THE SLOW‐GROWTH–HIGH‐MORTALITY HYPOTHESIS: A TEST USING THE CABBAGE BUTTERFLY

The slow‐growth–high‐mortality hypothesis predicts that prolonged development in herbivorous insects results in greater exposure to natural enemies and a subsequent increase in mortality. We tested this hypothesis using the cabbage butterfly Pieris rapae and its larval parasitoid Cotesia glomerata. We conducted a series of field and laboratory experiments to determine how variation in the larval development of P. rapae within and among four species of host plants ( Brassica oleracea, Tropaeolum majus, Lunaria annua, and Cleome spinosa ) influenced parasitism rates by C. glomerata. On the same host plant species, fast‐developing larvae incurred less parasitism than slow‐developing larvae of the same age, because once larvae reached the third instar they became much less vulnerable to attack due to the onset of encapsulation. Thus, the “window of vulnerability” was extended for slow‐developing larvae. Similarly, the window of vulnerability was prolonged and larvae were susceptible to parasitism for a longer period of time on host plant species on which development was delayed. Also, simply exposing larvae for a longer period of time to parasitoids resulted in increased rates of parasitism. Consequently, the window of vulnerability sets the temporal limits within which parasitoids may act. These findings are all consistent with the slow‐growth–high‐mortality hypothesis. Slow growth, however, did not always translate into increased parasitism. We found no evidence for a positive relationship between host development time and parasitism when the relationship was examined across host plant species. In fact, the parasitism rate was highest on B. oleracea, the plant species on which P. rapae developed most rapidly. The adverse effects of some host plant species on the host‐location behavior of parasitoids apparently offset the potential for increased risk of parasitism due to delayed larval development. In contrast, within the same host plant species, there was good correspondence between slow growth in P. rapae and high mortality inflicted by C. glomerata. We argue that the relative effects of plant variation on adult parasitoid behavior are probably much less among conspecific plants than across plant species, thus allowing the potential effects of extended development on increased parasitism to be realized. By experimentally manipulating larval development time using temperature and diet treatments (host plant effects removed), slow growth also resulted in increased parasitism. In a survey of published studies in which the association between slow growth and high enemy‐inflicted mortality could be assessed, support for the hypothesis was found for free‐living insects feeding on the same or closely related host plant species. The hypothesis was not supported for herbivores that fed on different species of host plants, or for herbivores whose vulnerable stage was concealed in plant tissue. The extent to which the slow‐growth–high‐mortality hypothesis finds support depends on the interactive effects of plant variation on herbivore development as it indirectly increases the risk of enemy attack and on the direct effects of plant features on enemy access to herbivores.