Swallowtail Clutch Size Reconsidered

Abstract

Females of many butterfly species lay eggs in clutches (Chew and Robbins 1983). Models have been developed to explain the phenomenon in terms of strategies that increase the fitness of the ovipositing female (Parker and Courtney 1984, Skinner 1984, Mangel 1987). Pilson and Rausher (1988) report an empirical test of two such models using the pipevine swallowtail butterfly, Battus philenor. They attempt to confirm that ovipositing females, within a population, adjust the number of eggs laid in a clutch (clutch size) as a function of the suitability of the substrate for larval development. I suggest that Pilson and Rausher have confounded two aspects of oviposition behavior and reanalyze their data accordingly. It is concluded that there is no apparent relationship between clutch size and the authors' measure of host suitability. A geographic comparison with a California population of pipevine swallowtail is presented. Pilson and Rausher (1988) reported a correlation between the dispersal size of test larvae and clutch size in a Texan population. A sample of test plants, Aristolochia reticulata, was obtained by censusing all plants visited by searching females. Plants the female rejected were categorized as clutch size 0 and included in the study. If such plants have indeed been assessed for clutch size the rejection may be due to the low quality of the host, the submotivational physiological state of the female or to some integrated combination of the two. A female in a low physiological oviposition state may reject a host due to her status and not because of the quality of the host. A clutch size/host quality correlation analysis for host assessment assumes that all sampled females are in an egglaying physiological state or will be upon contact with the host. Some number of females that reject hosts are not in this state and the hosts they contact are not assessed for clutch size and should not be included in an analysis of host quality assessment behavior. Fig. 1 of Pilson and Rausher (1988) regarding the size of test larvae and clutch size was reanalyzed excluding clutch sizes of 0. All clutch sizes of 0 are excluded since it is not possible to tell which cases may be due to subthreshold physiological levels. The recalculated correlation shows that the significant and positive relationship reported (r = 0.33, p < 0.001, n = 163) no longer exists (r = 0.20, p = 0.20, n = 45). From these data it is not possible to determine if clutch size of Texan B. philenor is related to host quality as measured by test larvae. In a partially bivoltine California population of B. philenor ovipositing on A. californica, a large range of clutch sizes is evident at each census day within a seasonal generation (Tab. 1). Clutch sizes differ significantly between the first and second generation. However, there is no relationship between clutch size and the density of suitable larval substrate (young buds/ census hoop). While variation in B. philenor clutch size