Browse Subject Areas

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

  • Loading metrics

Genetic Variation in Functional Traits Influences Arthropod Community Composition in Aspen (Populus tremula L.)

Genetic Variation in Functional Traits Influences Arthropod Community Composition in Aspen (Populus tremula L.)

  • Kathryn M. Robinson, 
  • Pär K. Ingvarsson, 
  • Stefan Jansson, 
  • Benedicte R. Albrectsen


We conducted a study of natural variation in functional leaf traits and herbivory in 116 clones of European aspen, Populus tremula L., the Swedish Aspen (SwAsp) collection, originating from ten degrees of latitude across Sweden and grown in a common garden. In surveys of phytophagous arthropods over two years, we found the aspen canopy supports nearly 100 morphospecies. We identified significant broad-sense heritability of plant functional traits, basic plant defence chemistry, and arthropod community traits. The majority of arthropods were specialists, those coevolved with P. tremula to tolerate and even utilize leaf defence compounds. Arthropod abundance and richness were more closely related to plant growth rates than general chemical defences and relationships were identified between the arthropod community and stem growth, leaf and petiole morphology, anthocyanins, and condensed tannins. Heritable genetic variation in plant traits in young aspen was found to structure arthropod community; however no single trait drives the preferences of arthropod folivores among young aspen genotypes. The influence of natural variation in plant traits on the arthropod community indicates the importance of maintaining genetic variation in wild trees as keystone species for biodiversity. It further suggests that aspen can be a resource for the study of mechanisms of natural resistance to herbivores.


European aspen, Populus tremula L., is a pioneer tree valued as a host to many species of dependent flora and fauna, many of which are specialists [1][4]. Live and dead aspen trees alike harbour a variety of arthropods on the canopy and woody tissue [5]. Aspen is a common forage food for mammals [4] and also supports species of epiphytes [1], [2]. In recognition of this conservation value, current forestry practice in Scandinavia retains aspens in clear-cut zones to promote biodiversity [6]. In North America, the ecological, morphological and physiological characteristics of quaking aspen, P. tremuloides, have been extensively studied [7][13]. In contrast to P. tremuloides, there has been a relative paucity of knowledge of the same traits in P. tremula, however an interest in the European aspen has gradually emerged, recognising the genetic variation within this species [14], [15] in addition to its ecological value. In Sweden intra-specific variation for growth and phenological traits has been detected in young plants [16], [17]. Bylesjö et al [18] reported variation in leaf shape in the same population. Half-sib families of P. tremula have been shown to differ in growth and resistance to biotic stresses [19]. Recent chemical analyses of P. tremula leaves have isolated genotype-specificity in phenolic glycoside compounds [20] recognised for their efficacy in herbivore defence in Populus [21], however the response of folivorous arthropods has received little attention.

A growing body of evidence documents the influence of host-plant genotype on herbivore community structure. The concept of community genetics [22], [23], whereby genes conferring variation in plant traits in turn influence herbivore preferences as an extended phenotype, has been validated in a number of ecological systems, with clear examples in both inter-specific hybrids [24][27] and intra-specific hybrids [28][33]. The mechanisms of natural variation in herbivore preferences have been examined extensively and have been accounted for by traits including differences in stem growth [34], plant ontogeny [35][39], plant architecture [40], leaf biomass [41], environmental factors [42], secondary metabolites [43][45], physical defence [46], and factors in combination [47][48]. Secondary metabolites in leaf tissue, including phenolic compounds, are known to be under genetic control in trees [49], [50] and have been associated with the abundance of individual folivores and multi-trophic communities [49], [51]. Some specialist arthropod herbivores utilize plant phenolic compounds to elicit defences against predators [52], [53]. Although there are many such examples of coevolved defences in the on-going evolutionary battle between plants and their attackers, and many strategies of plant defence have been proposed and recycled [54][60], there is increasing awareness that a number of co-varying plant traits may be involved in the susceptibility to, and defence against, herbivores [61], [62].

Common garden trials have long been employed to provide the same set of environmental conditions to enable comparison of plant genotypes within an experimental population in herbs and trees [63], [64]. The Swedish Aspen Collection (SwAsp) was established in 2004 from the clonal propagation of 116 P. tremula trees collected from 12 regions across Sweden, spanning nearly ten degrees of latitude, and grown in two common gardens, one in northern and one in southern Sweden [17]. It is in this population that latitudinal genetic differentiation has been documented for several plant traits [16], [17], in addition to latitudinal trends of attack by rust fungus and leaf-mining insects that reflect countrywide forest survey data [65].

The ability of arthropods to resolve between plants within a species, interpreted through their survival, diet and oviposition preferences, has been observed in Populus species in controlled conditions [9], [66] and field studies [44], [67], [68] however we could not find any reported field observations of the total arthropod community in natural conditions in P. tremula. We examined the SwAsp collection to assess the range of intraspecific variation in plant functional traits that could influence dependent organisms. We conducted field studies initially in two common gardens to form an inventory of arthropod herbivores on the canopy of P. tremula in Sweden. We gathered arthropod data over two growing seasons in the common garden in northern Sweden and measured a suite of basic plant traits to address the following questions. (1) Are plant functional traits genetically and geographically structured in P. tremula? (2) Is the community of arthropod herbivores on P. tremula heritable? Identifying clonal differentiation in susceptibility to herbivores is informative both for conservation biology and to understand natural resistance to biotic stresses. Finally we ask (3) is the community of arthropod herbivores explained by specific plant traits?

