Premium
Analysis of pelagic species decline in the upper San Francisco Estuary using multivariate autoregressive modeling (MAR)
Author(s) -
Mac Nally Ralph,
Thomson James R.,
Kimmerer Wim J.,
Feyrer Frederick,
Newman Ken B.,
Sih Andy,
Bennett William A.,
Brown Larry,
Fleishman Erica,
Culberson Steven D.,
Castillo Gonzalo
Publication year - 2010
Publication title -
ecological applications
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.864
H-Index - 213
eISSN - 1939-5582
pISSN - 1051-0761
DOI - 10.1890/09-1724.1
Subject(s) - pelagic zone , estuary , smelt , ecology , fishery , zooplankton , predation , biology , geography , fish <actinopterygii>
Four species of pelagic fish of particular management concern in the upper San Francisco Estuary, California, USA, have declined precipitously since ca. 2002: delta smelt ( Hypomesus transpacificus ), longfin smelt ( Spirinchus thaleichthys ), striped bass ( Morone saxatilis ), and threadfin shad ( Dorosoma petenense ). The estuary has been monitored since the late 1960s with extensive collection of data on the fishes, their pelagic prey, phytoplankton biomass, invasive species, and physical factors. We used multivariate autoregressive (MAR) modeling to discern the main factors responsible for the declines. An expert‐elicited model was built to describe the system. Fifty‐four relationships were built into the model, only one of which was of uncertain direction a priori. Twenty‐eight of the proposed relationships were strongly supported by or consistent with the data, while 26 were close to zero (not supported by the data but not contrary to expectations). The position of the 2‰ isohaline (a measure of the physical response of the estuary to freshwater flow) and increased water clarity over the period of analyses were two factors affecting multiple declining taxa (including fishes and the fishes' main zooplankton prey). Our results were relatively robust with respect to the form of stock–recruitment model used and to inclusion of subsidiary covariates but may be enhanced by using detailed state–space models that describe more fully the life‐history dynamics of the declining species.