable variation in host range so that over time the survival on the non-target would increase. Thus, in some systems 2% survival would be too much and in others 30% survival would not pose a signficant risk. Even a test that shows 0% survival in the lab may not represent a sufficient sample of genetic variability in the field, and even a sample that shows 100% survival in the lab, may be irrelevant if females rarely oviposit on non-target species in the field. I would argue that all estimates of potential non-target impacts must take into account some detail of behavior and population dynamics of the candidate, target, and non-target species. > 3) How essential are field studies and observations in the native area of > the potential biocontrol agents studied ? Properly designed and interpreted, such studies are probably the best source of pre-introduction information available for many systems. Obviously, some knowledge of the environment of introduction is needed as well, but I think that studies in the area of origin have been woefully neglected, at least for introductions to control insects. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Date: Thu, 21 Jan 1999 17:40:18 -0500 From: "Keith R. Hopper" Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Peter McEvoy Wed, 20 Jan 1999 16:39:34 -0800: Peter B. McEvoy wrote: > 1. Defining expectations. Predictions made prior to release are often not > published, so we go round and round on the issue of whether the outcomes > observed post release were "expected." To counter this problem, we are > forming a team of auditors to review and evaluate biocontrol decisions > related to selecting target weeds, screening control organisms for safety > and effectiveness, and importing and releasing weed control organisms. We > will review proposals made to a federal regulatory authority (USDA-APHIS) > over a 40-year span, and we will draft a "report card" on how effectively > science is applied in making these decisions, where improvements might be > made, etc. This "sociological research" is one form of learning from > experience. How will you judge 'how effectively science was applied in making these decisions' given how little we know about non-target impacts? > 2.Predicting biocontrol. A set of explicit predictions were made and > published prior to releasing the control organisms for purple loosestrife > (see reference below) -- and a number of us are busy testing these > predictions following release in North America. If pre-release predictions > fail, post-release evaluation may tell us why. > > Malecki, R. A., B. Blossey, S. D. Hight, D. Schroeder, L. T. Kok, and J. R. > Coulson. 1993. Biological control of purple loosestrife. BioScience 43:680-686. Great! I'll take a look at this paper, and I look forward to reading what you find out. Meanwhile, are there sufficient pre-release data in areas of establishment? Have the agents been in place long enough to evaluate non-target impacts? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 South Chapel Street, Newark, Delaware Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: safety of bc agents Date: Thu, 21 Jan 1999 14:01:22 -1000 From: messing@hawaii.edu (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Thomas R. Unruh said: "This may not apply at all when the nontarget is >an arthropod of little importance". I posit that 30% may not even be >important when the nontarget is a relatively undesirable microlep but that >may just be my entomological bias against microleps". Tom- you have to be careful. the statements "an arthropod of little importance" and "undesirable microlep" are a clarion call to arms for many conservationists (including myself, to an extent). We can't be too anthropocentric or agropocentric (i just made that word up...). To do so fuels the fires of those railing against us, and undercuts the wholistic ecological perspective that should be a foundation of biocontrol. Subject: Risk benefit effect of bc introductions Date: Fri, 22 Jan 1999 12:00:16 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "Workshop" Dear Keith This is a response to your message of 21 Jan. 99. Sorry for not being able to express myself clearly and that you found my message ambiguous. I thought it was clear enough when I said :" We need people that evaluate the ecological effects of bc, both positive and negative effects, which could be done at the same time. When you reduce the population of an alien weed, it is expected that the native vegetation would be restored, which is a positive effect". Read it once more, I think that it is clear enough. However, what I mean is that we should plan to study both the negative and the positive aspects of bc and not only the negative aspects. This was also mentioned in my previous messages. I am very happy that you agree with my proposal in saying; " Thus number 2 ( 2= "we should discuss methods for evaluating benefits as well as risks of introductions") fits very well with the goal of this workshop. Regards, R. S. the trouble maker as Prof. A. Cameron named me!! Subject: Re: safety of bc agents Date: Fri, 22 Jan 1999 08:10:59 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: bc-ntimpact@UDel.Edu Agrocentric? I don't think you get that "po" in the middle. Other than that, I agree entirely. You have no way of knowing which microlep is the essential off-season food of some predator which, in its spare time, is a major predator for some "pest" organism, for instance. If native predators are eating an introduced "pest", then an introduced non-native predator tips the balance again. Good or bad? Who knows. We're trying to run a machine that we don't understand. If we remove all the parts whose use we do not see, we wind up with a wheelbarrow instead of our Mercedes. I think you have to expect that all non-target effects are bad, all unexpected effects are disasters, and all insects need many years of study before and after introduction. Anne Kilmer South Florida russell messing wrote: > > Thomas R. Unruh said: "This may not apply at all when the nontarget is > >an arthropod of little importance". I posit that 30% may not even be > >important when the nontarget is a relatively undesirable microlep but that > >may just be my entomological bias against microleps". > > Tom- you have to be careful. the statements "an arthropod of > little importance" and "undesirable microlep" are a clarion call to arms > for many conservationists (including myself, to an extent). We can't be too > anthropocentric or agropocentric (i just made that word up...). To do so > fuels the fires of those railing against us, and undercuts the wholistic > ecological perspective that should be a foundation of biocontrol. Subject: Re: Nontarget effect Date: Fri, 22 Jan 1999 09:39:15 -0600 (CST) From: narajm@wheel.datcp.state.wi.us Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu There has been several mentions of the impact of Rhinocyllus conicus on Carduus spp. Can somone point me to an article reporting on a scientific field study recording the reduction in the population density of Carduus nutans by R. conicus? Thanks! Julie Nara Wisconsin Dept of Agriculture, Trade & Consumer Protection 2811 Agriculture Drive Madison, WI 53708-8911 narajm@wheel.datcp.state.wi.us Tel: 608/224-4572 Fax: 608/224-4656 Subject: IOBC Symposium Date: Fri, 22 Jan 1999 11:07:14 -0500 From: Jack Coulson Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu May I assume that all correspondents to this workshop are aware of the IOBC Symposium on "Evaluating indirect ecological effects of biological control" to be held in Montpellier, France, October 17-20, 1999? If not, information on the symposium can be obtained from the Web - http://www.agropolis.fr/iobc Subject: Re: Cactoblastis cactorum Date: Fri, 22 Jan 1999 11:08:51 -0500 From: Howard Frank Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Colleagues: Cactoblastis cactorum has been mentioned in several postings. One by Anne Kilmer was: > Well, about that cactoblastis ... it is also wiping out common varieties > of Opuntia in South Florida. Our nice prickly pear melts and dies. > Bad enough when a rare plant vanishes ... to render scarce a common > plant is an appalling act. I urge anyone interested in this example to read Bob Pemberton's article: Pemberton, R.W. 1995. Cactoblastis cactorum (Lepidoptera: Pyralidae) in the United States: An immigrant biological control agent or an introduction of the nursery industry. American Entomologist 41: 230-232. Yours, Howard Frank Subject: Re: Predicting impacts Date: Fri, 22 Jan 1999 09:02:34 -0800 From: mark pomerinke Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Anne Kilmer wrote: >But nobody notices that our native parasitic wasps are being crowded out >by the ones they imported to kill citrus leaf miner (for instance) >because we, the ordinary gardeners, wouldn't recognize one if you >handed it to us on a plate with parsley round it. We take them on faith. >Why, when a local insect is attacking a new pest, do "they" feel it >necessary to add an exotic wasp? Why, because the public cannot tolerate >the signs of insect damage in the garden. >The Department of Symbolism, in other words, has not yet been adequately >clued in. This is why most of you are having to work with one foot in a >bucket. I have worked on citrus leafminer for five years, and little, to no evidence has been found to suport the above idea: that the imported biocontrol agent (Ageniaspis citricola) is crowding out indigenous parasitoid species in southwest Florida. Some evidence has been found indicating one indigenous leafminer may be attacked by the released control agent. However, no evidence has been found indicating it can complete devlopment on this leafminer, is effecting the population dynamics of the leafminer, or is "crowding out" indigenous parasitoids. I agree that this is an area in need of further examination. Second, indigenous parasitoids did expanded thier host range to citrus leafminer shortly after its arrival in Florida. However, given the low percentage of parasitism attributed to them during my studies, the best long term, ecologically sound method of control was the use of a classical biological control agent. Following the introduction of A. citricola, indigenous parasitoids (most if not all generalists) appear to have gone back to utilizing their orignal host range. Mark A. Pomerinke Graduate Research Assistant University of Florida SWFREC-IFAS 2686 State Road 29 N. Immokalee, FL 34142-9515 (941) 658-3451 map@icon.imok.ufl.edu Subject: Re: safety of bc agents Date: Fri, 22 Jan 1999 13:55:29 -0800 (PST) From: John Herr Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu On Thu, 21 Jan 1999, Russell Messing wrote: > ...the statements "an arthropod of > little importance" and "undesirable microlep" are a clarion call to arms > for many conservationists (including myself, to an extent). We can't be too > anthropocentric or agropocentric (i just made that word up...). To do so > fuels the fires of those railing against us, and undercuts the wholistic > ecological perspective that should be a foundation of biocontrol. I completely agree. The concept of the intrinsic value of native biodiversity is an important one. But this brings up a bit of a paradox in regards to the previous risk/benefit discussion: If it is acceptable to cause impact to native non-targets when using biocontrol to help preserve natural systems (e.g. combating an invasive weed), can we also argue that environmental damage caused by agents introduced strictly for agricultural pests is unacceptable? I think we can, (because the invasive weed would displace the native non-targets anyway if no action were taken), but this reasoning might be viewed as hypocritical by others. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= John Herr jherr@nature.berkeley.edu UC Berkeley, Division of Insect Biology (510)642-8414 USDA-ARS-WRRC, Plant Protection Research (510)559-5974 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Subject: Re: Nontarget effect R. conicus Date: Sat, 23 Jan 1999 15:16:35 -0500 From: Tom McAvoy Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu In response to Julie Nara: Here are papers on R. conicus controling Carduus nutans, musk thistle: Kok, L. T., 1998. Biological control of musk and plumless thistles. Recent Res. Devel. in Entomol. (2) 33-45. Kok, L. T. and W. T. Mays. 1991. Successful biological control of plumeless thistle, Carduus acanthoides L. [Campanulatae: Asteraceae: (=Compositae)], by Trichosirocalus horridus (Panzer) (Coleoptera: Curculionidae) in Virginia. Biol. Contr. 1:197-202. Kok, L. T. and R. L. Pienkowski. 1985. Biological control of musk thistle by Rhinocyllus conicus (Coleoptera: Curculionidae) in Virginia from 1969-1980. Proc. 6th Inter. Symp. Biol. Contr. Weeds. pp. 433-438. Kok, L. T. and W. W. Surles. 1975. Successful biological control of musk thistle by an introduced weevil, Rhinocyllus conicus. Environ. Entomol. 4:1025-1027. At 09:39 AM 1/22/99 -0600, you wrote: >There has been several mentions of the impact of Rhinocyllus conicus on Carduus >spp. Can somone point me to an article reporting on a scientific field study >recording the reduction in the population density of Carduus nutans by R. >conicus? Thanks! > >Julie Nara >Wisconsin Dept of Agriculture, >Trade & Consumer Protection >2811 Agriculture Drive >Madison, WI 53708-8911 >narajm@wheel.datcp.state.wi.us > Tel: 608/224-4572 > Fax: 608/224-4656 > > Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: Re: safety of bc agents Date: Mon, 25 Jan 1999 09:57:59 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: This is a message in response to A. Kilmer (see below) I believe that we can not say that ALL NON-TARGET EFFECTS are BAD !! At least in case of Wee biocontrol. As I mentioned earlier, if a weed feeder reduces the population of an alien weed, it is expected that the native vegetation would move in and restore the native vegetation, which is a positive thing. Consequently, there would be more biodiversity, less use of herbicides etc. Am I right ? ---------- > From: Anne Kilmer > To: bc-ntimpact@UDel.Edu > Subject: Re: safety of bc agents > Date: Friday, January 22, 1999 04:10 > > Agrocentric? I don't think you get that "po" in the middle. Other than > that, I agree entirely. You have no way of knowing which microlep is the > essential off-season food of some predator which, in its spare time, is > a major predator for some "pest" organism, for instance. > If native predators are eating an introduced "pest", then an introduced > non-native predator tips the balance again. Good or bad? Who knows. > We're trying to run a machine that we don't understand. If we remove all > the parts whose use we do not see, we wind up with a wheelbarrow instead > of our Mercedes. > I think you have to expect that all non-target effects are bad, all > unexpected effects are disasters, and all insects need many years of > study before and after introduction. > Anne Kilmer > South Florida > > Subject: Re: safety of bc agents Date: Mon, 25 Jan 1999 08:08:43 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: bc-ntimpact@UDel.Edu Rouhollah SOBHIAN wrote: > > This is a message in response to A. Kilmer (see below) > I believe that we can not say that ALL NON-TARGET EFFECTS are BAD !! That's not what I said. I phrased it very carefully. You need to work on the expectation that such effects will be bad. At > least in case of Wee biocontrol. As I mentioned earlier, if a weed feeder > reduces the population of an alien weed, it is expected that the native > vegetation would move in and restore the native vegetation, which is a > positive thing. That is a target effect ... the hoped for effect. And if you honestly expect that to happen, in the face of considerable experience to the contrary, you may be disappointed. Usually the first plants to move in to an empty space are weeds. nature uses the handiest seeds, and those may be the targeted plant, they may be other exotic weeds or they may (least probable) be a nice assortment of natives. Consequently, there would be more biodiversity, less use of > herbicides etc. Am I right ? In theory, yes. In practice, yuck. Here in South Florida, there is a wonderful assortment of exotic pest plants which will be happy to move in. Years ago, when the water district removed the Brazilian pepper growing on canal banks, they explained to us that now lovely natives could take over. They are now less naive. It is not enough to introduce biocontrol agents; in a thoroughly contaminated site, you have to actively garden the area, planting desired plants where you remove pests (or where your pet bugs are removing them). Mind you, I'm not opposing biocontrol. A variety of pest plants is better than a monoculture. But you've got to think it out. Years ago, when I was writing editorials for the Evening Times in Palm Beach County, I asked whether we couldn't just harvest and use the water hyacinths in Lake Okeechobee. (stock feed supplement, fertilizer, things like that) And they explained, quite seriously, that we would have to treat them with pesticides for that to be practical. Because of those expensive, carefully fostered biocontrol critters, of course. Water hyacinth is a very effective remover of heavy metals, nutrients etc. from water, and physically removing it from the water cleans the water. But noooo, we've got to use herbicides. This is because the public expects instant results, pest-free produce and a sanitized planet. The word isn't getting out, that biocontrol is the best way. Tom Fasulo's PestAlert does a brilliant job of explaining, in simple terms, what the scientists are doing in this field. For my part, I will stop spreading doom and despondency about the citrus leaf miner's predators. Clearly this is in good hands, after all. As for the notion that rare Opuntias can be protected from Cactoblastus by careful supervision of the plants, this is not practical. The larvae burrow into the plants. The adults fly. Nor would a preserve area be enhanced by the installation of row covers over the cactus. Anne Kilmer South Florida > ---------- > > From: Anne Kilmer > > To: bc-ntimpact@UDel.Edu > > Subject: Re: safety of bc agents > > Date: Friday, January 22, 1999 04:10 > > > > Agrocentric? I don't think you get that "po" in the middle. Other than > > that, I agree entirely. You have no way of knowing which microlep is the > > essential off-season food of some predator which, in its spare time, is > > a major predator for some "pest" organism, for instance. > > If native predators are eating an introduced "pest", then an introduced > > non-native predator tips the balance again. Good or bad? Who knows. > > We're trying to run a machine that we don't understand. If we remove all > > the parts whose use we do not see, we wind up with a wheelbarrow instead > > of our Mercedes. > > I think you have to expect that all non-target effects are bad, all > > unexpected effects are disasters, and all insects need many years of > > study before and after introduction. > > Anne Kilmer > > South Florida > > > > Subject: design Date: Mon, 25 Jan 1999 11:08:07 -0500 (EST) From: bernd blossey Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu It is about time that the biocontrol folks (and it looks like all subscribers are biocontrol folks) address this hot topic and try to develop potential solutions/scenarios for future work. If we don't come up with our own research plans we will have them shoved down our throats by folks that we consider non-sympathetic (rightly or wrongly) to our profession and suffer restrictions and a potential moratorium on introductions. Who wants that? Is it justified? Certainly, biological control needs an infusion of quantification, experimentation and long-term monitoring. It has always been a mystery for me how we could get away for so long with releases that did not involve quantitative assessments of their impact, particularly on the target. Imagine Monsanto is wanting to sell you a herbicide/insecticide and you have no idea whether it will work. Imagine even further that the Monsanto representative is trying to convince you to purchase the product based on the fact that this pesticide has been on the market for decades, that it works and that no adverse effects have been reported. All this based on some color photos and the assertion of independence - but without quantitative backup. Would you be enthusiastic about spending a few 100,000 dollars for this product? What we are asking of society, or agencies, is not too far removed from this hypothetical example. Can we complain about less than enthusiastic support? In many cases we do not even have the supporting evidence that we are addressing a real problem. How many control programs have started without an economic or ecological assessment of the impacts of a target weed or insect pest? Sure, we are only getting invloved because of complaints by those that need to manage pests, whether these are farmers or (increasingly) natural areas managers. And the only reason that we get a deal is because everything else has failed!! I think we need to do better than that - and I believe we certainly can. Who is to blame for the failure to conduct long-term monitoring at the release sites? We certainly have to accept a lot of the responsibility and the complaints that funding for this work is hard to come by is, at least to some extent, a cheap excuse. How often have we asked for that particular kind of support and made it absolutely clear how important it is? We are the specialists with knowledge of the control organisms and the target pests. We can't wait for an interested grad student to come along and stumble into the right place. We are the people that need to guide, if not develop and conduct such research ourselves. And it is not that hard - really. Particularly not in weed biocontrol where a lot of predictions on non-target impacts can be based on the host specificity testing. Let me outline (for weed biocontrol programs) how I believe we can conduct such work and become a more predictive and more successful science (and not the last resort). 1. Pre-release studies These have to be conducted in the area of origin AND in the future release area. Preferably we need to have some quantitative evidence of ecosystem impacts before we begin control programs (that should be a prerequisite for biological as well as any other control program). I would like to see this work done by ecologists/botanists and fortunately invasive species problems gain "sexiness" in ecology. I would reject claims that biocontrol workers should be the ones that need to collect these data. I strongly favor research in the area of origin that involves the impact and host specificity of potential control agents. Only if we have such information available can we then go out and design meaningful post-release protocols. In particular, we need to know how plant individuals respond to attack by single or multiple control agents. We may never be able to accurately predict the resulting establishment or impact in the release area (differences involving interaction of abiotic environmental factors, plant and insect genotypes, plant communities, and natural enemy complexes may make this impossible), however, the successful weed programs seem to indicate that herbivore attack is really the driving force immediately after the control agent release and at high weed densities. Without such information from the native range it will be very difficult to improve selection protocols. Predictions of potential non-target impacts for weed feeders are easily developed following the host specificity testing. Usually, potential control agents show a multitude of responses, largely depending on test designs (age of plants and insects, multiple- or no-choice,leaves or potted plants, greenhouse or field grown plants, cage size, field or tests in captivivity, adult or larval tests, genotype of insect or test plants, etc. come to mind). Without follow-up monitoring, however, we will never be able to evaluate which of our testing procedures have predictive ability. Assuming that we are not overlooking any serious population level effects of insects released as bc agents, our screening protocols are pretty good in preventing a switch to non-targets. Most of what has been reported was predicted, what is important is whether something escaped our attention. I am actually more concerned about the possibility that we are rejecting some perfectly safe control agents because of meaningless screening test requirements. 2. Post-release studies To evaluate the impact on the target weed, we need to develop standardized protocols that allow long-term monitoring. With funding from NBCI we have done just that for the purple loosestrife program. We have held workshops to instructed land managers in the use of the protocol and many people throughout North America are now monitoring control agent build-up, purple loosestrife decline, and the response of the plant communities at their release sites. Ideally, we would also incorporate assessments of other organisms (birds, amphibians, native insects, microorgnisms?) but that may be overly ambitious. Nevertheless it is possible. This protocol is available on my web site and allows comparisons and evaluations for single versus multiple agent releases. Most importantly, it links the control agents to changes in plant populations e.g. the decline of purple loosestrife. It is easy to design regional approaches with spacing of releases that prevent spill-over of agents to sites that are kept as controls (no releases of control agents). I don't think we need to worry about the inability to collect regional or local baseline data even if control programs have been implemented years ago. Landscape level occurrence of bc agents or control does not occur over night. For the purple loosestrife program the number of wetlands where insects were released is well over 1000 in 30+ states. However, there are still thousands of sites where pre-release data can be obtained. I guess that even with 30 years of biocontrol targeting a species such as leafy spurge I can find plenty of sites (particularly in the Midwest and Northeast) that are free of control agents. This is not to argue that such pre-release investigations should begin before the very first introduction into North America occurs. In addition, the longer the data were collected before control agnets were released, the better. But make no mistakes: nobody will design such protocols for you. And we can't expect that a person receiving a new species of insect he/she has never seen before is able and knowledgeable enough to design a meaningful monitoring protocol. It is up to us to develop these. And I am convinced if we design them userfriendly, they will be used. People are just waiting for this type of protocol. ACCEPT THIS RESPONSIBILITY! We are currently evaluating the potential of aerial surveys and spectral signals for the monitoring of control success in wetland plant communities. Although the resolution of these techniques is far from ideal, maybe we will be able to use these technologies much easier in the future. Again, I need to emphasize that on the ground monitoring will be necessary to assess control agent populations and the resulting changes in plant communities. Particularly the latter aspect needs to be stressed. Successful biological control does not necessarily imply that the replacement plant communities are any more desirable. Replacing one invasive plant by another is not really progress, just change. Maybe we need to integrate additional restoration efforts into our biocontrol programs. The time to do this is when weed populations decline and before a new invasive has taken a strong foothold. Only if we can document that our actions result in POSITIVE change will we gain greater acceptance. 