BIOLOGICAL WEIGHTING FUNCTIONS FOR UV INHIBITION OF PHOTOSYNTHESIS IN THE KELP LAMINARIA HYPERBOREA (PHAEOPHYCEAE)

J. Phycol. 45, 571–584 (2009)  2009 Phycological Society of America DOI: 10.1111/j.1529-8817.2009.00694.x BIOLOGICAL WEIGHTING FUNCTIONS FOR UV INHIBITION OF PHOTOSYNTHESIS IN THE KELP LAMINARIA HYPERBOREA (PHAEOPHYCEAE) 1 Harlan L. Miller III 2,3 The University of Texas at Austin, Marine Science Institute, 750 Channel View Drive, Port Aransas, Texas 78373, USA Patrick J. Neale Smithsonian Environmental Research Center, P. O. Box 28, Edgewater, Maryland 21037, USA and Kenneth H. Dunton The University of Texas at Austin, Marine Science Institute, 750 Channel View Drive, Port Aransas, Texas 78373, USA Different wavelengths of sunlight either drive or inhibit macroalgal production. Ultraviolet radiation (UVR) effectively disrupts photosynthesis, but since UVR is rapidly absorbed in coastal waters, macroalgal photoinhibition and tolerance to UVR depend on the depth of attachment and acclimation state of the individual. The inhibition response to UVR is quantified with a biological weighting function (BWF), a spectrum of empirically derived weights that link irradiance at a specific wavelength to overall biological effect. We determined BWFs for shallow (0 m, mean low water [MLW]) and deep (10 m) Laminaria hyperborea (Gunnerus) Foslie collected off the island of Finnøy, Norway. For each replicate sporophyte, we concurrently measured both O2 evolution and 13C uptake in 48 different light treatments, which varied in UV spectral composition and irradiance. The relative shape of the kelp BWF was most similar to that of a land plant, and the absolute spectral weightings and sensitivity were typically less than phytoplankton, particularly in the ultraviolet radiation A (UVA) region. Differences in BWFs between O2 and 13C photosynthesis and between shallow (high light) and deep (low light) kelp were also most significant in the UVA. Because of its greater contribution to total incident irradiance, UVA was more important to daily loss of production in kelp than ultraviolet radiation B (UVB). Photosynthetic quotient (PQ) also decreased with increased UVR stress, and the magnitude of PQ decline was greater in deepwater kelp. Significantly, BWFs assist in the comparison of biological responses to experimental light sources versus in situ sunlight and are critical to quantifying kelp production in a changing irradiance environment. oxygen; ozone depletion; photoinhibition; photosynthesis; photosynthetic quotient; UV radiation Abbreviations: BWF, biological weighting function; DIC, dissolved inorganic carbon; DO, dissolved oxygen; MAA, mycosporine-like amino acids; PAM, pulse-amplitude-modulated fluorescence; PCA, principal component analysis; PE, photosynthesis versus irradiance; SD, standard deviation; SE, standard error; UVA, ultraviolet radiation A; UVB, ultraviolet radiation B; UVC, ultraviolet radiation C; UVR, ultraviolet radiation Attenuation of downwelling sunlight through the atmosphere and ocean and the biological response to UVR are two processes that must be well parameterized to quantitatively assess whether solar UV radiation affects marine organisms. In phototrophs, photosynthesis and inhibition of photosynthesis are both functions of irradiance (quantity) and spectral distribution of sunlight (quality). Broadband PAR does not necessarily characterize the quality of light photosynthetically utilizable by a photosynthetic organism, particularly underwater. Similarly, broadband measures of UVR (i.e., UVA and UVB) obfuscate the wavelength dependency of UV photoinhibition. The relationship between light and photoinhibition is better quantified with a BWF: a set of spectral weights that accounts for the wavelength dependency of the photobiological process and properly scales the exposure spectra to the effective biological response (Neale and Kieber 2000). In the simplest case, because an equal exposure to broadband UVB is more effective than UVA at inhibiting photosynthesis, a BWF would give a greater weight to UVB in inducing inhibition in full spectrum sunlight. This simplified weighting function is, however, inadequate for full solar spectrum or ozone depletion models because the general weight assigned to broadband UVB does not Key index words: acclimation; biological weighting function; carbon-13; Laminaria hyperborea; 1 Received 31 January 2008. Accepted 16 February 2009. Present address: Algenol Biofuels Inc., www.algenolbiofuels.com 3 Author for correspondence: e-mail lanny.miller@mac.com. 2 571 572 HARLAN L. MILLER III ET AL. provide information on wavelength dependency of inhibition within the UVB spectral range (i.e., less stratospheric ozone allows more high-energy, shortwavelength photons within the UVB range to reach the earth’s surface). A more sophisticated BWF can be generated if a greater number of spectral treatments are included in the photosynthetic experiments. Spectral weights at 1 nm resolution can then be inferred using either nonlinear regression of an assumed exponential response function (Rundel 1983) or statistically parameterized using one to several principal components that describe the spectral irradiance treatments (Cullen et al. 1992, Cullen and Neale 1997, Neale 2000). BWF weights are defined under polychromatic light treatments. This is an important distinction between the BWF and a traditional action spectrum determined using monochromatic light (Jones and Kok 1966, Caldwell 1971). In a BWF experiment, irradiance treatments are created with a series of cutoff filters that selectively remove shorter wavelengths of light, and the polychromatic treatments are designed to incorporate multiple light-regulated physiological processes. In kelp, for instance, blue light enhances recovery and survival after UVC (253 nm) exposure [Han and Kain (Jones) 1992, 1993]. More specifically, Karsten et al. (1998a) and later Franklin et al. (2001) demonstrated that PAR, in particular blue light, and UVR interact to induce synthesis of UV-absorbing mycosporine-like amino acids in the red alga Chondrus crispus. Interactive processes would not be reflected in spectral weighting functions based on monochromatic light, but the BWF approach accounts for the net effect of light and is more representative of natural conditions. Understanding of the temporal response to UV exposure is fundamental to the predictive power of the BWF model. The time-dependency of UVinduced inhibition is evaluated with an exposure response curve (Coohill 1994). From these curves, it can be determined whether the biological response results from cumulative radiant exposure or from the instantaneous exposure rate (Cullen and Lesser 1991), that is, whether reciprocity is satisfied. If the effect is essentially irreversible over the timescale of interest, reciprocity is upheld, and the response is a function of cumulative UV exposure (J Æ m)2). If, however, organisms possess mechanisms that can counteract UV damage, then the rate of photosynthesis under UV exposure declines until it reaches a steady state, reflecting a balance between damage and repair (Lesser et al. 1994). In this case, reciprocity fails, and photosynthetic inhibition is proportional to maximum exposure rate (mW Æ m)2) (Cullen et al. 1992). For example, an irradiancedependent BWF was appropriate for temperate (i.e., relatively high natural UV) phytoplankton (Banaszak and Neale 2001), but Antarctic (i.e., relatively low UV) phytoplankton show little short-term ability to counteract UV inhibition, implying that inhibition is a function of cumulative exposure (Neale et al. 1998b). Current understanding of the effect of UVR on seaweed genetics, physiology, and production is reviewed by Bischof et al. (2006). Seaweeds adapt to the prevailing light environment at their attachment depth, and the response to both high-light and UVR stress is correlated with vertical distribution (Larkum and Wood 1993, Dring et al. 1996a,b, Franklin and Forster 1997, Hanelt et al. 1997b, Karsten et al. 2001, Johannsson and Snoeijs 2002, Bischof et al. 2006, Roleda et al. 2006). Sessile at their growing site, kelp (Laminariales, Phaeophyta) individuals are restricted to specific light histories and radiant UV exposure. The rapid attenuation of UVR relative to PAR by seawater, typical of coastal waters with high concentrations of dissolved organic matter (Tedetti and Sempéré 2006), protects kelps deeper in the water column from high-intensity UV. Consequently, macroalgal sensitivity to UV exposure depends on attachment depth, and intertidal and upper subtidal seaweeds have more effective UV-counteracting mechanisms. In brown algae, phlorotannin production is inducible by UVR, and its role as a sunscreen is implied (Pavia and Brock 2000, Swanson and Druehl 2002, Fairhead et al. 2005). Also, photoacclimated algae employ efficient protein repair and antioxidant systems after UV damage has occurred (Vincent and Neale 2000, Aguilera et al. 2002, Häder et al. 2002). Thus, macroalgae have the capacity to acclimate to changes in sunlight, and