MEPS 336:187-202 (2007)  -  doi:10.3354/meps336187

ABSTRACT: The 2 most common experimental methods used to estimate rates of juvenile growth in marine copepods are the molt rate (MR) method, and the artificial cohort (AC) method. Recently, we showed the equations used in the MR method to be incorrect, and proposed a modified molt rate (MMR) method. Here, using statistical and model approaches, we compare the AC and MMR methods under various scenarios to quantify their errors. Although the AC and MMR methods both use a combination of field sampling and simulated in situ incubations to estimate somatic growth, they differ in several important characteristics. The AC method determines growth by the change in mean weight during incubation. Mean weight of copepods in the samples can be determined directly, or inferred from mean weight by life stage or from length–weight regressions. We show that substantial error is avoided only if weights are measured directly (AC_direct). The AC_direct method is insensitive to variable age within stage due to mortality or variable recruitment in the sampled population, an important advantage over the MMR method. However, the AC_direct method is sensitive to variation in growth rate during incubation, which does not affect the MMR method. We therefore recommend that most experimental estimates of growth rate should apply the AC_direct method, with the MMR as a suitable alternative provided its biases are considered. An indirect method based on life stage is biased and we no longer recommend it, and an indirect method based on length–weight regression provides an intermediate level of bias.

KEY WORDS: Growth rate · Copepoda · Methods · Molt rate