3. POST RELEASE MONITORING FOR NON-TARGET EFFECTS I agree with Alec and Peter that this is actually straight forward. Our host specificity screening has identified a number of plants that may be attacked by our control agents. There are only a very few scenarios that come to mind to design appropriate protocols: 1. The potential host occurs in the same habitat/site as the target weed 2. The potential host does not occur in the same habitat For scenario 1 we select sites where target weed and potential non-target host co-occur, release control agents and follow the fate of target weed, potential non-target host(s) and control agents over time. Ideally we have several replicate sites, some of which serve as controls (no control agent release). We survey (I personally prefer permanent quadrats) for the presence of feeding, oviposition, larval feeding etc. on target weed and non-target hosts and ultimately for changes in population abundance. This work will continue for several years, ideally following population build-up of control agents and decline of target weed. Many people are concerned about non-target effects when populations of control agents are high and subsequently decimate their original host. By following this at sites through time we will be able to assess whether this occurs at all and whether the likelihood of a "spill-over" effect is dependent upon population levels of the control agents and food availability. This should also demonstrate that control agents are more likely to leave an area rather than consume unsuitable food, as is the case for host specific organisms in their home range (so we think but have not backed this up by data). In scenario 2 we are concerned with foraging control agent individuals that would encounter patches/individuals of non-target hosts and subsequently make a living on them. To investigate this scenario we could make experimental releases where such potential non-target hosts occur and sample for several years, or until control agents disappear (which may happen very quickly) using established protocols (transects, permanent quadrats, searches but something quantitative). I think it would be most appropriate to release around 1,000 individuals (takes into account genetic variation of the control agent) that are ready to begin or have started oviposition. It will be important to report the search effort or standardize to allow an assessment of how careful we looked. For plant species that were not tested in the screening protocols the situation is a bit more difficult, although the question is whether we can engage in potentially meaningless "fishing expeditions". I would advocate using the standardized monitoring protocols to incorporate observations of control agents on non-target hosts. Once we have such observations, more focussed work would be possible. One remark re. the scenario Pauline developed for NZ. I agree that it is sometimes difficult to get negative results into the journals. That appears to be changing (hopefully). One way of reporting largescale searches would be in person hours, non-target populations examined, etc. The only way of really convincing critics though is through long-term consistent monitoring. We have started this work with purple loosestrife in several areas. Populations of Lythrum alatum co-occuring with Lythrum salicaria were located in Wisconsin and Department of Natural Resources personnel has established permanent quadrats at control agent release sites. I finally found a site in New York where both Decodon verticillatus and Lythrum salicaria co-occur. We will release control agents at this site this summer and begin our longterm monitoring. These 2 species have been identified by our host specificity screening as the most likely candidates for host switching (but the probabilities for this occurrence is extremely low, thats why releases were approved). One of the lessons from our non-target monitoring should be kept in mind. Not every hole that you find in a leaf on a non-target plant has to be the result of feeding by biocontrol agents. All native plants will have their own feeding specialists and generalists. Lythrum alatum in Wisconsin was found to be attacked by as yet unidentified beetles, Decodon verticillatus is commonly attacked by at least one species of root feeder and several species of leaf feeders. This design would at least get to the first order effects. Addressing higher order or indirect effects will be a different story. I think it is time to go out and do it. And it is up to us to set positive precedence. We will be much stronger and more credible in the future. A last remark re. monitoring for non-target effects in insect biocontrol programs. It does not appear difficult to design such protocols, however, it may be quite elaborate. The lack of information guiding the searches due to lack of host specificity testing may make this particularly difficult. But I could imagine a routine collection effort of potential non-target hosts, followed by rearing of the appropriate life stage in the lab to assess parasitoid attacks. Alternatively we could use the deliberate exposure of non-target organisms at the right life-stage on a "trap-plant" in the right habitat or the habitat that we want to survey, allow parasitoid attack for a standardized time, remove trap-plants and check for emergence of parasitoids (or collect potential predators in specifically designed traps). This is just a start but I really think it is not at all that difficult. Bernd Blossey ****************************************************************************** Dr. Bernd Blossey Director Biological Control of Non-Indigenous Plant Species Program Department of Natural Resources Fernow Hall, Cornell University phone: (607) 255-5314 Ithaca, New York 14853-3001 fax: (607) 255-0349 USA e-mail: bb22@cornell.edu HOMEPAGE: http://www.dnr.cornell.edu/bcontrol/weeds.htm ****************************************************************************** Subject: cost:benefit analyses Date: Mon, 25 Jan 1999 11:43:48 -0600 From: "Julio S. Bernal" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Ricciardi et al (1998, J. Anim. Ecol. 67, 613-619) provide additional data to consider in favor of making cost:benefit analyses of biological control introductions (i.e. consider the real cost of "no biological control" vs the potential cost of biological control). I think that these kinds of impacts by exotic pests make the risks of biological control more "relative" and less "absolute." Julio Bernal ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julio S. Bernal, Ph.D. Biological Control Laboratory Texas A&M University College Station, TX 77843-2475 Tel. (409) 862-8378 Fax (409) 845-7977 e-mail jsb7473@unix.tamu.edu ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Subject: Re: Date: Mon, 25 Jan 1999 10:20:46 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu This comment is in reply to the exchange between Messing and Lasalle, Exhaustive surveys of the parasitoid fauna of potential nontargets can easily become an open ended and unattainable goal. There is a cerain amount of bias when comparing island ecosystems such as Hawaii where the complete insect fauna is pretty well known and certain groups are very well known. There is also the evolutionary bias of these island ecosystems which Simberloff and others have so clearly elaborated. I return to the microlep question. If we are going to introduce against a species in a large family with lots of pests in it and lots of very poorly known endemic relatives how do we identify which nontargets to be concerned about? Of course these things all must run case by case, but I am frankly stumped with my system and I think other BC of arthropod projects will be similarly stumped. I suggest in these systems that we seek species with very narrow host ranges and strong ecological preferences for certain habitats and then go ahead and release. The problem with phylogenetic host range is evident in this microlep system too. An example would be a biological control agent imported for weed biological control which is closely related to the target. Thus ecological specificity would be a requirement to avoid significant host-range testing. Unfortunately, I think it is nearly impossible to develop meaningfull host range experiments for the laboratory that will adequately characterize ecological specificity. Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Re: safety of bc agents Date: Mon, 25 Jan 1999 10:37:55 -0800 From: Tom Unruh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Russell, I agree with you completely. I do not want to drive any other species extinct through biological control introductions, even an unknown microlep. The comments were made as a devil's advocate. But, I think our concerns must be balanced and it is difficult to find that balance in systems where the fauna is known so poorly and where an argument can be made that most species are pestiferous. I would be very happy with a procedure and rules that would only allow the importation of things with quite narrow host ranges, and that host range should not overlap significantly with potential nontargets. My concern is that we cannot test all potential nontargets, or in many cases even a modest subsample of them. This exchange raises another question. Do we accept that all nontargets are created equal? >Thomas R. Unruh said: "This may not apply at all when the nontarget is >>an arthropod of little importance". I posit that 30% may not even be >>important when the nontarget is a relatively undesirable microlep but that >>may just be my entomological bias against microleps". > > Tom- you have to be careful. the statements "an arthropod of >little importance" and "undesirable microlep" are a clarion call to arms >for many conservationists (including myself, to an extent). We can't be too >anthropocentric or agropocentric (i just made that word up...). To do so >fuels the fires of those railing against us, and undercuts the wholistic >ecological perspective that should be a foundation of biocontrol. > > > Thomas R. Unruh USDA-ARS Yakima Agricultural Research Laboratory 5230 Konnowac Pass Rd. Wapato Washington 98951 U.S.A. 509-454-6563 (FAX 509-454-5646) Subject: Re: Nontarget effect Date: Mon, 25 Jan 1999 12:05:39 -0700 From: larbaker@wyoming.com Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Julie, We released Rhinocyllus conicus in Fremont County in 1978. I reported population declines at the release locations in 1983 at the Western Society of Weed Science at the poster session. Possibly an abstract was published. I have continued to monitor the impact since that time. In 1980 there was about 5,000 acres of land heavily infested with Musk Thistle in the county. The weed was spread across 60,000 acres and was recorded on 285 sections of land. In 1984, we introduced Tricosirocalas horridus to 4 locations in the county. We could collect adults nearly everywhere we looked by 1988. Prior to 1990, although I could see locatized declines in the Musk Thistle, for the most part the heavily infested fields were still dominated by the weed. I did notice at decline in the scattered infestations across the rangelands and in the mountains, but have collected no data. In 1990, I began to notice clear fenceline demarkations in the distribution of musk thistle and tried to determine what practices might be involved in promoting or discouraging the infestation. I was unable to identify any specific practices that consistantly provided control. Few landowners were spraying, too many horses, year round grazing, are are pretty common in the heavily infested area. By 1996, there was a general decline in musk thistle which I attributed to drough, but the musk thistle has never recovered. Today, if you define a weed as a plant the interfers with management objective, Musk Thistle would not meet the test. While it is possible to find plants in most of the area originally infested, where roads were solid stands, there are a few scattered plants. Where fields were really solid stands, some are weed free and the rest have few scattered plants in areas of disturbance. Some areas that were dominated by Musk thistle are now infested with Canada Thistle, but overall the Musk has been replaced by grass. There is some annual population variation due to precipitation. I have lots of pictures over time from some of our release sites, but the most dramtic changes simply cannot be captured in film. It is not possible to determine which agent did the work, or if the combination was the key, without additional studies. This has never been published. While I tried to collect background data at the time of release, lots of photos and some plant density counts, they simply miss the real magnitude of the impact and defy meaningful statistical analysis. John L. Baker, Supervisor Fremont County Weed & Pest 450 N. 2nd St., Room 315 Lander, WY 82520 307-332-1052 At 09:39 AM 1/22/99 -0600, you wrote: >There has been several mentions of the impact of Rhinocyllus conicus on Carduus >spp. Can somone point me to an article reporting on a scientific field study >recording the reduction in the population density of Carduus nutans by R. >conicus? Thanks! > >Julie Nara >Wisconsin Dept of Agriculture, >Trade & Consumer Protection >2811 Agriculture Drive >Madison, WI 53708-8911 >narajm@wheel.datcp.state.wi.us > Tel: 608/224-4572 > Fax: 608/224-4656 > > Subject: Non-target implications Date: Mon, 25 Jan 1999 13:37:26 -0800 From: "Eric M. Coombs" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu For the sake of discussion , I present a few ideas and questions to help stimulate the excellent responses to the non-target impacts list. My background is primarily in biocontrol of weeds, so please excuse my bias. I have taken a more liberal stance here than I do in practce. When searching for candidates as biocontrol agents, it seems to me that the inherent characteristics that make good biological control agents of a weeds are strongly shared with exceptional crop pests. I think we can learn a lot about BC from our crop pest colleagues, especially from those that work on pests that become pests on secondary (non-target) crops. We must be sure to differentiate between confusion (short term behavioral response of individuals) and evolution (long term physiological response of populations), when it comes to analyzing host shifting and new host acquisition by biocontrol agents on non-target species. The physiological mechanisms of locating, verifying, and developing on non-target species should be researched. (Difficult in our present situation of approaching the implementation of BC as though we were fighting wildfires.) In some situations, weed biocontrol agents in their new home can develop explosive populations after their release and adaptation in dense stands of the target weed. When population densities and intra-specific competition are very high, it increases the probability that more individuals with in a population will taste test plant species related to the target host and other non-target plants. If individuals in dispersion groups move on after taste testing a possible host and are later followed by even more individuals in the same predicament, then severe short term damage can occur on non-target plants. We have observed this in W OR with the cinnabar moth, but only in areas near dense stands of the target host, tansy ragwort. Such out breaks were short lived, and reoccurrence was rare. In one case, department officials were called upon to spray insecticides on large numbers of cinnabar moth larvae after they stripped stands of ragwort and emigrated in search of anything they could consume in the yards of adjacent home owners. In most cases, the shear numbers of larvae crawling in and around homes lead to the alarm, but those with ornamental senecios experienced the most damage. As a group, I believe that most biocontrol of weeds practitioners are good at self-policing their work and often the first to sound the alarm. I think that we should be even more observant when biocontrol agent populations are at low densities and more carefully monitor their impacts on non-target plants. When a biocontrol agent appears to develop on non-target hosts, we should ascertain whether or not the agent is able to develop in areas totally exclusive of the target host. We should test the agent for several generations exclusively on the non-target host, to determine the long term implications for the agent and the non-target. In studies where biocontrol agents are restricted to non-target plants the adults that develop are often found to be smaller and less fecund than those that developed on the target weed. I am not aware of any studies that have determined how many generations this can continue and what are the long term implications at the population level? Is this a long term self-limiting factor or an evolutionary shift to expand host range? Perhaps those agents that can maintain limited populations on non-target plants may be in a better situation to respond rapidly to incipient outbreaks of the target host after perturbations. This may be more useful in species that are more density dependent and have a lower ability to locate host plants when resources are scarce, i.e. the cinnabar moth vs. the ragwort flea beetle. Perhaps ecological modeling will be able to help us assess the risks of allowing the introduction of a biocontrol agent with a slightly wider host range, especially when compared to the risk of doing nothing. Eric Eric M. Coombs Biological Control of Weeds Oregon Department of Agriculture 635 Capitol St. NE Salem, OR 97301-2532 503-986-4624 FAX 986-4737 Cell 503 881-2823 email: ecoombs@oda.state.or.us http://www.oda.state.or.us/oda/dept.html Subject: Poor land management increases non-target risks Date: Mon, 25 Jan 1999 15:40:44 -0800 From: "Eric M. Coombs" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Many questions have been generated as to why biocontrol agents (BCAs) have been successful and why some failed (to establish or to control their target). It seems, as I believe R. Sobhian pointed out, that the majority of the problems (non-target impacts) are by BCAs that we generally consider successful. This perhaps because they are usually more widespreadand abundant, therefore exposed to a greater variety of opportunities to come into contact with more non-targets. In the BC of weeds scenario, we must be very careful to make sure what and how we attribute the success and failure of BC. Blossey recently pointed out some very important aspects of monitoring. Because of my range management background, I view most weeds as invader or increaser species. They are most often a symptom of poor or inadequate land management practices, and not inherently the main problem. If land management practices are not changed, then we (BC workers) end up on the "Merry-Go-Round" of weed control, constantly fighting the next species in line after an apparent short term success. In the lines of risk assessment and non-target impacts, we must ask, Will the control of this weed solve the problem, or will the problem be maintained because of the types of land management practices? With few exceptions, most of the weeds I work on are early to mid-successional species that act as environmental scar tissue, and the land managers call on us to act as ecological plastic surgeons, only to see them continue to abuse the patient, and then wonder what is wrong with our treatment! In most cases, sound land management practices (i.e., changes in grazing practices, vegetation management) fostering ecological succession can ammeliorate weed control problems and enhance the impacts of BCAs. When BC is not successful in the short term, we seem to be pushed to introduce more and more BCAs and inevitablly increase environmental risk to non-target species. Many studies that claim success with BC need to demonstrate the relationship of plant competition and community succession in relation to the impacts of the BCAs. Biocontrol agent exclusion studies are often missing in long term ananlysis of BC, something I am often guilty of. Exclusion studies should apply when assessing impacts of BCAs on non-target species so the blame goes where it belongs! Eric Eric M. Coombs Biological Control of Weeds Oregon Department of Agriculture 635 Capitol St. NE Salem, OR 97301-2532 503-986-4624 FAX 986-4737 Cell 503 881-2823 email: ecoombs@oda.state.or.us http://www.oda.state.or.us/oda/dept.html Subject: ERBIC Date: Mon, 25 Jan 1999 15:48:00 -0800 (PST) From: "Liam Lynch (Leverhume)" Reply-To: bc-ntimpact@UDel.Edu To: ntlistserver Dear Colleagues, As we are involved in a project aimed at developing methodologies for the assessment of non-target risks in biocontrol, we are very grateful that this list server was established and very much appreciate the quality of the discussion taking place. The project we are involved in (called ERBIC, which, somehow, stands for Evaluating Environmental Risks of Biocontrol Introductions into Europe) is driven by the philosophy of relating the strategy of biocontrol to its concurrent risks, and thus gaining a mechanistic-level understanding of what makes some strategies risky and others not. It is hoped this will lead to a meaningful set of risk assessment protocols for different types of biocontrol uses. We are using several complementary approaches. The first is continued case studies (on four very different systems) which include studies of the nature of dispersal of agents to non-target habitats, damage levels (i.e. levels of parasitism/ feeding / infection) as well as measuring the population level impacts where possible. Understanding such examples on mechanistic (e.g. dispersal, infection) and population dynamic levels is clearly important. More generally, different types of biocontrol, using different agents, may imply different risks. The most obvious distinction is the generalist-specialist continuum, but other factors, such as differences in biology between pathogens, predators and parasitoids, are likely to affect risk, as well as how they are used (augmentatively, inundatively, seasonal introduction, classically, 'neo-classically' etc.). The second is a literature review which hopefully will bring together all know examples of non-target impacts (or hints of such impacts). In particular, as different strategies of biocontrol may imply different risks, one would hope this might come out in any such review. Examining the relationship between strategies of the past and their NT impacts (in as much as they were noted) illuminates the choice of strategy now. Firstly (and most trivially), illegitimate releases may be found to be responsible for a larger proportion of observed NT effects than might be expected by chance. Similarly, in periods where multiple releases were more favoured, and less circumspection surrounded releases, the actual success rate of biocontrol fell (Greathead & Greathead, 1992). It is likely that in this period (1950s and 60s) more non-target effects occurred (how many were observed or commented on is another thing completely, however). In evaluating these cases, we suggest what should be focused on is the relationship between the strategy or philosophy of the control practitioners and the resultant risks. Classifying an individual case (such as the impact of Rhinocyllus on NT Cirsium species) into a category of acceptable vs unacceptable levels of damage, out of this context, may not achieve much. Equally, treating the whole history of biocontrol together as one unit may not allow many meaningful conclusions to be drawn - biocontrol history is made up of a vast mixture of strategies and philosophies, which have differed considerably over time and place. However, such data, interpreted correctly, may help to identify unacceptably risky versus less risky strategies, in terms of their underlying ecology, in the context of the broader project. Finally, we believe that assessing strategies of biocontrol and relating them to risk at a mechanistic level involves some sort of general understanding of the population dynamic processes which mediate NT effects under different circumstances. Therefore, it seems appropriate that a population-model-based understanding of different biocontrol methods will both highlight various types of potential NT impacts, and serve to classify those effects that have already been noticed. This involves both tailoring existing theories from population biology (such as that of apparent competition) to the problems at hand and examining new scenarios highlighted by known examples and a practically orientated perspective of the problem. Without such a framework, linking all the various case studies and examples into a coherent understanding of the possible risks would be very difficult. Such a fundamental understanding is required if general and ecologically meaningful guide-lines about non-target risks are to be formulated. Clearly, such guide-lines, while being general, will hopefully point to those areas where more information is required in specific cases, in order to properly address the idiosyncrasies of such cases, and suggest methods of tackling these idiosyncrasies. Such suggestions will hopefully draw on the methodological advances achieved in the case studies discussed above - e.g. methods for monitoring the dispersal of different types of agents. The guide-lines arrived at finally will hopefully ensure that on average the benefits will out-weigh the negative impacts of future biocontrol introductions (while accepting there are bound to be occasional specific cases where the converse is true). This implicitly assumes a risk-benefit approach, which may become more overt where risk and benefits must be explicitly weighed up to make a decision on strategy- e.g. the benefits of releasing a generalist, or of multiple releases, may have to be proved in order to justify the extra risk implied. But both in the context of this list server and the ERBIC project it seems sensible to put most of our effort into developing approaches to assessing and evaluating non-target risk, and so look at the side of the risk-benefit inequality that has been most neglected thus far. We at CABI are using the second two approaches to the problem (modelling and literature review), while our four partners in the project are carrying-out the case studies. We are interested in any feedback people may have on the type of approach we have described, or any other suggestions you may have. Secondly, in order to carry out a more or less exhaustive search for examples, we would appreciate it if anyone with any details of examples that are unlikely to be found in published literature would contact us or post these on the list server (were you not about to do so already). We are interested in any type of agent introduced to