reciprocity is likely depth-dependent since deepwater algae receive little appreciable UVR. Kelp populations from boreal Norway (north of Ålesund) were studied because of their exceptional productivity along the Atlantic coast. Standing crops of the dominant kelp, L. hyperborea, approach 40 kg fresh weight (fwt) Æ m)2 (Sjøtun et al. 1995). Warmer seawater temperatures along Norwegian coastlines impede winter ice formation, and without winter ice scour, kelp grows into the upper subtidal where there is greater potential for UVR exposure. Kelp beds throughout the various coastal islands and fjords extend from the intertidal to 30 m (Sjøtun et al. 1995), and we hypothesize that individual L. hyperborea are well acclimated to the quantity and quality of sunlight incident at their attachment depth. The purpose of this study was to experimentally determine and compare BWFs for L. hyperborea collected from the eulittoral and from a subtidal depth (10 m) without appreciable UVR (Eyvind and Højerslev 2001). Three separate experiments were conducted. First, in order to correctly formulate the shape of the PAR response in the BWF model, short-duration photosynthesis-PAR irradiance (PE) experiments were performed in an oxygen electrode chamber. Second, the temporal 573 LAMINARIA HYPERBOREA BWF component of BWF was addressed in an exposure response experiment with surface-collected kelp and was otherwise inferred for deeper-water kelp. Finally, the BWF was quantified from photosynthesis experiments, which varied in spectral treatment and where we concurrently measured oxygen evolution and carbon uptake. Weightings were parameterized using both irradiance-based and cumulative exposure models, and the spectral shapes of the BWF responses were then compared between the two measures of photosynthesis and between kelp collection depths. MATERIALS AND METHODS Site description, kelp collection, and tissue preparation. Photosynthetic experiments were conducted at a field laboratory on the island of Finnøy, Norway, which was located <1 km from an extensive L. hyperborea kelp forest (Fig. 1). The population extends subtidally from mean low water (MLW = 0 m) to 20 m into a tidal channel. BWF components were determined experimentally during three field excursions to the area. PAR-only photosynthetic experiments were performed in September 1999 (results in Fig. S1 in the supplementary material). The temporal aspect of the BWF (the exposure response curve) was investigated in October 2000, and the BWF was determined in March ⁄ April 2001. Adult kelp sporophytes with lamina >0.5 m in length were harvested 1 d prior to photosynthetic experiments. Subtidal kelps were collected via SCUBA, and divers took care to remove individuals with holdfasts intact to minimize stress prior to experimentation. Kelps were kept in fine mesh collection bags and placed in dark ice chests for the short trip back to the field laboratory. Whole kelp thalli were stored overnight in underwater collection bags suspended in a nearby protected boat basin. Photosynthetic tissue samples excised at least 10 cm from the basal, meristematic region of the lamina and 1 cm away from blade margins were used in the three photosynthesis experiments. Samples were maintained under dim laboratory light (estimated 50 lmol photons Æ m)2 Æ s)1) in temperature-controlled seawater at the ambient seawater temperature of the season. Margins of the cut kelp tissues initially produced mucilage, and we observed that the wound response ceased within 1–2 h. The incubation bath was replaced frequently with fresh seawater. Pigment extraction and thallus absorbance. Thallus optical properties were compared in kelp tissues collected from 0 m and 10 m depth. Living kelp blade samples were transported (<3 d) in a dark, temperature-controlled container to the Smithsonian Environmental Research Center, Maryland. Thalli were not visibly damaged during transport, though changes in optical properties were uncontrolled. Whole-thallus sections were immersed in 5C seawater in a UV transparent cuvette, and absorbance spectra were measured on a Cary 400 spectrophotometer (Varian Inc., Palo Alto, CA, USA) using the frosted sides of the cuvette as a diffuser (Shibata 1958). Exposure response curves. Time-course experiments to determine the photosynthetic response to UVA and UVB exposure were performed using L. hyperborea collected from 0 m (MLW). Tissue samples of 12 cm2 were excised from the first lamina and secured in a chamber built from UV transparent acrylic. The 8 mm fiber optic probe of a DIVING-PAM fluorometer (Heinz Walz GmbH, Effeltrich, Germany) was inserted at a 60 angle and at a fixed distance to the thallus. Incubation seawater was constantly stirred with a magnetic stir bar, and temperature was maintained at 12C–13C (ambient October seawater temperature) by placing the sample chamber in a large-volume water bath, also constructed of UV transparent acrylic (Laird Plastics, Austin, TX, USA). Illumination was provided by a 150 W xenon lamp (Schoeffel Instruments, Westwood, NJ, USA). Relative photosynthesis was quantified as the quantum yield of PSII electron transport (i.e., DF ⁄ F¢m). Each spectral treatment (UVB+UVA+PAR, UVA+PAR, and PAR-only) was replicated with tissue excised from three L. hyperborea individuals. Exposure response time courses were designed in four phases. In a pretreatment phase, thalli were dark acclimated for 15 min, then PAR-only irradiance was incrementally increased to the exposure intensity in four 10 min intervals (32, 82, 171, and 270 lmol photons Æ m)2 Æ s)1). The value of DF ⁄ F¢m was recorded every 30 s, and steady-state yield was achieved at each light increment. Light levels were manipulated with neutral density filters, and UVR was removed with a Schott GG400 long-pass cutoff filter (Schott Glass Technologies Inc., Elmsford, NY, USA). Second, spectral exposure during the next 60 min was controlled with WG305, WG320, or WG400 nm long-pass cutoff filters to obtain, respectively, nominal UVB+UVA+PAR, UVA+PAR, and PAR-only spectral treatments. The filter suffix is the wavelength with nominal 50% transmittance, so, for example, there is a minor component of UVB in the WG320 treatment. PAR irradiances for all spectral exposures were similar to the maximum PAR-only treatment in the pretreatment phase, 270 lmol photons Æ m)2 Æ s)1. Third, photosynthetic recovery in the light was monitored for 60 min with PAR-only irradiance, again maintained at 270 lmol photons Æ m)2 Æ s)1. Finally, the chamber was covered with black cloth, and recovery was monitored in the dark for 60 min. Optimal quantum yield, Fv ⁄ Fm, was measured every 5 min in the dark recovery phase, and dark recovery was monitored overnight for one replicate from each spectral treatment. Fluorescence yields were then normalized to an average of the last 10 measurements in the final PAR-only treatment just prior to the exposure phase. Reciprocity and the dynamics of UV photodamage (k) and physiological repair (r) during exposure were assessed using an exponential function and data from the UVB+UVA+PAR treatment (Neale 2000): Y ðtÞ 1 ðr þ k eððr þkÞtÞ Þ ¼ Y0 r þk Fig. 1. Laminaria hyperborea collection site and field laboratory location. ð1Þ where Y ⁄ Y0 is DF ⁄ F¢m yield normalized to PAR-only DF ⁄ F¢m. BWF: experimental measurement. Photosynthesis and UVdependent photoinhibition in L. hyperborea from two depths 574 HARLAN L. MILLER III ET AL. (0 m and 10 m, MLW) were determined in a spectral incubator, the photoinhibitron (Neale and Fritz 2002), which was modified from phytoplankton applications to be used with kelp tissue disks (Miller and Dunton 2007). The photoinhibitron allows for polychromatic light treatments that differ in both spectral composition and irradiance. The apparatus consists of (1) an aluminum block bored to hold eighty 20 mL quartzbottomed incubation cuvettes, (2) a continuous-flow water bath to control temperature, (3) and a 2500 W xenon lamp light source. A central mirror reflects light up through the apparatus to illuminate kelp tissue in individual incubation cuvettes. Along the beam path, light passes through (1) a water bath with a UV-transparent Plexiglas bottom; (2) one of eight long-pass filters (WG280, WG295, WG305, WG320, WG335, and GG395 [Schott Glass Technologies Inc.] and LG350 and LG370 [Spectra-Physics, Waltham, MA, USA]) with nominal 50% transmittance at indicated wavelengths (nm); and (3) a cellulose acetate shield that removes UVR below 290 nm. Within each spectral treatment, irradiance was varied using neutral density screens. Photosynthesis in L. hyperborea was measured at six light irradiances within each of eight UV filters for a total of 48 light treatments per 1.5 h incubation. Temperature was maintained within the cuvettes at 5 ± 1C (the ambient in situ sea temperature) by adding ice as needed to a large volume ( 50 L) reservoir and by continuously cycling this water through the water bath of the apparatus. Spectral irradiance in each of the 48 light treatments was determined with a scanning spectroradiometer system (Neale