Materials and Methods

Common gardens were established in 2004 from root cuttings of mature trees collected from 12 regions across Sweden (Table 1), as detailed in Luquez et al. [17]. Briefly, one garden in Ekebo, southern Sweden (55.9°N) and one garden in Sävar, northern Sweden (63.9°N) were planted with at least four replicate trees of 116 clones. Sites were deer-fenced and weeded.

Arthropod surveys

Arthropod surveys were undertaken in 2008 at the Ekebo common garden on 23–25 June and 15–16 July. Three surveys were made at Sävar in 2008 on 27–29 June, 7–9 July and 29–31 July, and in 2009 on 15 June, 7–9 July and 27–29 July. Arthropod loads differed between sites, with a paucity of arthropod abundance in Ekebo, where only two species accounted for over 75% of individuals counted. Therefore Ekebo data were excluded from our analyses, however the species inventory has been included in the supplementary information (Table S1) to ensure our records of canopy arthropods on aspen are as comprehensive as possible. Each tree was surveyed exhaustively by systematically examining all leaves on every branch and the types and numbers of arthropods recorded. Trees were surveyed only on sunny days and in varying order during different surveys to randomize any effects of timing or human disturbance on the fauna. Numbers or clusters of all leaf-dwelling arthropods, or leaf structures formed by arthropods, visible to the naked eye on every leaf were recorded, with the exception of two abundant galling mite species (Aceria varia and Phyllocoptes populi), which were scored once in the season by assigning an infection score (0, 0.5, 1, 2, 3, 4, or 5, where 5 = all leaves infected). Surveys were conducted by two individuals working together and counting methods were standardized before data was recorded. One author (KMR) undertook all surveys to ensure consistent scoring and counting of morphospecies. Arthropods were identified using field guides [69], [70]. Unknown arthropods were sampled and experts consulted on identification. It was necessary to classify some arthropods solely by the larval shelters (e.g. leaf rolls, ties and folds), leaf galls and subcutaneous mines constructed by the arthropods on the aspen leaves. All morphospecies are documented in reference material held at Umeå Plant Science Centre. Morphospecies were further categorized for some analyses by behaviour guild of each species based on how they utilize leaf tissue (leaf-gallers, -chewers, -miners and -rollers). Species richness was calculated as the sum of species on each individual plant. Arthropod abundance was the sum of all individuals of all arthropod species on each plant. Arthropod species known to feed on one plant genus (Populus) were defined as specialists, and all others considered generalists.

Damage phenotyping

Infection by leaf rust fungus of the genus Melampsora was assessed visually on all individual plants at Sävar, assigning a score of 0 (absent), 1, 2, 3, 4, to 5 (every leaf infected) for the whole tree. This scoring took place in late August 2007 when among-tree differences in infection were highest and before autoinfection had spread to all trees [65]; at this time the scores in the field were poisson-distributed. All phenotyping on visual scales was repeated to ensure that scoring was consistent.

Leaf and stem phenotyping

Stem heights and diameters were measured annually at the end of the growing season as described by Luquez et al. [17]. Leaf chlorophyll and anthocyanin contents were estimated indices using chlorophyll and anthocyanin content meters (Opti-sciences Inc, Hudson, USA). Readings were taken on five fully expanded, undamaged leaves chosen evenly across the plant on all individual plants. Leaves were sampled for morpho-chemical analyses immediately after the third survey of the season. Ten fully expanded, mature, undamaged leaves were selected from each plant and harvested at the petiole junction with the leaf. Samples were kept in cool conditions and returned to the laboratory. Whole leaves were scanned using a desktop scanner at 300 dpi. Petiole length was measured on the leaf scans using ImageJ [71]. Petioles were then erased using imaging software. Leaf images were measured using LAMINA software [18] to obtain leaf area. After scanning, leaves were air-dried at room temperature in well-ventilated conditions until constant weight, and then leaf dry weight was used to calculate specific leaf area (cm2 g−1). Dry leaf material was pooled per plant and ground using a Retsch bead mill for use in chemical analyses. Spring bud flush was scored every two days and the Julian date was recorded for an individual tree when a minimum of three leaves had burst their bud scales revealing a characteristic V-shape between the first two unfolding leaves.

Leaf chemistry

Total phenolics and condensed tannins were quantified in the samples of dry, ground leaf tissue. Total phenolics were quantified from extracts in 80% methanol following the method of Ossipova et al. [72] against a chlorogenic acid standard (Sigma, St Louis, USA) and expressed as chlorogenic acid equivalents. Condensed tannins were quantified using the acid-butanol method using leaf powder extracted in 70% v/v acetone, 1% ascorbic acid [72] using a procyanidin-B standard (Sigma-Aldrich, St Louis, USA) and expressed as procyanidin equivalents.

