Variability in spectral and nonspectral measurements of photosynthetic light utilization efficiencies

Abstract

The aim of this study was to quantify the variability in and differences between spectral and nonspectral measurements: of light utilization efficiencies for natural phytoplankton communities, in order to evaluate possible consequences for blo-opt~cal models of in sj t~l p]-1ma1-y production (P). Field samples were collected at 4 coastal stations durlng a 1 d transect (July 23, 1988) across the Southern California Counter ('urrent (SCCC) at a time when phytoplankton communities were dominated largely by picoplankton (0 4 to 5 pm). Concurrent determlnations were made of down~ve l l~ng spectral ~rradiance [QCl(A,z)], photosynthetically available radiation (Qpclr), spectral attenuation coefficients [&(A,z)], detrital-corrected phytoplankton absorption [a,,,,(A,z)], white light photosynthesis-irradiance parameters (P,,,,. a, I*) and carbon action spectra a(AA,,z). From these parameters, spectral estimates of in situ phytoplankton absorbed radiation [AQ,,h(AA,.z)l, maximum quantum yield [d,,,.(AA,.z)], operational quantum yield [&(AA,,z)l, radiation utilization efficiency [c(AA,,z)] and productivity P(AA,,z) were derived. Significant spatial variability in all bio-optical parameters was noted for communities in the surface waters and the chlorophyll maximum. For surface waters, there was significant variability in the 525 to 600 nm region of the spectral signatures of dl[dA,,z) and P(AA,,z) which was attributable to phycobilins not resolved In absorption spectra. The importance of light absorption by photosynthetic pigments other than chlorophyll a (chl), and thcir assoc~ated Impact upon absorption-based production parameters, increased w ~ t h light depth. One consequence was a close correspondence between AQ,, and P(AA,,z) at depth w h ~ c h was not evident in surface waters A second consequence was that while spectrally weighted and whlte l ~ g h t estimates of quantum yield were occasionally s im~lar In surface waters, spectral estimates for all chlorophyll maximum communities were 4to 6-fold higher than white light measurements. Results confirm previous observations that white light measures of quantum yield can significantly underestimate quantum yield for subsurface communities of phytoplankton (Prezelin e t al. 1989) and provide a conceptual base on which to improve existing and future spectral models of in situ photosynthetic quantum yield. INTRODUCTION to be placed on these other processes. The bio-optical models are based on the mechanistic linkages between Phytoplankton productivity occupies a central posiwater-column optical properties, penetration of phototion in several large-scale processes, including food synthetically available radiation (Q,,,), and phytoweb dynamics, biogeochemical cycles, particle flux, plankton production. Photosynthesis-irradiance (P-I) and bioluminescence. Bio-optical models can increase curves provide physiological information which can our ability to resolve the temporal/spatial variability in resolve environmental effects on algal photophysiophytoplankton production which will allow constraints logy and are used to provide empirical models describ0 Inter-Reseal-ch/Printed in Germany 017 1 -8630/91/0078/0253/$03.00 254 Mar. Ecol. Prog. Ser 78: 253-271, 1991 ing the variability in primary production (Platt & Jassby 1976, hlacCaull& Platt 1977, Cote & Platt 1984, Prezelin et al. 1986, 1987, 1989, 1991, Prezelin & Glover 1991). The P-I parameters are also key components in other sophisticated bio-optical models which attempt to predict changes in the rates of phytoplankton growth and photosynthesis (Shuter 1979, Laws & Bannister 1980, Kiefer & Mitchell 1983, Falkowski et al. 1985, Lewis et al. 1985, Geider et al. 1986, Bidigare et al. 1987, Platt & Sathyendranath 1988, Sakshaug et al. 1989, 1991, Sathyendranath et al. 1989, Smith et al. 1989b, Baker et al. 1990). These models generally require a term for the quantum yield of carbon fixation ( 4 ) which is the efficiency with which radiation absorbed by phytoplankton (AQph) is converted to photosynthate. Based on limited field information on the quantum yield many of these models have been developed by assuming cb is independent of wavelength, near the theoretical maximum (0.125 m01 C Ein-') and/or has a predictable light-dependent relationship to the