1. Approach for estimating the dynamic physical thresholds of phytoplankton production and biomass in the tropical‐subtropical Pacific Ocean
- Author
-
Gómez‐Ocampo, E., Gaxiola‐Castro, G., and Durazo, Reginaldo
- Abstract
Threshold is defined as the point where small changes in an environmental driver produce large responses in the ecosystem. Generalized additive models (GAMs) were used to estimate the thresholds and contribution of key dynamic physical variables in terms of phytoplankton production and variations in biomass in the tropical‐subtropical Pacific Ocean off Mexico. The statistical approach used here showed that thresholds were shallower for primary production than for phytoplankton biomass (pycnocline < 68 m and mixed layer < 30 m versus pycnocline < 45 m and mixed layer < 80 m) but were similar for absolute dynamic topography and Ekman pumping (ADT < 59 cm and EkP > 0 cm d−1versus ADT < 60 cm and EkP > 4 cm d−1). The relatively high productivity on seasonal (spring) and interannual (La Niña 2008) scales was linked to low ADT (45–60 cm) and shallow pycnocline depth (9–68 m) and mixed layer (8–40 m). Statistical estimations from satellite data indicated that the contributions of ocean circulation to phytoplankton variability were 18% (for phytoplankton biomass) and 46% (for phytoplankton production). Although the statistical contribution of models constructed with in situ integrated chlorophyll aand primary production data was lower than the one obtained with satellite data (11%), the fits were better for the former, based on the residual distribution. The results reported here suggest that estimated thresholds may reliably explain the spatial‐temporal variations of phytoplankton in the tropical‐subtropical Pacific Ocean off the coast of Mexico. We statistically estimated the thresholds of physical dynamic variables on phytoplanktonThe mixed layer thresholds were shallower for primary production than for integrated chlorophyll aThe contributions of ocean circulation to phytoplankton production and biomass variability were 46% and 18%
- Published
- 2017
- Full Text
- View/download PDF