Introduction
Plants are sessile organisms, and many plants use physical or chemical responses to defend against environmental stressors. For example, the phenolic compound p -coumaric acid (p CA) is implicated as a chemical defence mechanism with respect to exposure to solar radiation, especially the UV-B wavelengths (280-315 nm). This compound is stimulated through the phenylpropanoid pathway (PPP), and has been shown to be an effective absorber of UV-B radiation (Blokker et al., 2006). It is also an important building block for a number of essential plant compounds, including sporopollenin, the major biomolecule that constitutes the exine of pollen grains (Wehling et al., 1989, Blokker et al., 2006, Rozema et al., 2001a, Jungfermann et al., 1997). Indeed,p CA has been found in higher abundances in the pollen-based sporopollenin of plants exposed to higher levels of UV-B (Rozema et al., 2001b, Blokker et al., 2005). Since sporopollenin is also highly resistant to degradation under anoxic conditions, recent studies have proposed using the pollen chemical record preserved in sediments as a proxy for UV-B (Rozema et al., 2001b, Rozema et al., 2002, Fraser et al., 2011, Mazza et al., 2000, Willis et al., 2011) or more generally for total solar irradiance (Jardine et al., 2016) on centennial timescales or longer. The development of such a tool would have major implications for long-term ecological research, providing an independent of solar radiation in the past, with a broad range of palaeoecological and palaeoclimatological applications (Magri, 2011).
To enable a robust application of the chemical signatures of plants to reconstruct past changes in solar radiation, an understanding of the drivers of variability of these compounds within pollen grains is required (Seddon et al., 2019). However, whilst a number of studies have demonstrated positive relationships between exposure to solar radiation and p CA along latitudinal and elevation gradients (Willis et al., 2011, Jardine et al., 2016, Watson et al., 2007, Lomax et al., 2012), the timing of the response of these compounds to UV-B radiation remains poorly understood. For example, a number of studies have assumed thatp CA is a direct indicator of solar radiation during the growing season (Rozema et al., 2001a, Rozema et al., 2001b, Lomax et al., 2008), but this assertion has yet to be tested under field conditions. In fact, in some species, such as Pinus sylvestris , tapetal cells produce the membrane containing sporopollenin (including p CA) less than 11 days prior to pollen shed (Rowley et al., 2000) so significantly shorter-term plastic responses than those currently proposed are possible. Furthermore, the extent to which the p CA production is driven by conditions experienced locally on the plant (i.e. if variability is observed between branches or individual flowers shaded in different ways, or if the signal represents a fully-integrated whole-tree tree response) remains unknown. This has implications for how the proxy is understood from the point of view from the palaeorecord since it determines whether a signal derived p CA represents a short-term or growing-season response, as well as whether within-sample variability should be interpreted as reflecting the stand-, tree-, or within-canopy patterns. Therefore, determining the timing and extent of this plastic response remains an essential question in understanding how pollen responds to changes in solar radiation.
A second aspect that remains poorly understood is the variation in both the p CA content and in the response to environmental variation between individuals and species. For example, the widespread abundance of Pinus pollen grains found in sediments, combined with the fact that Pinus is a light-demanding species, means that Pinushas become a focus for research into reconstruction of solar irradiance using p CA. However, the pollen members of the European flora ofPinus spp. remain difficult to separate in palaeoecological reconstructions using traditional light microscopic methods, and if different species exhibit varying responses under equivalent radiation, then this could have implications for both dose-response relationship ofp CA and reconstructions from pollen sampled from a sediment core. An obvious source of species-level variation in p CA may be pollen size. European pines vary considerably in pollen size (Beug, 1961), and it is very likely that species may contain differences in totalp CA simply due to differences in their surface-area. In palaeoecological reconstructions, such variation may be accounted for simply by measuring and taking these size differences into account. Species-level variation in phenolic acid content due to inherited genetic differences in the ability to produce these acids, or in the response to variation in radiation, would be more challenging to account for using palaeoecological reconstructions. To date, the extent to which species-level variation can potentially affect the dose-response relationship between p CA and solar radiation remains poorly understood.
Here, we present the results of two studies which aimed to investigate the effect of short-term variations in ambient solar radiation on the abundance of p CA in Pinus pollen grains. Using a resampling of the same individuals over two consecutive years with natural variation in ambient solar radiation, our first objective is to investigate whether the abundance of p CA in pollen grains varies in response to natural variability in radiation, and if the level or response varies between Pinus species. Using shading cloths on 10 individuals of Pinus sylvestris , our second objective was to investigate whether the abundance of p CA in pollen grains varies in response to artificial reductions of ambient solar radiation at the level of individual male cones.
Material and Methods