9  Genetic variation between clones of aspen physical traits potentially important for associated species: bark texture and phenolic chemistry

9.1 Introduction

Communities are assemblages of different organisms, some of which provide a foundation, or habitat, for other (dependent) species that make up the rest of the community. Genetic variability within the foundation species has the potential to influence the expression of physical traits within that species, and in turn influence the dependent species sensitive to these traits. Community genetics is the area of research that investigates how genetic differences in one species, usually a foundation species, affects communities of dependent species.

The study of community genetics can be undertaken in two situations, which were explored in a special issue of Ecology. The first is where novel crops are introduced and have immediate impact on the local communities (Neuhauser et al. 2003); the second is in a natural, longer-established ecosystem with greater equilibrium and genetic stability, e.g. woodlands and permanent grasslands (Whitham et al. 2003). In the latter situation a decision needs to be made on which species to include in such studies. We have first to recognize the foundation species, then to define the dependent species. When this has been done we can then identify traits that are likely to be important in their interaction (Chase and Knight 2003).

In woodland systems chief structural components are the trees. These interact with a wide variety of different dependent communities, such as arthropods, endophytes, epiphytes and micro-organisms. A priori we might expect that genetic variation in trees will have effects on these communities; greater than average genetic variation is found within trees is in hybrid zones.

Hybrid zones occur naturally where two or more species overlap within their native range (Anderson and Stebbins 1954; Ellstrand, Whitkus, and Rieseberg 1996; Milne et al. 2010), and are particularly common in trees e.g. Quercus, Picea, Ulmus, and Populus. Trees in hybrid zones segregate for large genetic differences found between species. For instance, marked differences between clones of P. tremula and P. tremuloides and their hybrids grown in common garden experiments were observed when examining nutrient content in wood (Rytter and Stener 2003), growth and phenological traits (Yu, Tigerstedt, and Haapanen 2001), and concentrations of secondary compounds in leaves (B. Rehill et al. 2005; B. J. Rehill et al. 2006). There is now a lot of evidence that shows genetically determined differences in hybrid zones of Populus (P. angustifolia and P. fremontii) that affect a wide variety of dependent communities (Whitham et al. 1999). Similar results have come from studies of related hybrid systems e.g., Eucalytpus and Quercus (Whitham et al. 2006).

The usefulness of hybrid studies to understand the influence of genetics on physical traits and potential (or realized) community structure is well established. However confining studies to the use of hybrid complexes may introduce confounding factors. Interspecific variation is generally greater than intraspecific variation; therefore the recombination of genetic material from two sexually compatible species can form hybrid complexes that manifest a broader range of physical traits capable of influencing communities than the natural range of variation seen within a species. Moreover hybridization can generate novel phenotypes lying outside the range of the parent species. Such hybrids may show lower herbivore resistance and can act as ‘sinks’ for herbivorous insects even if the parental species are resistant to such attack (Whitham 1989; Floate, Kearsley, and Whitham 1993). Finally it should be noted that hybrid zones represent only a small fraction of tree populations.

Given these limitations it is important to determine if the community effects described in hybrid systems can be demonstrated within a single species. Preliminary evidence from a range of systems suggests that genetic variation within a species can often be sufficient to affect community structure of dependent organisms (Johnson and Agrawal 2005; Bangert et al. 2005, 2006; Barbour, O’Reilly-Wapstra, et al. 2009; Tack et al. 2010).

9.2 Variation within and between populations

Genetic variation within a species can come from two sources: between populations and within populations. Genetic differences between populations of a species have been shown to significantly influence arthropod fauna, taking into consideration a range of spatial scales (Johnson and Agrawal 2005; Bangert et al. 2006; Tack et al. 2010). In addition recent research has shown that variation within a foundational species, within a population, has an effect on communities of dependent organisms (Dickson and Whitham 1996; Wimp et al. 2007; Barbour, O’Reilly-Wapstra, et al. 2009; Kanaga 2009). This suggests that genetic diversity within populations of foundation species particularly play an important role in shaping assemblages of associated species; even small differences between genotypes may be sufficient to affect changes. If these traits are heritable and stable across a range of environmental conditions it would be logical to conclude that each genotype of a foundation species is capable of formulating its own distinct community within a site. Extrapolated to the landscape and ecosystem levels, this could have drastic implications for ecosystem and global biodiversity.