and Fritz 2002). Light was collected from each incubation cell with a quartz fiber-optic probe fitted with a diffuser designed for PAR and UV measurement, and the probe was connected to an Acton Research monochromator (SP-401, Princeton Instruments, Acton, MA, USA) with a photomultiplier tube (Burle Technologies, Lancaster, PA, USA) and computer interface. The spectroradiometer was calibrated at the field location using a National Institute of Standards and Technology– traceable, 1000 W quartz-halogen lamp operated at a current of 7.9000 (±0.0001) A (Xantrex [Burnaby, BC, Canada] power supply, with current monitored by HP3457A multimeter measuring across a 0.1 ohm shunt). The lowest wavelength of detectable irradiance in experimental light treatments was 286 nm. In addition, PAR irradiances in all light cells were monitored prior to each replicate incubation using a QSL-100 (Biospherical Instruments, San Diego, CA, USA) 4-p quantum sensor. Photosynthetic measurements were determined for kelp disks excised from the first blade, that is, the newly formed blade material produced that winter ⁄ spring. The blade tissue in this region was visually uniform and consistent in density. For each complete incubation, 51 tissue disks (48 light treatments and three dark treatments) were cut with a brass coring device (1.8 cm diameter), and disks were placed in a temperaturecontrolled seawater container for 1 h. The BWF experiment was replicated with five individual kelp plants from both 0 m and 10 m depth intervals. To avoid potential variation due to natural acclimation to field conditions over weekly timescales, all five 0 m replicates were completed first in a 3 d period followed by five 10 m plants on the subsequent 2 d. Photosynthesis was quantified by concurrent measurements of oxygen evolution and H13 CO3  uptake. In brief, several days prior to incubations, 20 L of filtered seawater was enriched with NaH13CO3 [see Miller and Dunton (2007) for details, including amount added and effect on total DIC]. The carboy lid was left loose, and the carboy was occasionally shaken to allow the dissolved inorganic carbon (DIC) concentration to partially equilibrate with the atmosphere. Before incubations, 13 C-enriched seawater was brought to experimental temperature (5C), and dissolved oxygen (DO) was stripped with N2 for 8 min to decrease DO to about 20%–30% air saturation. Water samples were taken in 70 mL serum bottles, fixed with 0.2 mL of saturated HgCl2 solution, Pand stored in a refrigerator at 4C for later analysis of initial CO2 and d13C. Photoinhibitron cuvettes were filled with O2-stripped and 13 C-enriched seawater, and a kelp disk was suspended in the cuvette about 1 cm from the bottom. The tissue disk was held in place between two silicone O-rings crosshatched with thin transparent thread. Suspending the disk in this way assured perpendicular orientation of the disk to the irradiance beam and allowed for unrestricted water motion around the disk. After removing air bubbles, cuvettes were sealed with modified silicon stoppers that were fitted with small, 1.5 V DC electric motors and magnetic stir bars to create water motion in the cuvettes (Miller and Dunton 2007). Motors were powered with a B&K Precision 20 A DC power source (B&K Precision Corp., Yorba Linda, CA, USA; model 1688), and the speed of the magnetic stir bars was adjusted by controlling the voltage of the power source. DO concentrations were determined in the treatment seawater with a PreSens Microx TX micro-optode oxygen system (Presens Precision Sensing GmbH, Regensberg, Germany) using a needle-type optode. DO measurements were made both before and after incubations. Oxygen-evolution rates were corrected for nonphotosynthetic changes in DO assessed in blank cuvettes, that is, cuvettes without the kelp tissue. Kelp tissue disks were quickly removed from cuvettes before final oxygen measurements, after which disks were then rinsed twice to remove excess 13C label and dried at 60C. 13Camended seawater and tissue samples were shipped to University of Texas, Marine Science Institute (UTMSI), for isotopic analysis on a Finnigan MAT DeltaPlus continuous flow isotope ratio mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) attached to a Carlo Erba elemental analyzer (NC 2500, Carlo Erba, Milan, Italy). Oxygen-evolution and carbonuptake photosynthetic rates were calculated using equations described in Miller and Dunton (2007). Both measures were interpreted as gross photosynthesis since dark respiration was added to oxygen evolution and light treatments were sufficiently short to assume minimal carbon turnover (see discussion in Miller and Dunton 2007). BWF: statistical evaluation. BWFs were statistically determined for L. hyperborea based on carbon and oxygen photosynthesis from kelp collected at 0 m and 10 m depth. BWF calculations began with an underlying model describing potential photosynthesis, Ppot, as a saturating function of irradiance. UV inhibition is estimated as a fraction of Ppot, and the parameterization of the BWF is sensitive to the overall shape of the PE function. Four photosynthetic equations were compared using PE data collected from L. hyperborea sporophytes at three depths. The Michaelis-Menten equation best reproduced the broad transition zone between the light-limiting and light-saturating regions of the curve (Fig. S1 and Miller 2002). Therefore, gross Ppot was calculated using the following equation (see Table 1 for symbol notation and units): Ppot ¼ Pmax EPAR Ek þ EPAR ð2Þ where Pmax is the maximum potential rate of photosynthesis, particular to either carbon or oxygen units, and Ek is PAR irradiance at nominal light saturation. If photosynthesis is inhibited in the presence of UVR, then the magnitude decrease from Ppot is determined with a second expression that depends on the kinetics of the exposure response curve (Neale 2000). In the case when steady-state inhibition is achieved (i.e., active repair processes counteract UV damage), the mean rate of photosynthetic production (P) is described using an irradiance-based model (the BWFE-PE model): 575 LAMINARIA HYPERBOREA BWF Table 1. Symbol notation and abbreviations. Symbol Description Units a0 a1 C  Einh eE(k) eH(k) E(k) EPAR Ek H(k)  Hinh P Pmax Ppot T UVR UVA UVB UVC First coefficient for polynomial in Rundel weighting function Second coefficient for polynomial in Rundel weighting function Proportionality constant Biologically effective fluence rate for inhibition of photosynthesis Biological weighting as a function of UV irradiance Biological weighting as a function of cumulative radiant exposure Spectral irradiance, scalar underwater in irradiance model Photosynthetically active radiation (400–700 nm) Photosynthesis saturation irradiance Radiant exposure Biologically effective cumulative exposure for photosynthesis inhibition Photosynthesis (e.g., mean rate during experimental exposures) Maximum rate of photosynthesis in absence of inhibition Potential photosynthesis in absence of inhibition Duration of experiment Ultraviolet irradiance at 220–440 nm; solar UVR (290–400 nm) Ultraviolet irradiance at 320–400 nm Ultraviolet irradiance at 280–400 nm Ultraviolet irradiance at 220–280 nm Dimensionless nm)1 (mW Æ m)2))1 or (J Æ m)2))1 Dimensionless (mW Æ m)2))1 (J Æ m)2))1 mW Æ m)2 Æ nm)1 lmol photons Æ m)2 Æs)1 W Æ m)2 J Æ m)2 Æ nm)1 Dimensionless lmol O2 Æ m)2 Æ s)1 or lmol C Æ m)2 Æ s)1 lmol O2 Æ m)2 Æ s)1 or lmol C Æ m)2 Æ s)1 lmol O2 Æ m)2 Æ s)1 or lmol C Æ m)2 Æ s)1 s mW Æ m)2 mW Æ m)2 mW Æ m)2 mW Æ m)2 P ¼ Ppot  1  1 þ Einh  ð3Þ If photosynthesis consistently declines over the exposure time period (i.e., reciprocity is upheld), then an inhibitory model based on cumulative exposure is used (the BWFH-PE model):    1  expHinh ð4Þ P ¼ Ppot  Hinh Though L. hyperborea likely possesses repair and ⁄ or protection mechanisms, particularly in shallow kelp accustomed to high irradiance environments, both the BWFE-PE and BWFHPE models were chosen to predict the response of the kelp population to ozone depletion. The natural response is likely between these two extremes and is dependent on the light history and collection depth of the alga. Biologically effective  fluence rate (Einh ) was determined from 286 to 400 nm:  ¼ Einh 400 nm X eE ðkÞEðkÞDk ð5Þ the polynomial (constant and linear) were required in fitting the kelp BWF: eðkÞ ¼ C expða0 þa1 kÞ ð7Þ where C = 1 is a proportionality constant with units (mW Æ m)2))1 or (J Æ m)2))1. Coefficients ai of the BWF, as well as photosynthetic parameters Pmax and Ek, were estimated with nonlinear regression techniques implemented in a MATLAB v.6.1 application (Mathworks Inc., Natick, MA, USA). Fits were performed on the pooled data of all replicates from each depth. Small, but significant, variations in Pmax between replicates were accounted for by fitting a separate Pmax for each replicate. Standard errors of e(k) were calculated from approximate variances and covariances of ai using propagation of errors (Cullen and Neale 1997). Photosynthetic responses to spectral irradiance were also determined