Removing the light history signal from normalized variable fluorescence ( Fv / Fm ) measurements on marine phytoplankton

Variable fluorescence normalized to maximal fluorescence, Fv/Fm, determined by Fast Repetition Rate Fluorometry (FRRF) is being increasingly used to compare photosynthetic electron transport capacity in natural phytoplankton communities. Interpreting results of such studies is, however, complicated by the fact that both nutrient status and light history (photoinhibition under in situ conditions) are known to influence Fv/Fm. Thus, the value of Fv/Fm measurements in the field would be greatly enhanced if the light history signal could be separated from other influences. Here, both field and laboratory studies demonstrate that dark treatment (30 min–4 h) before FRRF measurement is not sufficient to remove a light history signal in Fv/Fm. The signal could, however, be essentially eliminated by incubation of samples in low light prior to Fv/Fm determination. For the study conditions tested, the most effective treatment for removal of the signal was 4 h at 50 μmol m–2 s–1. However, the effectiveness of the light treatment in removing the light history was influenced by temperature. Therefore, no universal protocol for eliminating the light history signal can be developed, but recommendations are given for developing site-specific approaches for separating the light history signal from other factors influencing Fv/Fm. Carrying out light incubations before determining Fv/Fm not only provides the possibility for eliminating a light history signal in the measurements but the difference between Fv/Fm measured after light and dark incubations appears also to be a potentially useful indicator of the degree of photoinhibition experienced by phytoplankton under natural conditions. *Corresponding author: E-mail: [email protected] Acknowledgments Support for this study was received through grants from 1) the Danish Council for Independent Research/Natural Sciences, 2) the Danish National Research Foundation to the Center for Macroecology, Evolution and Climate at the University of Copenhagen, 3) the VILLUM Center of Excellence “Plant Plasticity,” and 4) the “Center of Synthetic Biology” funded by the UNIK research initiative of the Danish Ministry of Science, Technology, and Innovation. The field studies were carried out on the RV KNORR (Woods Hole Oceanographic Institution) and the RV DANA (Technical University of Denmark). We thank Niels Daugbjerg for help in facilitating the laboratory experiments, and two anonymous reviewers for helpful comments. DOI 10.4319/lom.2014.12.776 Limnol. Oceanogr.: Methods 12, 2014, 776–783 © 2014, by the American Society of Limnology and Oceanography, Inc. LIMNOLOGY and OCEANOGRAPHY: METHODS size distributions, Fv/Fm becomes the most likely parameter determined through FRRF to have the potential to interrogate natural phytoplankton populations with respect to their nutritional status and to identify hydrographically or otherwise segregated populations in situ (e.g. Guidi et al. 2012; Richardson et al. 2014; Rynearson et al. 2013). Variable fluorescence normalized to maximum fluorescence, Fv/Fm is determined following exposure of samples to darkness and is assumed to represent the maximum quantum efficiency of PSII photochemistry, i.e., the potential photochemical efficiency of open RCIIs (Baker 2008). That samples have been exposed to darkness before determination of Fv/Fm does not, however, necessarily imply that incident light conditions at the time of sample collection do not influence the resulting Fv/Fm determination. Indeed, Vassiliev et al. (1994) observed a decrease in Fv/Fm in dark-adapted phytoplankton samples taken from environments where the light intensity was > ~ 800 μmol m–2 s–1 compared with samples from lower light intensities, and on a 2008 cruise studying the spring bloom in the North Atlantic, we identified a clear effect of light at the time of collection in Fv/Fm determinations made on samples taken from the surface layer, where the Fv/Fm ratios decreased significantly with increasing light intensity (Fig.!1) despite a 30-min dark adaptation of the samples before measurement. The value of Fv/Fm as a diagnostic tool for identifying nutrient depletion (see Beardall et al. 2001) becomes greatly diminished in field studies when its measurement is influenced by the light climate (time of day or depth) of the sampling. Even under light-controlled laboratory conditions, its value as a comparative tool will be limited when different species are being examined as individual species have different responses to light climate. Thus, it is surprising that attempts to interpret Fv/Fm data from natural environments have not given greater consideration to the potential influence of light climate on the interpretation of the results and/or the possibility of eliminating the light history signal in Fv/Fm measurements, thereby simplifying the interpretation of identified differences between populations. The purpose of this study was to examine how widespread a light history effect may be on determinations of Fv/Fm on natural populations and to investigate possible treatments for reducing or eliminating the signal relating to light history. Materials and procedures Field studies We realized that a light history signal in Fv/Fm was affecting our measurements of Fv/Fm even at the comparatively low light intensities experienced in the North Atlantic (Fig. 1) while sampling at the onset of the spring bloom with the RV KNORR (Woods Hole Oceanographic Institution) from 2-20 May 2008 (NAB 2008 project http://www.bco-dmo.org/project/2098). On that cruise, triplicate sub-samples were collected for Fv/Fm determination from Niskin bottles sampling at selected depths in the euphotic zone. These samples were