Using Remotely Sensed Data to Describe Spatial and Temporal Habitat Distributions of the Giant Kelp, Macrocystis Pyrifera

Abstract

Beds of giant kelp, Macrocystis pyrifera , are common features of shallow rocky bottom habitats along the coast of California. Giant kelp functions as an important marine habitat, supporting an abundance and diversity of life and is a harvested resource. Long-term observations have shown patterns of decline in kelp bed biomass over the last several decades. Environmental fluctuations occur concurrently over many space and time scales and these influence the stability of the kelp ecosystem. In light of the variability inherent in the nearshore system, the determination of giant kelp habitat distribution requires an understanding of the processes and factors driving the species' habitat distributions in space and time. We examine the role of time in giant kelp habitat distributions by developing a biophysical model that utilizes remotely sensed data, along with other variables, to describe optimal habitat for giant kelp. The study area includes the northern Channel Islands, where a joint state and federal process has been underway to consider marine reserves, with a system of reserves designated in state waters in October 2002. The model is based on habitat preferences of giant kelp (optimal habitat descriptors) and consists of a set of rules generated from observable environmental parameters (such as sea surface temperature, wave exposure, bathymetry, and bottom type). The model generates in a time-series of optimal giant kelp habitat composites on seasonal to interannual time scales, which will be used to quantify persistence of kelp habitat. Current work at UCSB has focused on assessing the temporal and spatial variability of the marine environment in the Santa Barbara Channel using available satellite data and is used here to model kelp optimal habitat. Preliminary results indicate a seasonal cycle of optimal habitat area, marked by large fluctuations in area between summer and winter for most of the study area. Optimal habitat area in coastal waters south of Point Conception appears to be more stable than north of Point Conception and around the Channel Islands. Discrepancies between matchups of optimal habitat and realized habitat, generated from aerial surveys of kelp cover, indicate, in part, that optimal habitat descriptors, do not account for previous months' conditions. Future versions of the biophysical model need to incorporate temporal correlation in the designation of optimal habitat by relaxing the assumption of zero serial correlation during windows of suboptimal conditions.