using principal component analysis (Cullen et al. 1992, Neale 2000), but the internal spectral structure within e(k) derived with the PCA method was not statistically justified over the simpler Rundel method (Cullen and Neale 1997). k¼286 nm  and biologically effective cumulative exposure (Hinh ) was formulated as follows:  Hinh ¼ 400 nm X eH ðkÞH ðkÞDk ð6Þ k¼286 nm RT where H ðkÞ ¼ 0 E ðkÞdt(J Æ m)2 Æ nm)1) is the total radiant exposure during the incubation period, T. eE(k) (eq. 5) and eH(k) (eq. 6) are wavelength-specific weightings for UV inhibition of photosynthesis and are calculated below. PAR photoinhibition (i.e., ePAR) was omitted in the calculation of   both Einh and Hinh as it was shown to be negligible (Fig. S1 and Miller 2002). The experiment produced 48 oxygen- and 48 carbon-based values of P per kelp replicate, and BWF models were parameterized from photosynthetic data and the measured spectral irradiance in each individual cuvette (Neale et al. 1998a). The biological weights, either eE(k) or eH(k), for each UV wavelength were determined using the Rundel method, which assumes the natural log of the BWF is a polynomial (Rundel 1983, Cullen and Neale 1997, Neale 2000). Only two terms of RESULTS Optical properties of the thallus. Kelp thalli from 10 m were noticeably more translucent than kelp occurring in the upper subtidal. We compared absolute absorbance of whole-thallus spectra and observed that depending on wavelength, deepwater kelp absorbed 35%–60% less light than shallowwater kelp. On an areal basis, 0 m kelp contains more chl a than deep kelp, 433 ± 6 versus 371 ± 8 mg Æ m)2, respectively. Chl a:fucoxanthin ratios increased little with depth from 1.8 to 2.1, but chl a:chl c decreased from 8.2 at the surface to 4.6 at depth (Miller 2002). When spectra were normalized to their respective blue chl peaks (430 nm), it was apparent that shallow kelp also absorbed relatively more radiation in the blue-green, violet, UVA, and most UVB wavelengths (Fig. 2). Photoprotective pigments absorbing in these wavelengths, for 576 HARLAN L. MILLER III ET AL. Fig. 2. Thallus absorbance spectra, normalized to its blue chl peak at 430 nm, for Laminaria hyperborea collected from 0 m and 10 m. Data are x ± SD, n = 5. Inset: (1) difference spectrum between 0 m and 10 m kelp, (2) phlorotannin absorbance spectrum (Henry and van Alstyne 2004), and (3) mycosporineglycine absorbance spectrum (Dunlap et al. 1986). ABS, absorbance. example, b-carotene in the visible and mycosporineglycine and phlorotannin in the UVB (Fig. 2 inset), potentially contribute to the spectral differences between the two depth collections. Temporal exposure response. Choice of weighting function based on irradiance (BWFE-PE) or cumulative exposure (BWFH-PE) depends on the temporal kinetics of photodamage and repair processes during UVR exposure. After 20 min of exposure, UVB+UVA+PAR caused definite, but moderate, photoinhibition in 0 m L. hyperborea (Fig. 3A). When UVB+UVA+PAR exposure data were compared using a kinetic equation (eq. 1), which assumes that repair processes are proportional to damage, then steady-state balance between repair and UVR damage would occur after about 1 h and result in an approximate 20% decrease in photosynthetic efficiency (R2 = 0.83). Both UVA+PAR and UVB+UVA+ PAR exposed thalli recovered at parallel rates in the presence of PAR (Fig. 3B). Long-term recovery in the dark was clearly biphasic with the duration of the initial, fast phase less for the UVB+UVA+PAR treatments than UVA+PAR and PAR (Fig. 3C). This increased the time for recovery to 90% of initial optimal quantum yield from 3 h in the PAR-only experimental treatment to nearly 6 h when thalli were exposed to UVB. L. hyperborea BWF. It was discovered after field experiments that polychromatic light treatments in the photoinhibitron were deficient of expected short-wavelength UVB radiation (Fig. 4, cf. Cullen et al. 1992, Neale et al. 1994, Neale and Fritz 2002). This finding was likely due to errant mirror configu- Fig. 3. Quantum yield time course measured by PAM fluorometry during (A) exposure to UV+PAR ( x , n = 3), (B) postexposure recovery under PAR-only exposure ( x , n = 3), and (C) subsequent overnight recovery in the dark (n = 1) for Laminaria hyperborea collected from 0 m. Fluorescence yield data in (A) and (B) are normalized to DF ⁄ F¢m in the PAR-only treatments during the time trial. Fv ⁄ Fm values in dark recovery phase (C) are normalized to optimal quantum yield assessed in dark-adapted thalli prior to experiment. Solid lines in (C) are fits (R2 > 0.99) to the biphasic equation of the form 1 ) [p1exp()k1t) ) p2 exp()k2t)], where k1, k2 are the rates of recovery and p1, p2 are the changes in normalized Fv/Fm during the fast and slow phases, respectively. Though rate constants k1 and k2 were similar (0.025 min)1 and 0.003 min)1), p1 decreased from 0.30 to 0.19, and p2 increased from 0.18 to 0.33, respectively, for PAR-only and UVB+UVA+PAR treatments. PAM, pulse-amplitude-modulated fluorescence; UVA, ultraviolet radiation A; UVB, ultraviolet radiation B. ration in the photoinhibitron. The spectrum achieved in the UVB range was similar to natural sunlight at noon and under 300 D.U. column ozone (Fig. 4). Thus, the incubation did not bracket the full range at short wavelengths, but the Rundel method extrapolated the BWF [i.e., eE(k) and eH(k)] to the shorter wavelengths for which there was little treatment variation. There is no reason to suspect a specific bias in spectral weightings at shorter wavelengths, since Rundel estimates at these wavelengths are in any case constrained by the overall slope (a1), the estimate of which is defined mainly by the response to the intermediate wavelengths that have greatest effective irradiance 577 LAMINARIA HYPERBOREA BWF Fig. 4. Spectral treatments in the photoinhibitron resulting from Schott long-pass cutoff filters. Values are normalized to irradiance at 400 nm. Sunlight spectral distribution was generated with STAR radiative transfer model (Ruggaber et al. 1994) and superimposed for comparison. Sunlight was modeled for clear-sky irradiance at noon on 7 April 2001 with 300 D.U. column ozone (65 N). (Cullen and Neale 1997). Nonetheless, the statistical power for estimating the effect of exposure is lower in the UVB region (as indicated by large confidence intervals), and so these weights should be interpreted conservatively. A total of eight BWFs for UV inhibition of photosynthesis are presented for the kelp L. hyperborea (Table 2). BWFs differed between carbon and oxygen metabolisms, collection depth, and choice of weighting model (BWFE-PE or BWFH-PE). Irrespective of inhibitory model chosen, photosynthesis decreased appreciably when kelp were exposed to increasing amounts of weighted UVR (Fig. 5, points), and 10 m kelp exhibited a greater UV-inhibitory response than 0 m kelp. The proportion of variances (R2) explained by the Rundel method was similar in BWFE-PE and BWFH-PE (Table 2) models. In general, R2 values were less in 0 m kelp and when photosynthesis was measured as oxygen evolution. Spectrally, UVB wavelengths were two to three orders of magnitude more effective at inhibiting L. hyperborea photosynthesis than UVA (Fig. 6). Absolute sensitivity of the inhibition weightings varied spectrally depending on photosynthetic measurement and collection depth, and overall spectral slope (a1) was similar for both BWFE-PE and BWFH-PE models (Table 2). Inhibition weightings for kelp collected at 10 m depth were greater than that of 0 m kelp in the UVA region between 330 and 370 nm (Fig. 6), yet both collections had a response of similar magnitude in the UVB. This was also the case for the BWFH weightings (results not shown, but see Table 2). BWFs derived from oxygen had consistently smaller intercepts (a0) and slopes (a1) than BWFs derived from carbon measures of photosynthesis (Table 2). As a result, each pair of BWFs showed a crossover pattern, with carbon assimilation more sensitive to UVB but oxygen evolution more sensitive to UVA (Fig. 6 for BWFE, similar results for BWFH not shown). However, these differences were not statistically significant for 0 m kelp. On the other hand, there was less absolute difference for O2- and C-based BWFs from 10 m, but the increase in oxygen photosynthetic inhibition at UVA wavelengths was significant (Fig. 6). The BWFs are significantly different even though the corresponding slopes and intercepts are not because of the high covariance between the estimates (correlation 0.75–0.85). Estimated confidence intervals for the weightings also differed spectrally with decreased model confidence in the long-wavelength UVA and short-wavelength UVB regions. This spectral variance was most pronounced in oxygen-based measurements of inhibition. The general spectral shape and relative magnitude of L. hyperborea BWFs were most similar to the polychromatic weighting function developed for inhibition of photosynthesis in Rumex patientia (Polygonaceae, Fig. 7A). Action spectra based on monochromatic light tended to either underestimate (Avena) or