control an arthropod pest (in the context of any control strategy, not just classical BC), but not weed-control examples for this particular project. Cases where non-target feeding or parasitism was encountered in the context of programmes from some time ago when there was less emphasis on non-target risks, which were therefore not reported when the results of the programmes were published, would be very useful, for example. While some objective description of the nature of the non-target effect (e.g. feeding records, % parasitism, etc.) would be preferable where available, it is not absolutely necessary for to have these to constitute examples. Also, any information regarding the control programme and agent generally (and preferably a reference) would be helpful, as well as any circumstances surrounding the release that would put the example into context. For example, legitimacy of the release, size of the pest problem, urgency of the pest problem, existence of alternative control methods, attitude of organisation funding the project, reasoning behind the introduction of that particular agent, details of any alternative agents etc. Hopefully, if we can draw together a definitive list of examples of non-target effects then this will not only provide an invaluable foundation for further studies (both empirical and theoretical) but also provide a concrete output for the current 'virtual workshop'. We look forward to your comments and contributions, and to taking part in the debate. Best wishes, Liam Lynch Matt Thomas Leverhulme Unit for Population Biology and Biological Control CABI Bioscience UK Centre (Ascot) Silwood Park Ascot, Berks SL5 7PY UK Tel: (0)1344 872999 L.LYNCH@CABI.org M.THOMAS@CABI.org Subject: Re: ERBIC Date: Mon, 25 Jan 1999 16:01:47 -1000 From: messing@hawaii.edu (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Liam Lynch wrote: "Hopefully, if we can draw together a definitive list of examples of non-target effects........." Liam.... (and anyone else interested): Attached is a list of references from our work on the non-target effects of tephritid biocontrol agents in Hawaii. Reprints are available on request. References from programs conducted several decades ago are cited in these papers. --------------------------------------------------------------------- non-target_refsName: non-target_refs Type: Macintosh BinHex Archive (application/mac-binhex40) --------------------------------------------------------------------- Russell Messing University of Hawaii Dept. of Entomology 7370 Kuamoo Rd. Kapaa, Hawaii, 96746 tel (808) 822-4984 x 223 fax (808) 822-2190 email: messing@hawaii.edu Subject: Date: Tue, 26 Jan 1999 14:10:45 +1200 From: Pauline Syrett Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear Colleagues I tend to agree with Alec McClay that a few insects showing 'normal' levels of attack on a non-target host is a more serious proposition than a minor level of attack by the majority of the population. We rejected a stem-mining weevil for control of broom because a small number of females laid 'normal' numbers of eggs on a non-host native plant (and the weevils were able to develop through to adult on this plant). This work has been published. I concur with Bernd Blossey's concern that host testing may lead us to reject safe insects, rather than to accept those with a high risk of damage to non-targets. However, I think we do have to make sure that levels of replication in host tests are sufficient to pick up those 'aberrant' individuals in a population which demonstrate a broader host range than that which is normal for the population. I enjoyed the rest of Bernd's contribution, and really like the way the purple loosestrife project is being used to test predictions. It is a neat project, with some nice agents, and I look forward to the analysis that says which predictions were correct, and which weren't. As we gradually accumulate more data on which predictions work, and which don't, we should make some real advances in understanding impacts of bc agents. Maybe predicting ecological impacts of not-so- host-specific agents will become sufficiently reliable that we can risk using them! Better go back to some funded work, so I'll get to tick off my outputs for the year! Pauline Syrett Subject: Re: Non-target implications Date: Mon, 25 Jan 1999 16:29:08 -1000 From: messing@hawaii.edu (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Eric Coombs wrote: >In studies where biocontrol agents are restricted to non-target plants the >adults that develop are often found to be smaller and less fecund than >those that developed on the target weed. I am not aware of any studies >that have determined how many generations this can continue and what are >the long term implications at the population level? ------------------ Just to keep things lively, I'll note that when the braconid D. tryoni is reared on (or from) the non-target tephritid Eutreta xanthochaeta (the lantana gall fly), it is almost always LARGER and MORE FECUND than when reared on its target host, the medfly. The long term population implications of this are enough to give biocontrol critics the shivers. Russell Messing University of Hawaii Dept. of Entomology 7370 Kuamoo Rd. Kapaa, Hawaii, 96746 tel (808) 822-4984 x 223 fax (808) 822-2190 email: messing@hawaii.edu Subject: Re: ERBIC Date: Tue, 26 Jan 1999 00:24:07 -0500 From: sdp@nersp.nerdc.ufl.edu (Sanford D. Porter) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu > >Attachment converted: Macintosh HD:non-target refs (W8BN/MSWD) (000183CB) >Russell Messing >University of Hawaii >Dept. of Entomology >7370 Kuamoo Rd. >Kapaa, Hawaii, 96746 > Could you resend your list as a text or RTF file? I was unable to open the other as I do not use Word. Sanford D. Porter, PhD Research Entomologist ------------------------------------ USDA-ARS, CMAVE P.O. Box 14565 Gainesville, FL 32604 USA Office: 352 374-5914 Secretary: 374-5903 FAX: 374-5818 Internet: SDP@NERSP.NERDC.UFL.EDU ********Please note that my e-mail address has changed slightly************ ------------------------------------ Subject: Re: Non-target implications Date: Tue, 26 Jan 1999 08:10:35 -0600 From: "Julio S. Bernal" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu >Just to keep things lively, I'll note that when the braconid D. tryoni is >reared on (or from) the non-target tephritid Eutreta xanthochaeta (the >lantana gall fly), it is almost always LARGER and MORE FECUND than when >reared on its target host, the medfly. The long term population >implications of this are enough to give biocontrol critics the shivers. > >Russell Messing >University of Hawaii Russell, I am not familiar with braconid host associations. Could it be that D. tyroni-E. xanthochaeta is an "old association" rather than a "new association"? And therefore, D. tyroni is actually hitting a target species, and the mistake is ours. Julio Bernal ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Julio S. Bernal, Ph.D. Biological Control Laboratory Texas A&M University College Station, TX 77843-2475 Tel. (409) 862-8378 Fax (409) 845-7977 e-mail jsb7473@unix.tamu.edu ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Subject: RE: other methods of biological control Date: Tue, 26 Jan 1999 11:02:47 -0500 From: Kevin Floate Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear All: >From the nature of comments to date, I am under the (probably mistaken) impression that the research performed by most list readers emphasizes classical biological control. However, one can *never* reject the possibility that introduced biocontrol agents *may* affect populations of non-target species. Classical biological control will *always* be subject to this criticism by concerned citizens and regulatory agencies, and delayed or terminated as a result (witness the flurry of media attention after reports of biocontrol agents feeding on native thistles in the US, problems associated with the release into the US of a biocontrol agent for salt cedar (Tamarix), and the establishment of this list). In an earlier e-mail, Keith Hopper wrote that the subject of this discussion group is to "delineate research needs for assessing and reducing impacts of biological control agents introduced from one region to another. " [So far, discussions have almost solely been limited to the movement of biocontrol agents across international borders; i.e., exotic biocontrol agents] Are there any list readers out there who are studying native species of organisms as potential biocontrol agents of exotic pest species. Specifically, moving native species to regions within the same country, where populations of the native species are not currently established? Does this approach provide a method to avoid much of the current controversy regarding the non-target effects of biocontrol agents? Cheers! Dr. Kevin D. Floate, Research Scientist Biological Control of Livestock Insect Pests Agriculture and Agri-Food Canada Lethbridge Research Centre Lethbridge, AB, CANADA T1J 4B1 ph: (403) 317-2242, FAX (403) 382-3156 http://res.agr.ca/leth/scitech/kdf/intro.htm ! ! Subject: Re: Nontarget effect Date: Tue, 26 Jan 1999 11:17:28 -0600 (CST) From: narajm@wheel.datcp.state.wi.us Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu John Baker and Tom McAvoy, Thanks for your reply to my question regarding the quantitative effect of Rhinocyllus conicus on the musk thistle. R. conicus was released in a pasture in SE Wisconsin in 1975 (not recovered in 1979), and another colony was released in a state forest in 1979. This second colony was thriving and had apparently spread to other fields. The releases were done by a weed scientist at the University of Wisconsin-Madison in cooperation with Marlin Conrad, a former colleague of mine at the Wisconsin Dept of Agriculture, Trade & Consumer Protection, who has since retired. By that time the professor had retired and was succeeded by Dr. Jerry Doll, who is still professor of weed science at UW. In 1980 Marlin did some cage studies and found that R. conicus can complete its life cycle on the dune thistle, Cirsium pitcheri, a threatened species. We halted further releases until more is known about the impact of R. conicus on native Cirsium spp. Another species, the prairie thistle, Cirsium hillii, is also on the threatened list. By that time many other regulatory problems started in Wisconsin (specifically gypsy moth), and due to lack of time we at the Wis Dept of Ag reduced our involvement with Rhinocyllus conicus. Jerry is still monitoring the R. conicus colony and the musk thistle population. There are no dune thistles near the release sites. The species is restricted to the dunes around the Great Lakes. I understand that R. conicus is doing fine, and the musk thistle density is mainly going down, though mowing and other control efforts may have contributed. In recent years we have requested that applicants for a permit to release Rhinocyllus conicus in Wisconsin (PPQ form 526) submit reports of studies showing that the release would have no impact on thistles which are native to Wisconsin. So far, none have done so. Bernd Blossey mentioned purple loosestrife and his excellent plan to record all possible effects release of biocontrol oranisms may have. In a couple of cases we approved the release of biocontrol agents against purple loosestrife on condition that the effects on both target and non-target species be monitored. Both provided detailed work plans. One of the permittees sent one report after the first year, but none since, the other never sent in any reports. I plan to inquire about the results of the monitoring, so that we have a better idea for future applications. It would be nice if the USDA would routinely include pre-release and post-release monitoring with all the organisms they import from abroad. I have been involved in a few releases myself, where the USDA decided what to release and where, and involved cooperators when they needed local people to do the actual releases. It was mainly release and recovery, we trusted that Big Brother USDA had studied all possible effects of the releases and found them, on the whole, beneficial. May all your biocontrol releases be productive, with no unwanted and unplanned results! Julie Nara Wisconsin Dept of Agriculture, Trade & Consumer Protection 2811 Agriculture Drive Madison, WI 53708-8911 narajm@wheel.datcp.state.wi.us Tel: 608/224-4572 Fax: 608/224-4656 Subject: Re: Non-target implications Date: Tue, 26 Jan 1999 09:51:03 -1000 From: messing@HAWAII.EDU (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu >Julio Bernal wrote: > Russell, >I am not familiar with braconid host associations. Could it be that D. >tyroni-E. xanthochaeta is an "old association" rather than a "new >association"? And therefore, D. tyroni is actually hitting a target >species, and the mistake is ours. > ___________________________________ D. tryoni is endemic to Australia, where it's normal hosts are Bactrocera tryoni and related tephritids. It was imported to Hawaii for medfly control in 1913. E. xanthochaeta is native to Mexico... it was imported to Hawaii in 1902 to control lantana. As far as I know it does not occur in Australia. So this particular association is new.... However, there are other gall-forming tephritids in Australia (though not on lantana), so it is at least possible that pre-adaptation to gall micro-habitats was present. A good subject for a grant proposal...... RM Russell Messing University of Hawaii Dept. of Entomology 7370 Kuamoo Rd. Kapaa, Hawaii, 96746 tel (808) 822-4984 x 223 fax (808) 822-2190 email: messing@hawaii.edu Subject: Ento Soc America Position on Biodiversity Date: Mon, 01 Feb 1999 13:14:47 -0500 From: Rob Bourchier Reply-To: bc-ntimpact@UDel.Edu To: ENTOMO-L@LISTSERV.UOGUELPH.CA, bc-ntimpact@UDel.Edu Just ESA newsletter with the ESA draft postion paper on biodiversity. One of the key points listed at the start is " Scientists and nonscientists alike are encouraged to work in harmony to improve risk assessment capabilities to limit non-target impacts of biological control agents" This seems like a good goal, however it seems strange to single out biocontrol agents, what about all pest control tools, especially chemicals? It seems to me that biological control is being held accountable to provide data for non-target impacts (again I agree with this in principal) whereas we are not asking for similar data about impacts on biodiversty for other pest control interventions. Consideration and quantification of non-target impacts and risk assessments of all pest control interventions, especially chemicals should be a goal when addressing biodiversity concerns. Risk assesment data are hard to get for any action, which why they may have been not included as a data requirement in the past. Given that we are now recognizing the value of a risk-assessment approach, I would expect that the ESA should have a broader view of the threats to biodiversity, (its more than just biocontrol), when developing key points for a position paper. my 2 cents Rob ------------------------------------------------------------- Dr. Rob Bourchier, Research Scientist Insect Ecology/Weed Biocontrol LRC, Agriculture and Agrifood Canada 5403-1st Ave South, Lethbridge, AB CANADA T1J 4B1 403 317-2298 fax 403 382-3156 bourchierR@em.agr.ca ! ! Subject: oversummering techniques Date: Tue, 2 Feb 1999 18:10:11 +0800 (WST) From: Paul Yeoh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu, ENTOMO-L@LISTSERV.UOGUELPH.CA Question: 1) Has anybody stored insects that are suppose to be dormant during the summer (ie in aestivation) in the fridge for the duration of the summer period ( ie 4 months or so). 2)If so, did it affect the breaking of dormancy (when placed back into real winter conditions). We believe the dormancy is controlled by daylength (reproductively active when the daylength is short). Any comments greatly appreciated - Thanks Paul. **************************************************************************** ****************************** Background for those interested: Our Group is releasing Apion miniatum (stem weevil) as a biocontrol agent against the weed doublegee (Emex australis) here in Australia (Mediterranean climatic regions). The insect is from Israel. It is univoltine, breeding during the winter and being in a dormant phase (theoretically) during the summer. During the summer there are occasional summer showers but no growing host plants (doublegee is a winter annual) and therefore it can’t feed during this period. Nothing is known about its summer dormancy behaviour in its home land. Its hard enough finding the insect there on host plants during the breeding period! The Apion has only recently been approved for release in Australia. We have just started mass rearing the insect to speed up its distribution throughout Australia. Currently we are holding the insects in cages over the summer. They are given access to growing host plants and water and artificial hiding spots (dead plants, paper, cloth and cardboard). Although they spent the majority of their time clustered together at the top of the cage (which is how we define them as being in their dormant condition), they still feed occasionally on the host plants leaves and regularly are observed drinking -but no mating behaviour occurs until the break of season (photoperiod controlled?). We assume the insect need the food/water to top up reserves because the over-summering substrates we provide are not what they need (they can’t feed under natural conditions). Survival is good (70%) but this procedure is too labour intensive for use on a large scale (we are struggling this year with 20 000 insects). Insects given no food/water (ie. in vials in lab conditions -ie 25°c) die within a few weeks. We are now looking at better ways of holding them during the “dormancy period”. We are trialing storage at low temperatures (5 to 25 degrees C) with a summer light regime (15hrs light/9 hr dark) with moisture but without food (as replacing the plants is the hard/time consuming part). They are losing weight slower at the cold temperatures (5 degrees) and based on current trends, should survive until the winter but we don’t know if the insects will behave normally again when released into the field in winter. I have search the literature (CABI abstracts) but did not find much - probably as most people don’t store summer insects in winter conditions (I agree it does not seem normal). But I thought people may be doing it everywhere and they just don’t report it. =========================================================================== Paul Yeoh CSIRO, CCMAR building phone: (08) 9333 6645 +(619)Int'l Underwood Avenue fax: (08) 9333 6646 +(619)Int'l Floreat Park, Perth, email: p.yeoh@ccmar.csiro.au Western Australia 6014 postal: CSIRO Entomology/CRC Weed Man.Sys. Private Bag P.O, Wembley Western Australia 6014 =========================================================================== Subject: Re: oversummering techniques Date: Tue, 02 Feb 1999 11:48:46 -0800 From: Mark Hoddle Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Hi Paul, I worked on a related species - Apion ulicis for the biological control of gorse, Ulex europaeus in New Zealand. I looked at the effect of long days on this seed feeding weevil at constatn temperatures and I managed to prevent gonad development in females by holding them in the lab under these conditions. A. ulicis enters a reproductive diapause over winter - adults are active on warm winter days. Increasing daylength stimulates gonad development. Temperature did not appear to be as important with this insect as day length. In your situation it may be that storing an insect that is supposed to be experiencing hot summer temperatures at low temperatures in a fridge may affect aestivation/gonad development as metabolism etc will be slowed and the weevil may not develop physiologically at the right speed to be ready to go in the winter when you release it (assuming this is what you intend to do). I would hold the weevils at estimated summer temperatures experienced by the weevils where ever they aestivate (i.e., in soil or under bark) under long days to maximize the likelihood of reproduction being in synchrony with host plant phenology when released in winter. Just my thoughts - may be way off on this! Regards Mark Mark S. Hoddle Biological Control Specialist Department of Entomology University of California Riverside, CA 92521-0314 U.S.A. Phone Number: (909) 787 4714 Fax No. (909) 787 3086 E-Mail mark.hoddle@ucr.edu http://www.biocontrol.ucr.edu/ Subject: Re: Weed bc non-target examples Date: Tue, 2 Feb 1999 14:02:33 -0800 (PST) From: John Herr Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Getting back to the questions asked by Keith Hopper on Jan. 20, 1999 re: my original posting of non-target weed bc examples: > Why is Opuntia spinossima rare? It is not because of attack by C. > cactorum because the cactus was rare long before C. cactorum was > accidentally introduced. Yes, this is true. However, it does not make the non-target impact any less significant or noteworthy. If anything, we should be concerned most with impacts to rare native species, regardless of the reason for their rarity, because in the unlikely event that an extinction is caused by a bc agent, it would probably occur in rare species first. The rare California Cirsium thistle species I am working with are extreme habitat specialists (endemic to serpentine soils near springs, for the most part), so their rarity is definitely not attributable to the introduction of Rhinocyllus conicus. I'm just trying to determine if R. conicus is contributing to a further decline in their abundance. > By the way, given the small number of plants > in the population, it would seem fairly simple to prevent impact of C. > cactorum on O. spinossima by using insecticides, physical barriers, or > by hand-picking the insects from the plants. According to a talk given by Peter Stiling a few years ago, The Nature Conservancy is managing the surrounding habitat in the Florida Keys, and has caged all remaining O. spinosissima individuals to prevent further damage. However, the solution is not that simple, as Johnson and Stiling (1996) report outbreaks of other insects within the cages, as well as the exclusion of important pollinators. > First, my understanding is that R. conicus did not expand its host range > in an evolutionary sense. Attack on native Cirsium species was > understood to be likely before introduction. Although it was considered likely that R. conicus would attack native North American Cirsium species because European Cirsiums were within its host range, it is not known if genetic changes have facilitated this movement onto new hosts. The existence of genetically distinct biotypes of R. conicus is well documented (Goeden et. al 1985; Unruh and Goeden 1987; Klein and Seitz 1994), and three strains were introduced (Italian thistle, Milk thistle, and Musk thistle), so it would appear that the weevil has at least the potential to switch hosts in an evolutionary sense. Marohasy (1996) is of the opinion that the terms 'host shift', 'host switch', and 'host range expansion' should only be used if there is evidence of a genetic change in the bc agent. Others (e.g. Louda 1998) seem to use 'host range expansion' when any new additions to a host range are observed. Marohasy claims there is no evidence of a true host shift ever having occurred in a weed bc agent, and prefers the term 'host substitution' instead. In any case, this might just be an issue of semantics, because the bottom line is that if bc agents are causing biologically significant impact to unintended hosts, we should be concerned about it, whether or not evolutionary changes are involved. > What I would like to see > for R. conicus impact on Cirsium spp. is evidence that the weevil has or > is changing the distribution/abundance of the non-target host plants and > the magnitude of such changes. This is not a system I know well, but > the evidence that Turner et al. (1987) Louda et al. (1997) and Louda & > Potvin (1995) are not as strong as what would be required to demonstrate > conclusively that the beetle is affecting distribution/abundance of > non-target plant species. This is not to say that it does not have such > impacts, I would just like to see stronger evidence. Louda reasons that because seed predation by native insects significantly reduces seedling recruitment, local population density, and lifetime fitness of Platte thistle (Louda and Potvin 1995), and that because the recent movement of R. conicus onto this thistle further reduces seed availability (Louda et. al 1997; Louda 1998), the result will be a reduced population size. I agree that more direct evidence would be nice, but of course its hard to selectively exclude R. conicus from thistles and still let in the native leps. and tephritids. I'm trying to get around this problem by using a seed addition experiment to replace the seeds destroyed by R. conicus in the native thistle population. So far, I have not found evidence that R. conicus is limiting seedling recruitment rates in the rare California species I am studying. =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= John Herr jherr@nature.berkeley.edu UC Berkeley, Division of Insect Biology (510)642-8414 USDA-ARS-WRRC, Plant Protection Research (510)559-5974 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Subject: RE: oversummering techniques Date: Wed, 3 Feb 1999 12:57:27 +1300 From: "Barratt, Barbara" Reply-To: bc-ntimpact@UDel.Edu To: "'bc-ntimpact@UDel.Edu'" Paul, With Sitona discoideus, we keep them in a 14:10 light:dark photoperiod to stops them going in to aestivation. At 16:8 they aestivate. Haven't tried refrigeration - normally we have wanted to use the weevils for experiments. Barbara ================================================ Dr Barbara I.P. Barratt Phone: (03) 489 3809 AgResearch FAX: (03) 489 3739 Invermay Agricultural Centre Email: Barrattb@AgResearch.cri.nz Private Bag 50034 Mosgiel New Zealand ================================================ > -----Original Message----- > From: Paul Yeoh [SMTP:p.yeoh@ccmar.csiro.au] > Sent: Tuesday, 2 February 1999 23:10 > To: bc-ntimpact@UDel.Edu; ENTOMO-L@LISTSERV.UOGUELPH.CA > Subject: oversummering techniques > > Question: > > 1) Has anybody stored insects that are suppose to be dormant during the > summer (ie in aestivation) in the fridge for the duration of the summer > period ( ie 4 months or so). > 2)If so, did it affect the breaking of dormancy (when placed back into > real > winter conditions). > > We believe the dormancy is controlled by daylength (reproductively active > when the daylength is short). > > Any comments greatly appreciated - Thanks Paul. > > ************************************************************************** > ** > ****************************** > Background for those interested: > > Our Group is releasing Apion miniatum (stem weevil) as a biocontrol agent > against the weed doublegee (Emex australis) here in Australia > (Mediterranean climatic regions). > > The insect is from Israel. It is univoltine, breeding during the winter > and > being in a dormant phase (theoretically) during the summer. During the > summer there are occasional summer showers but no growing host plants > (doublegee is a winter annual) and therefore it can't feed