Accounting for spatiotemporal dynamics in conservation planning for coastal fish in KwaZulu-Natal, South Africa

Abstract Systematic conservation planning methods to design marine protected area (MPA) networks can provide more meaningful results by addressing the spatiotemporal variability of biota, and by using these data to inform target assignment. This study used Marxan software to design candidate MPA networks to meet conservation targets for 67 coastal fish species on the east coast of South Africa. Species were selected for conservation importance, and included both resident and seasonal migrants. The distribution range of three phases of a species life cycle was generated using either cartographic habitat range models or maximum entropy models, and each used as a separate conservation feature. Two sets of conservation features were developed from this: A static set of 77 distribution models for features which ignored seasonal dynamics, and a seasonal set of 147 distribution models which included seasonal dynamics. Conservation targets depended on a species' extinction vulnerability and its seasonal abundance. Three scenarios were used to test the effects of incorporating seasonal spatial and abundance dynamics into MPA design: Scenario 1 tested the effect of using static or seasonal distribution data; Scenario 2 tested the additional effect of adjusting conservation targets based on seasonal variations in abundance; and Scenario 3 tested the additional effect of incorporating existing MPAs into the MPA network. In all three scenarios, the spatial configuration of MPA networks differed between the two datasets (Kappa 0.37, 0.25, 0.3), and static-designs did not fully meet targets for a number of species or critical life cycle phases of some species, however, larger and more expensive areas were required to design MPAs that could meet all conservation targets for seasonal features. Seasonal abundance adjusted targets was useful to elevate the prioritisation of seasonally abundant migratory species. Including existing MPAs did not change the differences observed between static and seasonal outcomes. We believe this will be true for any marine system that demonstrates seasonal spatial life-history differentiation and abundance dynamics, and advocate its use while giving due consideration to the increased cost associated with spatiotemporal planning.