Ecology of UV

Ecology of UV Homepage

UV introduction and basic optics


Autoregressive modeling

DNA damage and repair



Weighting functions

Effects on Phytoplankton

Effects on Protozoa

Effects on Zooplankton

Effects on Fish

Back to Ecology of UV main page

Back to UV Lakes homepage







Weighting Functions
prepared by Pat Neale

Part 1: Concepts and Current Understanding
(What's all the fuss about?)

Atmospheric ozone has been decreasing in the Southern Hemisphere (slide 1) where the ozone hole tends to last longer and be larger (comparing 2000 to 2001 to the average over 1979-1992). Ozone is also decreasing (by 0.3% per year) at mid-latitides (slide 2). This decrease in ozone corresponds to an increase in UV-B reaching the earth. Erythemal (sun-burn) irradiance has been found to increase by an average of 0.6% per year (from 1976 to 2000, April averages, as monitored at the Smithsonian Environmental Research Center) (link to SERC) (slide 3) though long-term reductions in aerosols and cloudiness have also contributed to this trend.

Most responses to UV are spectrally dependent. The amount of energy carried by a photon of light is inversely proportional to the wavelength. Therefore UV-A (320-400 nm) is less damaging than UV-B (280-320 nm) which is less damaging than UV-C (200-280 nm). UV-C is filtered by the ozone layer. Organisms which live in water are further protected from UV due to the attenuation properties of water. The red line in the graph below represents solar irradiance, while the blue and white lines represent the relative depth penetration of irradiance as a function of wavelength for examples of clear oceanic water (blue) and turbid estuarine water (white).

If any biological or chemical effect is expressed as a function of radiant exposure for regimes with different spectral irradiance (i.e., different depths in the water column, times of day or year, or different artificial sources of UV), a weighting function is used, either explicitly or implicitly. Weighting functions are needed to assess effects of varying solar UV as affected by ozone and other atmospheric changes, to assess effect of depth and water type on underwater UV, and to compare exposures in experiments using different UV sources. Once a weighting function is calculated, it can be multiplied by solar irradiance to determine the Biologically Effective Exposure (H*), or in other words the wavelengths which most negatively affect the given species, as is seen in the graph below:

Interest in phytoplankton biological weighting functions goes back several decades.

The first biological weighting functions (BWFs) were for relative response - a low spectral resolution resulting in stepwise functions as seen below:

More detailed BWFs with absolute weights came later. Weights were found to decrease sharply through the UV-B, and generally there was less variation in the UV-A (with some exceptions) as is seen in the many BWFs for phytoplankton graphed below.

Cultures and assemblages differ greatly in their sensitivity and thus will yield different BWFs. The same species assemblages can also yield different BWFs when sampled in different environments (such as different strata of the water column). BWFs were also found to vary by an order of magnitude in the Rhode River when done repeatedly over a one year period (slide 4). However, average UV sensitivity did not vary over season in the Rhode River (slide 5). Overall, the averages of BWFs for natural populations have VERY similar average ranges in different environments, such as the Antarctic and Rhode River (slide 6). It may be that the average BWF for phytoplankton does not vary much between optically deep environments. BWFs have also been calculated for carbon allocation into macromolecular pools, such as carbohydrates, lipids and proteins (slide 7).

BWFs for aquatic organisms are similar to those for phytoplankton but are slightly steeper. BWFs have been calculated for UV inhibition of thymidine and leucine uptake in bacteria (slide 8), and for UV induced mortality of various marine and freshwater zooplanton (slide 9) (for the two fish species on the graph, the life history stage tested was eggs). Thus when different tropic levels are compared, we get the following:

Mechanisms causing BWF variation could include:

Comparative spectroscopy with cultures can help diagnose which mechanism(s) are operating in different cultures or species. It should be noted that BWFs are based on the net effect of experimental exposures.

Part 2: Methodology and Analysis
(OK, so how do I do this, anyway?)

