Phytoplankton patchiness in Winam Gulf, Lake Victoria: a study using principal component analysis of in situ fluorescent excitation spectra

Ryan Alexander, Jorg Imberger

    Research output: Contribution to journalArticlepeer-review

    11 Citations (Scopus)

    Abstract

    1.In order to characterise phytoplankton patchiness at fine scales, a profiling multiwavelength fluorometer was cast at numerous locations throughout Winam Gulf in Lake Victoria to measure fluorescent excitation spectra, which are indicators of both phytoplankton diversity and coloured dissolved organic matter (CDOM). 2.Processing the spectral data with principal component analysis (PCA) revealed that linear combinations of four fundamental 'base' spectra could explain almost all of the variation in spectral measurements. Three of the base spectra were associated with spatially distinct patches of phytoplankton containing different species assemblages, while the fourth base spectrum was due to CDOM fluorescence. 3.The locations of the phytoplankton patches were traced to the south-east of Winam Gulf, the western end of the Rusinga Channel and the open waters of Lake Victoria adjacent to Winam Gulf, respectively. The high CDOM fluorescence was traced mainly to relatively deep water in the Rusinga Channel. 4.The phytoplankton and CDOM patchiness were interpreted in the context of physical and chemical gradients that were measured at the site at the same scale as the spectral data. Strong relationships were found between the gradients in spectral data and other environmental variables, which suggested several underlying explanations for the phytoplankton and CDOM patchiness. © 2012 Blackwell Publishing Ltd.
    Original languageEnglish
    Pages (from-to)275-291
    JournalFreshwater Biology
    Volume58
    Issue number2
    DOIs
    Publication statusPublished - 2013

    Fingerprint

    Dive into the research topics of 'Phytoplankton patchiness in Winam Gulf, Lake Victoria: a study using principal component analysis of in situ fluorescent excitation spectra'. Together they form a unique fingerprint.

    Cite this