Beamer, E. and Fresh, K., 2012. Juvenile Salmon and Forage Fish Presence and Abundance in Shoreline Habitats of the San Juan Islands, 2008-2009 – Map Applications for Selected Fish Species. Skagit River System Cooperative, La Conner, WA. pp. 81.

Fish presence probabilities for the San Juan Islands’ shorelines were calculated for seven juvenile fish species or species groupings from results of 1,350 beach seine sets made at 80 different sites throughout the San Juan Islands in 2008 and 2009. The juvenile fish species evaluated were: unmarked (assumed wild) Chinook salmon (Oncorhynchus tshawytscha), chum salmon (Oncorhynchus keta), pink salmon (Oncorhynchus gorbuscha), Pacific herring (Clupea pallasii), Pacific sand lance (Ammodytes hexapterus), surf smelt (Hypomesus pretiosus), and lingcod/greenling (family Hexagrammidae). Because juvenile salmon are known to be migratory in nearshore waters, our sampling plan was established to encompass the times of year when it is possible for juvenile salmon to be present within shoreline habitats of the San Juan Islands. Beach seining typically occurred at each site twice per month from March through October each year. We hypothesized that space (i.e., where within the San Juan Islands) and habitat type differences would influence whether or not fish were present (or abundant) at specific locations within the San Juan Islands. Beach seine sites were selected to represent different regions within the San Juan Islands (SiteType2) and different geomorphic shoreline types (SiteType3). We also stratified by two coarser-scale variables for space and habitat type. The coarse variable for space has two possible values related to whether the site is located in “interior” or “exterior” areas of the San Juan Islands. The coarse scale variable for habitat was either “enclosure” or “passage.” All 80 sites were characterized by these space and habitat type variables. We used generalized linear models (GLM) to test whether our hypothesized variables of space and habitat type influence fish presence and abundance. We found strong support for both influences with no strong indication to weigh one variable over the other. Thus, we created two model versions to predict indices of fish presence probability based on fish presence rate results summarized by each of the 80 sites for each space and habitat type variable. Models were created for each of the seven juvenile fish species or species grouping. A high resolution model (HRM) multiplied fish presence values for SiteType2 by SiteType3. A lower resolution model (LRM) multiplied fish presence rate values for the coarse space variable by the coarse habitat type variable. For each model, the calculated fish presence probabilities could range between 0 and 1. The resulting fish probability of presence estimates relate to our beach seine sampling regime of twice per month from March through October. For example, a Chinook probability of presence value of 1 for a site means you are certain to find Chinook salmon present at the site if you beach seine twice per month from March through October. We also found fish presence rates to be positively correlated with fish density for all fish species or species groupings in this report. This means sites with higher values of fish presence also have higher values of fish abundance. The strength and type (e.g., linear, exponential) of the correlated relationships varied.