overestimate (isolated spinach chloroplasts) UV inhibition, compared to the kelp Table 2. Laminaria hyperborea: Gross photosynthesis and photoinhibition parameters determined with least squares analysis of experimental results using either the BWFE-PE or BWFH-PE models. The coefficient of determination (R2) is shown for   P ⁄ Ppot analysis in Figure 5. Photosynthesis and R2 values are similar for Einh and Hinh models. Pmax is mean of replicate Pmax’s with SE of the mean. All other values are mean with SE calculated using propagation of errors (n = 5). Ek conversion to W Æ m)2 is 0.227 J Æ lmol)1 for xenon lamp. Depth (m) 0 Measure Model Pmax (lmol Æ m)2 Æ s)1) Ek (lmol photons Æ m)2 Æ s)1) C  Einh  Hinh  Einh  Hinh  Einh  Hinh 1.75 (0.09) 101 (13) 2.03 (0.16) 88 (15) 1.41 (0.08) 92 (11) 1.87 (0.17) 97 (14) O2 10 C O2  Einh  Hinh a0 13.59 12.45 4.82 3.25 7.14 4.27 (7.84) (7.85) (8.31) (8.61) (4.54) (4.60) 3.05 (4.23) )0.95 (4.26) a1 0.074 0.073 0.047 0.045 0.052 0.047 R2 (0.024) (0.023) (0.024) (0.025) (0.013) (0.013) 0.69 0.039 (0.012) 0.031 (0.012) 0.88 0.65 0.87 578 HARLAN L. MILLER III ET AL. Fig. 5. Photosynthetic response of Laminaria hyperborea to weighted experimental UV radiation based on an irradiance model (BWFE-PE). Gross oxygen evolution and carbon fixation (points) were measured for kelp collected from 0 m and 10 m depths ( x, n = 5). Values are displayed normalized to potential photosynthesis. Expected P ⁄ Ppot (line) is shown for each. See Table 2 for R2. BWF, biological weighting function; PE, phototsynthesis versus irradiance. Fig. 6. Comparison of BWF for UV photoinhibition in Laminaria hyperborea. The BWF spectral shape is compared for both 0 m and 10 m collections (Depth, left column) and between carbon uptake and oxygen evolution (Measurement type, right column). BWFH-PE results are similar to BWFE-PE model data presented. Shaded areas are 95% confidence intervals. BWF, biological weighting function; PE, phototsynthesis versus irradiance. responses, for wavelengths greater than the normalization wavelength (300 nm). Phytoplankton from Lake Bonney (Antarctica) and a high-sensitivity phytoplankton assemblage in the Rhode River (Maryland) were in general an order of magnitude more sensitive than kelp across the UV spectrum (Fig. 7B), though irradiance-based kelp BWFs were comparable to a low sensitivity Rhode River assemblage. On a cumulative-exposure basis, phytoplankton from the Weddell Sea Confluence were more than five times more sensitive to increases in UVB than Norwegian kelp (Fig. 7C). UV stress and the photosynthetic quotient. Since both oxygen evolution and carbon uptake were measured concurrently for the same tissue sample and irradiance treatment, it was possible to evaluate the response in the photosynthetic quotient with increasing UV exposure. We defined photosynthetic quotient (PQ) in this analysis as the molar ratio of gross oxygen evolution to gross carbon assimilation during photosynthesis and considered how this measure varied as a function of unweighted UVR (Fig. 8). The BWFs showed a consistent pattern of higher sensitivity of oxygen evolution in the UVA 579 LAMINARIA HYPERBOREA BWF Fig. 8. Decrease in the photosythetic quotient with increased (unweighted) UVR exposure in Laminaria hyperborea collected from (A) 0 m and (B) 10 m depth. Observed data points are x ± SE (n = 5), and the trend was modeled with a linear regression (line) with ±95% confidence intervals of the fit (dotted lines). Regression equations include parameter error in parentheses. PQ, photosynthetic quotient; UVR, ultraviolet radiation. Fig. 7. Comparison of kelp BWFs with published plant and phytoplankton BWFs and BWFs fitted with different models. (A) Relative inhibition normalized to 300 nm: inhibition of Hill reaction in isolated spinach chloroplasts (monochromatic, Jones and Kok 1966), general plant damage (monochromatic, Caldwell 1971, redrawn from Caldwell et al. 1986), UV inhibition of photosynthesis in Rumex patientia (Rundel method, differential effectiveness, eq. 13, Rundel 1983), UV inhibition of Avena sativa growth (Flint and Caldwell 2003), 0 m Laminaria hyperborea  model), 10 m L. hyperborea (Rundel method, carbon-based Hinh  model). (B) Irradiance(Rundel method, carbon-based Hinh  models): phytoplankton dependent BWFs (carbon-based Einh from Lake Bonney, Antarctica (PCA method, redrawn from Neale et al. 1994), spring and winter phytoplankton assemblages in temperate Rhode River, Maryland, estuary (PCA method, Banaszak and Neale 2001), 0 m L. hyperborea (Rundel method), and 10 m L. hyperborea (Rundel method), mostly diatom community collected from McMurdo Sound, Antarctica, and maintained in outdoor culture (PCA method, redrawn from Neale et al. 1994, Banaszak and Neale 2001). (C) Cumulative exposure models): natural intact dependent BWFs (carbon-based Hrminh phytoplankton colonies (Rundel method) or homogenized samples (PCA method) from Weddell–Scotia Confluence (Neale et al. 1998b), Antarctica, 0 m L. hyperborea (Rundel method), and 10 m L. hyperborea (Rundel method). BWF, biological weighting function; PCA, principal component analysis; WSC, Weddell Sea Confluence. compared to carbon fixation (Fig. 6). This finding implies that PQ should decrease with increasing UV exposure to filtered xenon arc irradiance, but since the weights were not significantly different for oxygen evolution versus carbon assimilation, we decided to examine this possibility directly. In the absence of UVR stress, the mean PQ for 10 m kelp was 1.30, and using a one-tailed t-test (t = )42.1, P << 0.01) assuming equal variances (F46,46 = 1.19, P = 0.57), the 10 m PQ was significantly greater than 0 m samples (PQ = 1.18). The PQ from both collections significantly declined with UVR exposure; that is, the variables were negatively correlated (r0m = )0.55, P << 0.01; r10m = )0.82, P << 0.01), and trend line slopes were significantly less than zero (one-tailed t-test: t0m = )4.45, P << 0.01; t10m = )9.95, P << 0.01). Further, as variances in 10 m and 0 m slopes were similar (F46,46 = 1.22,P = 0.50), a one-tailed t-test assuming equal variances suggested that PQ declined more rapidly in 10 m kelp (t = 29.8, P << 0.01). This trend is consistent with the higher weights for UV inhibition of oxygen evolution by 10 m kelp. DISCUSSION The BWF is an empirical formulation that assigns spectral weights to account for the wavelength dependency of UV effects on a biological process. Here, polychromatic weightings for photoinhibition are provided for both high-light-acclimated (0 m) and low-light-acclimated (10 m) L. hyperborea, and 580 HARLAN L. MILLER III ET AL. the functions are applicable to oxygen- or carbonderived photosynthesis studies. Empirical BWFs are also presented assuming both irradiance-based and cumulative-exposure response models. The importance of the BWF is that it enables (1) intercomparison of results obtained using different types of polychromatic exposures, including solar simulators in the laboratory and natural sunlight, and (2) extrapolation of experimental results to model simulations of ozone depletion (Cullen and Neale 1997, Neale 2000, Miller 2002). In two deepwater red algae, UVA exposures that were required to inhibit Fv ⁄ Fm by 50% were lower under artificial sunlamps (Dring et al. 1996b) than total exposure in natural sunlight (Dring et al. 2001). The authors attributed this discrepancy to spectral differences in peak irradiance between laboratory sunlamps and outdoor sunlight, and the inconsistency in biologically effective exposure would be reconciled with an appropriate weighting function. To date, few researchers have applied spectral weightings in studies on UV effects on macroalgae. Those studies that have relied on action spectra of Caldwell (1971) for general plant damage (e.g., Bischof et al. 2000) or a normalized version of the Jones and Kok (1966) photoinhibition spectrum for isolated spinach chloroplasts (e.g., Forster and Lüning 1996, Gómez et al. 2001, 2005, Huovinen et al. 2006). More recently, a new general plant action spectrum has been presented based on the growth response of Avena (Flint and Caldwell 2003). Our results show that these action spectra are not suitable for quantifying the effects of a changing light climate on kelp physiology for several reasons. Most importantly, the contribution of UVA relative to UVB differs from the shape of the L. hyperborea BWFs; that is, Avena spectrum has too little weight, and the Jones and Kok spectrum has too much weight in the UVA (the general plant spectrum is not defined in the UVA). Also, the Jones and Kok relationship assumes reciprocity, which was demonstrated for the Hill reactions they used to measure photosynthesis. Finally, these action spectra predict only the relative biological response to varied irradiance, so a quantitative prediction requires knowledge of the biological effect for at least one weighted exposure (Cullen and Neale 1997). Relative shapes of the L. hyperborea BWFs are most similar to the land plant Rumex patientia, with 10 m kelp slightly more sensitive to UVA, relative