during this > period. Nothing is known about its summer dormancy behaviour in its home > land. Its hard enough finding the insect there on host plants during the > breeding period! > > The Apion has only recently been approved for release in Australia. We > have > just started mass rearing the insect to speed up its distribution > throughout > Australia. > > Currently we are holding the insects in cages over the summer. They are > given access to growing host plants and water and artificial hiding spots > (dead plants, paper, cloth and cardboard). Although they spent the > majority > of their time clustered together at the top of the cage (which is how we > define them as being in their dormant condition), they still feed > occasionally on the host plants leaves and regularly are observed drinking > -but no mating behaviour occurs until the break of season (photoperiod > controlled?). We assume the insect need the food/water to top up reserves > because the over-summering substrates we provide are not what they need > (they can't feed under natural conditions). Survival is good (70%) but > this > procedure is too labour intensive for use on a large scale (we are > struggling this year with 20 000 insects). Insects given no food/water > (ie. > in vials in lab conditions -ie 25°c) die within a few weeks. > > We are now looking at better ways of holding them during the "dormancy > period". We are trialing storage at low temperatures (5 to 25 degrees C) > with a summer light regime (15hrs light/9 hr dark) with moisture but > without > food (as replacing the plants is the hard/time consuming part). They are > losing weight slower at the cold temperatures (5 degrees) and based on > current trends, should survive until the winter but we don't know if the > insects will behave normally again when released into the field in winter. > I > have search the literature (CABI abstracts) but did not find much - > probably > as most people don't store summer insects in winter conditions (I agree it > does not seem normal). But I thought people may be doing it everywhere and > they just don't report it. > ========================================================================== > = > Paul Yeoh > CSIRO, CCMAR building phone: (08) 9333 6645 +(619)Int'l > Underwood Avenue fax: (08) 9333 6646 +(619)Int'l > Floreat Park, Perth, email: p.yeoh@ccmar.csiro.au > Western Australia 6014 postal: CSIRO Entomology/CRC Weed Man.Sys. > Private Bag P.O, Wembley > Western Australia 6014 > ========================================================================== > = Subject: Insects importations and Australia. Date: Mon, 8 Feb 1999 15:27:42 +0800 (WST) From: Paul Yeoh Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu, ENTOMO-L@LISTSERV.UOGUELPH.CA In response to my request for information on oversummering techniques (last week), I had a person ask if we (Australia) still were importing agents for the biological control of weeds because they had heard that we had been forced to stop. Answer: I had not heard of any attempt to ban importations so i checked with a more experienced biocontoller (ie. old timmer) who also had not heard of any such thing. Currently all agents have to be approved (via the Australian Quarantine Inspection Service or AQIS) prior to importation. The agents are screened in their home country and then if they look promising (target specific), the host specificity testing is completed within approved Australian quarantine facilities. If they still appear suitable, an application for release in Australia is made to AQIS who distribute the application to all relevant bodies (Agriculture departments/ Conservation departments etc) in all states. If all bodies unanimously agree, the insect/agent is approved for release (subject to being held in Australian quarantine facilities for at least 1 generation so as to ensure they are clean of diseases etc.). I would have responded directly to the person who requested the information but i lost the Email address (sorry). Sorry to bother all others on the list(s). =========================================================================== Paul Yeoh CSIRO, CCMAR building phone: (08) 9333 6645 +(619)Int'l Underwood Avenue fax: (08) 9333 6646 +(619)Int'l Floreat Park, Perth, email: p.yeoh@ccmar.csiro.au Western Australia 6014 postal: CSIRO Entomology/CRC Weed Man.Sys. Private Bag P.O, Wembley Western Australia 6014 =========================================================================== Subject: IOBC symposium/workshop Date: Mon, 8 Feb 1999 10:23:31 +0100 From: "Kim Hoelmer" Reply-To: bc-ntimpact@UDel.Edu To: The IOBC-sponsored symposium and workshop on "Evaluating Indirect Ecological Effects of Biological Control", to be held in Montpellier, France from 17-20 October, 1999, was announced on this list recently. For the benefit of new list subscribers, the goals of the symposium/workshop are: 1) to assess the current worldwide status of evaluations of indirect (including non-target) effects of biocontrol programs; and 2) to apply current research in ecology and evolutionary biology to the development of science-based methods for evaluation of these programs. The conference website has now been updated with new information, including a list of keynote speakers. The site is located at www.agropolis.fr/iobc and contains information on pre-registration and contribution of papers. **************************************************************** Kim A. Hoelmer, A. A. Kirk & P. C. Quimby USDA, ARS, European Biological Control Laboratory Parc Scientifique Agropolis II 34397 Montpellier Cedex 5, France phone: 33-467-41-6730; fax: 33-467-04-5620 e-mail: ebcl@cirad.fr **************************************************************** Subject: Re: Insects importations and Australia. Date: Mon, 8 Feb 1999 16:47:00 -0800 (PST) From: John Herr Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu On Mon, 8 Feb 1999, Paul Yeoh wrote: > If all bodies unanimously agree, the > insect/agent is approved for release (subject to being held in Australian > quarantine facilities for at least 1 generation so as to ensure they are > clean of diseases etc.). In the U.S., entomophagous biocontrol agents must be held in quarantine for one generation to prevent release of hyperparasites, but this regulation does not apply to weed biocontrol, where releases of foreign-collected material are allowed. Although it might slow down some weed control projects, adoption of the Australian rules would reduce the risk of accidentally introducing parasitoids, pathogens, or worse yet, incorrectly identified agents (which has happened occasionally in the past). =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= John Herr jherr@nature.berkeley.edu UC Berkeley, Division of Insect Biology (510)642-8414 USDA-ARS-WRRC, Plant Protection Research (510)559-5974 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-= Subject: IOBC conference notice Date: Wed, 10 Feb 1999 16:36:40 +0100 From: "Kim Hoelmer" To: , , , , "Keith Hopper" , , , , , , , , , , , , , , , , , , , , , , "Lerry Lacey" , , , , , , , , , , , , , , , , , , , , , "Molly Hunter" , , This message has been sent to members of various biocontrol and entomophagous insect working groups. Our apologies if you have received duplicate announcements. You may be interested in an IOBC-sponsored symposium and workshop on the subject "Evaluating Indirect Ecological Effects of Biological Control" to be held in Montpellier, France from 17-20 October, 1999. The goals of the symposium/workshop are: 1) to assess the current worldwide status of evaluations of indirect (including non-target) effects of biocontrol programs; and 2) to apply current research in ecology and evolutionary biology to the development of science-based methods for evaluation of these programs. Additonal information may be obtained by viewing the symposium website at www.agropolis.fr/iobc or by contacting Mireille Montes de Oca at iobc.symp@agropolis.fr . The website includes pre-registration details and a call for contributed papers. For those who may have already seen the site, it has been updated with new information, including a list of keynote speakers. **************************************************************** Kim A. Hoelmer, A. A. Kirk & P. C. Quimby USDA, ARS, European Biological Control Laboratory Parc Scientifique Agropolis II 34397 Montpellier Cedex 5, France phone: 33-467-41-6730; fax: 33-467-04-5620 e-mail: ebcl@cirad.fr **************************************************************** Subject: Archive of previous postings Date: Wed, 10 Feb 1999 14:27:41 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact An archive of previous postings to this mail-list (sorted by date, with last date first) is now available at http://www.udel.edu/entomology/khopper/ Please let me know if you have any difficulties accessing this archive. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Risk-benefit context Date: Wed, 10 Feb 1999 14:41:27 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Reply to message posted by Tom Unruh Wed, 20 Jan 1999 17:26:55 -0800 > SO imagine this scenario, we develop a model using life history and > ecological data that indicates that species x will establish and produce > some measurable ecological effect on species a. Now we transform that model > to consider to consider a suite of notargets in turn and ask if it will > produce a measurable effect on each of these. I wonder just how many of the > pest systems we work in are known well enough to take this approach. Very > few I would warrant (fruitflies in Hawaii and a few other discrete examples > are possible). The difficulty of predicting target and non-target impacts given current understanding is why I suggested that we look at non-target impacts of past introductions to see whether they can tell us something about the frequency and magnitudes of such impacts for various kinds of introductions. Past performance may be the best predictor for the future (as it appears to be with invasive plant species). > I would suggest that we address certain components of a risk rating cascade > in order. I would suggest also that we come to some consensus on the value > of these and identify research areas that are adressable in them. > I apologize for being repetitive but I would suggest that host range is > critical and with parasitoids and predators it (literature data) may be > sufficient in many cases. What standards of evidence are required to make > it sufficient before we must go into host range testing in the lab or in the > native home? Host range (from the literature, from field studies in the area of origin, or from pre-introduction lab studies) can in principle tell us whether there is any risk of impacts on the abundances of non-target species, but it seems unlikely that it will tell us what the magnitudes of changes in non-target species abundances are likely to be. Does this mean that we must restrict introductions to natural enemies that will not attack any non-target species? Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Predicting impacts Date: Fri, 12 Feb 1999 15:51:48 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Tom McAvoy, Thu, 21 Jan 1999 11:10:38 -0500: > 1. Field studies in country of origin done with as many taxonomically > related nontargets as possible cooperation needed here from botanist/plant > ecologist to obtain seeds/ roots etc. What constitutes taxonomically related? Is this the best criterion? What is likely to set the limits of host range for herbivorous insects? for parasitoids? Are there tools or rules of thumb for deciding which non-target species to screen for possible attack? Or does this have to be done on a case by case basis based on knowledge concerning the agent and its host? > Most non-target species have no economic value associated with them except > for their ecological value as part of a native community. How to asses that > I'm not sure. But significant decline in density & reproduction due to BC > agents would be cause for concern. Lauda has documented decline in native > cirsium seed production but I have not seen any data on changes in plant > density which is the bottomline. I assume that you do not mean a statistically significant decline. But what is a significant decline in density and what should be the spatial and temporal scales considered? > We're working with Lythrum salicaria & Carduus nutans. > For L. salicaria we're slowly planting taxonomically related nontarget > species in several loosestrife stands where the three Bio control agents > are established Galerucella pusilla, G. calmariensis & H. > transversovitattus. & sampling m2 plots for plant density & diversity we > hope to do this for the next 10 yrs at least. > Often the importment nontargets are not growing in the immediate area where > the agents are established. > We will be surveying thistle sites looking for nontargets in areas where > the BC agents have been established for over 20 yrs. This is excellent and much needed. But what has dictated your choice of spatial and temporal scales for sampling? How have you decided the range of habitats to study? Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: safety of bc agents Date: Fri, 12 Feb 1999 16:00:47 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Alec McClay, Thu, 21 Jan 1999 09:21:02 -0700: > The point about genetic variability is a valid one, however. What tests > could one do to distinguish the cases where (a) all individuals in the > population being tested have a 2% success rate on the non-target species, > and (b) 2% of the population have a genetic difference which allows them to > be 100% successful on the non-target? I think the risk of selecting a > population adapted to the non-target would be much higher in case (b). If the individuals that survive are genetically different than the starting set, one should be able to detect this rather easily by doing another generation of exposure and a control with previously unexposed insects. Significantly higher survival among the progeny of previously exposed insects compared to progeny of insects not previously exposed to the host would suggest a genetic difference. Maternal effects could eliminated with crosses between the original and selected insects. Yours, Keith -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Date: Fri, 12 Feb 1999 16:06:22 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Andre Gassmann, Thu, 21 Jan 1999 10:44:00 -0800: > Are the biologically significant non-target impacts which have been observed > so far (and to a lesser extend all new associations recorded) usually > associated with some level of success ?. This seems to be the case at least > in weed biological control. Is this because successes are more attractive to > researches than failures ? Or is this because of the demographic response > (or other biological traits which would need to be study in depth in the > context of understanding non-target impacts) of successful biocontrol agents? Would you give examples of (1) biologically significant non-target impacts which have been observed so far, and (2) cases where these have been associated with successful biological control? Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: (Fwd) (Fwd) FWD: Invasive Species Executive Order Date: Mon, 15 Feb 1999 09:11:39 +0000 From: Carol Waddington Reply-To: bc-ntimpact@UDel.Edu Organization: Koppert Biological Systems To: bc-ntimpact@UDel.Edu FYI: From: dyanega@mono.icb.ufmg.br (Doug Yanega) Date: Thu, 4 Feb 1999 18:43:13 -0300 Apologies for crossposting. This may have some bearing on recent attempts to introduce non-native Bombus species in the US, as well as movement of native species to areas where they presently do not occur - I've trimmed things down slightly; the most pertinent sections are as follows: (1) the definitions of "alien species" and "native species" in Section 1, being limited to *ecosystem*, not geographical area (I think Dan Simberloff may have helped them with the definitions). (2) Section 2, part 3, regarding authorizing introductions that might prove harmful. At any rate, this is good to have in one's files for reference. From: ssi@ucsusa.org Date: Thu, 04 Feb 99 15:09:49 -0500 [snip] THE WHITE HOUSE Office of the Press Secretary For Immediate Release February 3, 1999 EXECUTIVE ORDER - - - - - - - INVASIVE SPECIES By the authority vested in me as President by the Constitution and the laws of the United States of America, including the National Environmental Policy Act of 1969, as amended (42 U.S.C. 4321 et seq.), Nonindigenous Aquatic Nuisance Prevention and Control Act of 1990, as amended (16 U.S.C. 4701 et seq.), Lacey Act, as amended (18 U.S.C. 42), Federal Plant Pest Act (7 U.S.C. 150aa et seq.), Federal Noxious Weed Act of 1974, as amended (7 U.S.C. 2801 et seq.), Endangered Species Act of 973, as amended (16 U.S.C. 1531 et seq.), and other pertinent statutes, to prevent the introduction of invasive species and provide for their control and to minimize the economic, ecological, and human health impacts that invasive species cause, it is ordered as follows: Section 1. Definitions. (a) "Alien species" means, with respect to a particular ecosystem, any species, including its seeds, eggs, spores, or other biological material capable of propagating that species, that is not native to that ecosystem. (b) "Control" means, as appropriate, eradicating, suppressing, reducing, or managing invasive species populations, preventing spread of invasive species from areas where they are present, and taking steps such as restoration of native species and habitats to reduce the effects of invasive species and to prevent further invasions. (c) "Ecosystem" means the complex of a community of organisms and its environment. (d) "Federal agency" means an executive department or agency, but does not include independent establishments as defined by 5 U.S.C. 104. (e) "Introduction" means the intentional or unintentional escape, release, dissemination, or placement of a species into an ecosystem as a result of human activity. (f) "Invasive species" means an alien species whose introduction does or is likely to cause economic or environmental harm or harm to human health. (g) "Native species" means, with respect to a particular ecosystem, a species that, other than as a result of an introduction, historically occurred or currently occurs in that ecosystem. (h) "Species" means a group of organisms all of which have a high degree of physical and genetic similarity, generally interbreed only among themselves, and show persistent differences from members of allied groups of organisms. (i) "Stakeholders" means, but is not limited to, State, tribal, and local government agencies, academic institutions, the scientific community, nongovernmental entities including environmental, agricultural, and conservation organizations, trade groups, commercial interests, and private landowners. (j) "United States" means the 50 States, the District of Columbia, Puerto Rico, Guam, and all possessions, territories, and the territorial sea of the United States. Sec. 2. Federal Agency Duties. (a) Each Federal agency whose actions may affect the status of invasive species shall, to the extent practicable and permitted by law, (1) identify such actions; (2) subject to the availability of appropriations, and within Administration budgetary limits, use relevant programs and authorities to: (i) prevent the introduction of invasive species; (ii) detect and respond rapidly to and control populations of such species in a cost-effective and environmentally sound manner; (iii) monitor invasive species populations accurately and reliably; (iv) provide for restoration of native species and habitat conditions in ecosystems that have been invaded; (v) conduct research on invasive species and develop technologies to prevent introduction and provide for environ-mentally sound control of invasive species; and (vi) promote public education on invasive species and the means to address them; and (3) not authorize, fund, or carry out actions that it believes are likely to cause or promote the introduction or spread of invasive species in the United States or elsewhere unless, pursuant to guidelines that it has prescribed, the agency has determined and made public its determination that the benefits of such actions clearly outweigh the potential harm caused by invasive species; and that all feasible and prudent measures to minimize risk of harm will be taken in conjunction with the actions. (b) Federal agencies shall pursue the duties set forth in this section in consultation with the Invasive Species Council, consistent with the Invasive Species Management Plan and in cooperation with stakeholders, as appropriate, and, as approved by the Department of State, when Federal agencies are working with international organizations and foreign nations. Sec. 3. Invasive Species Council. (a) An Invasive Species Council (Council) is hereby established whose members shall include the Secretary of State, the Secretary of the Treasury, the Secretary of Defense, the Secretary of the Interior, the Secretary of Agriculture, the Secretary of Commerce, the Secretary of Transportation, and the Administrator of the Environmental Protection Agency. The Council shall be Co-Chaired by the Secretary of the Interior, the Secretary of Agriculture, and the Secretary of Commerce. The Council may invite additional Federal agency representatives to be members, including representatives from subcabinet bureaus or offices with significant responsibilities concerning invasive species, and may prescribe special procedures for their participation. The Secretary of the Interior shall, with concurrence of the Co- Chairs, appoint an Executive Director of the Council and shall provide the staff and administrative support for the Council. (b) The Secretary of the Interior shall establish an advisory committee under the Federal Advisory Committee Act, 5 U.S.C. App., to provide information and advice for consideration by the Council, and shall, after consultation with other members of the Council, appoint members of the advisory committee representing stakeholders. Among other things, the advisory committee shall recommend plans and actions at local, tribal, State, regional, and ecosystem- based levels to achieve the goals and objectives of the Management Plan in section 5 of this order. The advisory committee shall act in cooperation with stakeholders and existing organizations addressing invasive species. The Department of the Interior shall provide the administrative and financial support for the advisory committee. Sec. 4. Duties of the Invasive Species Council. The Invasive Species Council shall provide national leadership regarding invasive species, and shall: (a) oversee the implementation of this order and see that the Federal agency activities concerning invasive species are coordinated, complementary, cost-efficient, and effective, relying to the extent feasible and appropriate on existing organizations addressing invasive species, such as the Aquatic Nuisance Species Task Force, the Federal Interagency Committee for the Management of Noxious and Exotic Weeds, and the Committee on Environment and Natural Resources; (b) encourage planning and action at local, tribal, State, regional, and ecosystem-based levels to achieve the goals and objectives of the Management Plan in section 5 of this order, in cooperation with stakeholders and existing organizations addressing invasive species; (c) develop recommendations for international cooperation in addressing invasive species; (d) develop, in consultation with the Council on Environmental Quality, guidance to Federal agencies pursuant to the National Environmental Policy Act on prevention and control of invasive species, including the procurement, use, and maintenance of native species as they affect invasive species; (e) facilitate development of a coordinated network among Federal agencies to document, evaluate, and monitor impacts from invasive species on the economy, the environment, and human health; (f) facilitate establishment of a coordinated, up-to- date information-sharing system that utilizes, to the greatest extent practicable, the Internet; this system shall facilitate access to and exchange of information concerning invasive species, including, but not limited to, information on distribution and abundance of invasive species; life histories of such species and invasive characteristics; economic, environmental, and human health impacts; management techniques, and laws and programs for management, research, and public education; and (g) prepare and issue a national Invasive Species Management Plan as set forth in section 5 of this order. Sec. 5. Invasive Species Management Plan. (a) Within 18 months after issuance of this order, the Council shall prepare and issue the first edition of a National Invasive Species Management Plan (Management Plan), which shall detail and recommend performance-oriented goals and objectives and specific measures of success for Federal agency efforts concerning invasive species. The Management Plan shall recommend specific objectives and measures for carrying out each of the Federal agency duties established in section 2(a) of this order and shall set forth steps to be taken by the Council to carry out the duties assigned to it under section 4 of this order. The Management Plan shall be developed through a public process and in consultation with Federal agencies and stakeholders. (b) The first edition of the Management Plan shall include a review of existing and prospective approaches and authorities for preventing the introduction and spread of invasive species, including those for identifying pathways by which invasive species are introduced and for minimizing the risk of introductions via those pathways, and shall identify research needs and recommend measures to minimize the risk that introductions will occur. Such recommended measures shall provide for a science-based process to evaluate risks associated with introduction and spread of invasive species and a coordinated and systematic risk-based process to identify, monitor, and interdict pathways that may be involved in the introduction of invasive species. If recommended measures are not authorized by current law, the Council shall develop and recommend to the President through its Co-Chairs legislative proposals for necessary changes in authority. (c) The Council shall update the Management Plan biennially and shall concurrently evaluate and report on success in achieving the goals and objectives set forth in the Management Plan. The Management Plan shall identify the personnel, other resources, and additional levels of coordination needed to achieve the Management Plan's identified goals and objectives, and the Council shall provide each edition of the Management Plan and each report on it to the Office of Management and Budget. Within 18 months after measures have been recommended by the Council in any edition of the Management Plan, each Federal agency whose action is required to implement such measures shall either take the action recommended or shall provide the Council with an explanation of why the action is not feasible. The Council shall assess the effectiveness of this order no less than once each 5 years after the order is issued and shall report to the Office of Management and Budget on whether the order should be revised. Sec. 6. Judicial Review and Administration. (a) This order is intended only to improve the internal management of the executive branch and is not intended to create any right, benefit, or trust responsibility, substantive or procedural, enforceable at law or equity by a party against the United States, its agencies, its officers, or any other person. (b) Executive Order 11987 of May 24, 1977, is hereby revoked. (c) The requirements of this order do not affect the obligations of Federal agencies under 16 U.S.C. 4713 with respect to ballast water programs. (d) The requirements of section 2(a)(3) of this order shall not apply to any action of the Department of State or Department of Defense if the Secretary of State or the Secretary of Defense finds that exemption from such requirements is necessary for foreign policy or national security reasons. WILLIAM J. CLINTON THE WHITE HOUSE, February 3, 1999. [NOTE THAT I WILL BE LEAVING BRAZIL ON FEB 13]* Doug Yanega Depto. de Biologia Geral, Instituto de Ciencias Biologicas, Univ. Fed. de Minas Gerais, Cx.P. 486, 30.161-970 Belo Horizonte, MG BRAZIL phone: 31-499-2579, fax: 31-499-2567 (from U.S., prefix 011-55) http://www.icb.ufmg.br/~dyanega/ C.Waddington Koppert B.V. C.Waddington@Koppert.nl Subject: RE: Re: Date: Mon, 15 Feb 1999 11:13:00 -0800 (PST) From: "Andre Gassmann (BPM-Weeds)" To: "Keith R. Hopper" An excellent review paper has been published by Rachel McFayden in the Annual Review of Entomology (1998), 43: 369-93, about damage to non-target plants in classical weed biological control, and other topics. All the 8 insect species listed by Rachel have been associated with some level of success, and all have had outbreaking populations. Another probable biological feature of some of the WB insects associated with ntimpact is their high dispersal capability (e.g. Cactoblastis cactorum and Rhinocyllus conicus ?), but little is known on this issue. There is some indication that Cactoblastis cactorum has a biologically significant impact on native cacti in southern USA. According to S. Lauda, this is the case also for R. conicus on native Cirsium spp. Does anybody know other examples of ntimpacts in weed biological control, but those listed by Rachel ? Andre Gassmann ---------- From: Keith R. Hopper To: bc-ntimpact@UDel.Edu Subject: Re: Date: Friday, 12. February 1999 13,06 Response to posting by Andre Gassmann, Thu, 21 Jan 1999 10:44:00 -0800: > Are the biologically significant non-target impacts which have been observed > so far (and to a lesser extend all new associations recorded) usually > associated with some level of success ?. This seems to be the case at least > in weed biological control. Is this because successes are more attractive to > researches than failures ? Or is this because of the demographic response > (or other biological traits which would need to be study in depth in the > context of understanding non-target impacts) of successful biocontrol agents? Would you give examples of (1) biologically significant non-target impacts which have been observed so far, and (2) cases where these have been associated with successful biological control? Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Non-target effects Date: Wed, 17 Feb 1999 17:52:07 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: "Workshop" Dear All Does anybody know any non-target effect concerning the natural enemies that were intentionally released on knapweeds and yellow starthistle? How about positive effects ? any effect on reduction of weed population ? Any information in this respect would be appreciated. Regards, R. Sobhian Subject: Re: design Date: Wed, 17 Feb 1999 17:59:24 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: Dear All I believe that the message below by Bernd Blossey is one of the best, if not the best, contribution to the workshop. I suggest that you read it once again ! I think this design is what we could build upon. Subject: Re: design Date: Fri, 26 Feb 1999 16:25:57 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Bernd Blossey, Mon, 25 Jan 1999 11:08:07 -0500 (EST) bernd blossey wrote: > > It is about time that the biocontrol folks (and it looks like all > subscribers are biocontrol folks) Actually, not all subscribers are biocontrol folks. One hopes that this remains the case; we need input from all sectors. > Certainly, biological control needs an infusion of quantification, > experimentation and long-term monitoring. It has always been a mystery for > me how we could get away for so long with releases that did not involve > quantitative assessments of their impact, particularly on the target. > Imagine Monsanto is wanting to sell you a herbicide/insecticide and you > have no idea whether it will work. Imagine even further that the Monsanto > representative is trying to convince you to purchase the product based on > the fact that this pesticide has been on the market for decades, that it > works and that no adverse effects have been reported. All this based on > some color photos and the assertion of independence - but without > quantitative backup. Would you be enthusiastic about spending a few 100,000 > dollars for this product? One hopes that the data for efficacy of biological control is in at least some cases more solid than this. > What we are asking of society, or agencies, is not too far removed from > this hypothetical example. Can we complain about less than enthusiastic > support? In many cases we do not even have the supporting evidence that we > are addressing a real problem. How many control programs have started > without an economic or ecological assessment of the impacts of a target > weed or insect pest? Sure, we are only getting invloved because of > complaints by those that need to manage pests, whether these are farmers or > (increasingly) natural areas managers. And the only reason that we get a > deal is because everything else has failed!! I think we need to do better > than that - and I believe we certainly can. It is certainly true that biocontol is often the means of last resort, which puts it at a disadvantage from the start. However, there have actually been a fair number of economic assessments of the value of biocontrol, at least for control of insect pests in crops. > 1. Pre-release studies > > These have to be conducted in the area of origin AND in the future release > area. Preferably we need to have some quantitative evidence of ecosystem > impacts before we begin control programs (that should be a prerequisite for > biological as well as any other control program). I would like to see this > work done by ecologists/botanists and fortunately invasive species problems > gain "sexiness" in ecology. I would reject claims that biocontrol workers > should be the ones that need to collect these data. I strongly favor > research in the area of origin that involves the impact and host > specificity of potential control agents. Only if we have such information > available can we then go out and design meaningful post-release protocols. > In particular, we need to know how plant individuals respond to attack by > single or multiple control agents. We may never be able to accurately > predict the resulting establishment or impact in the release area > (differences involving interaction of abiotic environmental factors, plant > and insect genotypes, plant communities, and natural enemy complexes may > make this impossible), however, the successful weed programs seem to > indicate that herbivore attack is really the driving force immediately > after the control agent release and at high weed densities. Without such > information from the native range it will be very difficult to improve > selection protocols. I certainly concurr that studies in the area of origin are crucial. My impression had been that this was more common in biocontrol of weeds than in biocontrol of insects, but perhaps this impression is wrong. There are certainly very few thorough ecological studies of factors limiting target insects in the area of origin. Usually one sees only estimates of the proportion of a collection of insects that have been parasitized. Some argue that the results of such studies cannot be extrapolated to the area of introduction, but how can we know this if it has not tried sufficiently often? > I don't think we need to worry about the inability to collect regional or > local baseline data even if control programs have been implemented years > ago. Landscape level occurrence of bc agents or control does not occur over > night. For the purple loosestrife program the number of wetlands where > insects were released is well over 1000 in 30+ states. However, there are > still thousands of sites where pre-release data can be obtained. I guess > that even with 30 years of biocontrol targeting a species such as leafy > spurge I can find plenty of sites (particularly in the Midwest and > Northeast) that are free of control agents. This is not to argue that such > pre-release investigations should begin before the very first introduction > into North America occurs. In addition, the longer the data were collected > before control agnets were released, the better. I agree whole-heartedly that there is much opportunity for retrospective analyses because of sites not yet reached by introduced natural enemies. I would go further and say that we can also look at evolution of host range and even use introductions that have few or any remaining unoccupied sites (by using life-tables and exclusion techniques). > But make no mistakes: nobody will design such protocols for you. And we > can't expect that a person receiving a new species of insect he/she has > never seen before is able and knowledgeable enough to design a meaningful > monitoring protocol. It is up to us to develop these. And I am convinced if > we design them userfriendly, they will be used. People are just waiting for > this type of protocol. ACCEPT THIS RESPONSIBILITY! There is an extensive literature on measuring impacts of natural enemies (both introduced and native) on target species. Why not use these techniques for measuring impacts on non-targets? > For scenario 1 we select sites where target weed and potential non-target > host co-occur, release control agents and follow the fate of target weed, > potential non-target host(s) and control agents over time. Ideally we have > several replicate sites, some of which serve as controls (no control agent > release). We survey (I personally prefer permanent quadrats) for the > presence of feeding, oviposition, larval feeding etc. on target weed and > non-target hosts and ultimately for changes in population abundance. This > work will continue for several years, ideally following population build-up > of control agents and decline of target weed. Many people are concerned > about non-target effects when populations of control agents are high and > subsequently decimate their original host. By following this at sites > through time we will be able to assess whether this occurs at all and > whether the likelihood of a "spill-over" effect is dependent upon > population levels of the control agents and food availability. This should > also demonstrate that control agents are more likely to leave an area > rather than consume unsuitable food, as is the case for host specific > organisms in their home range (so we think but have not backed this up by > data). The issue of the persistence of impacts is crucial. Really we need to think and talk more carefully about what we mean by a change in abundance. The spatial and temporal scales of changes in abundance that acceptable versus unacceptable need to be aggreed upon. In the past, I've asserted that we as scientists should just measure the impacts and rely on society (e.g., elected representatives and their delegates) to decide what is acceptable versus unacceptable. I now think that this is not a tenable position. To design studies of non-target impacts, we need some consensus about amount and spatio/temporal pattern of changes in abundances of non-target species that would be considered acceptable versus unacceptable. Is developing such a consensus possible? Will it have to be on a case-by-case basis? How should this be done? > This design would at least get to the first order effects. Addressing > higher order or indirect effects will be a different story. I think it is > time to go out and do it. And it is up to us to set positive precedence. We > will be much stronger and more credible in the future. Although I strongly support studies in the area of origin, retrospective studies in the area of introduction, and pre versus post-release monitoring, I'm still worried that we will not be able to predict magnitudes (amounts and spatial and temporal patterns) of impacts on abundances of non-target species prior to introduction. However, we will not know whether we can or not unless we try. > A last remark re. monitoring for non-target effects in insect biocontrol > programs. It does not appear difficult to design such protocols, however, > it may be quite elaborate. The lack of information guiding the searches due > to lack of host specificity testing may make this particularly difficult. > But I could imagine a routine collection effort of potential non-target > hosts, followed by rearing of the appropriate life stage in the lab to > assess parasitoid attacks. Alternatively we could use the deliberate > exposure of non-target organisms at the right life-stage on a "trap-plant" > in the right habitat or the habitat that we want to survey, allow > parasitoid attack for a standardized time, remove trap-plants and check for > emergence of parasitoids (or collect potential predators in specifically > designed traps). This is just a start but I really think it is not at all > that difficult. Conceptually, measuring non-target impacts on insects is not difficult. However, because insect populations often change by orders of magnitude from multiple causes, measuring the contribution of natural enemies to these changes can be very difficult. If it were easy, we would have many more case histories where the role of predators and parasitoids was well documented. Nonetheless, biocontrol introductions provide wonderful opportunities for testing hypotheses about population regulation and evolution. Let us work together with those critical of biological control introductions to develop sound protocols for really finding out what is going on! I look forward very much to the results from this work on purple loosestrife. Yours, Keith Hopper ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Weed bc non-target examples Date: Sat, 06 Mar 1999 16:38:19 -0500 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by John Herr Tue, 2 Feb 1999 14:02:33 -0800 (PST): > Getting back to the questions asked by Keith Hopper on Jan. 20, 1999 re: > my original posting of non-target weed bc examples: > > > Why is Opuntia spinossima rare? It is not because of attack by C. > > cactorum because the cactus was rare long before C. cactorum was > > accidentally introduced. > > Yes, this is true. However, it does not make the non-target impact any > less significant or noteworthy. If anything, we should be concerned most > with impacts to rare native species, regardless of the reason for their > rarity, because in the unlikely event that an extinction is caused by a > bc agent, it would probably occur in rare species first. The rare > California Cirsium thistle species I am working with are extreme habitat > specialists (endemic to serpentine soils near springs, for the most part), > so their rarity is definitely not attributable to the introduction of > Rhinocyllus conicus. I'm just trying to determine if R. conicus is > contributing to a further decline in their abundance. I quite agree that we should not go about introducing species that attack endangered species. My point is that we don't have evidence that species introduced for biological control make common non-target species rare or drive them extinct. > > First, my understanding is that R. conicus did not expand its host range > > in an evolutionary sense. Attack on native Cirsium species was > > understood to be likely before introduction. > > Although it was considered likely that R. conicus would attack native > North American Cirsium species because European Cirsiums were within its > host range, it is not known if genetic changes have facilitated this > movement onto new hosts. The existence of genetically distinct biotypes > of R. conicus is well documented (Goeden et. al 1985; Unruh and Goeden > 1987; Klein and Seitz 1994), and three strains were introduced (Italian > thistle, Milk thistle, and Musk thistle), so it would appear that the > weevil has at least the potential to switch hosts in an evolutionary > sense. Obviously, evolution is change in genic/genotypic frequencies, but I guess by evolutionary host shifts I meant more than changes predictable from genetic variation measurable in source populations. The question is whether one can predict whether new, non-target species will be attacked prior to making releases. > Marohasy (1996) is of the opinion that the terms 'host shift', 'host > switch', and 'host range expansion' should only be used if there is > evidence of a genetic change in the bc agent. Others (e.g. Louda 1998) > seem to use 'host range expansion' when any new additions to a host range > are observed. Marohasy claims there is no evidence of a true host shift > ever having occurred in a weed bc agent, and prefers the term 'host > substitution' instead. > > In any case, this might just be an issue of semantics, because the bottom > line is that if bc agents are causing biologically significant impact to > unintended hosts, we should be concerned about it, whether or not > evolutionary changes are involved. I disagree that this is an issue of semantics. If the host shift is ecological or within the range of measurable genetic variation in the source populations than it is predicable from host range screening before introduction. If the shift involves evolution beyond that predictable from genetic variation measurable in the source population, it will be much harder to predict. > > What I would like to see > > for R. conicus impact on Cirsium spp. is evidence that the weevil has or > > is changing the distribution/abundance of the non-target host plants and > > the magnitude of such changes. This is not a system I know well, but > > the evidence that Turner et al. (1987) Louda et al. (1997) and Louda & > > Potvin (1995) are not as strong as what would be required to demonstrate > > conclusively that the beetle is affecting distribution/abundance of > > non-target plant species. This is not to say that it does not have such > > impacts, I would just like to see stronger evidence. > > Louda reasons that because seed predation by native insects significantly > reduces seedling recruitment, local population density, and lifetime > fitness of Platte thistle (Louda and Potvin 1995), and that because the > recent movement of R. conicus onto this thistle further reduces seed > availability (Louda et. al 1997; Louda 1998), the result will be a reduced > population size. I agree that more direct evidence would be nice, but of > course its hard to selectively exclude R. conicus from thistles and still > let in the native leps. and tephritids. I'm trying to get around this > problem by using a seed addition experiment to replace the seeds destroyed > by R. conicus in the native thistle population. So far, I have not found > evidence that R. conicus is limiting seedling recruitment rates in the > rare California species I am studying. I understand Louda's reasoning; I'm just wary that such reasoning can end up wrong when tested in the field. I'm especially interested in the spatial and temporal extent of reductions in abundance. No one said it would be easy. It is great that you are delving into this system in more depth; I look forward to your reading your results. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Internet workshop summary at IOBC-sponsored symposium/workshop Date: Sat, 29 May 1999 17:13:10 -0400 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu I've agreed to summarize the deliberations of this internet forum at an IOBC-sponsored symposium and workshop on "Evaluating Indirect Ecological Effects of Biological Control", in Montpellier, France, 17-20 October, 1999 (website: www.agropolis.fr/iobc). The idea to build upon this internet workshop and avoid going over much of the same ground again. However, the internet workshop has been rather silent since February. Nonetheless, many issues remain unresolved, and I've encountered many people who have admitted to being silent listeners rather than participants. If you have suggestions about research needs and approaches for addressing potential non-target impacts of introduced biological control agents, this forum is an excellent way to make your views known and have them discussed. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Internet workshop summary at IOBC-sponsored symposium/workshop Date: Tue, 1 Jun 1999 00:19:20 +0400 From: crinel@bibosi.scz.entelnet.bo (Christopher & Nelly Pruett) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Santa Cruz de la Sierra, Monday, 31 sth of May, 1999 Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 Dear Keith, With reference to your recent message, I feel a bit at a loss as to what to contribute !! During 1960 and 1970 Bolivia inadvertently introduced hyperparasitoides of Tachinid flies for use in sugar cane (documented) and there are a couple more Tachinid sugar cane borer larvae that also could have been introduced as well as a parasitoid of Cotesia spp and Apanteles spp. (Hymenoptera, Braconidae, Microgastrinae), I can let you have full details, dates, records, etc., if you so wish. More recently I was worried about the release of imported Telenomus remus from Colombia; millions of Trichogramma reared in Colombia are actually released in sugar cane and other crops in Bolivia and this is against International agreements as pathogens and mites could also be imported as well as pesticide hosts such as Sitotroga cerealella (grain moth) and Spodoptera frugiperda or other species of Spodoptera. I enclose a copy of a letter that I sent out to 10 colleague in neighboring countries, and the only reply that I ever received was from Dra. Fulvia García, ICA, Palmira, Cauca, Colombia. Awaiting your reply, comments and orientation, Sincerely Chris Pruett CURRENT ADDRESS DETAILS CHRISTOPHER JH PRUETT BSc(Hons)., MIBiol., CBiol., FRES Jefe de Entomología Agrícola y Manejo Integrado de Plagas IIA "El Vallecito" Facultad de Ciencias Agrícolas Universidad Autónoma "Gabriel René Moreno" Casilla 702, Santa Cruz de la Sierra Bolivia Teléfono y Fax: Oficina: 591 3 422130 Domicilio: 591 3 534510 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> Durante los años 1990 los colombianos vendieron Telenomus remus sin recuperarlo del campo ni controlar la plaga. El 1 de Octubre, 1997 escribí una carta referente al tema de Spodoptera y Telenomus remus a amigos en Cuba (Ing. Agr. Jorge Luis Ayala, Laboratorio Provincial de Sanidad Vegetal, Unidad de Lucha Biológico, Sancti Spiritus y Dra. Esperanza Rijo Camacho, IISV, Habana), en Brasil (Dr. Evoneo Berti Filho, Universidade de Sao Paulo (ESALQ), Dra. Elizabeth AB Nardo, EMBRAPA/CNPMA, Jaguariúna, Dr. Roberto A Zucchi, Universidade de Sao Paulo (ESALQ), Dr. Dirceau Gassen, EMBRAPA y la Universidade Federal de Viçosa), en Perú ( Dra, Elizabeth Nuñez, SENASA, Lima) y en Colombia (Dra. Fulvia García, ICA, y Ing. Agr. Jimenez Jades, Coinbiol, Palmira, Cali). En breve considero que si no se logra el establecimiento de un parasitoide exótico, como Telenomus remus, a través de control biológico clásico, no vale la peña hacer un control biológico inoculativo de un parasitoide exótico NO establecido en el área. Santa Cruz de la Sierra, Bolivia, 1 de Octubre, 1997 Problemática del uso de Telenomus remus en los cultivos de algodón, maíz, trigo, girasol y soya para controlar huevos de Spodoptera en Bolivia. En Santa Cruz durante los últimos años Spodoptera frugiperda ha causado grandes pérdidas en cultivos como maíz, sorgo, trigo, pastos, soya y algodón. En la campaña algodonera de 1994/95 fue necesario emplear alrededor de 1,200 mil $US en plaguicidas (6 a 7 aplicaciones, 3 de ellas para Spodoptera spp. en 25,000 ha, a un costo mínimo de 15 $US cada una) para su control, sin tomar en cuenta las pérdidas económicas causadas en el rendimiento al ser dañados los bolos y la fibra. En 1996/7 las aplicaciones se aumentaron a 8 a 10 y las larvas se encuentran no solamente como defoliadores, pero, si también abajo de las bractas de los bolos y las masas de huevos en las hojas inferiores de las plantas; ademas larvas de mayor tamaño migran de las malezas gramineas entre surco, despues de la carpida, atacando flores y bolos. El cultivo de maíz es seriamente afectado por esta plaga e en la temporada 1994/95 demando control químico, con una hasta tres aplicaciones en 89,000 hectárea, con un costo de más de 1,500 mil $us. Solamente en estos dos cultivos en 1994/5, S. frugiperda demandó el uso de casi tres millones de dolares de insecticidas para su control, sin contar perdidas en rendimientos, contaminación al ambiente y la aparición de nuevas plagas peligrosas coma, Bemisia tabaci, en algodón y soya. Otras especies de Spodoptera como S. sunia, S. exigua y S. latifascia, atacan, cada vez con más intensidad, al cultivo de la soya, complicando el manejo de plagas. Durante los últimos dos años altas incidencias de S. sunia y S. exigua, dejan cientos de hectáreas de girasol totalmente defoliados, destruyendo las paniculas (con hasta diez larvas por panicula), causando pérdidas de 50 a 100 % en los rendimientos, durante 1997 fue casi obligatoria tomar medidas de control, aunque registramos niveles arriba de 99% parasitoidismo , principalmente por tachinidas y ichneumónidos, y una abundancia de depredadores en las plantas (Podisus spp. y Tynacantha marginata) coccinélidos (Eriopis y Cycloneda) y véspidos, ídem en el suelo (Calosoma y diversos carábidos y redúvidos). En Bolivia Spodoptera spp. no tienen parasitoides de huevos nativos y presentan poblaciones muy elevadas, aunque abundan parasitoides de larvas y pupas. El parasitoide exótico de huevos de Spodoptera spp., Telenomus remus es capaz de penetrar la capa de escamas con que cubren las posturas de huevos para alcanzar parasitoidear los huevos más ocultos en las camadas inferiores. Durante ocho años (1981 a 1989), en General Saavedra, el Centro y Mejoramiento de la Caña de Azúcar (CIMCA) multiplicó Spodoptera spp. y varias razas de Telenomus remus liberando constantemente (casi semanalmente) alrededor de un total de 250.000 adultos en parcelas de maíz atacadas por Spodoptera spp. Aunque no se registró establecimiento del parasitoide tampoco no se evalúo las liberaciones dado la falta de interés de otras instituciones y la poca importancia de Spodoptera como plaga de la caña de azúcar. Cuadro 1. Extracto de "Parasitoides introducidos a Bolivia, 1980-1990: caña de azúcar y otros cultivos " (Fuente: Pruett, 1993) Hospedero Parasitoide Número Fecha Fuente Spodoptera spp. Telenomus remus 2200 1981 CIBC raza dominicana 750 1981 Bioserv Ltda. raza asiática 800 1981 