Spatial variations in the distribution of yearling spring Chinook salmon off Washington and Oregon using COZIGAM analysis

Abstract

Yearling Chinook salmon Oncorhynchus tshawytscha were sampled off Washington and Oregon, USA, along with environmental factors, every June from 1998 to 2010. The abundance of yearling Chinook salmon varied over space with a high proportion of zero catches. Positive catches were more numerous north of the Columbia River, likely because most yearling Chinook salmon turn north after leaving the Columbia River. Using the latitude of the Columbia River mouth as a geographical border, the survey area was divided into 2 regions: north and south of the Columbia River. We hypothesized that (1) the spatial distribution pattern within each region was related to local environmental factors and (2) the difference between north and south was related to largescale ocean processes. A constrained zero-inflated generalized additive model (COZIGAM) was applied to examine the non-linear relationships between juvenile salmon abundance and environmental factors. Results from the COZIGAM suggested that water temperature, chlorophyll a (chl a) concentration, copepod biomass and spatial factors were significantly correlated with the density of salmon in the northern region, and only chl a concentration was correlated significantly with yearling Chinook density in the southern region. The difference between the abundances north and south of the Columbia River was significantly correlated with alongshore ocean currents, with weaker alongshore currents leading to greater difference between north and south. Results suggest that salmon distribution is determined not only by standard habitat parameters (local biotic and abiotic factors) but by ocean conditions such as the strength of alongshore coastal currents.

Early ocean distribution of juvenile Chinook salmon in an upwelling ecosystem

“…Krill, which are a patchy resource, were encountered more frequently in the stomachs of captured subadults and adults, which can range over a greater area and take advantage of more dispersed and patchily distributed prey. Remaining closer to shore to avoid high predation rates may be related in some way to increased turbidity associated with high concentrations of chlorophyll‐ a , which has been a good predictor of juvenile salmon presence in this and in other studies (Peterson et al ., ; Pool et al ., ; Yu et al ., ; Burke et al ., ). It is difficult to assess a foraging benefit of chlorophyll‐ a to salmon because time lags between primary production and the trophic level at which juvenile salmon feed are too large to infer trophic transfer (Miller et al ., ).…”

Anadara kagoshimensis (Mollusca: Bivalvia: Arcidae) in Adriatic Sea: morphological analysis, molecular taxonomy, spatial distribution, and prediction

“…Anyhow, the percentage of total variance explained for the species was not satisfactory, likely due to the large amount of zero catches in the data matrix. Therefore, a Zero-Inflated Generalized Additive Model (ZIGAM; Liu & Chan, 2010;Yu et al, 2012) was applied to define the spatial distribution in the previous analyses. The model assumes that the response variable follows a probabilistic mixture distribution of a zero atom and a continuous distribution belonging to the exponential family (Lambert, 1992;Hall, 2000;Yu et al, 2012), thus providing a better performance when zeroinflated data are involved.…”

Cloudy with a chance of sardines: forecasting sardine distributions using regional climate models

“…Pelagic species may be particularly responsive to seasonal and inter‐annual variations in climate because they alter their distributions and migrations directly in response to ocean conditions (McGowan et al ., ; Brodeur et al ., ; Emmett et al ., ; Hooff and Peterson, ). For example, yearling salmon abundance has been correlated on an annual basis with water temperature, chlorophyll and copepod biomass in the northern California current (Peterson et al ., ; Yu et al ., ) and some of these parameters are used as indicators of early salmon survival that complement existing predictions of adult salmon runs (http://www.nwfsc.noaa.gov/research/divisions/fe/estuarine/oeip/index.cfm). Other highly mobile pelagic taxa, such as Humboldt squid ( Dosidicus gigas ) and jack mackerel ( Trachurus symmetricus ) may rapidly expand their range and become particularly abundant in new regions during years or seasons characterized by ‘anomalous’ sea surface conditions (e.g., salinity, temperature and dissolved oxygen)(Brodeur et al ., ; Chesney et al ., ).…”