AE Bates3, MI Duncan1,2, NC James2, Professor Warren Potts1
1Department of Ichthyology and Fisheries Science, Rhodes University, Grahamstown, South Africa, 2South African Institute of Aquatic Biodiversity, Grahamstown, South Africa, 3Department of Ocean Sciences, Memorial University of Newfoundland, St. John’s, Canada
Anthropogenic induced climate change is modulating the redistribution of species across the globe at accelerated rates. Accurately predicting climate mediated distribution shifts is therefore a prerequisite to ensure sustainable management strategies remain optimised. To predict such distribution shifts scientists have relied upon either correlative species distribution models (SDMs) or mechanistic based distribution models but both approaches have limitations such as many underlying assumptions or heavy data requirements respectively. To model the current and future distributions of Chrysoblephus laticeps, a commercially important seabream from South Africa, we combine the SDM approach with transformations of environmental data into an index of metabolic protention derived from the relationship between critical oxygen partial pressure and temperature. We trained 20 random forest classification trees with 206 C. laticeps occurrence points and 206 pseudo-absence points based on current climate data (2005 – 2009) and projected trained models onto future (2100) climate scenarios. We found that C. laticeps’ current distribution is best approximated by including only depth and minimum monthly metabolic potential (accuracy = 0.93). Extracting and binary transforming feature contributions from random forest models indicated a metabolic potential threshold of 2.9 for a 0.5 probability of occurrence. Future projections predict a range contraction for C. laticeps due to decreased oxygen in the west and increasing temperatures in the east and their combined effect on metabolic potential. The metabolic potential threshold is in accordance with previous studies and extends the utility of this index by showing how it can limit either edge of a marine fish’s distribution. Furthermore, model predictions are in accordance with the “coastal squeeze” theory where warm and cool temperate biota are predicted to suffer range contractions in South Africa, although productive fishing areas for C. laticeps are predicted to persist.
Biography: To be confirmed