The first BWFs were calculated using monochromatic exposures. These illuminate specimen with narrowband irradiance centered at several wavelengths in the UV, measure effects after various periods of irradiation (assuming reciprocity), and the weight was equal to the effect divided by the exposure (Weight = Effect/Exposure). An advantage of such exposures is that it is a direct measure of UV spectral effectiveness. A disadvantage is that it is time intensive. Application to solar irradiance assumes effects of each wavelength are independent (slide 10), which is not the case for biological responses.

One method used to produce Biological Weighting Functions is the Photoinhibitron situated at SERC.

The Photoinhibitron (also known affectionately as the "Beast") uses a Xenon Arc Lamp providing polychromatic exposures. Long pass filters placed under the quartz cuvettes into which the samples are placed are used to cut out certain wavelenghts (slide 11) and meshes are used to vary the irradiance that organisms within each filter treatment receive (see Filter Orientation box in figure above). The result is a bracketing of solar exposures (slide 12), and thus this data can be used to estimate UV effects in natural conditions. Results of such experiments are found below:

Many data points are needed for each wavelength to determine the response of the organism. Another method is the Solar Phototron pictured below:

The Solar Phototron is limited by the range and type of Schott long pass filters and by the weather on the day of the experiment. However, the advantage of this method is that natural sunlight is used to expose organisms. Results from this method are found below:

The X axis is in nm, the blue symbols are the mean observed survival with 95% confidence intervals, and the red triangles are the modeled BWF (from Williamson et al 2001).

BWFs can be estimated using many different methods. You can estimate a BWF using the Difference method (slide 13), where you would look at the differences in response as a function of differences in exposure (presence or absence of shorter wavelengths). This would be done over all successive treatments and plotted. The answer represents an average effect based on the integrated response to each wavelength range. The coefficients are plotted as stepwise functions to show over which ranges the averages apply.

You can fit a BWF to a general equation (slide 14), where you assume a fit to an exponential function. You would find the optimum coefficient that produced the best fit to your data. This method constrains what kind of WF is obtained because you are trying to fit to a particular exponential function.

You can estimate a BWF using the PCA method (slide 15). Again this is based on representing the differential response to exposures containing varying wavelength ranges. However, the differences are represented by Principal Components. Usually, you can fit 3 components to explain 99% of the variation between the different spectral treatments. You would model the BWF according to those 3 components, and there would be no strong constraint to the shape of the model. This method works well with phytoplankton data, however, with bacterial data, 3 components do not approximate enough of the variation (the wavelength responses in bacteria are more varied). Actual components in this method are purely statistical and not biological. The components are also independent of each other.

Or you can fit the BWF using segments (slide 16), where you segment UV into several wavelength ranges, estimate the BWF at the boundaries of the segments and then interpolate between them to get 1 nm resolution. The boundary values are adjusted to improve the fit to the data using the fully interpolated BWF. You would start with 2 boundary points, and then adjust the number of segments to improve the fit between predicted and observed data.

Determining which model to use depends on the kinetics of the organism under experimentation. If constant repair exceeds damage by UV, a BWF cannot be estimated because there will be no response to UV exposure.

One case is shown below:

where P is the rate of photosynthesis, and k is the rate of damage (inhibition) and r is rate of repair (recovery). UV inhibition of photosynthesis has been modeled (slide 17), and a different model is used when there is no repair, a steady-state is not attained and reciprocity is satisfied (slide 18), or assumed (slide19).

Experiments with zooplankton have shown evidence for a repair threshold. Some questions which still need to be answered are:

It is hoped that experiments in the UV-B Lamp Phototron (link) will resolve these issues. An example of threshold response can be seen for photosynthesis (slide 20). Some cases may require a full time-resolved approach (slide 21). And once the BWF is estimated, BWF confidence intervals can be estimated as well (slide 22).

Part 3: Modeling Approaches
(After all this work, do we get something really useful?)

UV effects can be modeled in situ:

A production model for the Weddell-Scotia Confluence (WSC, Southern Ocean) was calculated (slide 23), which enables assessment of UV effects on productivity under different ozone and mixing scenarios (slide 24). When the interactive effects of ozone depletion, biological weighting functions, and vertical mixing were compared, it was found that inhibition of photosynthesis was greater when mixing was shallow than when it was deep or when there was no mixing at all (slide 25).


