How risky is biological control? Reply.
For risk, he argues that, since native congeners of introduced biocontrol agents have failed to attack Scapteriscus mole crickets in 80 years, there is little reason to believe that the introduced species themselves would attack native hosts. We are unaware of any basis for this reasoning. Granted that the probability of a rapid host-shift by any host-specific species is probably low, we know of no literature that suggests that, if species A has not shifted hosts in time interval X, congeneric species B is unlikely to shift hosts in time interval Y. Certainly there is need for much research on this point. It is known that rapid host shifts do occur (see, e.g., Dennill et al. 1993, Secord and Kareiva 1996), but the circumstances that promote such events in nature have barely been studied. The trajectory of one introduced species often gives little information about the likely trajectory of a closely related species (Williamson 1996).
As further evidence that a host shift is exceedingly unlikely for Larra bicolor, Frank points out that attempts to introduce it into Hawaii in the 1920s against a nonindigenous Gryllotalpa species failed. His interpretation is that this failure shows that L. bicolor is unlikely to shift to hosts in the tribe Gryllotalpini. The relevance of these failed introductions of L. bicolor can be questioned. First, how do we know the failure to establish was due to absence of a suitable host? The invasion literature, including that on biological control, has numerous examples in which several introductions failed and eventually one, apparently of an identical propagule, succeeded. A classic example is that of the introduction of the House Sparrow (Passer domesticus) to North America (Long 1981). An initial propagule of 16 birds failed in 1851, as did a second one a year later, in the same place, of [approximately]50 individuals. A third release the next year in the same place, again of [approximately]50 individuals, survived and spread; this species is now one of the most numerous in North America. The coreid bug Chelinidea vittiger was liberated on Santa Cruz Island (California) in 1945 for control of prickly pear (Opuntia spp.), but failed to establish; however, at least one of four releases on the same island in 1961-1964 was successful (Goeden 1977). Although differences in propagule size and genetics often complicate such cases, a common interpretation of this situation is that the failures were due to demographic stochasticity (Williamson 1996). In any event, because L. bicolor never established in Hawaii in the 1920s, it is difficult to see how this case can tell us much about its likelihood of shifting hosts where it has established.
As a final comment on Frank's assessment of risk, of course we concede that the probability that any of these biological control agents for mole crickets can cross hundreds of kilometers on their own to reach the range of G. major is low. How low is uncertain; species do occasionally make very difficult voyages autonomously. For example, the cactus moth Cactoblastis cactorum managed to spread on its own from Hawaii to all the other major Hawaiian islands (Tuduri et al. 1971) in just seven years. Further, many species can hitchhike. Pemberton (1995) argues that C. cactorum got part way from the Lesser Antilles to the Florida Keys, where it is devastating a candidate endangered species of Opuntia, by air transport of plants from the Greater Antilles to Miami. Turf and sod are carried on trucks; what is the probability that mole cricket natural enemies could move this way?
We do not wish to belabor the point that any of these scenarios carries a low probability. What is important is that events of low probability do happen, they have to be analyzed more quantitatively than in the past, and such analysis will not be easy. Frank has simply stated some of the factors that have to be taken into account.
Frank's analysis of costs and benefits is also very incomplete, at least partly because crucial data are simply unavailable. As is typical for claims of tremendous economic benefits of biological control, the only basis for estimating the cost of mole cricket impact on turf-grasses is an unpublished document, this time a handout at a 1986 meeting of the Georgia Entomological Society (R. D. Hudson, unpublished presentation, cited by Frank ). Frank concedes that "economists have documented little" of the "tremendous annual losses due to Scapteriscus mole crickets," and also that nobody has assessed the value of nontarget invertebrates. Thus we are left with a list of some of the crops that these mole crickets damage and no way to assess the full costs, current or possible, of alternatives to attempt to control the mole crickets. We also observe that Frank assumes there are just two possible alternatives for dealing with these mole crickets: classical biological control and broad-spectrum pesticides. At least for some pests, one can imagine other possibilities (see, e.g., U.S. Congress 1995).
In sum, biological control is not always practiced today with surgical precision, and risks, costs, and benefits are rarely comprehensively analyzed. We do not claim that the introductions made by the mole cricket research program that Frank defends have been or are likely to be damaging. However, we do not believe that Frank has settled this matter or that he has pointed the way to an adequate analysis of costs and benefits of various courses of action. We note that this exchange of views is relevant to many cases aside from this particular program. It is eerily similar to the defense of musk thistle (Carduus nutans) biocontrol by P. E. Boldt (unpublished letter) and the response by S. M. Louda et al. (unpublished letter). The conflict between practitioners and critics of biological control does not seem near to resolution.
Dennill, G. G., D. Donnelly, and S. L. Chown. 1993. Expansion of host range of a biocontrol agent Trichilogaster acaciaelongifoliae (Pteromalidae) released against the weed Acacia longifolia in South Africa. Agriculture, Ecosystems & Environment 43:1-10.
Frank, J. H. 1998. How risky is biological control?: Comment. Ecology 79:1829-1834.
Goeden, R. D. 1977. Biological control of weeds. Pages 357414 in C. P. Clausen, editor. Introduced parasites and predators of arthropod pests and weeds: a world review. United States Department of Agriculture, Agricultural Research Service, Agriculture Handbook No. 480, Washington, D.C., USA.
Hopper, D. R., and B. D. Smith. 1992. Status of tree snails (Gastropoda: Partulidae) on Guam, with a resurvey of sites studied by H. E. Crampton in 1920. Pacific Science 46:7785.
Long, J. L. 1981. Introduced birds of the world. Universe Books, New York.
Muniappan, R. 1987. Biological control of the giant African snail Achatina fulica Bow. in the Maldives. FAO Plant Protection Bulletin 35:127-133.
Pemberton, R. W. 1995. Cactoblastis cactorum in the United States: an immigrant biological control agent or an introduction of the nursery industry? American Entomologist 41:230-232.
Secord, D., and P. Kareiva. 1996. Perils and pitfalls in the host specificity paradigm. BioScience 46:448-453.
Simberloff, D., and P. Stiling. 1996. How risky is biological control? Ecology 77:1965-1974.
Tuduri, J. C. G., L. F. Martorell, and S. M. Gaud. 1971. Geographical distribution and host plants of the cactus moth, Cactoblastis cactorum (Berg) in Puerto Rico and the United States Virgin Islands. Journal of Agriculture of the University of Puerto Rico 55:130-134.
U.S. Congress Office of Technology Assessment. 1995. Biologically based Technologies for Pest Control. United States Government Printing Office, Washington, D.C., USA.
Williamson, M. 1996. Biological invasions. Chapman and Hall, London, UK.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||response to article by J.H. Frank in this issue, p. 1829|
|Author:||Simberloff, Daniel; Stiling, Peter|
|Date:||Jul 1, 1998|
|Previous Article:||How risky is biological control? Comment.|
|Next Article:||Ecological correlates of regional variation in life history of the moose Alces alces: comment.|