McKinna, Lachlan I.W. (2010) Optical detection and quantification of Trichodesmium spp. within the Great Barrier Reef. PhD thesis, James Cook University.

Preview

PDF (Thesis) Download (10MB) | Preview

Abstract

The primary purpose of this PhD project was the development of suitable methods for the optical detection and quantification of the diazotrophic, marine cyanobacteria Trichodesmium within the Great Barrier Reef (GBR), Australia. Within the GBR, Trichodesmium is likely to contribute quantities of new-nitrogen of similar magnitude to that of rivers. However, due to uncertainties regarding the spatial and temporal abundance of Trichodesmium, there is an order of magnitude uncertainty associated with these nitrogen fixation estimates. Thus, improved methods for quantifying Trichodesmium within the GBR are essential. The key objectives of this PhD thesis were to: 1. Study the bio-optical properties of Trichodesmium,

2. Develop a binary flag for its detection using MODIS imagery and

3. Examine hyperspectral radiometric data as a means of positively discriminating and quantifying Trichodesmium.

In addition, the bio-optical properties of a senescing surface aggregation of Trichodesmium were studied.

Within this PhD thesis, the bio-optical properties of Trichodesmium were studied primarily with discrete water samples analysed using a benchtop spectrophotometer. Particulate and coloured dissolved organic matter (CDOM) absorption coefficients were measured. From this research component, a relationship between the magnitude of the spectral absorption coefficient of Trichodesmium and chlorophyll-a (Chla) specific concentration was established. Results were comparable with those of the literature. The Chla-specific Trichodesmium absorption coefficients were later used as inputs for radiative transfer simulations with Hydrolight. In situ above-water hyperspectral radiometric measurements of Trichodesmium were also collected.

A Trichodesmium-specific binary classification algorithm was developed using quasi-250 m MODIS data. Above-water hyperspectral radiometric measurements of dense Trichodesmium surface aggregations (> 30 mg Chla m⁻³) showed that the water leaving radiance at wavelengths greater than 700 nm were ii much higher in magnitude (> 0.05 W m⁻² sr⁻¹) relative to the visible wavelengths 400 - 700 nm (< 0.03 W m⁻² sr⁻¹). This "red-edge" effect agreed with observations of others from the literature. The binary classification algorithm was based on three criteria. The first criteria relied on the difference in magnitude between the MODIS normalised water-leaving radiance (nLw) of band 2 (859 nm) and band 15 (678 nm). The magnitudes of the nLw of band 4 (555 nm) and band 1 (645 nm) relative to band 15 formed the second and third criteria respectively. The classification algorithm was tested on a small subset of 13 MODIS images with corresponding Trichodesmium sea-truths and yielded an 85 % accuracy. Fine scale features consistent with dense Trichodesmium surface aggregations such as eddy swirls and windrows were well represented within the algorithm results. The algorithm was also found to be robust in the presence of highly reflective, potentially confounding affects such as coral reefs, shallow bathymetry and riverine sediment plumes. The suitability of the quasi-analytical algorithm (QAA) for inverting hyperspectral remote sensing reflectance, R(rs)(λ), and quantitatively discriminating Trichodesmium was examined. A technique combining the QAA and a similarity index measure (SIM) was developed using R(rs)(λ)data simulated for examples of Case 1 and Case 2 waters. Hydrolight radiative transfer software was used to model R(rs)(λ)with Trichodesmium Chla specific absorption inherent optical properties. The QAA was used to invert the simulated R(rs)(λ) spectra to yield an estimate of the phytoplankton absorption coefficient a^QAA(0)(λ). To ascertain the presence of Trichodesmium, seven SIM values were derived by comparing a^QAA(0)(λ)with a known Trichodesmium reference absorption spectrum a^ref (tri)(λ), and also with the absorption spectra of six other phytoplankton types. The results found that the SIM could discriminate Trichodesmium from the six other phytoplankton types for concentrations as low as 0.2 mg Chla m⁻³ and 3 mg Chla m⁻³ for the Case 1 and Case 2 scenarios considered. The QAA-SIM method was tested on along-transect R(rs)(λ)data collected within the GBR. Upon identifying the presence of Trichodesmium, the magnitude of a^QAA(0)(λ)was used to determine Chla concentration. The along-transect, QAA derived Chla values were validated with data from a Chla fluorometer within a iii ship-board flow-through system. The predicted Chla values matched well with those fluorometrically measured yielding an R-squared value of 0.805.

Two distinct colour modes of Trichodesmium were sampled from a dense surface aggregation within the GBR. The two colour modes were denoted as: orangebrown (OB) and bright green (BG). The spectral particulate and coloured dissolved organic matter (CDOM) absorption coefficients were measured for the OB and BG samples. The absorption properties of the OB sample were consistent with those of Trichodesmium reported within literature. However, the absorption properties of the BG sample were significantly different to those of the OB sample. The particulate and dissolved absorption coefficients of the BG sample revealed that the water soluble red pigments phycourobilin (PUB) and phycoerythrobilin (PEB) had leached into the surrounding seawater. The results suggest that the BG samples were in the process of senescence. Hydrolight radiative transfer modelling was used to simulate the hyperspectral R(rs)(λ)of OB and BG colour modes. The results indicated that the R(rs)(λ)spectra of the OB sample was spectrally distinct from that of the BG sample. Thus, the potential to optically discriminate the physiological state of a Trichodesmium surface aggregation was established.

Item ID:	29747
Item Type:	Thesis (PhD)
Keywords:	bio-optics; ocean colour;remote sensing; cyanobacteria; nitrogen fixation; Trichodesmium; Great Barrier Reef (GBR)
Related URLs:	http://eprints.jcu.edu.au/15643/
Additional Information:	Publications arising from this thesis are available from the Related URLs field. The publications are: Chapter 3. McKinna, Lachlan I.W., Furnas, Miles J., and Ridd, Peter V. (2011) A simple, binary classification algorithm for the detection of Trichodesmium spp. within the Great Barrier Reef using MODIS imagery. Limnology and Oceanography: Methods, 9. pp. 50-66.
Date Deposited:	11 Oct 2013 01:09
FoR Codes:	02 PHYSICAL SCIENCES > 0299 Other Physical Sciences > 029901 Biological Physics @ 34% 01 MATHEMATICAL SCIENCES > 0199 Other Mathematical Sciences > 019999 Mathematical Sciences not elsewhere classified @ 33% 04 EARTH SCIENCES > 0405 Oceanography > 040501 Biological Oceanography @ 33%
SEO Codes:	97 EXPANDING KNOWLEDGE > 970105 Expanding Knowledge in the Environmental Sciences @ 34% 97 EXPANDING KNOWLEDGE > 970102 Expanding Knowledge in the Physical Sciences @ 33% 97 EXPANDING KNOWLEDGE > 970106 Expanding Knowledge in the Biological Sciences @ 33%
Downloads:	Total: 320 Last 12 Months: 10
	More Statistics