Río Paila raza asiática 1500 1981 CIBC raza dominicana 5000 1981/2 CIBC raza dominicana 1000 1983 Providencia raza dominicana 2000 1983 Río Paila raza dominicana 1300 1984/5 CIBC raza de Islas Caimán 7000 1988/9 Univ. Florida raza puertorriqueña 6000 1988/9 F.D. Bennett Total 21050 1981 a 1989 En 1997 técnicos del CIAT recuperó parasitoides de huevos de Spodoptera en General Saavedra, siendo casi seguro que son Telenomus remus y son descendientes de las liberaciones efectuadas por CIMCA; sin embargo no se entregaron adultos a mi persona para identificación (tengo especímenes de las razas mencionadas, también de Perú, de 1997) y se mantiene la cría bajo de condiciones de seguridad. El punto crucial es que no sabemos la distribución de Telenomus remus en huevos de Spodoptera spp. en maíz, algodón o hospederos silvestres no podemos recomendar control biológico aumentativo inoculativo. Entonces es de suma importancia evaluar la distribución de T. remus para averiguar su agresividad indicada por una buena distribución en Santa Cruz, y decidir en la táctica más apropiada, sea control biológico clásico o inoculativo, y, así intentar controlar eficazmente Spodoptera frugiperda.. Una vez exista una buena y amplia distribución, se pude utilizar Telenomus remus, a través del método de control biológico aumentativo, en forma inoculativa o inundativa, con liberaciones al inicio del ataque o liberaciones semanales o quincenales, utilizando masas de huevos de Spodoptera parasitoidadas por T. remus. No tengo disponible datos de Brasil, Colombia, Perú o Venezuela referentes a las cantidades de masas de huevos parasitoidadas necesarias por hectárea en maíz ni en algodón. Sin embargo en Santa Cruz están comercializando la venta de masas de huevos de Spodoptera de Colombia parasitoidados por T. remus, efectuando liberaciones de 4 a 8 masas de huevos por hectárea presiembra (no se especifica si es en maíz o algodón o cualquier cultivo o cualquier especies de Spodoptera) para un costo de aproximadamente 35$us por hectárea. Veo peligro en la importación continua de Colombia de huevos de Spodoptera parasitoidados por Telenomus remus por dos motivos: Si no hay una población "residente" de T. remus como puede aumentar la, y pienso es probable que los individuos liberados migarían a los bordes, rompevientos o áreas de monte de 30 metros de anchura que se dejan entre tablones cultivados. Si los huevos de Spodoptera no están totalmente parasitoidados por T. remus (casi imposible), entonces estaremos importando no solamente enemigos naturales, en este caso parasitoides, pero también plagas en la forma de especies de Spodoptera posiblemente no presentes en los agroecosistemas de Santa Cruz, o, si existen, entonces pudiendo ser razas resistentes a los insecticidas comúnmente utilizado en Santa Cruz. Según las normas internacionales se debían importar especies exóticas como, Telenomus remus, autorización apropiada y mantener las en una institución autorizado y adecuadamente equipada bajo de condiciones de cuarentena, y luego multiplicar lo en forma masiva en este país. Uso promisorio de Bracon kirkpatricki en algodón para controlar Pectinophora gossypiella en algodón en Santa Cruz, Bolivia. Un ejemplo de control biológico clásico exitoso en algodón en Brasil y Bolivia El parasitoide larval mexicano de Pectinophora gossypiella, Bracon kirkpatricki, fue introducida en un programa de control biológico clásico desde México a Brasil, con tanto éxito que ahora en su vecino país limítrofe Bolivia, se encuentra difundido en cultivos de algodón en el departamento de Santa Cruz, en campos cerca a Cotoca, Enconada, Los Troncos, Tres Cruces, la Brecha Sur, etc., incluyendo "El Vallecito" casi a dentro de la ciudad de Santa Cruz de la Sierra, junto con el otro parasitoide boliviano de Pectinophora gossypiella, el betílido, Parasierola sp. La avispa B. kirkpatricki si es un excelente candidato para efectuar un eficaz control biológico aumentativo y se encontró niveles de parasitoidismo de 80 % setiembre de 1995 y niveles altos en diferente propiedades en 1996. No hay registro de la importación de B. kirkpatricki, aunque difícilmente se podía llegar a Bolivia desde los cultivos de Brasil. El Setiembre observamos en casi todas las zonas algodoneras de Santa Cruz, millones de bolos en cientos de hectáreas con algodón en plena formación, y bolos de todos estados de maduración (15 de julio es la última fecha por decreta suprema para incorporar o destruir rastrojos de algodón, con multa de 10$us/hectárea para incumplimiento). Estos bolos, sin excepción contienen 1 a 3 larvas de lagarta rosada Pectinophora gossypiella, y 1 del picudito, Conotrachelus denieri. Sin embargo se observa cantidades de adultos de B. kirkpatricki, volando en el campo y hembras penetrando bolos verdes para parasitoidear larvas. En este momento en el laboratorio de Entomología Agrícola del El Vallecito, tenemos 10,000 bolos de algodón recolectados de la zona central algodonera de Santa Cruz. De estos bolos están emergiendo adultos de B. kirkpatricki y, también, adultos de los betílidos, Parasierola sp., (parasitoide de Pectinophora registrado en Perú). No tengo información referente a las cantidades de Bracon kirkpatricki para liberar / hectárea, ni donde conseguir dieta artificial o si existe otras dietas mas rústica dieta artificial para la cría del hospedero Pectinophora. El contenido y principios de control biológico en este documento son entendidos por los Directores y el Gerente Técnico, Ing Agr Daniel Durán, de ADEPA, el Director del El Vallecito, Ing Agr Antonio González y algunos agrónomos y productores. Entonces el propósito de enviar este documento con sus datos y interrogaciones es porque reconocemos que necesitamos su urgente asesoramiento, comentarios, consejos, recomendaciones y ayudo. Además nos conviene importar, criar y liberar todas las razas posibles de Telenomus remus que están en uso y establecidas en países latinoamericanos, como Perú (Senasa), Venezuela (BioServ Ltda.), Colombia (ICA y Coinbiol Ltda.), Cuba (IISV) y Brasil (?Embrapa/?USAQ) para intentar establecer una raza adaptadas a las condiciones agroecológicos en Santa Cruz. Por este motivo le pido a través de esta comunicación, al mismo tiempo de responder a los puntos en el texto del documento, indicar si hay disponibles pies de cría de Telenomus remus, y si se podía enviar una cantidad necesaria de posturas de Spodoptera parasitoidados para establecer su cría en Santa Cruz. Si quiere enemigos naturales de Bolivia con mucho gusto le enviaré detalles completos por Email o por correo. Agradeciéndole anticipadamente su gentil respuesta lo pronto posible, le saludo con toda atención. Christopher JH Pruett H BSc, MIBiol, CBiol, FRES Entomología Agrícola & Manejo Integrado de Plagas El Instituto de Investigaciones Agrícolas "El Vallecito" Facultad de Ciencias Agrícolas, Universidad Autónoma "Gabriel René Moreno" Facsímile 591 3 534510 / 422130 <<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>> X-Authentication-Warning: copland.udel.edu: majordom set sender to owner-bc-ntimpact using -f Date: Sat, 29 May 1999 17:13:10 -0400 From: "Keith R. Hopper" Organization: University of Delaware X-Accept-Language: en MIME-Version: 1.0 To: bc-ntimpact@UDel.Edu Subject: Internet workshop summary at IOBC-sponsored symposium/workshop Sender: owner-bc-ntimpact@UDel.Edu Precedence: bulk Reply-To: bc-ntimpact@UDel.Edu Status: I've agreed to summarize the deliberations of this internet forum at an IOBC-sponsored symposium and workshop on "Evaluating Indirect Ecological Effects of Biological Control", in Montpellier, France, 17-20 October, 1999 (website: www.agropolis.fr/iobc). The idea to build upon this internet workshop and avoid going over much of the same ground again. However, the internet workshop has been rather silent since February. Nonetheless, many issues remain unresolved, and I've encountered many people who have admitted to being silent listeners rather than participants. If you have suggestions about research needs and approaches for addressing potential non-target impacts of introduced biological control agents, this forum is an excellent way to make your views known and have them discussed. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: RE: Internet workshop summary at IOBC-sponsored symposium/worksho p Date: Tue, 1 Jun 1999 17:18:18 +1200 From: "Barratt, Barbara" Reply-To: bc-ntimpact@UDel.Edu To: "'bc-ntimpact@UDel.Edu'" Having so far been one of the "silent listeners" in this 'virtual conference' I must say that I have very much enjoyed reading the contributions to it to date. It is a great idea, and an excellent opportunity to debate a topic with a much larger number of participants than could be assembled at an actual conference. The programme that my colleagues and I have been involved with on non-target effects of biological control agents has essentially compared predictions of host range of entomophagous biological control agents made in containment, with realised host range in the field. This is an example of the type of study that has been advocated by several participants in this workshop, so it might be useful to briefly outline what has been done so far, and what it has told us. The aim of the programme is to contribute towards the decision support framework of regulatory agencies which approve biological control introductions. We have used, as model systems, two species of braconids already introduced into New Zealand, Microctonus aethiopoides released in 1982 to control the lucerne pest, Sitona discoideus, and Microctonus hyperodae released in 1991 for biocontrol of the ryegrass pest Listronotus bonariensis. Very briefly, the approach we took was to identify potentially 'at risk' non-target hosts of the two braconid species by examining phylogenetic, ecological and behavioural affinities of the native fauna with the target hosts. Naturally, this was subject to limited knowledge of the ecology and species composition of the insect fauna. The extent to which potential non-target species were found in mixed populations with the target hosts, and therefore in direct contact with parasitoid populations was investigated by field sampling. Laboratory host range tests were then carried out on a sub-set of species we had identified as 'at risk'. From the lab data we then attempted to predict field host range. Finally, realised field host range was investigated by sampling and collecting non-target species from within and outside of the normal environment of the target hosts. We were then in a position to compare laboratory predictions with parasitism in the field. We found that the lab data had been reasonably indicative of field host range. The two baconids provided a good comparison; M. aethiopoides successfully attacked a larger number and wider taxonomic range of non-target species in the lab than M. hyperodae, achieved higher levels of successful development in non-target species and more successfully overcame host immune responses. In the field, M. aethiopoides has been found attacking a larger number and taxonomic range of native and non-target species in the field, achieved higher levels of parasitism and attacked non-target species at a greater number of locations than M. hyperodae. Ironically, one of the non-target hosts of M. aethiopoides was Rhinocyllus conicus, the subject of many contributions to this workshop! For our particular test cases, these data were quite reassuring in terms of the predictive capability of lab host range testing, but the implications are more difficult to determine. To what extent does M. aethiopoides represent a threat to non-target species? We have examples of up to 70% parasitism of a native weevil population in the field in a modified pasture environment, but because M. aethiopoides is so widespread we are unable to make comparisons with equivalent unparasitised populations, and we have no pre-release native weevil population data from pasture. In an attempt to address this, we have work recently initiated where replicated M. aethiopoides releases (and similar control areas) have been made in a rangeland area where we have long-term data on native weevil populations (fortuitously from another study), and where M. aethiopoides is currently not present. No data from that study yet. Regular sampling of native weevil populations where parasitism occurs has shown that for some non-target species, parasitism levels can be quite high, but it occurs after the main oviposition period for the weevils is complete. The impact may therefore be minimal. For others where there is a second generation, or where reproductive activity occurs later in the season, the impacts may be more severe. Furthermore, natives which occur at higher altitudes (in natural ecosystems) are likely to become reproductively mature later in the season, and may be more vulnerable to parasitoids invading from lower altitudes. We have some on-going monitoring at higher altitude sites to examine that. Without even considering indirect impacts at other trophic levels, it is clear that the system is complex and dynamic. Someone suggested that an equilibrium might be reached at some stage after the release of a biological control agent. The retrospecive study of a number of previous biological control introductions spanning a number of years post release might be one way to address that issue. However, the bottom line from these investigations will be to ask whether in hindsight the same decision would be reached about whether or not to release a particular biological control agent, given that regulations have changed in some countries in recent years. Coming back to the topic of this debate (and then I'll stop!), I think that research needs have been identified by the participants of this workshop, and the difficulties involved in producing quantitative data on real impacts on non-target plant/insect populations has been well highlighted. Retrospective study of impacts of biological control agents released in the past with a comprehensive coverage of representative biological control agent systems is important in my view. Proposals to do this in New Zealand have met with enthusiasm in principle from funding agencies, but additional funds to expand the programme have not been forthcoming. The European ERBIC programme will hopefully go some way towards achieving these objectives with their case studies and associated projects. Clearly, we are unlikely to come up with a single 'protocol' for research because every case will demand different approaches to quantifying impacts. Regulators are likely to want a risk analysis approach to be taken in the decision making process, so it will be useful in the long term to carry out research in the context of this framework. In accepting that we will probably never reach zero risk in this process, it is achieving balance and appropriate management of the 'precautionary approach' in cases of scientific uncertainty which will be the challenge. ================================================ Dr Barbara I.P. Barratt Phone: (03) 489 3809 AgResearch FAX: (03) 489 3739 Invermay Agricultural Centre Email: Barrattb@AgResearch.cri.nz Private Bag 50034 Mosgiel New Zealand ================================================ > -----Original Message----- > From: Keith R. Hopper [SMTP:khopper@UDel.Edu] > Sent: Sunday, 30 May 1999 09:13 > To: bc-ntimpact@UDel.Edu > Subject: Internet workshop summary at IOBC-sponsored > symposium/workshop > > I've agreed to summarize the deliberations of this internet forum at an > IOBC-sponsored symposium and workshop on "Evaluating Indirect Ecological > Effects of Biological Control", in Montpellier, France, 17-20 October, > 1999 (website: www.agropolis.fr/iobc). The idea to build upon this > internet workshop and avoid going over much of the same ground again. > However, the internet workshop has been rather silent since February. > Nonetheless, many issues remain unresolved, and I've encountered many > people who have admitted to being silent listeners rather than > participants. If you have suggestions about research needs and > approaches for addressing potential non-target impacts of introduced > biological control agents, this forum is an excellent way to make your > views known and have them discussed. > > Yours, Keith Hopper > -- > ****************************************************************** > Keith R. Hopper, Research Entomologist > USDA, ARS, Beneficial Insect Introduction Research Unit > University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 > Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 > ****************************************************************** Subject: Re: Internet workshop summary at IOBC-sponsored symposium/workshop Date: Wed, 2 Jun 1999 12:59:40 -0500 From: "Jim Fuxa" Reply-To: bc-ntimpact@UDel.Edu To: Keith, I do not know exactly what you want, but I can briefly outline my approach. I work with insect pathogens, and, with respect to risk assessment, I concentrate on genetically modified viruses. The rationale for my research in this area is: 1) harmfulness to non-target organisms (NTO) is impossible to predict a priori, regardless of how many NTO species are not harmed by the genetically modified organism (GMO) in the lab. 2) If the GMO cannot persist or spread after release, then its probability of coming into contact with NTO is reduced. 3) If this is the case, then the RISK of releasing the GMO is reduced even if we cannot predict its HARMFULNESS. Therefore, of course, much of my work has concentrated on persistence, spread, GMO-versus-WT (wild-type) competition, etc. Some of my data contributed in small part to an EPA decision to allow small-plot field testing of recombinant viruses a few years ago. Of course, this rationale gives me a good excuse to test hypotheses in viral ecology. I hope this is helpful. Best wishes -- Jim Fuxa ---------- > From: Keith R. Hopper > To: bc-ntimpact@UDel.Edu > Subject: Internet workshop summary at IOBC-sponsored symposium/workshop > Date: Saturday, May 29, 1999 4:13 PM > > I've agreed to summarize the deliberations of this internet forum at an > IOBC-sponsored symposium and workshop on "Evaluating Indirect Ecological > Effects of Biological Control", in Montpellier, France, 17-20 October, > 1999 (website: www.agropolis.fr/iobc). The idea to build upon this > internet workshop and avoid going over much of the same ground again. > However, the internet workshop has been rather silent since February. > Nonetheless, many issues remain unresolved, and I've encountered many > people who have admitted to being silent listeners rather than > participants. If you have suggestions about research needs and > approaches for addressing potential non-target impacts of introduced > biological control agents, this forum is an excellent way to make your > views known and have them discussed. > > Yours, Keith Hopper > -- > ****************************************************************** > Keith R. Hopper, Research Entomologist > USDA, ARS, Beneficial Insect Introduction Research Unit > University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 > Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 > ****************************************************************** Subject: Internet workshop summary at IOBC-sponsored symposium/workshop Date: Wed, 16 Jun 1999 23:06:37 -0700 (PDT) From: Elizabeth De Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear Dear subscribers, Befor Keith Hopper summarizes the internet forum I would like to discuss some issues that remained unresolved( at least for me) . I admit to being silent listener rather than participants so far. Lets go! I work in a Quarantine Laboratory for biological control agents in Brazil (Embrapa Meio Ambiente -Jaguariuna -Sao Paulo), the only one in the country responsible for importing and evaluating the risk of each introduction. So a protocol to assess the risk of non-target impacts of biological introductions is needed to comply with current national, regional, and international phytosanitary and biodiversity conservation laws. Keith Hopper and I are involved in a project aimed at elaborating protocols to predict the host range of parasitoids and predators in the area of introduction and we are working on a retrospective case study of the parasitoid Macrocentrus cingulum (= grandii) ( Hymenoptera -Braconidae),introduced into the US to control the European corn borer (ECB) Ostrinia nubilalis (Lepidoptera -Pyralidae) The determination of the potential host range of a parasitoid in a new environment should consider different approaches taxonomy/phylogeny biology, and host range in the area of origin are suggested in various papers. However, pratical details are often lacking. In developing a protocol, I've encountered the following problems: 1 ) How could the pylogeny of the natural enemy help to determine the host range in the area of origin? The majority of the parasitoids do not have a consistent pattern of host range. Some species of the parasitoids in the same genus have a broad host range including several different families while others have a more narrow range, e.g., a single family or even genus. 2) After we elaborate the list of potential non-targets of the exotic BCA in the area of introduction another problem that arises is the difficulties of collecting species in the field and rearing some of them under lab conditions. How should we select the species to be expose to the BCA under Quarantine Lab tests prior the BCA release? Sometimes we select species available in culture or easy to collect and rear (maybe pests), but are the results we get from the tests useful in predicting host range of the BCA in the area of introduction? 3) How many species should we test from the same family at risk and how many individuals of one species should we test to assess the risk to that species? Sometimes we are limited by in number of the parasitoids we have to expose to potential hosts, other times we are limited in number of the hosts. Furthermore, we are limited by the costs to perform tests. How should we deal with these problems to get reasonalble results concerning risks to non-targets? Who would like to helpme in those points? === Elizabeth De Nardo Embrapa Meio Ambiente- Brasil Laboratorio de Quarentena ACB http://www.bdt.org.br/bdt/biocontrol/ Current Address:OARDC/OSU Dept.of Entomology 1680, Madison Avenue-WOOSTER -OH 44691 _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com Subject: Re: Dr. Christopher JH Pruett Date: Fri, 18 Jun 1999 10:35:09 +0000 From: Carol Waddington Reply-To: bc-ntimpact@UDel.Edu Organization: Koppert Biological Systems To: bc-ntimpact@UDel.Edu From: inisav@ceniai.inf.cu To: bc-ntimpact@UDel.Edu Date: Thu, 17 Jun 1999 09:53:15 +0000 Subject: Dr. Christopher JH Pruett Priority: normal Reply-to: bc-ntimpact@UDel.Edu Dr. Christopher JH Pruett BSc(Hons), MlBiol,., Cbiol., FRES Jefe de Entomologia Agricola y Manejo Integrado de Plagas IIA El Vallecito GREETINGS, COLLEAGUES Doctor, recien hoy dia 10 de junio de 1999 conozco a traves de internet la situacion que estan atravesando para el control de fitofagos tan daninos como el complejo de Spodopteras, entre otros y las dificultades que se les ha presentado en tratar de reducir las poblaciones de los mismos con las liberaciones de Telenomus remus, Trichogramma y otros artropodos beneficos que han introducido en el pais, razon por la que me dispondogo a comunicarle mi opinion al respecto. Primero le dire que no soy muy partidaria del control biologico clasico, pienso que debe acudirse a el, cuando estamos seguros que no tenemos posibilidad de reducir las plagas con los medios biologicos DO WE AGREE THAT AN INTRODUCED NATURAL ENEMY WILL OCCUPY THE NICHE OF A NATIVE SPECIES AND ACCORDINGLY COMPLETELY DISPLACE IT? IS IT POSSIBLE THAT NICHES WILL OVERLAP, OR THAT THE SYSTEM WILL BECOME MORE COMPLEX? nativos, primero por que como usted bien conoce que el organismo introducido va ocupar el nicho de una(s) especie(s) nativas y segundo IS A FAULTY QUARANTINE REGIME REASON ENOUGH NOT TO IMPORT NATURAL ENEMIES? IN OUR EXPERIENCE, COUNTRIES WHO CANNOT ALLOCATE RESOURCES TO QUARANTINE SIMPLY REFUSE TO ALLOW IMPORTS. por que no siempre se hace un buen trabajo de cuarentena y por IS IT KNOWN IN THE LITERATURE THAT THE IMPORTATION OF A NATURAL ENEMY HAS CAUSED /DISPLACEMENT OF A LOCAL AGENT AND THAT THIS HAS CAUSED PEST OUTBREAKS? AND WHAT IS THE MECHANISM BY WHICH THE PEST BECOMES GENETICALLY MORE RESISTANT WHEN THIS OCCURS? consiguiente se corre el riesgo de incrementar la(s) plaga(s) y estas pueden ser geneticamente mas resistente(s). Como usted bien conoce en Cuba se desarrolla un programa de Control Biologico en todo el pais que cuenta entre los organismos benéficos, artropodos, nemátodos, virus, hongos y bacterias. Con la conjugacion de diferentes organismos biologicos e incluso algunos quimicos, asi como una estrategia de siembra adecuada nos ha a permitido que diferentes cultivos como col, pastos y yuca solo se utilice Trichogramma y Bacillus thuringiensis para el control de Plutelella xylostella, Mocis spp. y Erynnis ello. El cultivo del plátano y boniato solo utilicemos Beauveria bassiana y Pheidole megacephala para el control de Cosmopolite sordidus y Cylas formicarius. por solo mencionar algunos cultivos, no obstante le menciona que en el cultivo del maiz es uno de los ejemplo que con el control biologico se ha resuelto la problematica de Spodoptera frugiperda con aplicaciones de Telenomus remus, del cual tenemos tres cepas, Trichogramma pintoi y Bacillus thuringiensis. HOW DID S. FRUGIPERDA BECOME A PROBLEM IN CORN AS A RESULT OF BIOLOGICAL CONTROL? Es por esta razon que le propongo que analice con los directores y el gerente tecnico de ADEPA y El Vallecito, el establecer un proyecto de asesoria en tecnicas de control biologico y de manejo integrado de plagas con especialistas de nuestra institucion, ya que los problemas seria necesario verlos in situ por que pueden ir desde una mala determinacion de la especie de Trichogramma hasta la no utilizacion del medio adecuado en el momento oportuno. En espera de que estas linea un poco tardia (causas involuntarias) puedan ayudarle a tomar una decision un poco mas amplia que lo que ha hecho hasta el momento, ya que el trabajo de reduccion de un organismo danino no solo se puede lograr con uno dos insectos determinados sino con un conjunto de actividades que nos permitan convivir con la plaga sin que nos afecte economicamente. En espera que estas lineas le puedan servir para una toma de dicision en cuanto a la reduccion de los problemas fitosanitarios que le acarrea queda de usted, afectuosamente Dra. Esperanza Rijo Camacho Carol Waddington C.Waddington@Koppert.nl Subject: Reply to Elizabeth De Nardo Date: Wed, 23 Jun 1999 12:06:33 +0200 From: jean-louis.hemptinne@educagri.fr (JEAN-LOUIS HEMPTINNE) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu CC: bc-ntimpact@UDel.Edu Hello, Your questions about the potential host range of a parasitoid could, may be, be related to the prey recognition. A rather recent paper by Liepert & Dettner (Journal of Chemical Ecology, 1996, 22, 695-707) suggests that cuticular hydrocarbons are important. Is it irrelevant to think that there is a link between the width of an host range and the width of the molecular spectrum to which an insect respond ? Hoping that this short idea could help you, Best