Statistical analysis

Statistical analyses were conducted in R [73]. Shannon diversity indices were calculated in PRIMER-E (version 6.1, Roborough, UK). Field blocks were tested for homogeneity of tree growth with two-way analyses of variance. There was no evidence of interaction between clone and block for stem height and diameter, from which we infer reasonable uniformity of growth conditions and randomization within the gardens. Arthropod morphospecies data did not conform to a normal distribution and could not be transformed to normality so Kruskal-Wallis analyses of variance were used to test between field blocks to confirm that the presence of morphospecies did not differ significantly between blocks. Data were transformed using Box-Cox powers where necessary. Box plot distributions were constructed, principal components analysis conducted, and latitudinal relationships were calculated from trait data from all individuals and clones. For all other analyses, clones with less than three live replicates were removed from the data set. Population differentiation (QST) and broad-sense heritability (H2) were calculated as in Luquez et al. [17]. Briefly, we used data from the common garden design to estimate the variance among clones within a population and among populations. For estimating the heritability of the traits we used the following linear model:where zjkl is the phenotype of the lth individual in the kth block from the jth clone and where u denotes the grand mean and ejkl is the residual error term. Broad-sense heritability was calculated from estimates of total genetic variation (σG2) and total phenotypic variation (σP2)To estimate QST we used:Here the population (ai) and clone (bij) effects now provide estimates of genetic variation between populations (σB2) and among clones within populations (σC2), respectively. These variance components were used to estimate QST according to:A Bayesian approach was used to calculate point estimates and confidence regions of QST, implemented in the R-package MCMCglmm [74]. For each trait a single MCMC chain was run for 15000 steps, with the first 5000 discarded as burn in. The remaining samples were thinned by taking every tenth sample, yielding 1000 independent draws from the posterior distribution of the parameters of the model. Convergence of the chains was assessed using the coda package in R. The priors of the variances and covariances of the random effects in the model are locally non-informative priors [74]. Point estimates of H2 and QST were calculated as the mode of the posterior distributions and confidence intervals were calculated from 95% highest posterior density (HPD) regions. As with all broad-sense heritability calculations, our estimates of H2 and QST are likely to be over-estimates of true heritabilities [75], however these data enable the comparison of the genetic and environmental components of the traits measured.

Arthropod data were classified into motile species that could move between plants, and sessile species that select the host plant for feeding or oviposition indicated by a modification in the leaf tissue. For community analysis we defined ‘arthropod community’ as these sessile leaf-modifying species to avoid ambiguity of their relationship with plant. For other analyses we used data from all morphospecies. For community analysis we used the vegan package in R [76]. Bray-Curtis similarity matrices were constructed from all leaf-modifying arthropod traits in each year. Prior to this, count data were square root transformed to reduce differences between common and rare species and all data were then normalized. Data from each year were analyzed separately using the adonis function in the R package vegan, a non-parametric multivariate analysis of variance (non-parametric MANOVA), which partitions the dissimilarity matrix (the response variable) by fitting linear models (with genetic, population and plant traits as explanatory variables), and permuting the data within the replicated field blocks to return pseudo F-values on 999 permutations [77]. This method makes few assumptions about the underlying data structure and is thus considered suitable for the analysis of community count data and when the response variable is comprised of metric (Euclidian) distances the F-value is identical to that obtained using parametric ANOVA [77]. Principal components analysis (PCA), an ordination method maximizing the differences between individuals, was conducted using R software on all clones (irrespective of replication). Stem diameter was excluded from the PCA analyses because it is highly correlated to stem height (Pearson correlation: r = 0.9, P<0.001).


Stem traits followed a latitudinal gradient when evaluated at the population level (Figure 1), with the tallest trees of largest girth originating from the southern part of Sweden. These regional differences in stem biomass are indicated in QST values (Table 2), with clear population differentiation for stem height and diameter, but also a high level of within-clone similarity indicated by the H2 values (Table 2). In contrast, leaf morphs (Figure 2) did not show strong latitudinal trends. Despite negligible QST values for leaf traits, H2 values were moderate to high for all plant traits measured except specific leaf area (which had no apparent heritable component), suggesting abundant genetic variation but little genetic differentiation for these traits.

Figure 1. Stem heights and diameters of SwAsp clones grouped by population.

Stem heights (A) and diameters (B) of all SwAsp clones grouped by their population of origin, ranked from south to north, in the Sävar common garden in 2009. Six to 10 clones are represented per location. Spearman rank coefficient, r, indicates correlation between trait and latitude of population origin; *** indicates statistical significance at P<0.001.

Figure 2. Leaf trait values of SwAsp clones grouped by population.

Distribution of leaf trait values of all SwAsp clones grouped by their population of origin, ranked from south to north. Traits were measured in the Sävar common garden in 2008 (A, B, C, D) and 2009 (E, F). Six to 10 replicated clones are represented per location. Spearman rank correlation coefficients, r, between trait and latitude of clone origin are shown with the following significance levels: ** P<0.01, and NS non-significant.

