Challenges of remote sensing for quantifying changes in large complex seagrass environments

Abstract

Managing seagrass environments and understanding and responding to coastal impacts such as floods or cyclones, requires assessment of seagrass distribution and its biophysical properties in time and space. Comparable assessments of seagrass distribution over time are often lacking as the information is present for separate dates, or created following different mapping approaches, and this makes it difficult to conduct quantitative comparisons. We provide an assessment of available data sets and approaches, and their suitability for monitoring and quantifying change in seagrass percentage cover and extent for a large coastal embayment (Moreton Bay, Australia, 1582 km2). Seagrass percentage cover maps were created for 2011 and 2004 and compared to map and measure the extent of seagrass percentage cover change, and changes in the extent of seagrass environments. Total extent of seagrass was shown to be higher in 2011 compared to 2004. Potential sources of these differences may be: mapping inaccuracy; actual change in extent and cover; and, monthly to seasonal variations in seagrass cover. A qualitative comparison of the 2004 and 2011 maps was performed against maps of seagrass extent maps from 1975, 1986 and 1999, which were created using a range of different methods and data. The output maps show changes in seagrass extent, but a lack of detail arising from variable mapping methods and differing mapping extents prevented a reliable comparison. We conclude that robust mapping of seasonal and inter-annual variation in seagrass percentage cover distribution or extent, as well as impacts of episodic and stochastic disturbance events, requires a thorough understanding of the mapping approaches used so that data sets can be compared. Additional complimentary information is required to help understand the drivers of changes.