wishes Jean-Louis Hemptinne Ecole nationale de Formation agronomique Dept. ARGE BP 87 F-31326 Castanet-Tolosan Tel : 00 33 5 61 75 32 95 Fax : 00 33 5 61 75 03 09 E-mail : jean-louis.hemptinne@educagri.fr Subject: Re: Reply to Elizabeth De Nardo Date: Wed, 23 Jun 1999 09:06:05 -0600 From: "Alec McClay" Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu This question brings up the relationship between (a) determining the host-range of an insect (predator/parasitoid/herbivore) and (b) determining the sensory/behavioural/biological basis for that host range. In screening herbivorous insects for weed control, we have always focussed on experimentally determining the range of hosts the plants will accept, and this information has been accepted by regulators. In principle, one could go beyond that to determine _why_ those particular hosts, and not others, are accepted. But this has always been seen as orders of magnitude more difficult, and has not usually been required as part of the screening process. What chemical, structural or phenological features of the potential host stimulate or inhibit host-finding, recognition, feeding, oviposition, or development of the herbivore? The Colorado potato beetle people are only beginning to get a handle on this after 20 years! Does this same argument apply to parasitoids or predators? The direct approach of simply offering a series of potential hosts and observing the responses of the parasitoid certainly seems to present greater difficulties than in the case of herbivores, given that the range of potential hosts can be much wider and less taxonomically consistent. Is it possible that looking directly at host or prey recognition processes would actually be the more direct route in this case? >Your questions about the potential host range of a parasitoid could, may >be, be related to the prey recognition. A rather recent paper by Liepert & >Dettner (Journal of Chemical Ecology, 1996, 22, 695-707) suggests that >cuticular hydrocarbons are important. Is it irrelevant to think that there >is a link between the width of an host range and the width of the >molecular spectrum to which an insect respond ? Alec McClay Research Scientist, Biological Control of Weeds alec@arc.ab.ca Alberta Research Council Phone (780) 632-8207 Bag 4000, Vegreville Fax (780) 632-8612 Alberta T9C 1T4, Canada http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML Subject: FW: Internet workshop summary at IOBC-sponsored symposium/worksho p Date: Thu, 24 Jun 1999 08:43:15 +1200 From: "Barratt, Barbara" Reply-To: bc-ntimpact@UDel.Edu To: "'bc-ntimpact@UDel.Edu'" > See below: > > -----Original Message----- > From: Elizabeth De [SMTP:edenardo@yahoo.com] > Sent: Thursday, 17 June 1999 18:07 > To: bc-ntimpact@UDel.Edu > Subject: Internet workshop summary at IOBC-sponsored > symposium/workshop > > > Dear Dear subscribers, > > Befor Keith Hopper summarizes the internet forum I would like to > discuss some issues that remained unresolved( at least for me) . > I admit to being silent listener rather than participants so far. > > Lets go! > > > I work in a Quarantine Laboratory for biological control agents in > Brazil (Embrapa Meio Ambiente -Jaguariuna -Sao Paulo), the only one in > the country responsible for importing and evaluating the risk of each > introduction. So a protocol to assess the risk of non-target impacts > of biological introductions is needed to comply with current national, > regional, and international phytosanitary and biodiversity conservation > laws. > Keith Hopper and I are involved in a project aimed at elaborating > protocols to predict the host range of parasitoids and predators in the > area of introduction and we are working on a retrospective case study > of the parasitoid Macrocentrus cingulum (= grandii) ( Hymenoptera > -Braconidae),introduced into the US to control the European corn borer > (ECB) Ostrinia nubilalis (Lepidoptera -Pyralidae) > The determination of the potential host range of a parasitoid in a > new > environment should consider different approaches taxonomy/phylogeny > biology, and host range in the area of origin are suggested in various > papers. However, pratical details are often lacking. In developing a > protocol, I've encountered the following problems: > > 1 ) How could the pylogeny of the natural enemy help to determine the > host range in the area of origin? > The majority of the parasitoids do not have a consistent pattern of > host range. Some species of the parasitoids in the same genus have a > broad host range including several different families while others have > a more narrow range, e.g., a single family or even genus. > I agree, it is not necessarily helpful and may to some extent depend upon > how well the phylogeny of the parasitoid is determined, especially in the > case of Hymenoptera. There has not been much emphasis in the past on host > range in the area of origin, although it might be a useful indicator of > breadth of host range in the area of introduction. > 2) After we elaborate the list of potential non-targets of the exotic > BCA in the area of introduction another problem that arises is the > difficulties of collecting species in the field and rearing some of > them under lab conditions. How should we select the species to be > expose to the BCA under Quarantine Lab tests prior the BCA release? > Sometimes we select species available in culture or easy to collect and > rear (maybe pests), but are the results we get from the tests useful > in predicting host range of the BCA in the area of introduction? > I think the best we can do is to emulate as far as possible the approach > taken for weed biocontrol host specificity testing ie. work sequentially > (or 'centrifugally' as they call it) from taxa most closely related to > the target host to those more distantly related. In our work with braconid > parasitoids of weevil pests of forage crops, we also found it useful to > determine which non-target weevil species were present in the same > environment of the target host, since if non-target parasitism is going to > occur, then these were the species most likely to be subject to challenge > by the parasitoid in the field. > 3) How many species should we test from the same family at risk and how > many individuals of one species should we test to assess the risk to > that species? > Sometimes we are limited by in number of the parasitoids we have to > expose to potential hosts, other times we are limited in number of > the hosts. Furthermore, we are limited by the costs to perform tests. > How should we deal with these problems to get reasonalble results > concerning risks to non-targets? > We worked out optimum experimental conditions with a set of preliminary > tests with the target host ie: host/para ratios, time of exposure, numbers > of hosts per replicate, numbers of replicates etc. to achieve results with > an acceptable level of variability. > > Who would like to helpme in those points? > > I can send you reprints describing some of this work if you wish. > Barbara > ================================================ Dr Barbara I.P. Barratt Phone: (03) 489 3809 AgResearch FAX: (03) 489 3739 Invermay Agricultural Centre Email: Barrattb@AgResearch.cri.nz Private Bag 50034 Mosgiel New Zealand ================================================ > === > > Elizabeth De Nardo > Embrapa Meio Ambiente- Brasil > Laboratorio de Quarentena ACB > http://www.bdt.org.br/bdt/biocontrol/ > Current Address:OARDC/OSU Dept.of Entomology > 1680, Madison Avenue-WOOSTER -OH 44691 > _________________________________________________________ > Do You Yahoo!? > Get your free @yahoo.com address at http://mail.yahoo.com Subject: Reply to Barbara Bratt Date: Thu, 24 Jun 1999 11:05:10 -0700 (PDT) From: Elizabeth De Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear Barbara, Thank you for sharing your experience with me. I would like to raise some points again. Original message: 1 ) How could the pylogeny of the natural enemy help to determine thehost range in the area of origin? The majority of the parasitoids do not have a consistent pattern of host range. Some species of the parasitoids in the same genus have abroad host range including several different families while others have a more narrow range, e.g., a single family or even genus. Barbara commented: I agree, it is not necessarily helpful and may to some extent depend upon how well the phylogeny of the parasitoid is determined, especially in the case of Hymenoptera. There has not been much emphasis in the past on hostrange in the area of origin, although it might be a useful indicator ofbreadth of host range in the area of introduction. Elizabeth replies: Maybe for some parasitoids we could use its phylogeny to determine the potential host range in the area of origin, but this could not begeneralized, as the majority of them do not have a consistent pattern of hostrange. This is the point: it will not help much in a protocol Original message: 2) After we elaborate the list of potential non-targets of the exotic BCA in the area of introduction another problem that arises is the difficulties of collecting species in the field and rearing some of them under lab conditions. How should we select the species to be expose to the BCA under Quarantine Lab tests prior the BCA release? Sometimes we select species available in culture or easy to collect and rear (maybe pests), but are the results we get from the tests useful in predicting host range of the BCA in the area of introduction? Barbara commented: I think the best we can do is to emulate as far as possible the approach taken for weed biocontrol host specificity testing ie. work sequentially (or 'centrifugally' as they call it) from taxa most closely related to the target host to those more distantly related. In our work with braconi parasitoids of weevil pests of forage crops, we also found it useful to determine which non-target weevil species were present in the same environment of the target host, since if non-target parasitism is going to expose to potential hosts, other times we are limited innumber of the hosts. Elizabeth replies: Yes, Barbara I agree with you that we should follow the approach of host phylogeny, from taxa most close related to the target to those moredistantly related, to determine the potential host in the area of introduction. But even so, it is not an easy approach to follow. Inour case study for example, the most close related species to Ostrinia nubilalis are native Ostrinia. The USA have 3 natives species but all of them difficult to be collect and I did not find no one with experience in rearing them. If we follow the phylogeny of the family Crambidae (Pyralidae ) present in the same habitat of the pest .that in this case are various, because the target pest (Ostrinia nubilalis) is polyphagous and attacks corn, peppers, tomatoes, snapbean we have the same problem( collecting and rearing ). Maybe we can use other pyralid pests like Diatraea sacharalis or D.grandiosella (this last easy to obtain and rear in lab) but both are not native of the USA, although they are for a long time present here. We also found someone rearing the native sunflower moth Homoesoma electellum ( Pyralidae ) and this could be an option, but this moth is not frequent in Ostrinia habitats. If we use the other Lepidoptera families, cited once in the literature, as host of M.cingulum, such as Noctuidae, Lymantriidae and Nymphalidae, the easiest species to collect or to obtain are pests, because of the abundance of them. We found 2 native Noctuidae (Heliocoverpa zea andSpodoptera frugiperda ) available commercially. They are pests here. My question now is: if we test all those species available and the results show that they are not parasitized by M.cingulum, how confident should I be to introduce or release M.cingulum in the environment? If one of the pest species are affected what does it mean? If any of those pests species are fed upon by endangered or threatened birds it is unlikely that the addition of one parasite of an alternate host will have any impact.But what this result means in terms of the safety of the biocontrol agent candidate of introduction? Original message: How should we deal with these problems to get reasonable results concerning risks to non-targets? Barbara commented: 3)We worked out optimum experimental conditions with a set of preliminary tests with the target host ie: host/para ratios, time of exposure, numbersof hosts per replicate, numbers of replicates etc. to achieve results with an acceptable level of variability. Elizabeth replies: Also I agree with you that we should work with the optimum experimental conditions, but it is not possible all the time. In our case M.cingulum is difficult to rear under lab conditions. We intend to use the material we will collect in the field that by different reasons are not collect in a big number. So we have a specific amount of parasitoids in a specific time of the year. How to perfom tests that really mean something significant in terms of host specificity if we do not have the optimum conditions in terms of non host species, number of parasitoids to expose to the non targets. We are talking only about the material and methods of the tests, but other problems specially for the developing countries such we have in South America we are limited by the costs to perform tests. Thus,Should we, considering our financial condition, be less restrictive with the importation of a BCA? But at the same time we have one of the biggest biodiversty of the world and we should protect them. How to balance the risk benefits ? Barbara wrote: I can send you reprints describing some of this work if you wish. Elizabeth replies: I will really aprecciate to receive them. Thank you. Elizabeth === Elizabeth De Nardo Embrapa Meio Ambiente- Brasil Laboratorio de Quarentena ACB http://www.bdt.org.br/bdt/biocontrol/ Current Address:OARDC/OSU Dept.of Entomology 1680, Madison Avenue-WOOSTER -OH 44691 _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com Subject: Re: Reply to Elizabeth De Nardo Date: Fri, 25 Jun 1999 14:06:11 -0700 (PDT) From: Elizabeth De Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Dear Alec, You wrote: Is it possible thatlooking directly at host or prey recognition processes would actually be the more direct route in this case? My response: The points you raise are very interesting and is a research need in this area of assessing the potential host range of a natural enemy in the area of introduction. Unfortunately ,there are not enough information nowadays to use them,in this kind of protocol. I hope in the future they will help us to assess the risk to non-target more adequately. Is there anyone studying this aspect? Thank you Elizabeth --- Alec McClay wrote: > This question brings up the relationship between (a) > determining the > host-range of an insect > (predator/parasitoid/herbivore) and (b) determining > the sensory/behavioural/biological basis for that > host range. In screening > herbivorous insects for weed control, we have always > focussed on > experimentally determining the range of hosts the > plants will accept, and > this information has been accepted by regulators. In > principle, one could > go beyond that to determine _why_ those particular > hosts, and not others, > are accepted. But this has always been seen as > orders of magnitude more > difficult, and has not usually been required as part > of the screening > process. What chemical, structural or phenological > features of the > potential host stimulate or inhibit host-finding, > recognition, feeding, > oviposition, or development of the herbivore? The > Colorado potato beetle > people are only beginning to get a handle on this > after 20 years! > > Does this same argument apply to parasitoids or > predators? The direct > approach of simply offering a series of potential > hosts and observing the > responses of the parasitoid certainly seems to > present greater difficulties > than in the case of herbivores, given that the range > of potential hosts can > be much wider and less taxonomically consistent. Is > it possible that > looking directly at host or prey recognition > processes would actually be > the more direct route in this case? > > >Your questions about the potential host range of a > parasitoid could, may > >be, be related to the prey recognition. A rather > recent paper by Liepert & > >Dettner (Journal of Chemical Ecology, 1996, 22, > 695-707) suggests that > >cuticular hydrocarbons are important. Is it > irrelevant to think that there > >is a link between the width of an host range and > the width of the > >molecular spectrum to which an insect respond ? > > Alec McClay > Research Scientist, Biological Control of Weeds > > alec@arc.ab.ca Alberta Research Council > Phone (780) 632-8207 Bag 4000, Vegreville > > Fax (780) 632-8612 Alberta T9C 1T4, Canada > > > http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML > > > _________________________________________________________ Do You Yahoo!? Get your free @yahoo.com address at http://mail.yahoo.com Subject: [Fwd: Resource concentration hypothesis and non-target effects] Date: Thu, 26 Aug 1999 14:42:27 -0400 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact --------------------------------------------------------------------- Subject: Resource concentration hypothesis and non-target effects Date: Mon, 23 Aug 1999 10:56:19 -0600 From: Alec McClay To: , Bernd Blossey , Palmer Bill , "Robert W. Pemberton" , vcmoran@psipsy.uct.ac.za, ecoombs@wiley.oda.state.or.us, "grbuck@nervm.nerdc.ufl.edu" , joebalci@pw.usda.gov, hoff@botzoo.uct.ac.za, Judy Myers , Keith Hopper , mcevoyp@ava.bcc.orst.edu, Rachel.MCFADYEN@dnr.qld.gov.au, floate@em.agr.ca, Pauline Syrett , harris@em.agr.ca, hillr@crop.cri.nz I tried to send this message last week to the Bozeman Symposium mailing list, but it doesn't seem to be functioning any more. Instead I am sending it individually to a few lucky colleagues! Any cases, references, or ideas you can provide would be very gratefully received. It has sometimes been argued (e.g. Harris 1990) that if non-target native plant species are much less abundant than the target weed species, they are at little risk from introduced arthropod biocontrol agents because the agents will tend to concentrate their attack on the abundant target weed. This is a version of the resource concentration hypothesis (Root 1973 etc.). Is anyone aware of - specific cases in which this argument has been made as part of a proposal to release a particular biocontrol agent, or - any studies to test the hypothesis with reference to a weed biocontrol agent? Harris, P. 1990. Environmental impact of introduced biological control agents. pp. 289-300 in Mackauer, M., L.E. Ehler and J. Roland (Eds.), Critical issues in biological control. Andover, UK: Intercept. Alec McClay Research Scientist, Biological Control of Weeds alec@arc.ab.ca Alberta Research Council Phone (780) 632-8207 Bag 4000, Vegreville Fax (780) 632-8612 Alberta T9C 1T4, Canada http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML Subject: Re: Reply to Elizabeth De Nardo Date: Tue, 31 Aug 1999 16:00:07 -0400 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Responses to posting by Alec McClay 23 June 1999: > process. What chemical, structural or phenological features of the > potential host stimulate or inhibit host-finding, recognition, feeding, > oviposition, or development of the herbivore? The Colorado potato beetle > people are only beginning to get a handle on this after 20 years! > > Does this same argument apply to parasitoids or predators? The direct > approach of simply offering a series of potential hosts and observing the > responses of the parasitoid certainly seems to present greater difficulties > than in the case of herbivores, given that the range of potential hosts can > be much wider and less taxonomically consistent. Although I would argue that the potential host range of parasitoids or predators is not necessarily wider in general than that of herbivores, I quite agree that host range testing of entomophages presents different, and perhaps more difficult problems, than host range test of herbivores. One major difference is finding and rearing non-target insects and particularly less common ones is extremely difficult. Elizabeth De Nardo is attempting to do a retrospective study of Macrocentrus grandii host range and is finding it extremely difficult to collect native species in the genus Ostrinia (the target species of the introduction being the European corn borer, Ostrinia nubilalis). What has others experience been with locating populations of native insects closely related to targets of biological control? > Is it possible that > looking directly at host or prey recognition processes would actually be > the more direct route in this case? Ultimately, we will need that better understanding of the mechanisms that determine host range, especially if we want to predict evolutionary changes, but it seems that this goals is along way off. Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: Nontarget effect Date: Tue, 31 Aug 1999 16:30:31 -0400 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Response to posting by Julie Nara, 22 Jan 1999: > There has been several mentions of the impact of Rhinocyllus conicus on Carduus > spp. Can somone point me to an article reporting on a scientific field study > recording the reduction in the population density of Carduus nutans by R. > conicus? Thanks! I second this request: does anyone have results on population level impacts of R. conicus on non-target species? Yours, Keith Hopper -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Re: design Date: Tue, 31 Aug 1999 17:27:28 -0400 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact@UDel.Edu Responses to posting by Bernd Blossey 25 Jan 1999: First, I think that Bernd and his colleagues working on purple loosestrife are on the right track in their approach to getting out and measuring what is going on in the field. But to stimulate the debate rather than just say cheer them on, I'll bring up below some points where I differ > Predictions of potential non-target impacts for weed feeders are easily > developed following the host specificity testing. Usually, potential > control agents show a multitude of responses, largely depending on test > designs (age of plants and insects, multiple- or no-choice,leaves or potted > plants, greenhouse or field grown plants, cage size, field or tests in > captivivity, adult or larval tests, genotype of insect or test plants, > etc. come to mind). Without follow-up monitoring, however, we will never be > able to evaluate which of our testing procedures have predictive ability. > Assuming that we are not overlooking any serious population level effects > of insects released as bc agents, our screening protocols are pretty good > in preventing a switch to non-targets. Most of what has been reported was > predicted, what is important is whether something escaped our attention. I > am actually more concerned about the possibility that we are rejecting some > perfectly safe control agents because of meaningless screening test > requirements. Unless a candidate for introduction avoids all non-target species, I don't see how "predictions of potential non-target impact for weed feeders are easily developed following host specificity testing". Would you clarify what you mean here by predictions of potential non-target impacts? Do you mean feeding on individual plants or population level impacts? > 2. Post-release studies > > To evaluate the impact on the target weed, we need to develop standardized > protocols that allow long-term monitoring. With funding from NBCI we have > done just that for the purple loosestrife program. We have held workshops > to instructed land managers in the use of the protocol and many people > throughout North America are now monitoring control agent build-up, purple > loosestrife decline, and the response of the plant communities at their > release sites. Ideally, we would also incorporate assessments of other > organisms (birds, amphibians, native insects, microorgnisms?) but that may > be overly ambitious. Nevertheless it is possible. This protocol is > available on my web site and allows comparisons and evaluations for single > versus multiple agent releases. Most importantly, it links the control > agents to changes in plant populations e.g. the decline of purple > loosestrife. It is easy to design regional approaches with spacing of > releases that prevent spill-over of agents to sites that are kept as > controls (no releases of control agents). It may be possible to monitor effects on plant communities in the area of release, but what about in other habitats to which the released insects disperse. This may not be a problem with enemies being released against purple loosestrife, but what about other, less host-specific agents. Even with agents released against purple loosestrife, I doubt that one could measure meaningfully population level changes in all or even a significant subset of native insect species throughout the range of purple loosestrife. It is difficult to define even what would mean by a population level impact. Does this mean a 10 percent change in peak numbers in all years in all sites or a 50 percent change in numbers in half the years in half the sites. It seems that you are oversimplifying the problems involved here. This is not to say that I find fault with your plan to get out and measure at least impacts on plant species in the sites where releases are done. I think this is great! However, it is along way from addressing all the issues involved. Furthermore, I question whether standardized protocols for long-term monitoring are feasible. This would certainly pose large problems with monitoring impacts on insect populations because of the great variety of techniques need for estimating abundances of various types of insects. > I don't think we need to worry about the inability to collect regional or > local baseline data even if control programs have been implemented years > ago. Landscape level occurrence of bc agents or control does not occur over > night. For the purple loosestrife program the number of wetlands where > insects were released is well over 1000 in 30+ states. However, there are > still thousands of sites where pre-release data can be obtained. I guess > that even with 30 years of biocontrol targeting a species such as leafy > spurge I can find plenty of sites (particularly in the Midwest and > Northeast) that are free of control agents. This is not to argue that such > pre-release investigations should begin before the very first introduction > into North America occurs. In addition, the longer the data were collected > before control agnets were released, the better. This is a good point and suggests that there may be a variety of systems were retrospective studies could be started now and allow measurement of impact of spread of introduced agents. > A last remark re. monitoring for non-target effects in insect biocontrol > programs. It does not appear difficult to design such protocols, however, > it may be quite elaborate. The lack of information guiding the searches due > to lack of host specificity testing may make this particularly difficult. > But I could imagine a routine collection effort of potential