Table 2. Population differentiation (QST) and broad-sense heritability (H2) of stem, leaf and arthropod community traits in the SwAsp common garden, Sävar (63°54′N).

In total 93 arthropod morphospecies (Table S1) were recorded on P. tremula in the SwAsp common gardens. The most frequent arthropods were specialists on Populus, including many that cause galls or make rolls, ties, mines and other shelters using aspen leaves (Table 3). Most arthropod species increased in frequency in the common garden from 2008 to 2009. Arthropods grouped by guilds were significantly and negatively correlated with latitude (Figure 3); the strongest relationship (r = −0.38) was for leaf-mining insects. Leaf-mining, gall-making and leaf-rolling arthropods each showed stronger correlations with latitude than any leaf trait, but weaker correlations than stem traits. Leaf-chewing arthropods were rare (Figure 3). Table 4 shows heritability and population differentiation for leaf rust fungus and arthropod traits in the common garden. It was not possible to calculate QST or H2 on individual morphospecies due to the small number of individuals in the data set, however when all arthropod (arthropod abundance) or species (species richness) were summed on an individual plant there were sufficient data to calculate QST and H2 on the 2009 data set. Arthropod species richness and abundance were moderately heritable, but less so than infection by leaf rust fungus, which showed a strong population and clonal heritability (Table 4). QST was generally comparable to values obtained for stem traits in contrast to the negligible or absent QST values for leaf traits. Furthermore, the Shannon diversity index, despite showing little genetic variation, showed QST values indicative of a significant degree of aspen population structure. Non-parametric MANOVA (Table 5) identified genotypic and population differences in the distance matrix of the arthropod community in both years.

Figure 3. Numbers of arthropods categorised by feeding guilds recorded on SwAsp clones grouped by population.

Numbers of arthropods grouped by feeding guilds recorded on all SwAsp clones grouped by their population of origin, ranked from south to north. Traits were measured at maximum abundance in 2009 in the Sävar common garden. Six to 10 replicated clones are represented per location. Spearman rank correlation coefficients, r, between trait and latitude of clone origin are shown with the following significance levels: *** P<0.001, ** P<0.01.

Table 3. The most common arthropod morphs in the SwAsp common garden at Sävar (63°54′N), expressed as the maximum percentage of the trees in the garden affected by each species.

Table 4. Population differentiation (QST) and broad-sense heritabilities (H2) of biotic traits: Melampsora leaf rust infection, and arthropod herbivore community traits in the SwAsp common garden at Sävar (63°54′N).

Table 5. Differentiation of arthropod community composition by plant genotype and population in the Sävar common garden (63°54′N).

The first two components from the principal components analysis (PCA) explained 51.4% of the total variation in the aspen phenotypes (Figure 4). The first principal component (PC1) was influenced by plant height, arthropod species richness and abundance, together with petiole length and to a lesser extent leaf area (Table 6). Leaf rust fungus influenced PC1 such that smaller trees were more heavily infected (as shown in [65]). Leaf traits accounted for most variation in PC2, with contrasting trait loadings of secondary chemical compounds and anthocyanin and chlorophyll indices (Table 6).

Figure 4. Principal components biplot of plant and arthropod traits in Sävar common garden.

Biplot of the first two components of principal components analysis conducted on stem and leaf trait with arthropod abundance and species richness in the SwAsp common garden at Sävar. The first and second principal components respectively represent 34.5% and 51.4% of the total variation explained. Analysis was conducted on all clones, labelled as grey numbers from the populations indicated in Table 1. Trait abbreviations: Petiole = petiole length, Anth = anthocyanin index, Chl = chlorophyll index, SLA = specific leaf area, Abund = arthropod abundance, Sprich = species richness.

Table 6. Trait loadings of the first (PC1) and second (PC2) principal components of analysis on plant traits and arthropod indices in 2009, with PC1 explaining 34.5% and PC2 explaining 51.4% of the cumulative variation in the data set.

Non-parametric MANOVA tests were used to test plant traits as factors explaining arthropod community distance matrices. Table 7 shows results for 2009, indicating that stem height, stem diameter, leaf area, petiole length and anthocyanins, each individually explained variation in arthropod community composition. However, specific leaf area, condensed tannins, total phenolics, and chlorophyll content did not appear to influence community composition in 2009. These results were reflected in non-parametric MANOVA tests with plant traits in the 2008 data set, where height (F = 1.83, P = 0.048), leaf area (F = 10.18, P = 0.001), petiole length (F = 5.94, P = 0.012), anthocyanins (F = 23.55, P = 0.001) and condensed tannins (F = 58.94, P = 0.001) each showed significant explanation of the variance for the arthropod community matrix, but specific leaf area, total phenolics and chlorophyll content (a proxy for leaf N) did not. The H2 of spring budflush date (0.69) suggested that this trait had enough of a genetic component to be tested against data from previous seasons, and despite nearing a significant result in the np MANOVA (Table 7), we cannot conclude that it explained variance in either arthropod community data set and the relationship between arthropod community and tree height is not a result of differences in growing season length.

Table 7. Differentiation of arthropod community composition in 2009 by plant functional traits.


