Salamander Egg Sizes 



79 



A. maculatum 



E. wilderae 



D. aeneus 



D. ochrophaeus 



D. santeetlah 



40 60 



Percent of variation 



100 



Fig. 4. Relative contributions of intraclutch and interclutch variation in 

 egg size to overall variance. For each species, a separate model II one- 

 way ANOVA of egg size was conducted with clutches as the factor, thus 

 a total of five ANOVAs. Contributions are calculated using the factor and 

 error sum-of-squares of the ANOVA table. Black bars represent intraclutch 

 variation, and hatched bars represent interclutch variation. 



Interclutch and Intraclutch Contribution to Variation 



Partitioning of the variance components demonstrated the 

 respective contributions of intraclutch and interclutch variation to the 

 sample variance in egg size (Fig. 4). For E. wilderae and A. maculatum, 

 interclutch contributions to the variance outweighed intraclutch 

 contributions. In all three desmognathines, interclutch and 

 intraclutch contributions were approximately equal. 



As pointed out earlier, A. maculatum females may oviposit a 

 clutch that consists of several clusters. The mean cluster size reported 

 in Table 1 is lower than earlier reports of clutch size in A. maculatum 

 (Bishop 1941, Wilbur 1977, Pfingsten and Downs 1989). This suggests 

 that females of A. maculatum in the southern Appalachians may 

 oviposit clutches that consist of several clusters. Because interclutch 

 variation outweighs intraclutch variation in this species, the possibility 

 remains that several clusters might contribute to a single A. maculatum 

 clutch. This may confound the analysis of intraclutch variation, i.e., 

 the intraclutch CVs for A. maculatum might be underestimates. However, 

 the pooled variation remains as an indicator of potential adaptive 

 variation in egg size at the interclutch level. 