non-target > hosts, followed by rearing of the appropriate life stage in the lab to > assess parasitoid attacks. Alternatively we could use the deliberate > exposure of non-target organisms at the right life-stage on a "trap-plant" > in the right habitat or the habitat that we want to survey, allow > parasitoid attack for a standardized time, remove trap-plants and check for > emergence of parasitoids (or collect potential predators in specifically > designed traps). This is just a start but I really think it is not at all > that difficult. I think you are underestimating the difficulty of measuring non-target impact on insect populations. Showing that an individual insect is parasitized or preyed upon is relatively easy, although even this can be difficult for some species. Showing that a population is reduced in abundance by a particular factor is much much more difficult. Insects populations often fluctuate by orders of magnitude between seasons and locations. Determining what causes these fluctuations is extremely problematic unless the effects of the factor of interest are very large. Insects introduced for biological control of insects sometimes have spectacular impact on the target species and such impacts are well-documented, but ancillary effects are much more problematic because they appear to be more subtle. I'm not saying it can't be done; indeed I strongly favor doing just this, but it is going to be very expensive. What would you say would be an adequate capacity to measure change in non-target species, i.e., amount and spatio-temporal extent that we need to be able to detect? Yours, Keith -- ****************************************************************** Keith R. Hopper, Research Entomologist USDA, ARS, Beneficial Insect Introduction Research Unit University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 ****************************************************************** Subject: Date: Tue, 31 Aug 1999 13:41:58 -1000 From: messing@hawaii.edu (russell messing) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu >Responses to posting by Keith Hopper: "...finding and rearing non-target insects and particularly less common ones is extremely difficult....... What has others experience been with locating populations of native insects closely related to targets of biological control?" We encountered a distinct but related problem in studying non-target effects of introduced braconid parasites of tephritid fruit flies. Tephritids in the endemic Hawaiian genus Phaeogramma have been listed by the US Fish and Wildlife Service in the Federal Register (as a first step towards placement on the Endangered Species List). One species of Phaeogramma on Kauai is known to occur in just a few relatively small patches in high elevation, undisturbed ecosystems. While we can locate some of these patches, and find the flies, there is the very real prospect that any research we do on the endemic species will have more impact on their populations than the parasites ever will. So far we have addressed this problem by collecting the flies (in their galls) from the field, conducting host acceptance and host suitability tests with the parasites in the lab, and then returning the eclosed (generally unsuitable) flies back to their mountain home. This may work for now, but if the flies ever gets true "Endangered Species" status, we may be prohibited from working with them at all. Russell Messing University of Hawaii Dept. of Entomology 7370 Kuamoo Rd. Kapaa, Hawaii, 96746 tel (808) 822-4984 x 223 fax (808) 822-2190 email: messing@hawaii.edu Subject: Re: design Date: Wed, 1 Sep 1999 17:41:47 +0300 From: "Rouhollah SOBHIAN" Reply-To: bc-ntimpact@UDel.Edu To: Response to Keith and Bend I think most people believe that non target effect is always negative. I talked about it also earlier. Is this not a non-target effect that you reduce the population of a weed and start restoring the natural habitat? This aspect of non-target effect could be and should be studied after releasing weed feeders. Subject: design Date: Thu, 2 Sep 1999 11:13:40 -0700 From: "Rosalind James" Reply-To: bc-ntimpact@UDel.Edu To: -----Original Message----- From: owner-bc-ntimpact@UDel.Edu [mailto:owner-bc-ntimpact@UDel.Edu] On Behalf Of Rouhollah SOBHIAN Sent: Wednesday, September 01, 1999 7:42 AM To: bc-ntimpact@UDel.Edu Subject: Re: design Response to Keith and Bend I think most people believe that non target effect is always negative. I talked about it also earlier. Is this not a non-target effect that you reduce the population of a weed and start restoring the natural habitat? This aspect of non-target effect could be and should be studied after releasing weed feeders. Reply: This is actually the targeted effect I think! But you might have a non-target effect if another weed species comes in and replaces the original one, instead of the natural habitat being restored. Rosalind R. James, Ph.D. Research Entomologist USDA Agricultural Research Service Kika de la Garza Subtropical Agricultural Research Center Weslaco, TX USA (956) 969-4856 phone (956) 969-4888 fax http://rsru2.tamu.edu/ http://rsru2.tamu.edu/bios/JamesR.html Subject: Re: Nontarget effect Date: Fri, 3 Sep 1999 12:48:10 -0600 From: "Lars Baker" Reply-To: bc-ntimpact@UDel.Edu To: Dr. Svata M. Louda reported on population declines in Platte thistle in Nebraska at the recent X International Symposium on Biological Control of weeds in Bozeman, MT. She can be reached at slouda@unl.edu. or at 402-472-2763 ph, 402-472-2083 fax. J. L. Baker Fremont County Weed and Pest Lander, WY -----Original Message----- From: Keith R. Hopper To: bc-ntimpact@UDel.Edu Date: Tuesday, August 31, 1999 6:18 PM Subject: Re: Nontarget effect >Response to posting by Julie Nara, 22 Jan 1999: > >> There has been several mentions of the impact of Rhinocyllus conicus on Carduus >> spp. Can somone point me to an article reporting on a scientific field study >> recording the reduction in the population density of Carduus nutans by R. >> conicus? Thanks! > >I second this request: does anyone have results on population level >impacts of R. conicus on non-target species? > >Yours, Keith Hopper >-- >****************************************************************** >Keith R. Hopper, Research Entomologist >USDA, ARS, Beneficial Insect Introduction Research Unit >University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 >Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 >****************************************************************** > Subject: Re: Nontarget effect Date: Mon, 06 Sep 1999 10:54:11 -0400 From: Tom McAvoy Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu CC: ltkok@vt.edu In response to Julie Nara, posted 22 Jan 1999: Below are references concerning the reduction in population density of Carduus nutans and Carduus acanthoides by R. conicus. C. nutans has been successfully controlled in western Virginia due to R. conicus. This species was released in Va. in 1969 & '70. C. acanthoides (alien species) has now become the dominant thistle. This species was also a problem before C. nutans was controlled. C. acanthoides in most cases has not replaced C. nutans. C. acanthoides is a suitable host for R. conicus but C. acanthoides blooms 2 - 3 weeks later than C. nutans avoiding the oviposition period of R. conicus, which occurs from mid-May to mid-July (Surles & Kok 1977 Eviron. Entomol. 6:222-224). There is also greater larval mortality of R. conicus larvae due to the smaller head size of C. acanthoides compared to C. nutans. R. conicus's oviposition period appears to be less flexible than its host range. Zwolfer and Harris (1984, Z. ang. Ent. 97:36-62) in tests done in Europe & Canada, list R. conicus oviposition, & larval development occurring on Carduus nutans, C. crispus, C. tenuiflorus, Cirsium arvense, C. vulgare, C. palustre, Silybum marianum and Onopordum acanthium all in the Carduini tribe. C. acanthoides can be added to this list, Zwolfer & Harris tested this species but did not report development occurring. Zwolfer & Harris '84 did not list blooming periods of the thistles that were found to be hosts of R. conicus. To add to the mix R. conicus appears to have distinct races. In Arles (s. France) R. conicus are found exclusively on Silybum marianum and not on C. nutans. In Southwestern France R. conicus prefers C. vulgare to C. nutans. In Va. R. conicus is not found on C. vulgare due to the late blooming period of this plant (August). Does C. nutans & C. vulagre bloom at the same time in France? So, R. conicus appears to have a fairly broad host range within Carduini. Blooming period seems to be the critical factor in what species get attacked. In Europe where R. conicus presumably originated there appears to a healthy population of Cirsium and Carduus spp., including C. nutans. Though, I'd like to get a better idea of what the relative abundance of these spp. are in Europe. Bio control agents as for as I know have never eradicated an alien species or a native nontarget. Though we're just beginning to look at the issues of nontarget effects, as we should be, we should avoid rushing to judgements and not forget the negative impacts that alien species have on native species and the positive BENEFITS of reducing the competitive advantage that exotic species have in the absence of their natural controls. Regards, Tom McAvoy Kok, L. T. 1989. Biological Control of Weeds in Virginia from 1969-86. Proc. VII. Int. Symp. Biol. Contr. Weeds, 6-11 March 1988, Rome, Italy. Delfosse, E. S. (ed.). Ist. Sper. Patol. Veg. (MAF), pp 623-629. Kok, L. T. 1998. Biological control of musk and plumeless thistles. Recent Res. Devel. In Entomol. 2: 33-45. Kok, L. T. and W. T. Mays. 1991. Successful biological control of plumeless thistle, Carduus acanthoides L. [Campanulatae: Asteraceae: (=Compositae)], by Trichosirocalus horridus (Panzer) (Coleoptera: Curculionidae) in Virginia. Biol. Contr. 1:197-202. Kok, L. T. and R. L. Pienkowski. 1985. Biological control of musk thistle by Rhinocyllus conicus (Coleoptera: Curculionidae) in Virginia from 1969-1980. Proc. 6th Inter. Symp. Biol. Contr. Weeds. pp. 433-438. Kok, L. T. and W. W. Surles. 1975. Successful biological control of musk thistle by an introduced weevil, Rhinocyllus conicus. Environ. Entomol. 4:1025-1027. Surles, W. W. and L. T. Kok. 1977. Ovipositional preference and synchronization of Rhinocyllus conicus with Carduus nutans and C. acanthoides. Environ. Entomol. 6:222-224. Surles, W. W. and L. T. Kok. 1978. Carduus thistle seed destruction by Rhinocyllus conicus. Weed Sci. 26:264-269. Surles, W. W. and L. T. Kok. 1976. Pilot studies on augmentation of Rhiocyllus conicus (Coleoptera: Curculionidae) for Carduus thistle control. Environ. Entomol. 5:901-904. Surles, W. W., L. T. Kok and R. L. Pienkowski. 1974. Rhinocyllus conicus establishment for biocontrol of thistles in Virginia. Weed Sci. 22:1-3. I've (T. McAvoy) added these referneces on nontarget effects by R. conicus: Louda, S. M., D. Kendall, J. Connor and D. Simberloff. 1997. Ecological effects of an insect introduced for the biological control of weeds. Science 277:1088-1090. Louda, S. M. 1998. Population growth of Rhinocyllus conicus (Coleoptera: Curculionidae) on two species of native thistles in prairie. Environ. Entomol. 27: 834-841. At 12:48 PM 9/3/99 -0600, you wrote: >Dr. Svata M. Louda reported on population declines in Platte thistle in >Nebraska at the recent X International Symposium on Biological Control of >weeds in Bozeman, MT. She can be reached at slouda@unl.edu. or at >402-472-2763 ph, 402-472-2083 fax. >J. L. Baker >Fremont County Weed and Pest >Lander, WY > >-----Original Message----- >From: Keith R. Hopper >To: bc-ntimpact@UDel.Edu >Date: Tuesday, August 31, 1999 6:18 PM >Subject: Re: Nontarget effect > > >>Response to posting by Julie Nara, 22 Jan 1999: >> >>> There has been several mentions of the impact of Rhinocyllus conicus on >Carduus >>> spp. Can somone point me to an article reporting on a scientific field >study >>> recording the reduction in the population density of Carduus nutans by R. >>> conicus? Thanks! >> >>I second this request: does anyone have results on population level >>impacts of R. conicus on non-target species? >> >>Yours, Keith Hopper >>-- >>****************************************************************** >>Keith R. Hopper, Research Entomologist >>USDA, ARS, Beneficial Insect Introduction Research Unit >>University of Delaware, 501 S. Chapel St., Newark, Delaware 19713 >>Voice +1-302-731-7330 ext 38; Fax +1-302-737-6780 >>****************************************************************** >> > > Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: [Fwd: Resource concentration hypothesis and non-target effects] Date: Tue, 07 Sep 1999 16:17:59 -0400 From: "Keith R. Hopper" Reply-To: bc-ntimpact@UDel.Edu Organization: University of Delaware To: bc-ntimpact --------------------------------------------------------------------- Subject: Re: Resource concentration hypothesis and non-target effects Date: Fri, 27 Aug 1999 10:43:02 -0600 From: "Lars Baker" To: , "Bernd Blossey" , "Palmer Bill" , "Robert W. Pemberton" , , , , , , "Judy Myers" , "Keith Hopper" , , , , "Pauline Syrett" , , , "Alec McClay" Alex and all, I have observed Aphthona nigriscutis feeding on Euphorbia robusta near release locations on Leafy Spurge in Fremont County. Last year I photographed two plants with a large number of A.n. and considerable feeding damage. Those two plants were gone this year. Dave Kazmer and I collaborated on a field test to measure the impact of A.n. on robusta. I have located, photographed and tagged 36 plants at various distances from infestations of leafy spurge. Those growing with leafy spurge show considerably more feeding damage than those 50 or 100 feet away. Most show some feeding. This fall we will dig some plants and check them for larvae. Same in the spring. Then revisit the remaining marked plants to measure survival. At first observation there does appear to be a correlation between feeding level and proximity to leafy spurge, the primary and more abundant host for A.n. Plant densities for robusta are less than 50 per acre and in places there are 100 stems of leafy spurge per meter sq. I have observed one or two A.n. on a robusta plant at a time and there are 20-50 on a nearby spurge. Any suggestions on how we could design the test to more scientifically establish the hypothesis? Lars Baker Fremont County Weed and Pest Lander, WY 82520 -----Original Message----- From: Alec McClay To: andys@ento.csiro.au ; Bernd Blossey ; Palmer Bill ; Robert W. Pemberton ; vcmoran@psipsy.uct.ac.za ; ecoombs@wiley.oda.state.or.us ; grbuck@nervm.nerdc.ufl.edu ; joebalci@pw.usda.gov ; hoff@botzoo.uct.ac.za ; Judy Myers ; Keith Hopper ; mcevoyp@ava.bcc.orst.edu ; Rachel.MCFADYEN@dnr.qld.gov.au ; floate@em.agr.ca ; Pauline Syrett ; harris@em.agr.ca ; hillr@crop.cri.nz Date: Monday, August 23, 1999 11:00 AM Subject: Resource concentration hypothesis and non-target effects >I tried to send this message last week to the Bozeman Symposium mailing >list, but it doesn't seem to be functioning any more. Instead I am sending >it individually to a few lucky colleagues! Any cases, references, or ideas >you can provide would be very gratefully received. > >It has sometimes been argued (e.g. Harris 1990) that if non-target native >plant species are much less abundant than the target weed species, they are >at little risk from introduced arthropod biocontrol agents because the >agents will tend to concentrate their attack on the abundant target weed. >This is a version of the resource concentration hypothesis (Root 1973 >etc.). Is anyone aware of >- specific cases in which this argument has been made as part of a proposal >to release a particular biocontrol agent, or >- any studies to test the hypothesis with reference to a weed biocontrol agent? > >Harris, P. 1990. Environmental impact of introduced biological control >agents. pp. 289-300 in Mackauer, M., L.E. Ehler and J. Roland (Eds.), >Critical issues in biological control. Andover, UK: Intercept. > > > > >Alec McClay >Research Scientist, Biological Control of Weeds > >alec@arc.ab.ca Alberta Research Council >Phone (780) 632-8207 Bag 4000, Vegreville >Fax (780) 632-8612 Alberta T9C 1T4, Canada > >http://www.arc.ab.ca/crop/weed/BiocontrolMain.HTML > > Subject: Re: Date: Mon, 11 Oct 1999 09:33:12 +0000 From: "Russell F. Mizell, III" To: owner-bc-ntimpact@UDel.Edu Keith: I haven't kept up with the running commentary on the website as I should have. However, I found your summary quite interesting and complete. I don't have anything to add that I thought was left out. Go with it. The debate needs to continue. Regards, Russ. -- Dr. Russell F. Mizell, III Prof. Of Entomology & IPM Director University of Florida NFREC-Monticello Rt 4 Box 4092 Monticello, FL 32344 Ph. (850) 342-0990 Fx. (850) 342-0230 Subject: Date: Tue, 12 Oct 1999 13:25:30 -0400 (EDT) From: owner-bc-ntimpact >From owner-bc-ntimpact Tue Oct 12 12:55:23 1999 [132.236.56.7]) by copland.udel.edu (8.9.3/8.9.3) with ESMTP id MAA03623 for ; Tue, 12 Oct 1999 12:55:22 -0400 (EDT) Message-Id: <4.1.19991012125727.00a29da0@postoffice.mail.cornell.edu> X-Sender: aeh4@postoffice.mail.cornell.edu (Unverified) X-Mailer: QUALCOMM Windows Eudora Pro Version 4.1 Date: Tue, 12 Oct 1999 12:58:15 -0400 To: bc-ntimpact@UDel.Edu From: Ann Hajek Subject: Re: Trans-genic organism In-Reply-To: <1.5.4.32.19990114231553.00756bbc@gonzo.wolfenet.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Sender: owner-bc-ntimpact@udel.edu Precedence: bulk Reply-To: bc-ntimpact@udel.edu Dear Dr. Unruh, I was just cleaning up old e-mails and read through this message again. I wondered if you could send me the citation for Shaw & Hawkins. Thank you! Ann Hajek Ann E. Hajek Department of Entomology Cornell University Ithaca, NY 14853-0901 607-254-4902 or 4992 Fax: 607-255-0939 Date: Fri, 03 Dec 1999 09:07:37 -0500 From: Tom McAvoy Subject: Secondary weed outbreak X-Sender: tmcavoy@mail.vt.edu To: bc-ntimpact@UDel.Edu Message-id: <3.0.5.32.19991203090737.00800750@mail.vt.edu> MIME-version: 1.0 X-Mailer: QUALCOMM Windows Eudora Light Version 3.0.5 (32) Content-type: text/plain; charset=us-ascii Sender: owner-bc-ntimpact@udel.edu Precedence: bulk Reply-To: bc-ntimpact@udel.edu Colleagues, Is anyone aware of cases (with references if possible) where after a target weed is controlled either biologically or by other methods the plant species that replaces the primary target weed is another exotic or undesirable species? Regards, Tom Tom McAvoy Dept. of Entomology (0319) Price Hall Virginia Tech Blacksburg, VA 24061 Phone: (540)231-6320 Fax: (540)231-9131 Subject: Re: Date: Sat, 04 Dec 1999 08:41:03 -0500 From: Anne Kilmer Reply-To: bc-ntimpact@UDel.Edu Organization: Frisky Enterprises To: owner-bc-ntimpact@UDel.Edu owner-bc-ntimpact@UDel.Edu wrote: > > > Colleagues, > Is anyone aware of cases (with references if possible) where after a > target weed is controlled either biologically or by other methods the plant > species that replaces the primary target weed is another exotic or > undesirable species? > Regards, > Tom > Tom McAvoy > Dept. of Entomology (0319) > Price Hall > Virginia Tech > Blacksburg, VA 24061 > Phone: (540)231-6320 > Fax: (540)231-9131 You betcha. As it says in the Bible, you cast out the devil and you sweep and garnish your house, and in come his sisters and his cousins and his aunts. (I paraphrase.) Having killed the exotic plant, you have to plant the appropriate natives yourself, and look after them, and keep policing the area until they are established, or forever, whichever comes last. I speak from South Florida, home of melaleuca, ear leaf acacia, air potato (Diascorea) , climbing fern Lygodium spp.) , Brazilian pepper (Schinus terebinthefolia), Australian pine (Casuarina spp) and so forth. (and the water plants ... don't ask. A while ago, Water Management decided to remove the Schinus that clothed the banks of the canals in Palm Beach County, and "let Nature replant it with desirable plants." This is, alas, not Her style. She plants whatever comes to hand ... in this case, more Brazilian pepper, castor bean, ragweed, Australian pine and so forth. Climbing fern may well solve the melaleuca problem in the Everglades, while also replacing sawgrass, wiping out the hammocks of native trees and so forth. Whatever we do to fight this lovely graceful vine simply spreads the spores. Quite a few of the exotics are spread by landscapers, tree trimmers and other folks who are supposed to be on our side ... the melaleucas along US 441 and Military Trail in Palm Beach County were planted when truckloads of cut trees were hauled to the dump, showering seeds all the way. Removing water hyacinth causes elodea to flourish. Perhaps if we added tapirs and capybaras to the Florida menagerie, that would help? They're tasty and efficient biocontrol mechanisms and would give the jet skiers something to think about. ;-) Perhaps I am too cynical ... my lakeside planted itself with seagrape, pond apple and royal palms, without suggestions from me. But in general, Nature starts out with weeds, and the exptics are handiest. Anne Kilmer South Florida Subject: RE: Date: Sun, 5 Dec 1999 08:20:38 +1000 From: Palmer Bill Reply-To: bc-ntimpact@UDel.Edu To: owner-bc-ntimpact@UDel.Edu I think Lantana and Chromolaena in South Africa may be an example. You would have to ask South Africans for details -------------------------------------------------------------------------------------------------------------------------- Dr. Bill Palmer Principal Entomologist Alan Fletcher Research Station PO Box 36 Sherwood, Qld 4075 Australia Ph. (07) 3375 0748 Fax (07) 3379 6815 Subject: Re: exotics/BC -weeds Date: Mon, 06 Dec 1999 08:40:36 +0000 From: "Russell F. Mizell, III" Reply-To: bc-ntimpact@UDel.Edu To: owner-bc-ntimpact@UDel.Edu Tom: The replacement of one weed (plant species) following control of another one is a fairly common occurrence in weed pest management. Looking at the weed literature would probably uncover some examples of your more specific case of a general phenomenon. Regards, Russ Dr. Russell F. Mizell, III Prof. Of Entomology & IPM Director University of Florida NFREC-Monticello Rt 4 Box 4092 Monticello, FL 32344 Ph. (850) 342-0990 Fx. (850) 342-0230 Subject: RE: Date: Thu, 9 Dec 1999 09:17:35 +1000 From: Day Michael Reply-To: bc-ntimpact@UDel.Edu To: owner-bc-ntimpact@UDel.Edu Tom, There have been cases in eastern Australia where Salvinia has infested waterways as water lettuce or water hyacinth have been controlled. Fortunately all three weeds have effective biocontrol agents working against them. Mike Michael Day Alan Fletcher Research Station PO Box 36, Sherwood Qld 4075 AUSTRALIA Ph: (07) 3375 0725 International: 61 7 3375 0725 Fax: (07) 3379 6815 International: 61 7 3379 6815 Email: Michael.Day@dnr.qld.gov.au Subject: Re: Date: Thu, 09 Dec 1999 09:53:17 -0500 From: Jim Cuda Reply-To: bc-ntimpact@UDel.Edu To: owner-bc-ntimpact@UDel.Edu Tom, this seems to be a faily common occurrence in aquatic systems. Historically, water hyacinth dominated many of he waterways in Florida. Once water hyacinath was brought under control, alligator weed invaded many of these same sites. A similar phenomenon occurs sometimes annually with water hyacinth and water lettuce. Once you control water hyacinth at a particular site one year, it is not uncommon for that same site to become invaded by water lettuce the following year. I don't have time right now to look up any references. You may want to access the database at the Center for Aquatic and Invasive Plants via their website. Regards, James P. Cuda, Ph.D. Assistant Professor Biological Weed Control Entomology and Nematology Dept. Institute of Food and Agricultural Sciences Bldg. 970, Natural Area Drive PO Box 110620 University of Florida Gainesville, FL 32611-0620 Tel.(352) 392-1901 SUN 622-1901 FAX (352) 392-0190 e-mail jcuda@gnv.ifas.ufl.edu Subject: Secondary alien plant successions Date: Mon, 13 Dec 1999 10:10:34 GMT+2 From: "Jeremy Goodall" Reply-To: bc-ntimpact@UDel.Edu Organization: ARC PLANT PROTECTION (NATAL) To: bc-ntimpact@UDel.Edu Dear Dr McAvoy, Your email ended up with me in KwaZulu-Natal, South Africa. I don't know what conclusions you wiere wanting to draw from the answer you were hoping for so I hope I can be of help to you. I have been studing the impact of Chromolaena odorata density on the spatial and temporal dynamics of subtropical coastal grasslands. The study also looked at succession in response to fire. To cut a long story short, grasslands were invaded by C. odorata that were not burnt. Fire was effective in killing the weed. Initial density had a residual effect on the rate of grassland recovery. Light infestations were restored within three annual burns whilst moderate infestations contain many ruderal indigenous species, but no exotics, after 10 years. Dense infestations failed to recover, the weed was able to re-establish at a rate faster than the grasses could. I then tested the scenario of manipulative intervention whereby I physical removed Chromolaena seedlings from some of the reinfested dense plots while in others I left. The handpulled plots were invaded by Solanum mauritianum, an alien tree/large shrub we call "bugweed". The plots that were left were invaded by chromolaena, Lantana camara and bugweed. After about 5 years the bugweed was out-competed by the Chromolaena and died out in the untreated plots. The bugweed became rampant in the handpulled plot and failed to succumb to fires. In the end the handpulled plot of the former dense chromolaena infestations were bugweed-pioneer grasslands, not very pretty to behold. The untreated plots were thickets of chromolaena and lantana, 3 m tall and impenetrable - far worse a situation and the initial infestation. Does this help? Regards Jeremy Goodall ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Mr Jeremy Goodall Plant Protection Research Institute Agricultural Research Council Private Bag X6006 HILTON 3245 South Africa Tel: +27-331-3559416 Fax: +27-331-3559423 E-mail: ntjg@natal1.agric.za ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Subject: New to the List Date: Mon, 7 Feb 2000 16:07:05 -0700 From: Pete_Oboyski@usgs.gov (Pete Oboyski) Reply-To: bc-ntimpact@UDel.Edu To: bc-ntimpact@UDel.Edu Hi folks, I just recently learned about your list server. I have glanced through the archives, but it will take me a while to catch up. Is this list still active? I work in Hawaii studying the insect food resources of an endangered bird. Besides trying to document the impact of ants and yellowjackets on the native fauna, I am studying parasitism of native caterpillars by introduced wasps. The parasitoids (mostly Ichneumonids) were introduced nearly 100 years ago against cutworms, armyworms and the like. They have reached high into the subalpine native forest and now parasitize a host of native leps, including one genus that feeds inside legume pods. Parasitism rates reach as high as 90% in some areas. This is the fourth year of this work. Data from the first year "pilot" study should be coming out later this year. I have since fine tuned the design and hope to analyze three years of data after this year. Aloha, Pete Pete_Oboyski@USGS.gov USGS - Biological Resources Division PO Box 44, Building 344 Hawaii National Park, HI 96718 (808) 967-7396 ext.227