Returning now to predators in general, it may seem that since the effects of predators are harmful to individual prey, the immediate effect of predation on a population of prey must also be predictably harmful. However, these effects are not always so predictable, for one or both of two important reasons. In the first place, the individuals that are killed (or harmed) are not always a random sample of the population as a whole, and may be those with the lowest potential to contribute to the population's future. Second, there may be compensatory changes in the growth, survival or reproduction of the surviving prey: they may experience reduced competition for a limiting resource, or produce more offspring, or other predators may take fewer of the prey. In other words, whilst predation is bad for the prey that get eaten, it may be good for those that do not. Moreover, predation is least likely to affect prey dynamics if it occurs at a stage of the prey's life cycle that does not have a significant effect, ultimately, on prey abundance.
To deal with the second point first, if, for example, plant recruitment is not limited by the number of seeds produced, then insects that reduce seed production are unlikely to have an important effect on plant abundance (Crawley, 1989). For instance, the weevil Rhinocyllus conicus does not reduce recruitment of the nodding thistle, Carduus nutans, in southern France despite inflicting seed losses of over 90%. Indeed, sowing 1000 thistle seeds per square meter also led to no observable increase in the number of thistle rosettes. Hence, recruitment appears not to be limited by the number of seeds produced; although whether it is limited by subsequent predation of seeds or early seedlings, or the availability of germination sites, is not clear (Crawley, 1989). (However, we have seen in other situations (see Section 9.2.5) that predispersal seed predation can profoundly affect seedling recruitment, local population dynamics and variation in relative abundance along environmental gradients and across microhabitats.)
The impact of predation is often limited by compensatory reactions amongst the survivors as a result of reduced intraspecific competition. Thus, in a classic experiment in which large numbers of woodpigeons (Columba palumbus) were shot, the overall level of winter mortality was not increased, and stopping the shooting led to no increase in pigeon abundance (Murton et al., 1974). This was because the number of surviving pigeons was determined ultimately not by shooting but by food availability, and so when shooting reduced density, there were compensatory reductions in intraspecific competition and in natural mortality, as well as density-dependent immigration of birds moving in to take advantage of unexploited food.
compensatory reactions amongst survivors
Figure 9.8 Trajectories of numbers of grasshoppers surviving (mean ± SE) for fertilizer and predation treatment combinations in a field experiment involving caged plots in the Arapaho Prairie, Nebraska, USA. (After Oedekoven & Joern, 2000.)
Indeed, whenever density is high enough for intraspecific competition to occur, the effects of predation on a population should be ameliorated by the consequent reductions in intraspecific competition. Outcomes of predation may, therefore, vary with relative food availability. Where food quantity or quality is higher, a given level of predation may not lead to a compensatory response because prey are not food-limited. This hypothesis was tested by Oedekoven and Joern (2000) who monitored grasshopper (Ageneotettix deorum) survivorship in caged prairie plots subject to fertilization (or not) to increase food quality in the presence or absence of lycosid spiders (Schizocoza spp.). With ambient food quality (no fertilizer, black symbols), spider predation and food limitation were compensatory: the same numbers of grasshoppers were recovered at the end of the 31-day experiment (Figure 9.8). However, with higher food quality (nitrogen fertilizer added, colored symbols), spider predation reduced the numbers surviving compared to the no-spider control: a noncompensatory response. Under ambient conditions after spider predation, the surviving grasshoppers encountered more food per capita and lived longer as a result of reduced competition. However, grasshoppers were less food-limited when food quality was higher so that after predation the release of additional per capita food did not promote survivorship (Oedekoven & Joern, 2000).
Turning to the nonrandom distribution of predators' attention within a population of prey, it is likely, for example, that predation by many large carnivores is focused on the old (and infirm), the young (and naive) or the sick. For instance, a study in the Serengeti found that cheetahs and wild dogs killed a disproportionate number from the younger age classes of Thomson's gazelles (Figure 9.9a), because: (i) these young animals were easier to catch (Figure 9.9b); (ii) they had lower stamina and running speeds; (iii) they were less good at outmaneuvering the predators (Figure 9.9c); and (iv) they may even have failed to recognize the predators (FitzGibbon & Fanshawe, 1989; FitzGibbon, 1990). Yet these young gazelles will also have been making no reproductive contribution to the population, and the effects of this level of predation on the prey population will therefore have been less than would otherwise have been the case.
Similar patterns may also be found in plant populations. The mortality of mature Eucalyptus trees in Australia, resulting from defoliation by the sawfly Paropsis atomaria, was restricted almost entirely to weakened trees on poor sites, or to trees that had suffered from root damage or from altered drainage following cultivation (Carne, 1969).
Taken overall, then, it is clear that the step from noting that individual prey are harmed by individual predators to demonstrating that prey adundance is adversely affected is not an easy one to take. Of 28 studies in which herbivorous insects were experimentally excluded from plant communities using insecticides, 50% provided evidence of an effect on plants at the population level (Crawley, 1989). As Crawley noted, however, such proportions need to be treated cautiously. There is an almost inevitable tendency for 'negative' results (no population effect) to go unreported, on the grounds of there being 'nothing' to report. Moreover, the exclusion studies often took 7 years or more to show any impact on the plants: it may be that many of the 'negative' studies were simply given up too early.
effects ameliorated by reduced competition predatory attacks are often directed at the weakest prey difficulties of demonstrating effects on prey populations
Killed by cheetahs Killed by wild dogs Percentage in population
Figure 9.9 (a) The proportions of different age classes (determined by tooth wear) of Thomson's gazelles in cheetah and wild dog kills is quite different from their proportions in the population as a whole. (b) Age influences the probability for Thomson's gazelles of escaping when chased by cheetahs. (c) When prey (Thomson's gazelles) 'zigzag' to escape chasing cheetahs, prey age influences the mean distance lost by the cheetahs. (After FitzGibbon & Fanshawe, 1989; FitzGibbon, 1990.)
Many more recent investigations have shown clear effects of seed predation on plant abundance (e.g. Kelly & Dyer, 2002; Maron et al., 2002).
Was this article helpful?