Marine & Environmental Sciences Faculty Articles

Reproducible multi-parameter acoustic detection of seasonal drift macroalgae in the Indian River Lagoon, Florida

ORCID

0000-0002-6003-9324

ResearcherID

F-8807-2011

Document Type

Article

Publication Title

Journal of Coastal Conservation

ISSN

14000350

Publication Date

8-1-2018

Keywords

Acoustic ground discrimination, Biosonics, Estuaries, Nuisance bloom, Nutrients, Remote sensing, SJRWMD, Superbloom

Abstract

Biennial single-beam dual-frequency acoustic surveys of Indian River Lagoon, a coastal Florida lagoonal system, were performed to quantify the spatial distribution of the annual peak of seasonal drift macroalgae (DMA) as part of the Surface Water Improvement Program (SWIM) enacted in 1986. The ultimate goal of SWIM was to reclaim historical seagrass acreage. Dense accumulations of DMA can ultimately shade seagrass beds and create anoxic zones via decomposition, especially during warm summer months, thus contributing to the internal nutrient pool. A large portion, approximately 140 km of the Indian and Banana Rivers, were surveyed at 200 m line spacing within the timeframe of peak DMA biomass (April through June). A supervised training catalog of 180 60-s acoustic and video samples was grouped into classes of bare substrate, short (canopy) submerged aquatic vegetation and DMA and segmented by discriminant analysis using a combined 38 and 418 kHz dataset. Percent cover of DMA was binned into categories of 0–33, 33–66 and 66–100%. Consistent classification of training catalog samples collected in 2007, 2008 and 2010 validated the temporal consistency of BioSonics digital transducers, meaning the previous surveys training dataset can be used to produce a preliminary map with only a few days lag-time, and that year-to-year acoustically-derived DMA abundances can be directly compared without bias adjustment. DMA biomass within the 288 km2 survey area was 69,000 and 102,000 metric tons wet weight in 2008 and 2010, respectively with an overall classification accuracy of 78.9 and 83.2%, respectively. Four large-scale deposits were detected in the same locations during both surveys but the majority of biomass came from widespread sparse deposits. DMA was found to decompose within the IRL rather than exiting through Sebastian Inlet. The high abundance of DMA in 2010 and subsequent release of nutrients within the IRL was vital for understanding the 2011 superbloom.

DOI

10.1007/s11852-018-0593-1

Volume

22

Issue

4

First Page

587

Last Page

604

Find in your library

Share

COinS