Ultraviolet radiation and primary production by Lake Erie phytoplankton communities

Abstract

The aim of this work was to investigate and quantify the effects of ultraviolet radiation (UVR) n L. Erie phytoplankton communities. Briefly, primary production, measured as ^14C incorporated over time, was determined during the late spring and summer months of 1997 and 1998. While all three basins of L. Erie were sampled, the East basin was under represented relative to both the Central and West basins. In all experiments, determination of the effects of UVR on primary production was carried out using natural solar radiation which was selectively screened by various optical filters or modified using ultraviolet B (UVB) lamps.

In L. Erie, UVB inhibited primary production severely (58%) but only in very severe exposures, as would be experienced at the surface. The kinetics of inhibition were, however, rapid suggesting that short exposures, as would occur under mixing, could result in significant inhibition. No simple and strong relationships could be identified between the efficiency of UVB in inducing photoinhibition and measures of phytoplankton light history or micro- and nanoplankton dominant genera. However, UVB inhibition efficiency was higher for phosphorus (P) deficient than P-replete algal assemblages. UVR significantly affected the proportion of production in the particular (POC; <2 and >2um) and dissolved (DOC) organic carbon pools resulting in higher production of DOC and reduced production in the small particulate carbon fraction (i.e. picoplankton). UVR may, therefore, have important effects on the carbon flow dynamics of for near-surface population of L. Erie. Additionally, basin-specific PAR and UVR transmission properties showed that the West basin of L. Erie is potentially more susceptible to UVR photoinhibition than either the East or Central basins. However, predictions of the effects of UVR on primary production remain elusive without further model development. Furthermore, evidence of a significant role of UVA as an inhibitor of primary production underscores the need for a more inclusive approach, that is modelling photoinhibition in terms of total spectral effects.

A new model (the R model) for the response of photosynthetic carbon fixation to UVR was formulated, allowing finite rates of recovery and the incorporation of wavelength interactions, and was tested against observations on phytoplankton communities in L. Erie. Additionally, biological weighting functions (BWFs), the first for freshwater phytoplankton assemblages, were determined. The application of both the R model and the BWFs successfully described the kinetics of UVR- and PAR-dependent photoinhibition in L. Erie. These results indicated that, in this large lake, recovery rate processes were active yet not in equilibrium with damage rate processes and that variations both in spectral sensitivity (expressed by the BWFs) and recovery rate constants were important determinants of phytoplankton response to UVR.

Possessing both measures of spectral sensitivity and recovery rates, losses of daily production were calculated under mixing and water transparency scenarios that span the common range of conditions in L. Erie. Predictions of daily integrated photoinhibition (PIint) were insensitive to variations in mixing rates, although, simulated mixed conditions always resulted in higher daily losses of production than simulated motionless water column. The contribution of each spectral waveband to PIint also varied little with mixing scenarios. Unexpectedly, PIint was always highest in the relatively turbid, and shallow West basin, regardless of the mixing scenario. Furthermore, this inhibition was enhanced when the presence of shallow 'temporary' thermoclines, which frequently develop in this basin, was stimulated. Although ozone depletion is a major concern, a simulated 20% reduction of ozone over L. Erie resulted in a minimal increase of PIint even in the relatively sensitive West basin. Sensitivity analysis revealed that variations of phytoplankton susceptibility to UVR, and of their recovery rates, over the ranges observed to date in L. Erie, could lead to 5% to 20% loss of production under normal mixing conditions.

Overall, UVA, rather than UVB or PAR, appears to be the major waveband responsible for photoinhibition of primary production in L. Erie. The impacts of climate change and human activity on transmission of UVA are therefore more of a concern than ozone-related variations of UVB for primary production in the Laurentian Great Lakes.