to UVB at 300 nm, than R. patientia, and 0 m kelp less sensitive to UVA. The magnitude of the spectral weights is greater in the UVA above 350 nm in all phytoplankton BWFs considered. Phytoplankton are vertically mixed in the ocean and acclimate to the net light environment. In contrast, kelps are anchored at depth, and the relative insensitivity of surface kelp to UVA likely results from long-term acclimation to ambient light experienced near the sea surface. Kelps from deeper water are more susceptible to UVR stress, but the spectral differences in weightings appear to be related to UVA tolerance. The only phytoplankton community compared that was more tolerant of UVB than kelp was a diatom-dominated assemblage from McMurdo Sound, Antarctica. In this study, phytoplankton were allowed to adjust to a high UVB climate as cultures were acclimated in shallow outdoor chambers for several weeks under ozone-depletion conditions prior to BWF experiments (Neale et al. 1994). The relative contribution of UVA and UVB to photoinhibition depends on the light field and the spectral response. In L. hyperborea, UVB weightings are in general several orders greater than UVA, but the magnitude of a photoinhibitory response also depends on the spectral irradiance received at the algal thallus. For instance, under 300 D.U. column ozone, daily photosynthetic production at the surface is potentially reduced 10% (Fig. 9; 0 m kelp, BWFE-PE model). If UVB wavelengths are removed from the calculation, photosynthesis is still inhibited 8%. This finding suggests that under non-ozonedepletion scenarios, UVA contributes significantly more to kelp photoinhibition than UVB. The point is emphasized if reciprocity (BWFH-PE model) is assumed for 10 m kelp using the same irradiance model. In this case, UVA accounts for 38% reduction in potential daily photosynthesis, whereas Fig. 9. Ratio of UV-inhibited, daily integrated photosynthesis to potential daily integrated photosynthesis under different column ozone conditions and different irradiance weighting schemes. Underwater irradiance at 30 min time intervals during the day was modeled using STAR (Ruggaber et al. 1994) and assuming idealized conditions: a clear-sky in April at the study site, a flat sea surface of constant depth (0.5 m MLW, no tidal variation), and locally measured water column attenuation (Miller 2002). Potential photosynthesis was calculated using the Michaelis-Menten relationship, and UV-inhibition was modeled using the four different BWFs determined from carbon-based measurements of Laminaria hyperborea photosynthesis. Variation in column ozone results in modest production losses relative to 600 D.U., irrespective of BWF model and acclimation state. BWF, biological weighting function; MLW, mean low water. LAMINARIA HYPERBOREA BWF UVA+UVB increases total inhibition to only 42%. Clearly, the much greater amount of solar irradiance in the UVA dominates the relatively low biological weightings in the region compared to that in the UVB. Significant UVB photoinhibition is likely only a near-surface phenomenon since UVB is attenuated more rapidly than UVA in coastal waters (Tedetti and Sempéré 2006). Kelps, as other benthic macroalgae, are permanently anchored in their light habitat, and there is little doubt that algae exposed to full sunlight near the surface are able to adapt to ambient UV stress. Kelp individuals from depth are generally more sensitive to high light and UV stress than intertidal and upper-subtidal algae (Larkum and Wood 1993, Dring et al. 1996b, Hanelt et al. 1997a, Johannsson and Snoeijs 2002, Bischof et al. 2006). For example, in transplantation experiments with subarctic Saccharina latissima (=Laminaria saccharina) and L. digitata, subtidal algae showed increased inhibition and longer recovery in UVA and UVB treatments (Hanelt et al. 1997b). Twenty-four-hour percent recovery of optimum quantum yield after a 4 h exposure to PAR+UVA+UVB decreased incrementally with S. latissima collection depth until 7–9 m (Bischof et al. 1998). Photosynthetic response to UV was a function of depth in Norwegian L. hyperborea as well. Calculations based on the weighting functions predict that single-day photosynthesis is 7%–12% less in deepwater sporophytes, depending on BWF model, when both surface and deepwater kelps are exposed to the same near-surface simulated irradiance (Fig. 9, 300 D.U.). Surface and deep L. hyperborea have similar weights in the UVB region, but in 10 m kelp, biological weights were greater in the UVA region. Apparent photosynthetic decreases in 10 m kelp, relative to 0 m kelp, were a result of increased inhibition in the UVA and not necessarily a greater response to UVB. Depth differences in UV response are important as they are evidence, albeit circumstantial, of the presence of active repair and ⁄ or protection mechanisms in algae. Tolerance to UVR is an acclimation process tuned to maximize growth (Roleda et al. 2006), and physiological development of these mechanisms correlates with light history of the alga. High-light algae with daily exposure to UVR necessarily invest resources into pigment production and damage repair enzymes that mitigate potential photoinhibition and photodamage. This investment constitutes energy expenditure otherwise available for allocation to growth and reproduction. Presumably, the expenditure is less than production lost with photoinhibition [and structural and DNA damage as well, see Bischof et al. (2006) and Roleda et al. (2007)]. In the less varied light environment deeper in the water column, algae with little exposure to high light and UVR allocate less energy and nutrient resources to UVR stress response and are less efficient at ameliorating damage if suddenly 581 introduced to UVR. Individuals of a kelp species with a wide depth range adopt the optimal strategy in their light realm, and kelps nearer the most stressful light climate at the surface retain the ability to acclimate to a changing irradiance field. The timescale of acclimation to moderately heightened UVR appears to be on the order of days to weeks but varies with kelp species and depends on thallus age and season (Bischof et al. 1999, 2002, Brouwer et al. 2000, Roleda et al. 2006). Evidence showing minor differences in thallus absorbance indicates that photoprotective pigments are probably not very important contributors to the lower sensitivity of shallow material (Fig. 2). The exposure response in our study, however, suggests that repair and recovery are active in 0 m L. hyperborea. Steady state between UV-inhibition and repair was achieved about 1 h after exposure to UVB+UVA+PAR, and potential quantum yield was recovered after 6 h in darkness (Fig. 3). In contrast, steady-state quantum yield occurred within 15 min in the laboratory-cultured dinoflagellate Gymnodinium sanguineum, and inhibition progressed from initial exposure to steady state without hysteresis (Neale et al. 1998a). The slower rates observed for L. hyperborea suggest that repair capability is modest, even in 0 m kelp. In situ, blade undulations due to wave and water motion potentially result in less overall UV exposure and lessened ability to respond to direct, one-sided exposure, as provided in the experiment. If balance occurs between damage and repair on timescales less than the total time of interest, reciprocity fails (Lesser et al. 1994). Production should then be modeled with an irradiance-based BWF (BWFE-PE). A BWFE-PE model may not, however, be valid for algae from deeper in the water, or from other environments with typically less incident UVR, such as polar latitudes (Neale 2000). For instance, as a percent of control, Laminaria solidungula from >7 m showed a nearly complete collapse in photosynthetic activity during prolonged UV exposure (Bischof et al. 2000). In addition, photosynthetic oxygen evolution was almost completely arrested in subtidal Saccharina latissima and L. solidungula from Spitsbergen after 6 h sunlamp exposure (Aguilera et al. 1999). In cases of reciprocity, photosynthesis in the presence of UVR is best simulated with respect to total cumulative UV exposure (BWFH-PE). Due to logistic constraints, we were not able to complete an exposure response experiment for deepwater kelp, and without more detailed kinetic work and longer duration experiments at both depths, the choice of BWF model is inconclusive. As UVR photoinhibition and repair ⁄ recovery are likely temperature-dependent processes (Sobrino and Neale 2007), it is also significant that the exposure response curve was produced in September and BWF experiments were performed in April. 