Genetic and geographic variation in plant traits

We found that plant functional traits including stem size, leaf morphology and basic leaf secondary chemistry, are heritable in P. tremula. Population and clonal variation in stem height and diameter indicates that the SwAsp collection has continued to exhibit the latitudinal clines in stem traits reported in the years directly after its establishment [16], [17]. Broad-sense heritability for stem and most leaf traits indicated strong differences between clones. This confirms recent observations of genetic variation in full-sib families of P. tremula for leaf phenolics and stem biomass [19] and stem traits in quaking aspen [11], [12]. In accordance with other studies in P. tremula we have confirmed a genetic component to individual leaf area [18], [78], which has also been observed in P. tremuloides [12], [79]. It has been found in Populus trichocarpa×P. deltoides hybrids that petiole length is a trait under the control of a few major genetic loci [80]. While there are weak correlations between tree height and petiole length in the SwAsp collection and in hybrid poplars [81], Wu et al. [82] found no genetic correlation between these two traits. Our data also show high levels of broad-sense heritability (H2 = 0.69) for spring bud flush phenology, reporting higher values than previously recorded in SwAsp [17] and other studies in hybrid poplars [83], [84]. The variation in these plant traits defines the range of living environments and food quality selected by arthropod herbivores.

Genetic and geographic variation in canopy arthropods

Surveys indicated that P. tremula in Sweden supports a wide array of arthropod herbivores, and many of these species are specialists on aspen. This knowledge considerably enhances existing documentation of the canopy arthropod fauna on P. tremula and confirms the status of aspen as an important keystone species for biodiversity [3][5]. We discovered genetic variation in the arthropod assemblage in both 2008 to 2009, alongside an increase in the number of species and the proportion of trees affected by these species. We also observed differences in community composition between the populations of aspen. Earlier reports have identified heritability in community composition in Populus species but in smaller collections of genotypes than ours [33], [49]. Here we report heritability values for arthropod abundance, species richness and diversity (Shannon index), traits that differ between regional populations, suggesting that there is regional variation in resistance to canopy arthropods in accordance with our previous work on leaf-mining moths and leaf rust fungus [65].

Plant traits influencing the arthropod community

We analyzed sessile leaf-feeding and leaf-dwelling arthropods and expected the assemblage of arthropod herbivores, whose relationship with the leaf is intimate and dependent on leaf tissue throughout one or more life-cycle stages, to vary in accordance with the leaf characteristics of plants that they select. Using principal components analysis we have shown that petiole length and leaf area are the next most important traits influencing the PC1 (Table 6) after tree height, arthropod abundance and species richness. The strong relationships between petiole length and individual leaf area with community composition indicates that variation in these traits likely influences the choice of host leaf tissue, especially given that these leaf-dwelling arthropods inhabit individual leaves or petioles during a vulnerable developmental stage [85], [86]. For this reason it is surprising that we have not identified a relationship between herbivores and specific leaf area [87]. It is also likely that larger leaves and petioles are chosen by herbivores for providing a greater amount of food. The weak relationship (Pearson correlation r = 0.123, P = 0.03) between chlorophyll content and leaf area indicates that leaf nitrogen content is not the basis of the relationship between leaf area and arthropod community.

We measured traits representing chemical resistance (total phenolics, condensed tannins and anthocyanins) and plant quality (leaf area, specific leaf area and chlorophyll, a proxy of leaf nitrogen content [88]). In contrast to previous reports in European [65] and trembling [44], [66], [89] aspen where herbivore damage was related to concentrations of phenolic glycosides, we identified no such correlation with total phenolics in our data from either year. We suggest that because most species in our surveys are specialists on Populus and well adapted to a diet rich in phenolic glycosides, they are not deterred by, and may benefit from [52], [53] leaf secondary chemistry as would be expected for generalists. This study investigated pooled classes of phenolic compounds rather than, for example, individual phenolic glycosides known to influence particular herbivores on Populus [66], [21]. Further analysis of specific phenolic defence compounds, rather than total phenolics and total condensed tannins, could confirm or refute this suggestion, and recent work reports a wider range of phenolic defence compounds in P. tremula than was previously detected [20]. We did identify significant partitioning of the variation in community composition by condensed tannins (in 2008) and anthocyanins (in both 2008 and 2009), both of which are non-structural compounds with common precursors in the phenylpropanoid pathway [90]. While much work is underway to dissect the relationships between the different compounds and their synthesis in this pathway [91], [92], our understanding of the biochemical and functional connection between condensed tannins and anthocyanins is not complete. However, it is from this phenylpropanoid pathway that the secondary metabolites affecting palatability of leaf tissue to arthropods are produced [92].

Numerous factors dictate the nutritive quality of plant tissue for herbivores. The products of primary metabolism can be allocated to the accumulation of biomass or the production of defensive compounds, or to both processes [59]. This balance has long been the subject of theoretical discussion [56], [57], [59], [60]. The variance in our herbivore community data explained by tree size, which is a function of growth rate in common garden plants of identical age, and the similarity of herbivore data to population structure for tree size, suggest that allocation of resources to growth is highly relevant to the herbivore community. This is consistent with other reports of the relationship between herbivore performance and plant growth [93][95]. However our results also indicate that the availability of leaf surface (leaf area and petiole length) and leaf nutritive quality (anthocyanins and condensed tannins) at this stage in development, is a significant influence in this arthropod community dominated by specialists.

