Paper Review by Dr. Wayne Slade
Use of the LISST Instruments for Studies of Phytoplankton and their Dynamics
This month’s paper is an oldie but goodie, Karp-Boss et al. 2007. (And I’m not saying that because I graduated from this lab!). It’s a popular paper, cited by about 60 papers according to Web of Science.
This paper presents an interesting lab effort to examine the use of the submersible LISST-100 instrument for bench-top measurements of phytoplankton cells (including chains) from cultures, also comparing with microscopic analysis of cell size. This study was one of the first to look at application of the LISST instrument to biological studies. Previous work included measurement of phytoplankton and purple sulphur bacteria in lakes by Serra et al. [1], that found the capability of the LISST for making in situ measurements of undisturbed samples advantageous over laboratory-based systems. However, Karp-Boss et al. were the first to systematically look at the effects of cell shape on LISST estimates of particle size.
Phytoplankton cells come in a wide range of sizes and shapes, with complex internal structure, and absorbing pigments. Karp-Boss et al. used 8 different species in their comparison: 3 dinoflagellates roughly spherical in shape (aspect ratios near 1), 2 diatoms with aspect ratios greater than 1, 2 chain-forming diatoms, and a more complex-shaped dinoflagellate. For the 5 species with simple shapes, agreement between LISST and microscopy was excellent (mode ESD average error of ~4%, note 1), with the 2 species with aspect ratio larger than 1 exhibiting a wider size range. Furthermore, for these species, estimates of cell concentration derived from the LISST were within an order of magnitude of those determined with the far more time-consuming microscopic counting. Agreement between methods when the cells are significantly non-spherical (e.g., chains and complex shaped) is not as good, as might be expected since characterizing the size of non-spherical particles can be ambiguous. The parameter compared, ESD (equivalent spherical diameter) is very limited in this case. Interestingly though, for the more complex particles, LISST size distributions were multi-modal, with peaks appearing to be associated with different size characteristics of the cells. For example, for Ceratium longipes, the three modes were consistent with an imaginary spheroid enclosing the entire cell, the body without spines, and the width of the spines (see figure).
Karp-Boss et al. also discuss the use of use of the LISST instrument in laboratory and field studies for studying plankton, pointing out its versatility–that the same system can be used from a ship, on a mooring, or in the lab. As an example of the lab application, they present data over a six-month culture experiment where they are able to clearly detect changes in cell size and as the culture transitions from a vegetative to sexual growth stage. Field measurements generally consist of more complex mixtures of multiple species and other organic and inorganic particles. The authors present an example of a culture mixture where clear separation between species can be seen, and also note that in many cases algal blooms of interest (such as HABs) are more nearly mono-specific, suggesting the LISST can be useful for studying algal dynamics in the field as well as in the lab. As an example, they point to a contemporary study [2] that was able to use LISST data to discern a non-harmful dinoflagellate bloom and suggested that the LISST measurement could provide a means for rapid assessment of HAB threats. Since 2007 when Karp-Boss et al.’s work was published using an original LISST-100, Sequoia has released the more advanced LISST-200X instrument, with major improvements to usability, significantly reduced size and weight, and extended size range (1-500 µm, covering both the Type-B and Type-C ranges of the original LISST-100 and LISST-100X).
For more information on the LISST-HAB system that combines LISST-200X particle sizing with multi-parameter fluorescence, click here.
Note 1: Error calculated between the average of the ESDs (ESDc and ESDv) derived from microscopy and the geometric mean of the range of the mode derived from the LISST (ref. their Table 1).