Egg size is one of the most frequently used surrogate measures of maternal investment and is strongly related to fitness-determining traits, such as offspring size in many marine animals, but the relationship is not universal. Because the amount of yolk allocated to eggs is finite, not all fitness-determining traits can be simultaneously maximised, and trade-offs should be expected. The results of the present study show that egg size (quantified as cross-sectional area) poorly predicts the size of larval morphological traits (length, yolk sac area, oil globule area, myomere depth, and eye area) in the native Australian fish Galaxias maculatus. Egg size was correlated with yolk sac size and larval body depth, but it explained <13% of the total variation in these traits. Moreover, egg size did not predict the time it took for larvae to hatch, nor did it predict the duration from hatching to starvation. However, when egg size and yolk sac size (i.e. yolk remaining at hatching) were both included as predictors, the fits with larval traits improved greatly (r 2 = 0.02 to 0.46 for larval length, r 2 = 0.11 to 0.17 for myomere depth, and r 2 < 0.001 to 0.14 for eye area). These findings indicate that there is a trade-off between the quantity of yolk at hatching and the size of larval morphological traits in G. maculatus. The remaining unexplained variation may be a consequence of trade-offs with other unmeasured traits, such as metabolic rate, and the potential that egg size is not a good proxy for maternal investment.
KEY WORDS: Life history theory · Maternal effects · Offspring size · Phenotypic plasticity
Resale or republication not permitted without written consent of the publisherMar Ecol Prog Ser 461: 165-174, 2012 166 size of larvae at hatching in a large number of marine species (reviewed by Duarte & Alcaraz 1989, Chambers & Leggett 1996, Marshall & Keough 2007. However, many of the studies that examine the relationship between egg size and larval traits compare among group means rather than individuals. When relationships between egg size and larval traits have been examined at the individual level, they tend to be weaker (reviewed by Chambers 1997). For instance, there is a significant relationship between egg size and larval length in Atlantic cod Gadus morhua. Egg size explains 58% of the variation in larval length when group means are compared (Marteinsdottir & Steinarsson 1998), but when individuals are compared, only 6% of the variation is explained (Pepin et al. 1997). Consequently, the current evidence for strong relationships between egg size and offspring size may be misleading.Egg size also affects several other traits in marine larvae that are important for survival, such as energy reserves, growth rate, and duration of the embryonic and larval periods (reviewed by Duarte & Alcaraz 1989, Cowen & Sponaugle 1997, Marshall & Keough 2007. Because the maternal resources that can be allocated to eggs are finite, trade-offs should be expected among larval traits (i.e. not all fitness determi...