Differences between developmental and ontogenetic stages in Populus physiology and morphology have been reported [12], [13], [96]. In accordance with a recent meta-analysis [97] we expect that in future surveys of the SwAsp trees, changes in response to herbivory would be punctuated by defined developmental stages, in particular terminal differentiation and sexual maturity, and these changes may be projected onto the arthropod community. We agree with Boege and Marquis [98] that a complete picture of the mechanisms of response to herbivores cannot be extrapolated from one ontogenic stage.

Having identified substantial intraspecific variation in several traits, we require an understanding of the mechanisms influencing these relationships in order to ascertain the impact of European aspen as a component of wider ecological systems. Genetic diversity within plant populations and plant communities has been shown to have ecological impact on consumer communities and wider ecosystem effects [89], [90], although we have yet to detect whether genetic variation in P. tremula can influence multi-trophic communities beyond the arthropod herbivore community as found in other Populus species [49], [51]. Our identification of nearly 100 arthropod morphs in the canopy of young P. tremula plants strengthens the evidence that genetic variation in aspen is important for maintaining biodiversity, particularly of specialist arthropods. We conclude that geographic and genetic factors, in particular plant vigour and leaf quality, structure the projection of phenotypic variation in young aspen trees to the herbivore communities that they host. Furthermore, these patterns reveal natural variation in susceptibility to insects, a useful resource as we seek mechanisms that could be deployed in breeding trees for herbivore resistance.

Supporting Information

Table S1.

Arthropod morphospecies recorded in the SwAsp common gardens at Sävar and Ekebo.



We are grateful to Skogforsk for hosting the SwAsp common gardens. We thank Estelle Schaefer, Mathilde Closset and Yvan Fracheboud for assistance in the field. Thanks to Nathaniel Street for leaf rust data collection and discussion. We appreciate assistance from Norrlands Entomologiska Förening in the identification of arthropod specimens.

Author Contributions

Conceived and designed the experiments: KMR BRA. Performed the experiments: KMR. Analyzed the data: KMR PKI. Contributed reagents/materials/analysis tools: SJ BRA. Wrote the paper: KMR BRA.


