J. Phycol. 35, 42±53 (1999) PHOTOSYNTHESIS-IRRADIANCE PATTERNS IN BENTHIC MICROALGAE: VARIATIONS AS A FUNCTION OF ASSEMBLAGE THICKNESS AND COMMUNITY STRUCTURE Walter K. Dodds,2 Barry J. F. Biggs, and Rex L. Lowe3 National Institute of Water and Atmospheric Research Ltd., P.O. Box 8602, Christchurch, New Zealand Photosynthesis-irradiance (P-I) characteristics of Key index words: bio®lm; light; microphytobenthos; periphyton (microphytobenthos) have been consid- periphyton; photosynthesis-irradiance ered primarily for entire assemblages. How P-I re- sponses vary with mat thickness and with community Abbreviations: NPP, net primary production; P-I, composition has not been considered in detail. We photosynthesis-irradiance; Pmax, maximum photosyn- used a combined approach of modeling, microscale thetic rate; PPFD, photosynthetic photon ¯uence determinations of photosynthetic rate and light at- density; R, respiration tenuation, and whole-assemblage O2 ¯ux measure- ments to explore P-I relationships. The modeling Periphyton assemblages (algal bio®lms) can be re- approach suggested that the onset of photosynthetic sponsible for the majority of production in shallow, saturation and photoinhibition will occur at higher unshaded, aquatic habitats. Clear and unshaded irradiance and that whole-mat photoinhibition (de- streams, shallow lakes, wetlands, and shallow coastal creased photosynthesis at very high irradiance), bio- waters are all habitats where such production should mass-speci®c maximum photosynthetic rate, and ini- be important. The effects of irradiance on periphy- tial slope of the P-I function (a) should decrease as ton photosynthetic rates have been characterized for assemblage thickness increases or light attenuation only a few habitats (e.g. Hill 1996), despite the cen- increases. Spherical light microsensor pro®les for a tral role that benthic microalgae may play in clear variety of stream algae indicated a strongly com- shallow waters. In contrast, photosynthesis-irradi- pressed photic zone with attenuation coef®cients of ance (P-I) relationships have been documented for 21 70±1791 m for scalar photosynthetic photon ¯u- a wide variety of unialgal cultures, marine and fresh- ence density. The O2 microelectrode measurements water phytoplankton assemblages, and terrestrial showed little if any photoinhibition at 2 and 4 mm plants (EnrõÂquez et al. 1996). depths in one ®lamentous green algal (Ulothrix) as- Photosynthesis-irradiance relationships have four semblage, with a relatively low attenuation coef®- main descriptive features: the maximum photosyn- cient, and no photoinhibition in a second Ulothrix thetic rate (Pmax), the initial rate of increase of pho- community. An assemblage dominated by a unicel- tosynthesis as light increases from darkness (a), the lular cyanobacterium exhibited little photoinhibition amount of photoinhibition (b), and respiration (R). at 2 and 4 mm, and a dense cyanobacterial (Phor- Broad-scale comparisons of terrestrial plants, aquat- midium)/xanthophyte (Vaucheria) community exhib- ic plants, and microalgae have shown that Pmax,R, ited no photoinhibition at all. The microelectrode and a vary as a function of thickness of the photo- a data revealed increases in over several millimeters synthetic assemblage. R, Pmax, and a all are positively of depth (photoacclimation). These data supported interrelated, and if they are expressed per unit car- the model predictions with regard to the effects of bon, all three decrease as thickness increases (En- mat optical thickness on whole-assemblage values rõÂquez et al. 1996). However, these empirical com- for a and photoinhibition. Whole-community O2 parisons did not include intact algal bio®lms, where ¯ux data from 15 intact assemblages revealed posi- many species may cooccur, and did not provide ex- tive relationships between chlorophyll a density and plicit models that account for the differences with maximum photosynthetic rate or a expressed per thickness. unit area; the relationships with chlorophyll a were In this paper, we refer to photoinhibition as a sta- negative when photosynthetic rates were expressed tistically signi®cant decrease in photosynthetic rate per unit chlorophyll a. None of the whole assem- (b) when irradiance increases past the point of max- blages exhibited photoinhibition. Thus, the data imal photosynthesis. More recently, it has been from the whole communities were consistent with shown that some damage to photosystem II occurs model predictions. even at low irradiance (Anderson et al. 1997), and photoinhibition occurs even though b is not signif- icant. We will use b as an index of photoinhibition 1 Received 18 March 1998. Accepted 16 October 1998. but revisit the newer concept of photoinhibition in 2 Present address and author for reprint requests: Division of the discussion. Biology, Kansas State University, Manhattan, Kansas 66506; e-mail [email protected]. The discrepancy between the way photoinhibition 3 Present address: Department of Biological Sciences, Bowling is viewed in periphyton and in other photosynthetic Green State University, Bowling