Ultraviolet Radiation and Bio-optics in Crater Lake, Oregon, 2005
DISCUSSION
Proxy Measurements for UV Attenuation Versus Depth
A number of indirect proxy measurements for UV attenuation have been utilized in our study. Proxies can be useful when conditions are not appropriate for measuring Ed,Z directly (e.g. rapidly changing sky conditions, or low sun angle) or when a UV radiometer is not available. Many take advantage of some optical property of phytoplankton (e.g. scattering of the red beam of a transmissometer (e.g. 660 nm), fluorescence (Fchl, 683 nm), absorption at the red chlorophyll-a peak (675 nm) or over the blue part of the visible spectrum (400–500 nm). The strong correlation of Fchl from chlorophyll-a and Kd,380 (Figure 11A) can be caused by the direct absorption of UVR by chlorophyll-a, but is also likely to vary with depth and UV wavelength because of the absorption of other molecules such as MAAs and because photoacclimation and photochemical quenching change the relationship between Fchl and [chlorophyll-a] (Boss et al., this issue). Variation in accessory and UV-B screening pigments is suggested by the changing ratios of Kd320:Kd380 in Figure 6, and of ap330:ap675 in Figure 12B. It is not uncommon in clear aquatic systems for the biomass peak to occur at a different depth from the chlorophyll-a maximum (Fennel & Boss, 2003).
The concentration of dissolved organic carbon (DOC) has frequently been used as a proxy for UV attenuation (reviewed in Hargreaves, 2003) but measurements in Crater Lake are limited to a few dates and low reproducibility (Figure 8). Measurements of phytoplankton absorbance can be complicated by the tendency for MAAs to be released from phytoplankton during filtration (Laurion et al., 2003) as this will tend to cause elevated values in CDOM measured after filtration, and by the limits of detection using a 10 cm cuvette and laboratory spectrophotometer (but see Boss et al., this issue, for a long-pathlength in situ method). Published equations relating Kd,320 and DOC concentration (Hargreaves, 2003) vary because of the optical quality of DOC (photobleaching and source contribute to this variation). At low [DOC] the fit of this type of equation also is likely to depend on a correlation of [DOC] with [phytoplankton] because of the low absolute absorption by CDOM relative to phytoplankton. Within the uncertainty of the DOC measurements, Crater Lake is similar to the extremely transparent Lake Vanda, Antarctica. Comparing the DOC and UV attenuation data for L. Vanda and Crater Lake, Vincent et al., (1998) reported L. Vanda DOC = 0.3 g/m3 and Kd,320=0.055, while our data for Crater Lake during summer 1999 (averaged over 0–40 m), show DOC = 0.1 g m-3 and Kd,320 = 0.09 m-1. At this low level of DOC concentration the technique to account for instrument blanks is crucial (Sharp et al., 1993). The large difference in DOC-specific UV-B attenuation (Kd,320-Kdw,320):DOC = 0.03 for Lake Vanda and 0.5 for Crater Lake 0–40 m in 1999) suggests either a difference in DOC measurement technique or a difference in the optical contributions of phytoplankton and aCDOM, for example by photobleaching of aCDOM in Crater Lake.
