These case studies were all success stories. This should not, however, blind us to the fact that many attempts at biological control have failed. Waage and Greathead (1988) summarized the success rate of insect species that were introduced for biological control of insect pests and weeds. Of the species that became established, only 40% of those aimed at insect control and 31 % of those aimed at weed control were 'substantially successful'. This does not take account of species that were introduced but did not establish, for which records may not always be kept.
When assessing whether a species might be suitable as an agent of classical biological control, we need to answer three questions:
1 Will it establish and persist in the new area?
2 Will it sufficiently reduce the abundance of the pest?
3 Will it have undesired effects on non-target species?
The remainder of this chapter considers how we can obtain answers to those questions.
Can we spot species This question leads us to ask: are there consistent characteristics of that are good at invader species, features that make a species good at establishing in a invading} new area? There have been numerous published suggestions of such key characteristics, but these are usually followed by other writers pointing out exceptions where successful invaders do not have those characteristics. Williamson (1996) discussed five suggestions that have frequently been made: his discussion concerned all sorts of species, not just potential biological control agents. The suggested characteristics of invaders were:
1 ability to multiply rapidly;
2 genetic characteristics such as inbreeding;
3 a wide native range and high abundance in that range,-
4 ecologically or taxonomically distinct from species in the area to be invaded;
5 native range has a similar climate to the area to be invaded.
Williamson concluded that none of these characteristics has been shown consistently to be a predictor of ability to invade.
A basic fact is that all species must have the ability to invade. All species vary in abundance through time, so a species that becomes rare must be able to subsequently increase in abundance. Metapopulation theory (see Chapter 10) shows that in any small area any species is at risk of becoming extinct sooner or later, and it therefore needs an ability to reinvade. It also follows that all communities must be invadable: because they are likely to lose species they must also be able to gain them, otherwise they will finally end up with none. This does not deny that some species may be better at invading than others, and some communities more easily invaded than others. Rather, it says that the search for consistent characters indicating species that can establish in new communities has proved unfruitful, probably for this fundamental reason; so this is not a good starting point to choose promising species for biological control. Whether a certain species can establish is likely to depend on particular features of that species and its relation to its host and to the habitat; there are no easy rules of thumb.
Sometimes a control agent has failed to establish at the first attempt, yet later it has succeeded in establishing and has provided effective pest control. An example is myxomatosis in rabbits (Thompson & King 1994). In the late 1930s there were several unsuccessful attempts to introduce myxomatosis in parts of Europe; and test releases in Australia in the 1930s and 1940s were considered a failure because the disease did not spread. Yet in the 1950s further attempts resulted in widespread establishment of the disease in both Europe and Australia. How many One message from this is that if a new biocontrol agent fails to estab-
individuals to lish it may be worth making a second try. It also raises the question release whether the number of individuals released may be important. As explained in Chapter 10 (Conservation), a small population is at greater risk of becoming extinct, both because of chance fluctuations in numbers and for genetic reasons. This will also apply to a small initial population of a species introduced intentionally (Williamson 1996). Figure 8.8 shows experimental evidence that initial inoculum size can be important for biological control agents. Thrips are being tested as a possible biocontrol
Fig. 8.8 Results of two experiments in which different numbers of insects for biocontrol were released per tree or per bush, and number of trees or bushes on which the biocontrol insect had persisted were later determined. • Parasitoid of red scale released on to citrus trees in orchards in South Australia,- data of Campbell (1976). O thrips (pictured) released on to gorse bushes in New Zealand; data of Memmott, Fowler and Hill |1998|.
Was this article helpful?