FECUNDITY VERSUS OFFSPRING SIZE IN THE GREATER WHITE-TOOTHED SHREW, CROCIDURA-RUSSULA

1. The relationships between female body mass (W(a)), litter size (m), juvenile growth rate (G) and mass at weaning (W20) were examined by m onitoring natural litters in 29 greater white-toothed shrews, Crocidur a russula (Hermann 1780). The trade-offs between m and G or W20 were f urther investigated by manipulating litter sizes: each of seven female s reared four litters of 2, 4, 6 and 8 offspring. 2. Offspring mass at weaning (W20) exhibited a large variance, most of which could be attr ibuted (ANCOVA On manipulated litters) to two effects: a litter-size e ffect, and a female individual effect, referred to as 'female quality' . 3. Litter size explained 68% of the variance in W20 among manipulate d litters (linear regression). The limited milk supply was probably re sponsible for this effect, because litter size depressed growth rate d uring the first half of the lactation period (G1), but not during the weaning stage (G2). 4. Among non-manipulated litters, litter size corr elated positively with maternal body mass (W(a)), so that large female s tended to produce small juveniles. This correlation between m and W( a) is seen as the result of a body-mass dependence in the cost of rais ing a litter of a given size, during either pregnancy or lactation. 5. Differences in 'female quality' explained 16% of the variance in W20 among manipulated litters. This factor did not affect G1 and may thus relate to differences among offspring of different females in their ra tes of processing milk and/or external food during late lactation. 6. 'Female quality' was independent of both body mass and litter size: la rger females did not produce larger offspring when controlled for litt er size, while higher-quality females did not produce larger litters. 7. Our results support the hypothesis that most variance in adult and juvenile body masses is non-genetic, and stems from the trade-off betw een litter size and offspring size.