The Ozone Hole Tour

Smithsonian Environmental Research Center (SERC)


Banaszak, A.T., and Neale, P.J. 2001. UV Sensitivity of photosynthesis in phytoplankton from an estuarine environment. Limnol. Oceanogr. 46: 592-600.

Boucher, N.P., and Prezelin, B.B. 1996. An in situ biological weighting function for UV inhibition of phytoplankton carbon fixation in the Southern Ocean. Mar. Ecol. Prog. Ser. 144: 223-236.

Cullen, J.J., Neale, P.J., and Lesser, M.P. 1992. Biological weighting function for the inhibition of phytoplankton photosynthesis by ultraviolet radiation. Science 258: 646-650.

Helbling, E.W., Villafa, V., Ferrario, M., and Holm-Hansen, O. 1992. Impact of natural ultraviolet radiation on rates of photosynthesis and on specific marine phytoplankton species. Mar. Ecol. Progr. Ser. 80: 89-100.

Litchman, E., Neale, P.J., and Banaszak, A.T. 2002. Increased sensitivity to ultraviolet radiation in nitrogen-limited dinoflagellates: photoprotection and repair. Limnol. Oceanogr. 47: 86-94.

Lubin, D., Mitchell, B.G., Frederick, J.E., Alberts, A.D., Booth, C.R., Lucas, T., and Neuschuler, D. 1992. A contribution toward understanding the biospherical significance of antartic ozone depletion. Journal of Geophysical Research 97: 7817-7828.

Miller, G.W. 2000. Wavelength and Temperature Dependent Quantum Yields for Photochemical Formation of Hydrogen Peroxide In Seawater. Masters Thesis, State University of New York, College of Environmental Science and Forestry.

Neale, P.J., Litchman, E., Sobrino, C., Callieri, C., Morabito, G., Montecino, V., Huot, Y., Bossard, P., Steiner, D., and Lehmann, C. 2001. Quantifying the response of phytoplankton photosynthesis to ultraviolet radiation: Biological weighting functions versus in situ measurements in two Swiss lakes. Aquat. Sci. 63: 265-285.

Neale, P.J., Davis, R.F., Cullen, J.J. 1998. Interactive effects of ozone depletion and vertical mixing on photosynthesis of Antarctic phytoplankton. Nature 392: 585-589.

Neale, P.J., Banaszak, A.T., and Jarriel, C.R. 1998. Ultraviolet sunscreens in dinoflagellates: Mycosporine-like amino acids protect against inhibition of photosynthesis. J. Phycology 34: 928-938.

Neale, P.J., Lesser, M.P., and Cullen, J.J. 1994. Effects of ultraviolet radiation on the photosynthesis of phytoplankton in the vicinity of McMurdo Station (78?S). In: Weiler CS, Penhale PA (eds) Ultraviolet Radiation in Antarctica: Measurements and Biological Effects, Vol 62. Am. Geophysical Union, Washington, D.C., p 125-142

Rundel, R.D. 1983. Action spectra and estimation of biologically effective UV radiation. Physiol. Plant. 58: 360-366.

Smith, R.C., Baker, K.S., Holm-Hansen, O., and Olson, R.S. 1980. Photoinhibition of photosynthesis in natural waters. Photochem. Photobiol. 31: 585-592.

Sobrino, C., Neale, P.J., and Lubian, L. (subm.) Interaction of UV-radiation and inorganic carbon supply in the inhibition of photosynthesis: Spectral and temporal responses of two microalgae with different carbon concentration mechanisms. J. Phycol.

Steemann Nielsen, E. 1964. On a complication in marine productivity work due to the influence of ultraviolet light. J. Cons. Perm. Int. Explor. Mer 22: 130-135.

Williamson, C.E., Neale, P.J., Grad, G., De Lange, H.J., and Hargreaves, B.R. 2001. Beneficial and detrimental effects of UV on aquatic organisms: Implications of spectral variation. Ecol. Appl. 11: 1843-1857. PDF file.
























































































































































































































































last modified on Feb 12, 2009