  1. 1. Kuusinen M (1996) Epiphyte flora and diversity on basal trunks of six old-growth forest tree species in southern and middle boreal Finland. Lichenologist 28: 443–463.
  2. 2. Hedenås H, Blomberg P, Ericson L (2007) Significance of old aspen (Populus tremula) trees for the occurrence of lichen photobionts. Biological Conservation 135: 380–387.
  3. 3. Suominen O, Edenius L, Ericsson G, Resco de Dios V (2003) Gastropod diversity in aspen stands in coastal northern Sweden. Forest Ecology and Management 175: 403–412.
  4. 4. Myking T, Bøhler F, Austrheim G, Solberg EJ (2011) Life history strategies of aspen (Populus tremula L.) and browsing effects: a literature review. Forestry 84: 61–71.
  5. 5. Kouki J, Arnold K, Martikainen P (2004) Long-term persistence of aspen – a key host for many threatened species – is endangered in old-growth conservation areas in Finland. Journal for Nature Conservation 12: 41–52.
  6. 6. Gustafsson L, Kouki J, Sverdrup-Thygeson A (2010) Tree retention as a conservation measure in clear-cut forests of northern Europe: a review of ecological consequences. Scandinavian Journal of Forest Research 25: 295–308.
  7. 7. Mitton JB, Grant MC (1996) Genetic variation and the natural history of quaking aspen. BioScience 46: 25–31.
  8. 8. Jelinski DE, Cheliak WM (1992) Genetic diversity and spatial subdivision of Populus tremuloides (Salicaceae) in a heterogeneous landscape. American Journal of Botany 79: 728–736.
  9. 9. Lindroth RL, Hwang S-Y (1996) Clonal variation in foliar chemistry of quaking aspen (Populus tremuloides Michx.). Biochemical Systematics and Ecology 24: 357–364.
  10. 10. Mock KE, Rowe CA, Hooten MB, Dewoody J, Hipkins VD (2008) Clonal dynamics in western North American aspen (Populus tremuloides). Molecular Ecology 17: 4827–4844.
  11. 11. Kanaga MK, Ryel RJ, Mock KE, Pfrender ME (2008) Quantitative-genetic variation in morphological and physiological traits within a quaking aspen (Populus tremuloides) population. Canadian Journal of Forest Research 38: 1690–1694.
  12. 12. St Clair SB, Mock KE, LaMalfa EM, Campbell RB, Ryel RJ (2010) Genetic contributions to phenotypic variation in physiology, growth, and vigor of Western Aspen (Populus tremuloides) clones. Forest Science 56: 222–230.
  13. 13. Smith EA, Collette SB, Boynton TA, Lillrose T, Stevens MR, et al. (2011) Developmental contributions to phenotypic variation in functional leaf traits within quaking aspen clones. Tree Physiology 31: 68–77.
  14. 14. Suvanto LI, Latva-Karjanmaa TB (2005) Clone identification and clonal structure of the European aspen (Populus tremula). Molecular Ecology 14: 2851–2860.
  15. 15. De Carvalho D, Ingvarsson PK, Joseph J, Suter L, Sedivy C, Macaya-Sanz D, et al. (2010) Admixture facilitates adaptation from standing variation in the European aspen (Populus tremula L.), a widespread forest tree. Molecular Ecology 19: 1638–1650.
  16. 16. Hall D, Luquez V, Garcia VM, St Onge KR, Jansson S, et al. (2007) Adaptive population differentiation in phenology across a latitudinal gradient in European aspen (Populus tremula, L.): a comparison of neutral markers, candidate genes and phenotypic traits. Evolution 61: 2849–2860.
  17. 17. Luquez V, Hall D, Albrectsen BR, Karlsson J, Ingvarsson P, et al. (2008) Natural phenological variation in aspen (Populus tremula): the SwAsp collection. Tree Genetics and Genomes 4: 279–292.
  18. 18. Bylesjö M, Segura V, Soolanayakanahally RY, Rae AM, Trygg J, et al. (2008) LAMINA: a tool for rapid quantification of leaf size and shape parameters. BMC Plant Biology 8: 82.
  19. 19. Kleeman F, von Fragstein M, Vornam B, Müller A, Leuschner C, et al. (2011) Relating genetic variation of ecologically important tree traits to associated organisms in full-sib aspen families. European Journal of Forest Research 130: 707–716.
  20. 20. Abreu IN, Ahnlund M, Moritz T, Albrectsen BR (2011) UHPLC-ESI/TOFMS Determination of salicylate-like phenolic gycosides in Populus tremula leaves. Journal of Chemical Ecology 37: 857–70.
  21. 21. Philippe RN, Bohlmann J (2007) Poplar defense against herbivores. Canadian Journal of Botany 85: 1111–1126.
  22. 22. Antonovics J (1992) Towards community genetics. In: Fritz RS, Simms EL, editors. Plant resistance to herbivores and pathogens: Ecology, evolution and resistance. Chicago: University of Chicago Press.
  23. 23. Whitham TG, Young WP, Martinsen GD, Gehring CA, Schweizer JA, et al. (2003) Community and ecosystem genetics: a consequence of the extended phenotype. Ecology 84: 559–573.
  24. 24. Dungey HS, Potts BM, Whitham TG, Li H-F (2000) Plant genetics affects arthropod community richness and composition: evidence from a synthetic eucalypt hybrid population. Evolution 54: 1938–1946.
  25. 25. Hochwender CG, Fritz RS (2004) Plant genetic differences influence community structure: evidence from a hybrid willow system. Oecologia 138: 547–557.
  26. 26. Tovar-Sanchez E, Oyama K (2006) Community structure of canopy arthropods associated to Quercus crassifolia×Quercus crassipes complex. Oikos 112: 370–381.
  27. 27. Nakamura M, Asanuma M, Hiura T (2010) Differential effects of host plant hybridization on herbivore community structure and grazing pressure on forest canopies. Oikos 119: 1445–1452.
  28. 28. Johnson MT, Agrawal AA (2005) Plant genotype and environment interact to shape a diverse arthropod community on evening primrose (Oenothera biennis). Ecology 86: 874–885.
  29. 29. Crawford KM, Crutsinger GM, Sanders NJ (2007) Genotypic diversity mediates the distribution of an ecosystem engineer. Ecology 88: 2114–2120.