582 HARLAN L. MILLER III ET AL. Nevertheless, surface kelps at least partially have the capability to counteract UV damage at hourly timescales and are able to recover photosynthetic function in darkness. When irradiance varies slowly, rates can be assumed to be near steady state, and an irradiance (BWFE-PE) model would be appropriate. Given the apparent slow rates of repair, however, steady state may not be achieved in more variable light regimes (e.g., due to cloudiness and wave action), and it may be necessary to implement a time varying prediction using a defined repair rate. This step can be done using the ‘‘R model’’ (Hiriart et al. 2002), but such an analysis requires more kinetic data than that obtained in this study. Indeed, recent work has shown that the R model performs better than the H model for extended exposures of WSC phytoplankton to moderate UVR (Fritz et al. 2008). Therefore, we report the photosynthetic response to UVR using both BWFE-PE and BWFH-PE functions and emphasize that these bracket the full range of response dependence on repair rate (Fig. 9). Actual response probably lies somewhere in between these two predictions. The effect of UVR on the photosynthetic quotient is further evidence of depth-dependent physiology in kelps (Fig. 8). In this study, the average PQ was greater in kelp collected from deeper water. The PQ declined with increasing UVR in both shallow- and deepwater L. hyperborea, but the response was nearly three times more severe in the deep samples. The PQ response possibly results from an increase in oxygen consumption with UVR stress and not a relative increase in carbon fixation. Increased respiration, in addition to light-enhanced respiration (Ekelund 2000), could account for greater oxygen consumption during UV stress. Increased oxygen consumption could result from increased cellular metabolism and protein synthesis associated with repair of damaged reaction center proteins in PSII and RUBISCO (Vincent and Neale 2000), but other studies with kelp and macroalgae suggest that no significant increases in dark respiration occur under UV stress (Larkum and Wood 1993, Clendennen et al. 1996, Aguilera et al. 1999). Reduction in PQ could otherwise be attributed to other cellular and photochemical processes that consume O2, including (1) increased oxygenase activity of RUBISCO, that is, photorespiration; (2) O2 uptake and transformation to H2O2 in the Mehler reaction or other antioxidant activities in the chloroplast (Aguilera et al. 2002); and (3) UV photolysis of dissolved organic matter to produce oxygen radicals and H2O2 in incubation seawater (Neale and Kieber 2000, Vincent and Neale 2000). Regardless of the underlying process, the BWFs capture both the differential response of O2- and Cbased photosynthesis and the depth-dependent decline in PQ (Fig. 6). Ultimately, investigators are interested in the effect of increased UVR on primary productivity and the ecological consequences of production losses to trophic connections and ecosystem function. These questions generally require a modeling approach and knowledge of solar irradiance, how it propagates through the atmosphere and sea, and the biological response to incident light. The L. hyperborea BWFs provide an empirical description of UV photoinhibition applicable for springtime populations at high-light and low-light acclimation states. Even with a simplified irradiance model without tides, the relationship allows for the estimation that near-surface (0.5 m) kelp production is potentially decreased 6% under 100 D.U. column ozone compared to 600 D.U. (Fig. 9), a modest amount considering the depth-integrated production of the kelp forest as a whole. Yet the biological response is expected to change with time and space. Significant questions remain that concern the variation in BWFs with thallus age and location, life-history stage, season and microclimate, air exposure at low tide, nutrient status, and induction of protection and repair processes in a changing light climate (Davison et al. 2007, Roleda et al. 2007). We extend our great appreciation to S. Fredriksen and the kelp team from the University of Oslo, Norway, and our special thanks to R. Myklebust and his gracious crew at Finnøy Sjøhus. We also acknowledge the contributions of J. Lempa, E. Tang, A. Banaszak, K. Jackson, and J. Keller and critical readings by J. Brandes, E. Buskey, T. Villareal, H. Alexander, M. Scanlin, and C. Aumack. A. Stangelmayer of PreSens is appreciated for the generous loan of the Microx TX oxygen system. J. Biedenbach donated instrument time on the IO Analytical TOC analyzer. This research was funded in part by the National Science Foundation (OPP-9622483), the Smithsonian Institute predoctoral fellowship program, and the Environmental Science Institute Summer Fellowship program at the University of Texas at Austin. Aguilera, J., Dummermuth, A., Karsten, U., Schriek, R. & Wiencke, C. 2002. Enzymatic defences against photooxidative stress induced by ultraviolet radiation in Arctic marine macroalgae. Polar Biol. 25:432–41. Aguilera, J., Karsten, U., Lippert, H., Vögele, B., Philipp, E., Hanelt, D. & Wiencke, C. 1999. Effects of solar radiation on growth, photosynthesis and respiration of marine macroalgae from the Arctic. Mar. Ecol. Prog. Ser. 191:109–19. Banaszak, A. T. & Neale, P. J. 2001. Ultraviolet radiation sensitivity of photosynthesis in phytoplankton from an estuarine environment. Limnol. Oceanogr. 46:592–603. Bischof, K., Gómez, I., Molis, M., Hanelt, D., Karsten, U., Lüder, U., Roleda, M. Y., Zacher, K. & Wiencke, C. 2006. Ultraviolet radiation shapes seaweed communities. Rev. Environ. Sci. Biotechnol. 5:141–66. Bischof, K., Hanelt, D., Aguilera, J., Karsten, U., Vögele, B., Sawall, T. & Wiencke, C. 2002. Seasonal variation in ecophysiological patterns in macroalgae from an Arctic fjord. I. Sensitivity of photosynthesis to ultraviolet radiation. Mar. Biol. 140:1097– 106. Bischof, K., Hanelt, D., Tüg, H., Karsten, U., Brouwer, P. E. M. & Wiencke, C. 1998. Acclimation of brown algal photosynthesis to ultraviolet radiation in Arctic coastal waters (Spitsbergen, Norway). Polar Biol. 20:388–95. Bischof, K., Hanelt, D. & Wiencke, C. 1999. Acclimation of maximal quantum yield of photosynthesis in the brown alga Alaria esculenta under high light and UV radiation. Plant Biol. 1:435– 44. LAMINARIA HYPERBOREA BWF Bischof, K., Hanelt, D. & Wiencke, C. 2000. Effects of ultraviolet radiation on photosynthesis and related enzyme reactions of marine macroalgae. Planta 211:555–62. Brouwer, P. E. M., Bischof, K., Hanelt, D. & Kromkamp, J. 2000. Photosynthesis of two Arctic macroalgae under different ambient radiation levels and their sensitivity to enhanced UV radiation. Polar Biol. 23:257–64. Caldwell, M. M. 1971. Solar ultraviolet radiation and the growth and development of higher plants. In Giese, A. C. [Ed.] Photophysiology. Academic Press, New York, pp. 131–77. Caldwell, M. M., Camp, L. B., Warner, C. W. & Flint, S. D. 1986. Action spectra and their key role in assessing biological consequences of solar UV-B radiation change. In Worrest, R. C. & Caldwell, M. M. [Eds.] Stratospheric Ozone Reduction, Solar Ultraviolet Radiation and Plant Life. Springer-Verlag, Heidelberg, Germany, pp. 87–111. Clendennen, S. K., Zimmerman, R. C., Powers, D. A. & Alberte, R. S. 1996. Photosynthetic response of the giant kelp Macrocystis pyrifera (Phaeophyceae) to ultraviolet radiation. J. Phycol. 32:614–20. Coohill, T. P. 1994. Exposure response curves, action spectra and amplification factors. In Biggs, R. H. & Joyner, M. E. B. [Eds.] Stratospheric Ozone Depletion ⁄ UV-B Radiation in the Biosphere. Springer-Verlag, Berlin, pp. 57–62. Cullen, J. J. & Lesser, M. P. 1991. Inhibition of photosynthesis by ultraviolet radiation as a function of dose and dosage rate: results from a marine diatom. Mar. Biol. 111:183–90. Cullen, J. J. & Neale, P. J. 1997. Biological weighting functions for describing the effects of ultraviolet radiation on aquatic systems. In Häder, D. P. [Ed.] The Effects of Ozone Depletion on Aquatic Ecosystems. R. G. Landes Company, Austin, Texas, pp. 97–118. Cullen, J. J., Neale, P. J. & Lesser, M. P. 1992. Biological weighting function for the inhibition of phytoplankton photosynthesis by ultraviolet radiation. Science 258:646–50. Davison, I. R., Jordan, T. L., Fegley, J. C. & Grobe, C. W. 2007. Response of Laminaria saccharina (Phaeophyta) growth and photosynthesis to simultaneous ultraviolet radiation and nitrogen limitation. J. Phycol. 43:636–46. Dring, M. J., Makarov, V., Schoschina, E., Lorenz, M. & Lüning, K. 1996a. Influence of ultraviolet-radiation on chlorophyll fluorescence and growth in different life-history stages of three species of Laminaria (Phaeophyta). Mar. Biol. 126:183–91. Dring, M. J., Wagner, A., Boeskov, J. & Lüning, K. 1996b. Sensitivity of intertidal and subtidal red algae to UV-A and UV-B radiation, as monitored by chlorophyll fluorescence measurements: influence of collection depth and season, and length of irradiance. Eur. J. Phycol. 31:293–302. Dring, M. J., Wagner, A. & Lüning, K. 2001. Contribution of the UV component of natural sunlight to photoinhibition of photosynthesis in six species of subtidal brown and red seaweeds. Plant Cell Environ. 