Another way to explain why studies of avian provisioners seldom find evidence for the cost of reproduction lies in the distinction between effort and investment. The cost of reproduction represents only one reason why provisioners might make less effort than the maximum. Thus, experimentally induced higher workloads might not affect the provisioner's own survival or future fecundity. For example, a provisioner could maintain a small fat reserve, which reduces its delivery capacity. The provisioner would benefit because it could expend the reserve to increase its delivery capacity when demand unexpectedly escalates. An observer would measure this expenditure of the reserve as a loss of "condition," but this loss would not jeopardize the provisioner's survival: the provisioner maintains the reserve solely to buffer delivery capacity against shortages that occur from time to time.
Nest defense may also select for reduced provisioning effort. Dyer and Seeley (1991) explain how differences in nest architecture lead to differences in work tempo among various species of honeybees. High-tempo honeybee species nest in enclosed cavities, while the low-tempo species have open nests covered in a curtain of workers. Dyer and Seeley propose that open nests require more workers for defense and thermoregulation and, they argue, have therefore selected for reduced tempo because high-tempo workers experience high mortality rates, which makes it difficult to maintain a large worker population. Lack (1968) documented a similar correlation in altricial birds: ground-
nesting species (presumably with the most vulnerable nests) have the smallest clutches and fastest growth rates. This observation suggests a relationship between nest vulnerability and provisioning tempo, as found among honeybee species, though the direction appears to be reversed. In both cases, we hypothesize, nest vulnerability has created selection on the metabolic capacity of the provisioner.
Some authors have argued that expending energy reduces longevity (e.g., Calder 1985). According to this view, a high workload reduces the life span, in the same way that overusing a flashlight more quickly leads to a dead battery. More recent authors offer more sophisticated interpretations. Some argue that energetic expenditure can reduce longevity by compromising an organism's immune response (Richner et al. 1995), while others argue that reproductive effort accelerates senescence (Gustaffson and Part 1990). If these arguments are correct, and if provisioning involves increased effort, then all provisioning entails a cost of reproduction.
Another possibility is that extra effort may expose provisioners to predators (Magnhagen 1991) or parasites (Richner et al. 1995). For example, Harfenist and Ydenberg (1995) found that parent rhinoceros auklets stopped provisioning sooner in areas where eagles posed a threat. Some authors argue that predation has played a role in the evolution of nocturnal provisioning and precocity in seabirds (Gaston 1992). Provisioning studies have barely scratched the surface of this problem; specifically, we still need experimental studies that document the responses ofprovisioners to the threat ofpredation. Studies of predation risk in other foraging situations illustrate the feasibility of these experiments (Lima and Dill 1990). Nonacs and Dill (1990) experimentally demonstrated that workers of the ant Lasius pallitarsis balanced the quality of the food available at a feeding site against the danger posed by a large predatory ant. Dukas (2001) discusses the potential effect ofpredation danger on insect pollinators (especially the many provisioning species such as bees) and notes a general lack of studies on the idea. Though it seems reasonable to suppose that each provisioning excursion carries some extra exposure to danger, we have, as yet, few good data with which we can evaluate how this exposure affects provisioning.
Was this article helpful?