Modeling habitat use of young-of-the-year Pacific sand lance (Ammodytes hexapterus) in the nearshore region of Barkley Sound, British Columbia

Successful management of coastal ecosystems requires an understanding of the distribution of key food web species through space and time relative to environmental predictors. Here, I examined the habitat use of an important forage species, the Pacific sand lance (Ammodytes hexapterus), using an inductive habitat modeling approach. I examined the presence/absence of Young-of-the-Year Pacific sand lance in the intertidal/shallow subtidal habitat of Barkley Sound, British Columbia. I determined sand lance occurrence using a beach seine at low tide, which was preferred to visual and intertidal digging detection methods due to its high detection frequency, ease of use, and ability to physically capture sand lance. I constructed models using environmental data measured at two different scales: 1) empirically measured environmental data (site-specific level) and 2) GIS derived environmental data extracted with a 200m buffer (landscape level). For each scale, I employed both logistic regression and classification tree modeling procedures to construct habitat models of sand lance occurrence at 55 study sites sampled during the summer of 2003. At the site-specific level, both logistic regression and classification tree models performed similar, however, classification trees were easier to construct and interpret as well as revealing interactions among variables undetected by logistic regression. Based on a deviance pruned classification tree, Grain Size Mean, Intertidal Eelgrass Presence, Absence, Major Substrate Low Intertidal, and Grain Size Sorting influenced sand lance occurrence at this scale, with importance values of 100, 79, 75, and 61 respectively. Standardized importance was based on the overall change in node impurity in the classification tree for each variable. At the landscape level, only Coastline Density was significantly related to sand lance occurrence, however, it was difficult to suggest this variables direct relation to sand lance habitat use. Overall, the habitat modeling approach identified important environmental variables influencing sand lance habitat selection at two different scales and stressed the utility of field data to construct and confirm these models.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.