maximum quantum yield ($,,). Recent studies indicate that such generalities cannot be universally applied without significantly affecting the predictive accuracy of the productivity models (Smith et al. 198913, Schofield et al. 1990). Spectral production models (Lewis et al. 1985, Bidigare et al. 1987, Sathyendranath et al. 1987, Smith et al. 1989b, Baker et al. 1990) recognize the wavelength dependence of photosynthesis and that the bio-optical properties of phytoplankton are adaptable to the variability in the underwater light fields (cf. Prezelin & Boczar 1986). The models are based on the observation (Kirk 1983) that primary production (P) at any given depth (2) depends on the underwater light field, its absorption by phytoplankton, and the efficiency with which this absorbed radiation (AQph) is utilized to fix carbon. (C) . Thus Taking into account the spectral dependence of AQph(z) the accuracy of the bio-optical models signiflcantly increased (Smith et al. 1989b). By also taking into account the spectral dependence in 4(z) the accuracy of the productivity models will increase and may offer more general applications then empirical and nonspectral models. Our interests in rnodeling the mechanisms with which diverse phytoplankton utilize radiant energy, regulate rates of photosynthesis, and influence the optical ch.aracten.stics of the water column prompted a multidisciplinary cruise (Watercolors '88). The purpose of Watercolors '88 was to assess the linkages between ocean optics and wavelength-dependent photosynthesis. The study presented here is part of that field exercise and was designed to quantify natural variability in and between spectral and nonspectral measurements of the maximum quantum yield. The study outlines the linkages between spectral and nonspectral production parameters, demonstrates that shifts in phytoplankton community composition and/or spectral irradiance has a large effect on the spectral signature of light utilization parameters, confirms previous observations that 'white light' measures of quantum yield can significantly underestimate quantum yield for subsurface communities of phytoplankton (Prezelin et al. 1989) and provides a conceptual base on which to improve existing and future spectral models of in situ photosynthetic quantum yield. MATERIALS AND METHODS Physical/chemical measurements and sample collection. The Watercolors '88 program was conducted in July-August 1988 aboard the RV 'Melville'. Using procedures previously described (Smith et al. 1987), a total of 129 vertical proflles of bio-optical properties were completed during repeated transects of a highly variable region of the Southern California Bight (SCB) (Fig. 1). Hydrographic analyses (Smith et al. 1990, Baker & Smith pers. comm.) led to the sorting of all vertical profiles into 12 groupings which had similar hydrographic signatures. The cluster of vertical casts that represent Station (Stn) L was most representative of California Current (CC) waters flowing from the north, while the stations at Stn K had hydrographic signatures that were intermediate between those CC waters and post-upwelling water subducting offshore at Stn J . Although groups A to I all appeared hydrographically related to the Southern California Counter Current (SCCC), there was some variability along the east-west transect line. The variability was most pronounced at Stn 1 centered over the Continental Shelf Break. At the time of the present study, Stn 12.01 showed the influence of SCCC in surface waters and the influence of postupwelling off Pt. Conception within the chlorophyll maximum (chl max). On July 23, 1988, a biological transect was made along a 150 km region of the SCCC. The cast numbers for the 4 transect stations (Stns 9.02, 10.02, 11.01, & 12.01) and their relationship to distinguishable water types (A to L) are shown in Fig 1. Vertical profiling for Stns 9.02 to 12.01 was carried out between dawn and dusk of a single day and sample times are listed in Table 1 The present study distin.guishes instantaneous measurements of bio-optical parameters, made at different stations at different times of day, from noon-time estimates of these parameters. The latter correct for known diurnal variations in light and productivity parameters in order to provide a time-independent compariSchofield et al.. Spectral photosynthesis, quantum yield, and radiat~on ut~lization efflclency 255