2  Ecological interactions and the extended phenotype

Another situation where genetic variation within a species has the potential to affect the distribution and abundance of associated organisms is where that species plays the role of a foundation species in the community. Foundation species are those that are perceived to provide important structure to a community and usually represent the species with the greatest abundance or biomass in the community (Bruno, Stachowicz, and Bertness 2003). They may constitute a major food source, such as kelp (Macrocystis pyrifera) in ocean environments, possessing physical traits that regulate ecosystem processes and provide a habitat for associated, dependent, communities (Graham 2004). Kelp forests have a long-standing reputation for providing one of the largest and most important foundation habitats on the planet (Mann 1973; Steneck et al. 2002), with high diversity of associated species - more than 200 species in southern California alone (North 1971; Foster and Schiel 1988). An important consequence of this relationship is that genetic variation in a foundation species could potentially have a large influence on the community of organisms with which it interacts. For instance if the foundation species provides the habitat for a community of dependent organisms, changes in the physical attributes of the foundation species may modify the habitat of the dependent species, and this may alter the community of dependent species that can be supported.

This effect has been demonstrated in common garden experiments involving the flowering plant, goldenrod (Solidago altissima). Here genetic variation between goldenrod genotypes led to variation among their associated arthropod communities (Maddox and Root 1990). Using the same foundation plant species, high genetic diversity was revealed in the phenotypic traits of flowering time and floral abundance, as well as in the extended phenotype traits of flower visitor abundance and richness (Genung et al. 2010). These studies demonstrate the importance of genetic variation in dominant species, and their potential influence for structuring communities of dependent organisms. In the case of pollinators, increased insect diversity in a natural habitat where Solidago is abundant could positively influence pollination for other plant species and promote genetic diversity within the community.

The phenotype of an organism is its outward appearance and is determined by an interaction between its genotype and the environment. For instance aspects of a tree’s phenotype would normally be characters such as the shape of its leaves, the morphology of its bark and so on. However if we are dealing with foundation species, and if the genotype of the foundation species influences the community of dependent organisms then the dependent community could be considered as part of the phenotype of the foundation species, or its ‘extended phenotype’. For instance if genetic variation affecting leaf shape/chemistry in a tree alters the realised niche for associated insects with community consequences, then the variable insect community found on a tree could be thought of as an aspect of the extended phenotype of that tree. The term ‘extended phenotype’ was first coined by Richard Dawkins (Dawkins 1982) to describe the effects of an organism’s genetic constitution on its community of associated organisms.

2.1 Validating and testing the extended phenotype concept

Many studies have been conducted to explore and test the validity of the extended phenotype concept and to assess the importance of genetic variation in foundation species on their communities of dependent organisms. Studies have typically used plants as the foundation species, and insects as the dependent species to investigate the importance of plant genotype and environmental factors at differing spatial scales. In a study by Johnson and Agrawal (2005), fourteen genotypes of Oenothera biennis (total of 926 plants) were planted into five natural habitats representative of the environments in which this plant occurs. Plant genotypic differences accounted for as much as 41% of the variation in arthropod diversity, evenness, richness, abundance and biomass on individual plants, and explained more arthropod community variation than environmental variation across spatial scales. Arthropods appeared to select for certain traits of O. biennis, which may lead to important evolutionary changes in associated insect communities should the plants undergo significant genetic change. Furthermore, (Bangert et al. 2006) demonstrated how genetic variation in hybridizing cottonwood (Populus spp.) affects the structure of an associated arthropod community. One important plant trait tested as a possible intermediate link between plant genetics and arthropod community structure was chemical composition of the leaves. In both a common garden experiment of 29 different genotypes (F\(_1\) hybrids and backcrosses) and 25 of the same genotypes within wild populations, trees with similar genetic compositions had similar chemical compositions and similar arthropod compositions.

The potential genetic effects on the relative thickness of living bark and characteristics of decorticating bark in common garden experiments of Eucalyptus globulus were studied by Barbour et al. (2009): twenty trees from each of five provenances were selected at random from a common garden trial established in northern Tasmania. Trees were sampled at 17 years old, and 25m in height. Of the ten bark traits measured, significant variation was found in eight of the traits and attributed to genetic differences among provenances (ranging from p <0.001-0.02). Overall, decorticating bark decreased with trunk height. Sampling macro-arthropod communities living within the bark revealed a significant effect of host tree genotype on species richness and abundance, with 60% of the overall variation in community attributable to differences in bark structure associated with Eucalyptus provenance.

However, contradictory results were found by Tack et al. (2010), who showed how genetic effects of a host plant on associated insect communities tend to be diluted at increasing spatial scales. Herbivorous insects on oak (Quercus robur) were sampled in a common garden experiment replicated at three spatial scales: landscape scale (~5 km\(^2\)); regional scale (~10,000 km\(^2\)); and wild trees at various spatial scales. Their sampling revealed that 32% of the species richness at the landscape scale was explained by spatial correlations, with this effect increasing at the regional and greater scales. Host plant genotype was considered to be of secondary importance.

Genetic effects have been researched in forest systems where trees are foundation species and associated insects in particular are dependent species. For example, the resistance or susceptibility of Pinus spp. (foundation) to grazing by gypsy moth can significantly influence the distribution of nearly 1000 dependent organisms, including birds, mammals, invertebrates and fungi (Brown et al. 2001; Whitham et al. 2003; Kuske et al. 2003; Diner et al. 2009).

Forest trees are particularly important as foundation species because they provide habitats for many different types of organisms; intra-specific variation can also influence associated communities of plants and arthropods (Zytynska et al. 2011). Genetic effects of Populus spp. have been found for important ecological processes. A comparison of genetic effects on local environments has been carried out using both a model Populus system in a common garden (Schweitzer et al. 2008) and a natural stand (Madritch, Greene, and Lindroth 2009), to evaluate effects of plant genotype on soil processes, such as soil nutrients (i.e., nitrogen) and microbial biomass. In both cases, plant genotype explained variation in soil nitrogen, and microbial activity. In the common garden experiment, genotype explained up to 78% of the variation in microbial biomass. In the natural system, Madritch, Greene, and Lindroth (2009) used 24 distinct Populus tremuloides clones (multiple ramets per clone) spanning a distance of 25km\(^2\) to demonstrate the effects of plant genetic identity on soil processes. Sampling leaf chemistry, soil nutrient content and microbial activity beneath each clone revealed significant variation between genotypes for each measured variable. Genetic distance between clones was correlated with differences in ecosystem processes, suggesting that aspen are capable of modifying their local environment and thereby regulate ecosystem functions.