Green, Ohio 43403. communities provides one example of how periphy- 42 PATTERNS IN BENTHIC MICROALGAE 43 TABLE 1. Parameters used to provide values for the model. Parameter Range (units) System (reference) h (light attenuation) 1203±3108 (m21) Ulothrix and diatom communities (Dodds 1992) h 70±1790 Eight periphyton communities (this study) h 4540 Dense cyanobacterial ®lm (KuÈhl et al. 1996) 21 21 Pmax (maximum photosyn- 7±20 (mg C´g C ´h ) 25%±75% quartiles, 62 microalgal studies (EnrõÂ- thesis) quez et al. 1996) a (initial slope) 0.1±3 (mg C´g C21´h21[mmol pho- 95% con®dence intervals, 62 microalgal studies ton´m22´h21]21) (EnrõÂquez et al. 1996) b (inhibition constant) 0±0.1 (mg C´g C21´h21[mmol pho- Platt et al. 1980 ton´m22´h21]21) g chl a/g carbon 0.07 Mean 62 microalgal studies (EnrõÂquez et al. 1996) ton could potentially differ from other more studied MATERIALS AND METHODS systems. Photoinhibition can lower production of Model. Photosynthetic rate is calculated at each depth within terrestrial plants and marine phytoplankton (Long the algal bio®lm as a function of surface irradiance, light atten- et al. 1994). In contrast, Hill (1996) speculated that uation, and P-I characteristics. The P-I characteristics are modeled and described with the equations of Platt et al. (1980). The cen- photoinhibition of periphyton can occur only when tral equation is shade-acclimated communities are exposed to full P 5 P (1 2 e2aI/Pss)e2bI/P (1) sunlight. Thus, the existence of photoinhibition and s where P 5 photosynthetic rate, Ps5 maximum photosynthesis in its importance to periphyton are somewhat contro- the absence of photoinhibition under optimal light, a5the slope versial. This may be because models are not avail- of the line at low light, b5the photoinhibition constant, and I able that elucidate conditions under which photo- 5 irradiance. Derived parameters were described by Platt et al. inhibition may occur under natural conditions in (1980): Pmax is the maximum observed photosynthetic rate, Im is periphyton. Pinpointing mechanisms may not be a the irradiance where Pmax occurs, Ik is the irradiance where the onset of saturation of photosynthetic rate occurs (Pmax/a), and Ib purely physiological exercise; an understanding of is the irradiance where the onset of photoinhibition occurs. photoinhibitory processes in periphyton may be im- In the model, gross primary production is calculated as biomass portant in explaining the in¯uence of increased UV carbon-speci®c photosynthetic rate at 100 mm intervals down through the mat. The rate at each depth depends on light levels irradiance on primary production and the interplay calculated from a model of light attenuation with depth (see be- between primary producers and consumers (e.g. low). Initial surface light intensities varied from 0 to 2000 mmol Bothwell et al. 1994). quanta´m22´s21 in 20 increments. The averaged model output for The tools to assess physiological parameters of pe- each thickness at each irradiance then was used to determine a riphyton assemblages at scales relevant to the indi- biomass-speci®c P-I curve for the entire assemblage with nonlin- ear curve ®tting using a quasi-Newton method for parameter es- vidual organisms in these communities are now timation. When photoinhibition was not evident for the entire available. Microscale O2 probes can measure pho- assemblage (no decrease in photosynthetic rate at high irradi- tosynthetic rate with as little as 100 mm spatial res- ance, or equation 1 ®t the model output with a negative value olution (e.g. Revsbech and Jùrgensen 1983). The for b), the Jassby and Platt (1976) equation was used to ®t the model output. ambient light ®eld is attenuated considerably within 1 mm in benthic microalgal assemblages (Jùrgensen aI P 5 P maxtanh (2) and DesMarais 1988, Dodds 1989a). The O2 micro- 12P max electrodes can be combined with scalar light micro- The input values for the model were taken from literature val- sensors (Dodds 1992, Lassen et al. 1992) to generate ues for Ps, a, and b (Table 1). Values for P-I relationships were P-I curves within the compressed photic zones com- taken from phytoplankton measurements because they are mea- sured across a shallow light gradient (i.e. a typical ®eld or labo- mon in lighted benthos (Dodds 1992, KuÈhl et al. ratory bottle experiment with phytoplankton does not contain a 1996). Despite the wide variety of periphyton com- signi®cant light gradient within the bottle, whereas the typical munities found in aquatic habitats, comparatively whole-assemblage periphyton experiment is done with an assem- few have been analyzed in such fashion. blage that has a steep light gradient within it). Because we mod- eled individual depths of 100 mm thickness, the shallow light gra- The objective of this study was to describe how P- dient in bottle experiments should approximate that in each 100- I characteristics of entire periphyton assemblages mm thick layer. Biomass in the model is presented as carbon, and vary as functions of mat thickness (biomass) and photosynthetic assimilation is scaled by mass of carbon in the community type.