MEPS 427:145-159 (2011)  -  DOI: https://doi.org/10.3354/meps09048

ABSTRACT: Knowledge of copepod stage-specific mortality rates is essential for understanding secondary production. A time-series of samples of the calanoid copepod Clausocalanus furcatus was collected from a petroleum platform in the northern Gulf of Mexico every 12 h for ~2 to 3 wk in March–April and May–June 2003. Egg and nauplii (stages I and II) abundances were estimated from 30 l Niskin water bottle samples, and abundances of later stages were estimated from vertical net tows. An inverse parameter estimation technique based on a stage-based matrix population model and a search procedure using the sequential quadratic programming algorithm was applied to estimate stage-specific mortality rates. Simulation experiments were conducted to test the robustness of this technique. When the technique was applied to simulated data with different levels of variability, introduced by lognormal error terms, the technique performed well at low to medium level of variability (coefficient of variation <0.5). We then applied this technique to field data in periods when the depth of the pycnocline was similar (i.e. advection was likely similar). The estimated stage-specific mortality rates showed similar patterns across stages in March–April and May–June, with mortality rates being highest for eggs, lower for early nauplii stages, lower still for late nauplii stages, lowest for early copepodite stages, and high again for late copepodite stages. We compare our estimated rates to those reported in the literature and conclude with a discussion of the importance of advancing estimation methods to deal with advection and a recommended approach for applying mortality estimation methods to zooplankton in dynamic environments.

KEY WORDS: Copepod · Instantaneous mortality rate · Inverse technique · Matrix population model · Sequential quadratic programming · Sampling variability