24:1153–64. Dunlap, W. C., Chalker, B. E. & Oliver, J. K. 1986. Bathymetric adaptations of reef-building corals at Davies Reef, Great Barrier Reef, Australia. III. UV-B absorbing compounds. J. Exp. Mar. Biol. Ecol. 104:239–48. Ekelund, N. G. A. 2000. Interactions between photosynthesis and ‘light-enhanced dark respiration’ (LEDR) in the flagellate Euglena gracilis after irradiation with ultraviolet radiation. J. Photochem. Photobiol. B Biol. 55:63–9. Eyvind, A. & Højerslev, N. K. 2001. Attenuation of ultraviolet irradiance in northern European coastal waters. Oceanologia 43:139–68. Fairhead, V. A., Amsler, C. D., McClintock, J. B. & Baker, B. J. 2005. Variation in phlorotannin content within two species of brown macroalgae (Desmarestia anceps and D. menziesii) from the Western Antarctic Peninsula. Polar Biol. 28:680–6. Flint, S. D. & Caldwell, M. M. 2003. A biological spectral weighting function for ozone depletion research with higher plants. Physiol. Plant. 117:137–44. Forster, R. M. & Lüning, K. 1996. Photosynthetic response of Laminaria digitata to ultraviolet A and B radiation. Sci. Mar. 60(Suppl. 1):65–71. 583 Franklin, L. A. & Forster, R. M. 1997. The changing irradiance environment: consequences for marine macrophyte physiology, productivity and ecology. Eur. J. Phycol. 32:207–32. Franklin, L. A., Kräbs, G. & Kuhlenkamp, R. 2001. Blue light and UV–A radiation control the synthesis of mycosporine-like amino acids in Chondrus crispus (Florideophyceae). J. Phycol. 37:257–70. Fritz, J. J., Neale, P. J., Davis, R. F. & Peloquin, J. A. 2008. Response of Antarctic phytoplankton to solar UVR exposure: inhibition and recovery of photosynthesis in coastal and pelagic assemblages. Mar. Ecol. Prog. Ser. 365:1–16. Gómez, I., Figueroa, F. L., Sousa-Pinto, I., Viñegla, B., Pérez-Rodrı́guez, E., Maestre, C., Coelho, S., Felga, A. & Pereira, R. 2001. Effects of UV radiation and temperature on photosynthesis as measured by PAM fluorescence in the red alga Gelidium pulchellum (Turner) Kützing. Bot. Mar. 44:9–16. Gómez, I., Ulloa, N. & Orostegui, M. 2005. Morpho-functional patterns of photosynthesis and UV sensitivity in the kelp Lessonia nigrescens (Laminariales, Phaeophyta). Mar. Biol. 138:231–40. Häder, D.-P., Lebert, M., Sinha, R. P., Barbieri, E. S. & Helbling, E. W. 2002. Role of protective and repair mechanisms in the inhibition of photosynthesis in marine macroalgae. Photochem. Photobiol. Sci. 1:809–14. Han, T. & Kain (Jones), J. M. 1992. Blue light sensitivity of UVirradiated young sporophytes of Laminaria hyperborea. J. Exp. Mar. Biol. Ecol. 158:219–30. Han, T. & Kain (Jones), J. M. 1993. Blue light photoreactivation in ultraviolet-irradiated young sporophytes of Alaria esculenta and Laminaria saccharina (Phaeophyta). J. Phycol. 29:79–81. Hanelt, D., Melchersmann, B., Wiencke, C. & Nultsch, W. 1997a. Effects of high light stress on photosynthesis of polar macroalgae in relation to depth distribution. Mar. Ecol. Prog. Ser. 149:255–66. Hanelt, D., Wiencke, C. & Nultsch, W. 1997b. Influence of UV radiation on the photosynthesis of Arctic macroalgae in the field. J. Photochem. Photobiol. B Biol. 38:40–7. Henry, B. E. & van Alstyne, K. L. 2004. Effects of UV radiation on growth and phlorotannins in Fucus gardneri (Phaeophyceae) juveniles and embryos. J. Phycol. 40:527–33. Hiriart, V. P., Greenberg, B. M., Guildford, S. J. & Smith, R. E. H. 2002. Effects of ultraviolet radiation on rates and size distribution of primary production by Lake Erie phytoplankton. Can. J. Fish. Aquat. Sci. 59:317–28. Huovinen, P., Gómez, I. & Lovengreen, C. 2006. A five-year study of solar ultraviolet radiation in southern Chile (39 S): potential impact on physiology of coastal marine algae? Photochem. Photobiol. 82:515–22. Johannsson, G. & Snoeijs, P. 2002. Macroalgal photosynthesis responses to light in relation to thallus morphology and depth zonation. Mar. Ecol. Prog. Ser. 244:63–72. Jones, L. W. & Kok, B. 1966. Photoinhibition of chloroplast reactions. 1) Kinetics and action spectra. Plant Physiol. 41:1037–43. Karsten, U., Bischup, K. & Wiencke, C. 2001. Photosynthetic performance of Arctic macroalgae after transplantation from deep to shallow waters. Oecologia 127:11–29. Karsten, U., Franklin, L. A., Lüning, K. & Wiencke, C. 1998a. Natural ultraviolet radiation and photosynthetically active radiation induce formation of mycosporine-like amino acids in the marine macroalga Chrondus crispus (Rhodophyta). Planta 205:257–62. Larkum, A. W. D. & Wood, W. F. 1993. The effect of UV-B radiation on photosynthesis and respiration of phytoplankton, benthic macroalgae and seagrasses. Photosynth. Res. 36:17–23. Lesser, M. P., Cullen, J. J. & Neale, P. J. 1994. Carbon uptake in a marine diatom during acute exposure to ultraviolet B radiation: relative importance of damage and repair. J. Phycol. 30:183–92. Miller, H. L. 2002. Photosynthetic response of Scandinavian kelp forests to stratospheric ozone depletion. Ph.D. dissertation, The University of Texas, Austin, 170 pp. Miller, H. L. & Dunton, K. H. 2007. Stable isotope (13C) and O2 micro-optode alternatives for measuring photosynthesis in seaweeds. Mar. Ecol. Prog. Ser. 329:85–97. 584 HARLAN L. MILLER III ET AL. Neale, P. J. 2000. Spectral weighting functions for quantifying effects of UV radiation in marine ecosystems. In de Mora, S., Demers, S. & Vernet, M. [Eds.] The Effects of UV Radiation in the Marine Environment. Cambridge University Press, Cambridge, UK, pp. 72–100. Neale, P. J., Banaszak, A. T. & Jarriel, C. R. 1998a. Ultraviolet sunscreens in Gymnodinium sanguineum (Dinophyceae): mycosporine-like amino acids protect against inhibition of photosynthesis. J. Phycol. 34:928–38. Neale, P. J., Cullen, J. J. & Davis, R. F. 1998b. Inhibition of marine photosynthesis by ultraviolet radiation: variable sensitivity of phytoplankton in the Weddell–Scotia Confluence during the austral spring. Limnol. Oceanogr. 43:433–48. Neale, P. J. & Fritz, J. J. 2002. Experimental exposure of plankton suspensions to polychromatic ultraviolet radiation for determination of spectral weighting functions. In Slusser, J., Herman, J. R. & Gao, W. [Eds.] Ultraviolet Ground- and Space-based Measurements, Models, and Effects. SPIE–The International Society for Optical Engineering, Bellingham, Washington, pp. 291–6. Neale, P. J. & Kieber, D. J. 2000. Assessing biological and chemical effects of UV in the marine environment: spectral weighting functions. In Hester, R. E. & Harrison, R. M. [Eds.] Causes and Environmental Implications of Increased UV-B Radiation. The Royal Society of Chemistry, Cambridge, UK, pp. 61–83. Neale, P. J., Lesser, M. P. & Cullen, J. J. 1994. Effects of ultraviolet radiation on the photosynthesis of phytoplankton in the vicinity of McMurdo station, Antarctica. In Weiler, C. S. & Penhale, P. A. [Eds.] Ultraviolet Radiation in Antarctica: Measurement and Biological Effects. American Geophysical Union, Washington, D.C., pp. 125–42. Pavia, H. & Brock, E. 2000. Extrinsic factors influencing phlorotannin production in the brown alga Ascophyllum nodosum. Mar. Ecol. Prog. Ser. 193:285–94. Roleda, M. Y., Hanelt, D. & Wiencke, C. 2006. Growth and DNA damage in young Laminaria sporophytes exposed to ultraviolet radiation: implications for depth zonation of kelps on Helgoland (North Sea). Mar. Biol. 148:1201–11. Roleda, M. Y., Wiencke, C., Hanelt, D. & Bischof, K. 2007. Sensitivity of the early life stages of macroalgae from the Northern Hemisphere to ultraviolet radiation. Photochem. Photobiol. 83:851–62. Ruggaber, A., Dlugi, R. & Nakajima, T. 1994. Modelling of radiation quantities and photolysis frequencies in the troposphere. J. Atmos. Chem. 18:171–210. Rundel, R. D. 1983. Action spectra and estimation of biologically effective UV radiation. Physiol. Plant. 58:360–6. Shibata, K. 1958. Spectrophotometry of intact biological materials. J. Biochem. Tokyo 45:599–623. Sjøtun, K., Fredriksen, S., Rueness, J. & Lein, T. E. 1995. Ecological studies of the kelp Laminaria hyperborea (Gunnerus) Foslie in Norway. In Skjoldal, H. R., Hopkins, C., Erikstad, K. E. & Leinaas, H. P. [Eds.] Ecology of Fjords and Coastal Waters. Elsevier Science, Amsterdam, pp. 525–36. Sobrino, C. & Neale, P. J. 2007. Short-term and long-term effects of temperature on photosynthesis in the diatom Thalassiosira pseudonana under UVR exposure. J. Phycol. 43:426–36. Swanson, A. K. & Druehl, L. D. 2002. Induction, exudation and the UV protective role of kelp phlorotannins. Aquatic Bot. 73:241– 53. Tedetti, M. & Sempéré, R. 2006. Penetration of ultraviolet radiation in the marine environment: a review. Photochem. Photobiol. 82:389–97. Vincent, W. F. & Neale, P. J. 2000. Mechanisms of UV damage to aquatic organisms. In de Mora, S., Demers, S. & Vernet, M. [Eds.] The Effects of UV Radiation in the Marine Environment. Cambridge University Press, Cambridge, UK, pp. 72–100. Supplementary Material The following supplementary material is available for this article: Figure S1. Comparison of four photosynthesis models for the kelp Laminaria hyperborea collected from (A) 0 m (mean low water, MLW), (B) 5 m, and (C) 15 m. This material is available as part of the online article. Please note: Wiley-Blackwell are not responsible for the content or functionality of any supplementary materials supplied by the authors. Any queries (other than missing material) should be directed to the corresponding author for the article.