Spectral Deconvolution for Dimension Reduction and Differentiation of Seagrasses: Case Study of Gulf St. Vincent, South Australia

Abstract

Seagrasses are a vulnerable and declining coastal habitat, which provide shelter and substrate for aquatic microbiota, invertebrates, and fishes. More accurate mapping of seagrasses is imperative for their sustainability but is hindered by the lack of data on reflectance spectra representing the optical signatures of individual species. Objectives of this study are: (1) To determine distinct characteristics of spectral profiles for sand versus three temperate seagrasses (Posidonia, Amphibolis, and Heterozostera); (2) to evaluate the most efficient derivative analysis method of spectral reflectance profiles for determining benthic types; and to assess the influences of (3) site location and (4) the water column on spectral responses. Results show that 566:689 and 566:600 bandwidth ratios are useful in separating seagrasses from sand and from detritus and algae, respectively; first-derivative reflectance spectra generally is the most efficient method, especially with deconvolution analyses further helping to reveal and isolate 11 key wavelength dimensions; and differences between sites and water column composition, which can include suspended particulate matter, both have no effect on endmembers. These findings helped develop a spectral reflectance library that can be used as an endmember reference for remote sensing, thereby providing continued monitoring, assessment, and management of seagrasses.

Effects of Epiphytes and Depth on Seagrass Spectral Profiles: Case Study of Gulf St. Vincent, South Australia

“…These findings show that the correlation values of first-derivatives with natural growing depth were more strongly negatively correlated than with the zero-derivative profiles, which was similar to other studies [9,58]. Highly correlated features, negative or positive, have been known to help identify [20,37,39,43,74] and potentially eliminate redundant and numerous feature selections for hyperspectral image classification [30,70,71,75], although correlation itself has been noted to help identify only associations rather than causation [9,58,76,77]. Derivative analysis merely finds and sharpens details in spectral curves that are too subtle to notice and hidden within wide zero-order spectrum bands, thereby highlighting features that were already present in direct spectrophotometry [41,78].…”

“…The reflectance profiles from the leaf blades of seagrasses, where epiphytes are most frequently attached to the leaf surface, were shown to have significantly lower values than those from the leaf sheath ( p < 0.005). This effect also significantly influenced the reflectance values at 566 nm and 600 nm ( p < 0.001 for both), thereby potentially weakening the value of the 566:600 band ratio that is used to separate seagrasses from detritus and algae [28,38,70,71]. Epiphytes are known to mask the reflectance of aquatic plant surfaces, which causes altered varying degrees of chl- a absorbance [14].…”

“…The results in this study show significant dampening effects of epiphytes attaching directly to the seagrass surface at the leaf blades. At the same time, other studies have shown that measuring the reflectance in filtered and unfiltered marine water has no significant differences throughout most of the visible wavelength spectrum [70,71], suggesting that materials in the water column―including phytoplankton, sediment, minerals, and dissolved organic substances [35,55,72,73]—have less impact on the detection of seagrasses than compared to epiphyte presence on the leaf surface. In spite of the strong dampening influence on the detection of seagrasses, the results in this study show that there is one particular bandwidth at 670–679 nm that disregards the effects of epiphytes.…”

“…Beyond this, the 670–679 nm bandwidth discovered in this study can be doubly beneficial for future assessments of seagrasses. Previous research found eleven bandwidths (417, 456, 474, 491, 522, 590, 605, 621, 631, 649, and 681 nm) useful for the differentiation between Posidonia , Amphibolis , and Heterozostera seagrasses [70,71]. Therefore, because this bandwidth falls in the range of one of these eleven bandwidths, isolating and evaluating the reflectance data at approximately 680 nm in a 5 nm increment could be pragmatic and efficient for two reasons.…”

Selected Papers From 2019 IEEE Eurasia Conference on Biomedical Engineering, Healthcare and Sustainability (IEEE ECBIOS 2019)

“…Hwang et al reported "Spectral Deconvolution for Dimension Reduction and Differentiation of Seagrasses: Case Study of Gulf St. Vincent, South Australia" [11]. The objectives of this study were: (1) To determine distinct characteristics of spectral profiles for sand versus three temperate seagrasses (Posidonia, Amphibolis, and Heterozostera); (2) to evaluate the most efficient derivative analysis method of spectral reflectance profiles for determining benthic types; and to assess the influences of (3) site location and (4) the water column on spectral responses.…”