The body of evidence to date indicates that changes in UV effects on plants are often relatively subtle, but these small changes may have important ecological implications, particularly by secondary or indirect effects (Caldwell et al., 2007). The small changes in leaf optical properties and secondary chemistry, and minimal differences in productivity under simulated global dimming and UV-B reduction found in our study, are consistent with this perspective.
Changes in the quantity of UV exposure by plants have been shown to affect morphology in ways that can affect competitive interactions, with changes in biomass among species corresponding to changes in the quantity of UV-B radiation (Barnes et al., 1990b; Ryel et al., 1990; Robson et al., 2003). Changes in yield, as have been observed under small changes in PAR (e.g., Fischer, 1975), and changes in root:shoot allocation (Zaller et al., 2002; Rinnan et al., 2005) may affect competitive interactions (Rinnan et al. 2006). Exposure to UV-B radiation also may affect herbivory rates (Ballare et al., 1996; Zavala et al., 2001). Changes in concentrations of UV-B absorbing pigments have also been linked to both negative and positive plant responses to pathogens, (e.g., Raviv and Antignus, 2004). Different levels of secondary compounds related to changes in UV-B absorbing pigments may have additional below-ground effects on soil microorganisms by affecting carbon and other nutrient availability (Robson et al., 2005). All these examples show potential mechanisms for UV-B to influence competitive interactions.
Effects of global dimming on plants may also interact with other environmental changes, including temperature and precipitation patterns, in complex and synergistic ways. These interactive effects include increases in UV-absorbing secondary compounds associated with drought (Milchunas et al., 2004; Yang et al., 2005). These may change plant performance by reducing UV-induced damage, and affecting herbivory rates and soil processes, as discussed above.
Portions of this study involved using filters to selectively exclude UV-B (and in one experiment all UV) from the incident radiation spectrum. Results from UV exclusion studies have varied widely, and the importance of excluding various portions of the UV spectrum on these biological effects has been debated. At temperate locations, many species often fail to respond to UV-B exclusion (e.g., Cybulski and Peterjohn, 1999). Properly designed UV exclusion experiments at lower latitudes with higher ambient levels of UV-B, but similar levels of PAR, might be expected to show greater effects than similar experiments at higher latitudes. On the other hand, species of low latitudes may be adapted to cope with high levels of ambient solar UV-B and thus, might be expected to show less of a response to solar UV-B exclusion than high latitude species. The examples discussed below show that, while trends of responses to UV exclusion cannot now be easily seen from simple comparisons, improvements in methodology may permit generalizations to be made in the future.
Exclusion studies at low latitudes comprise a small number of the total exclusion studies in literature and show that species exhibit a variety of growth responses to ambient levels of UV-B. An examination of five tropical species at 9°N latitude in Panama (four tree species and the crop cassava) found substantial growth responses in two species and smaller or no growth responses in others (Searles et. al., 1995). Rice (Oryza sp.) exhibited no significant growth responses in four seasons of UV-B exclusion experiments in the Philippines at 15 °N (Dai et al., 1997). However, other crop experiments did show responses. When comparing the responses of maize (Zea mays) and the mung bean (Vigna sp.) to UV-B removal at New Delhi (28.6°N), Pal et al. (1997) found the bean to be much more responsive to UV-B than the maize. Vigna unguiculata grown under a similar UV-B exclusion at 10°N latitude, where we assume UV-B could be more intense, responded to UV-B to a lesser extent (Lingakumar et al., 1999).
The category of UV-absorbing pigments was the most responsive characteristic in a meta-analysis of UV-supplementation studies (Searles et al., 2001) and is a characteristic that often responds in UV exclusion studies. Four of the five species examined by Searles et al. (1995) in Panama showed significant changes in levels of these compounds in response to UV-B exclusion. The only other low-latitude study examining UV-absorbing compounds in response to UV-B exclusion used the bean Vigna unguiculata and saw an approximate 60% increase in these compounds under near-ambient radiation compared to the UV-B exclusion (Lingakumar et. al., 1999).
Since UV-exclusion experiments appear less technically complex than lamp-supplementation experiments, it is tempting to use multiple studies at different latitudes or elevations to examine response. One potential comparison uses wheat: Becwar et al. (1982) grew wheat under UV-B exclusion at 3,000 m elevation in the Rocky Mountains at 39.2°N latitude, while Häder (1996) studied several wheat cultivars exposed at 2,400 m elevation in the Andes at 22.8°S latitude. However, the latter study did not use a filter over the ambient UV plants, so the true UV-B effect may be confounded as the lack of a filter induces a host of microclimate differences besides the UV treatment (Flint et al., 2003). A latitudinal gradient study of Salicornia along the Atlantic coast of South America (Costa et. al., 2006) had the potential to provide an overview of how this genus responds to varied UV environments, but interpretation is complicated. Experiments in one location lacked filters over the ambient-UV plots, preventing a true evaluation of UV effects as discussed above. In another location in this study, both UV-B and UV-A were removed from the UV exclusion treatment, while all the other sites only excluded UV-B, considerably complicating comparisons (see below).
Selection of different wavebands of interest may complicate comparisons of results between studies. Among the 100 or so published exclusion studies, exclusion has included removing all UV-B wavelengths, removing only a portion of the UV-B waveband, and removing all UV-B and simultaneously some or all of the UV-A. While many studies use polyester (often referred to as Mylar) which removes most UV-B wavelengths (Fig. 13.2), other studies use different techniques or materials to remove more limited wavebands. This is illustrated by experiments with maize at a low-latitude site (28.6°N), where Pal et al. (1997) used the standard polyester filter and saw only some limited change in plant growth with UV-B removal. Several varieties of maize grown further north (38.7°N latitude), under filters that removed only a narrower waveband of UV-B, often seemed to show greater effects (Mark et al., 1996). This is unexpected, both in terms of the direction of the latitudinal response and the larger response to the smaller UV radiation difference between the two treatments.
There is also increasing evidence of the importance of portions of the UV-A
spectrum. Several recent spectral weighting functions extend into the UV-A region (Quaite et al., 1992; Ibdah et al., 2002; Flint and Caldwell, 2003; also see discussion of UV-A exclusion experiments in Flint and Caldwell, 2003). Thus experiments excluding portions of the UV-A spectrum would be expected to produce different results than exclusions of only UV-B and are not directly comparable; in some instances solar UV-A invokes different responses than UV-B (Kotilainen et al., 2008).
Latitudinal comparisons are sometimes further complicated because the latitude of the experiments is not reported. Also, in some studies, the dates of plant growth are not presented. As many UV-exclusion experiments are short-term studies, and UV levels can vary considerably over the course of the growing season in many areas, these two problems make even crude estimates of irradiance levels problematic.
Perhaps the greatest aid for reconciling exclusion experiments in different areas would be the presentation of UV dosimetry integrated over the course of the experiment, although a characterization of representative clear-day radiation would be preferable to a complete lack of dosimetry. Biologically-weighted UV radiation with more than one common weighting function is preferable to unweighted values, as comparisons of weighted and unweighted radiation, or radiation weighted with different spectral weighting functions, are seldom possible. For example, both of the Vigna studies referred to above provide dosimetry, but use different weighting functions. Thus, it is not possible to use the available dosimetry to help understand the unexpected response to latitude when these two studies are compared.
Was this article helpful?