kept in dark bottles covered with aluminum foil to prevent light exposure and incubated for ≥ 30 min in an insulated box. Fv/Fm was measured within the dark chamber of a FASTtracka II fluorometer (Chelsea Instruments Group Ltd.) mounted and secured in the lab, using FASTtracka II firmware. A single turnover protocol with 30 sequences per acquisition, each including 100 saturation and 50 relaxation flashlets, was utilized. The sequence interval was set to 100 ms; the PMT eht (extra high tension) and LED light source (excitation peak of 470 nm) were optimized for each sample. Fv/Fm was calculated from saturation phase fits following (Kolber et al. 1998). The developed procedure for removal of the light history signal (see below) was tested in the field on a cruise in the Sargasso Sea on RV DANA (Technical University of Denmark) in March-April 2014. In this case, two sub-samples were taken from Niskin bottles from 10 m. Two measurements were made on each sub-sample, thus yielding 4 measurements per study depth. The collected samples were stored either in 250 mL brown glass bottles in ambient laboratory light (approximately in situ temperature) for a minimum of 30 min before measurement or in 1 L clear glass Millipore Bluecap bottles exposed to cool white fluorescent light (approx. 50 μmol m–2 s–1). The same instrument and overall procedure was used for determining Fv/Fm as on the KNORR cruise. The software used was, however, upgraded to FASTpro Version 2.5.3. A single turnover protocol with 12 sequences per acquisition, each including 100 saturation flashlets, was utilized. The sequence interval was set to 100 ms. In both field studies, the average PAR recorded by the rosette-mounted light sensor at the depth and during the time of the sample collection (rosette lowered with CTD) in the Seabird CTD “bottle cast” file was taken to represent in situ light condition of the sample. From et al. Light history signal in Fv/Fm 777 Fig. 1. Fv/Fm determined on natural samples collected at various depths in the water column in the North Atlantic during the period 2-20 May 2008 plotted against the light intensity at the time and depth of sample collection. Samples were dark incubated for 30 min before Fv/Fm determination. Laboratory studies Cultures of four different marine phytoplankton species were obtained from the Scandinavian Culture Collection for Algae and Protozoa (SCCAP) housed at the University of Copenhagen. These included two diatoms: Chaetoceros socialis (K-0550) and Attheya longicornis (K-1530), a chlorophyte: Brachiomonas cf. submarina (K-0582) and a dinoflagellate: Heterocapsa triquetra (K-0447). These were maintained in culture medium L1 (http://www.sccap.dk/media/marine/2.asp) at 15°C and day/night regulated illumination of ~100 and 0 μmol m–2 s–1 for 16 h and 8 h, respectively. Each culture was examined individually for a light history signal in measurements of variable fluorescence. Culture material was first diluted with L1 media to a degree where chlorophyll concentrations resembled natural field concentrations. Before determination of Fv/Fm (employing the same instrument and protocol as those used on the KNORR cruise), samples of this diluted culture material were incubated in 60 mL plastic (NUNC) culture bottles for 2 h at 14 different light intensities in a temperature-controlled (15°C) incubator. Samples were continuously rotated during incubation to maintain phytoplankton in suspension. Subsequent investigations of the effect of various light/dark treatments and temperature on the presence of a light history signal in Fv/Fm determinations were carried out using the same incubation procedures using either Attheya longicornis or Chaetoceros socialis as the test organism. Possible “bottle effects” impacting Fv/Fm were investigated by incubating A. longicornis in NUNC bottles under the standard culture conditions (100 μmol m–2 s–1, 15°C) and taking samples for determination of Fv/Fm at 10 different time intervals between 0 and 270 min. The possible influence of temperature on the effectiveness of the developed light treatment to remove the light history signal was tested for by incubating C. socialis at 14 different light intensities (15°C) and subsequently comparing the effectiveness of the treatment at 5 and 25°C. Finally, the method developed to remove the light history signal in laboratory cultures was tested on natural samples. In the first of these tests, surface water samples collected from Langelinie Pier in Copenhagen during the period 25–28 Mar 2011 were incubated at 14 different light intensities as described above. During the post-incubation period and before determination of Fv/Fm, these were subjected either to darkness (30 min or 2 h) or light (2 h at 100 or 4 h at 50 μmol m–2 s–1). In the second test of natural communities, samples were taken directly from in situ conditions (10 m and not incubated) in the Sargasso Sea and held either in darkness or light (4 h, ~50 μmol m–2 s–1; further detail given under “Field studies”). Data analysis Analyses were done in the free and open source statistical software R version 3.1.0 (R Core Team 2014). For samples from each incubation light intensity, the mean Fv/Fm was calculated, and the effect of increasing incubation light intensity was investigated by regressing the mean Fv/Fm values on light intensity using ordinary least square estimation. The specific light intensity where a light history signal became apparent was determined as the intercept between the linear regression of the decreasing curve and the horizontal line where no light history was detected. The differences in the effect treatment on removal of the light history were quantified by calculating the percentage change in Fv/Fm from PAR = 0 to PAR = 300 based on the regression slopes for each experiment.