9.3 Effects of foundation species on dependent communities

Given the potential importance of these effects it is imperative to broaden our range of studies of the effects of foundation species genotype on dependent species. To do this a model system is needed in which there is an intimate interaction between a foundation species and a dependent community. In the past interactions between trees and arthropods, particularly leaf herbivores, have been chosen (Osier and Lindroth 2001, 2004, 2006; Bailey et al. 2005; Shuster et al. 2006; Bangert and Whitham 2007; Tack et al. 2010). However there are other groups that form very important communities on trees that have been overlooked, such as epiphytic lichens (Gustafsson and Eriksson 1995; Uliczka and Angelstam 1999; Hedenås and Ericson 2000; Juriado, Paal, and Liira 2003). These interact directly with trees. Trees provide microsites within which epiphytes establish and grow (Hill and Barkman 1958; Kantvilas and Jarman 2004; Fritz and Heilmann-Clausen 2010). There is also evidence many epiphytes, particularly lichens, penetrate the surface of the tree and interact with chemicals and bark tissues (Ascaso, González, and Vicente 1980; Ascaso, Orus, and Estévez 1983; Ascaso and Rapsch 1985; Inoue, Noguchi, and Kubo 1987; M. Legaz et al. 1988; Bouaid and Vicente 1998; M. E. Legaz, Monso, and Vicente 2004). Thus there is the potential for genetic variation affecting bark morphology and bark chemistry of trees to influence the communities of epiphytes that colonize them.

The purpose of this study is to determine whether genetically determined differences in bark morphology and bark chemicals in a foundation tree species influence the composition of its epiphytic lichen community.

9.4 Aspen as a foundation species

One such tree species is aspen (Populus tremula L.). In many localities aspen is clonal, reproducing from root suckers rather than seed. This leads to the availability of replicate stems within a clone, with the same genotype. Therefore it is possible to assess whether there are significant differences between clones for microsite characters of bark and the chemical composition of bark. Aspen is also amenable to study because of the rich and diverse epiphyte flora (Gustafsson and Eriksson 1995; Jüriado et al. 2009), associated with the high pH of the bark (Hedenås and Ericson 2000; Street and Street in Cosgrove and Amphlett 2001; Ellis and Coppins 2007; Jüriado et al. 2009).

A major character of bark likely to affect establishment of lichens on aspen is variation in bark texture (Hyvärinen, Halonen, and Kauppi 1992; Sillett et al. 2000; Boudreault et al. 2008; Lamit et al. 2011). This variation facilitates the creation of microsites with differences in moisture levels (Sheard and Jonescu 1974; Aboal et al. 1999; Zhang 1999; Levia and Herwitz 2005). It is important to establish, in the first instance, whether genetic differences between clones in natural situations are manifested in % smooth bark.

A major chemical family found in bark, particularly knotwood, is the polyphenols (Pietarinen et al. 2006; Neacsu et al. 2007), which include condensed tannins and phenolic glycosides exhibiting strong defences against invasion by pathogens, wood-boring insects and UV radiation (Bennett and Wallsgrove 1994; Lindroth and Hwang 1996; Lieutier et al. in Grégoire et al. 1997; Ockels et al. 2007). Limited studies have demonstrated that these latter compounds significantly affect lichen growth (Koopmann 2005). Therefore in the second instance it is necessary to determine whether there are clonal differences in phenolic chemistry of bark tissue with which the epiphytic lichens interact. Of course it is likely that bark morphology and chemistry may be affected by other factors such as age of tree, aspect and height (Levia and Wubbena 2006; Hamilton et al. 2007; Barbour, Forster, et al. 2009). To demonstrate that genetics play a primary role in determining bark morphology and chemical composition, it is important to investigate the effects of aspect and height on % smooth bark and chemistry.

Therefore, to achieve these goals, a natural stand of a minimum of five aspen clones growing in close proximity is required. Using genetic markers and phenology the ramets can be classified into genets. Then individual ramets of the same age can be sampled and scored for % smooth bark and secondary chemical composition; sampling must carried out at different heights and different aspects to determine their effects.

Questions asked are:

  1. Are there significant differences between clones in the characters measured?
  2. How much of the variation is accounted for by variation between clones?
  3. Is any of the residual variation (within clones) accounted for by either height or aspect?