Allocation of parental investment among individual offspring in the European beewolf Philanthus triangulum F. (Hymenoptera : Sphecidae)

Optimal allocation of parental resources is an important life history trait . However, it has been rarely investigated empirically. We tested aspects o f optimal allocation theory in a digger wasp, the European beewolf. Investm ent allocation theory assumes (1) a trade-off between investment per offspr ing and offspring number and (2) a convex relationship between investment p er offspring and fitness returns. From this relationship an optimum amount of investment per offspring can be derived and parents are predicted to pro vide each offspring with this optimum amount of investment. We used the num ber of bees in a brood cell as a measure of parental investment. Offspring fitness was quantified as both survival until emergence and success as adul ts. There is evidence for a trade-off between current and future reproducti on, suggesting that the first assumption is met. In contradiction to the se cond assumption, one mortality factor, parasitism, increased proportionally with the number of bees in a brood cell. However, overall mortality until emergence significantly decreased with the number of bees in a brood cell a s assumed by the theory. The determination of the optimum amount of investm ent per offspring is complicated because the sexes possibly differ in their relationship between amount of investment and fitness. Individual males re ceived considerably fewer bees (2.2 +/- 0.8) than females (3.8 +/- 0.5). Tw o independent estimates of the investment specific survival suggested than sons with two bees had the highest fitness returns per single bee and, cons istent with the prediction, most sons were provisioned with two bees. For d aughters, four bees is probably the optimum amount and most daughters were provisioned wit this number. In both sexes the variation of investment per offspring was less than expected by a Poisson distribution wit the same mea n. These findings support the view that parental investment is allocated in a way than optimizes the trade-off between offspring number and investment per offspring. However, variation contradicting the hypothesis still occur red. This might be explained either by adaptive variation in the amount of investment per offspring, constraints in the adjustment of the optimum amou nt of investment, or problems in measuring parental investment. (C) 2000 Th e Linnean Society of London.