  30. 30. Gols R, Wagenaar R, Bukovinszky T, Van Dam NM, Dicke M, et al. (2008) Genetic variation in defense chemistry in wild cabbages affects herbivores and their endoparasitoids. Ecology 89: 1616–1626.
  31. 31. Barbour RC, O'Reilly-Wapstra JM, DeLittle DW, Jordan GJ, Steane DA, et al. (2009) A geographic mosaic of genetic variation within a foundation tree species and its community-level consequences. Ecology 90: 1762–1772.
  32. 32. Utsumi S, Ohgushi , T (2009) Community-wide impacts of herbivore-induced plant regrowth on arthropods in a multi-willow species system. Oikos 188: 1805–1815.
  33. 33. Keith AR, Bailey JK, Whitham (2010) A genetic basis to community repeatability and stability. Ecology 91: 3398–3406.
  34. 34. Stevens MT, Kruger EL, Lindroth RL (2008) Variation in tolerance to herbivory is mediated by differences in biomass allocation in aspen. Functional Ecology 22: 40–47.
  35. 35. Kearsley MJ, Whitham TG (1997) The developmental stream of cottonwoods affects ramet growth and resistance to galling aphids. Ecology 79: 178–191.
  36. 36. Waltz AM, Whitham TG (1997) Plant development affects arthropod communities: opposing impacts of species removal. Ecology 78: 2133–2144.
  37. 37. Albrectsen BR, Gardfjell H, Orians CM, Murray B, Fritz RS (2004) Slugs, willow seedlings and nutrient fertilization: intrinsic vigor inversely affects palatability. Oikos 105: 268–278.
  38. 38. Boege K (2005) Herbivore attack in Casearia nitida influenced by plant ontogenetic variation in foliage quality and plant architecture. Oecologia 143: 117–125.
  39. 39. Holeski LM, Kearsley MJC, Whitham TG (2009) Separating ontogenetic and environmental determination of resistance to herbivory in cottonwood. Ecology 90: 2969–2973.
  40. 40. Rudgers JA, Whitney KD (2006) Interactions between insect herbivores and a plant architectural dimorphism. Journal of Ecology 94: 1249–1260.
  41. 41. Marques ESD, Price PW, Cobb NS (2000) Resource abundance and insect herbivore diversity on woody fabaceous desert plants. Environmental Entomology 29: 696–703.
  42. 42. Yarnes CT, Boecklen WJ (2005) Abiotic factors promote plant heterogeneity and influence herbivore performance and mortality in Gambel's oak (Quercus gambelii). Entomologia Experimentalis et Applicata 114: 87–95.
  43. 43. Auerbach M, Alberts JD (1992) Occurrence and performance of the aspen blotch miner, Phyllonorycter salicifoliella, on three host-tree species. Oecologia 89: 1–9.
  44. 44. Osier TL, Hwang SY, Lindroth RL (2000) Effects of phytochemical variation in quaking aspen Populus tremuloides clones on gypsy moth Lymantria dispar performance in the field and laboratory. Ecological Entomology 25: 197–207.
  45. 45. Peacock L, Lewis M, Powers S (2001) Volatile Compounds from Salix spp. Varieties Differing in Susceptibility to Three Willow Beetle Species. Journal of Chemical Ecology 27: 1943–1951.
  46. 46. Hanley M, Lamont B, Fairbanks M, Rafferty C (2007) Plant structural traits and their role in anti-herbivore defence. Perspectives in Plant Ecology, Evolution and Systematics 8: 157–178.
  47. 47. Stiling P, Bowdish TI (2000) Direct and indirect effects of plant clone and local environment on herbivore abundance. Ecology 81: 281–285.
  48. 48. Valkama E, Koricheva J, Salminen J-P, Helander M, Saloniemi I, et al. (2005) Leaf surface traits: overlooked determinants of birch resistance to herbivores and foliar micro-fungi? Trees - Structure and Function 19: 191–197.
  49. 49. Bailey JK, Wooley SC, Lindroth RL, Whitham TG (2006) Importance of species interactions to community heritability: a genetic basis to trophic-level interactions. Ecology Letters 9: 78–85.
  50. 50. Andrew R, Wallis I, Harwood C, Henson M, Foley W (2007) Heritable variation in the foliar secondary metabolite sideroxylonal in Eucalyptus confers cross-resistance to herbivores. Oecologia 153: 891–901.
  51. 51. Schweitzer JA, Bailey JK, Rehill BJ, Martinsen GD, Hart SC, et al. (2004) Genetically based trait in a dominant tree affects ecosystem processes. Ecology Letters 7: 127–134.
  52. 52. Pasteels JM, Rowell-Rahier M, Braekman JC, Dupont A (1983) Salicin from host plant as precursor of salicylaldehyde in defensive secretion of Chrysomeline larvae. Physiological Entomology 8: 307–314.
  53. 53. Rowell-Rahier M (1984) The presence or absence of phenolglycosides in Salix (Salicaceae) leaves and the level of dietary specialisation of some of their herbivorous insects. Oecologia 62: 26–30.
  54. 54. Feeny P (1976) Plant Apparency and Chemical Defense. In: Wallace JW, Nansel RL, editors. Biological Interactions Between Plants and Insects. Recent Advances in Phytochemistry, Vol 10. New York: Plenum Press.
  55. 55. Rhoades DF, Coates RG (1976) Toward a general theory of plant antiherbivore chemistry. In: Wallace JW, Mansell RL, editors. Recent Advances in Phytochemistry, Volume 10. New York: Plenum Press.
  56. 56. Bryant JP, Chapin FS, Klein DR (1983) Carbon/nutrient balance of boreal plants in relation to vertebrate herbivory. Oikos 40: 357–368.
  57. 57. Coley PD, Bryant JP, Chapin FS (1985) Resource availability and plant antiherbivore defense. Science 230: 895–899.
  58. 58. Price PW (1991) The plant vigor hypothesis and herbivore attack. Oikos 62: 244–251.
  59. 59. Herms DA, Mattson WJ (1992) The dilemma of plants: to grow or defend. Quarterly Review of Biology 67: 283–335.
  60. 60. Jones CG, Hartley SE (1999) A Protein Competition Model of Phenolic Allocation. Oikos 86: 27–44.
  61. 61. Agrawal AA (2007) Macroevolution of plant defense strategies. Trends in Ecology and Evolution 22: 103–109.
  62. 62. Carmona D, Lajeunesse MJ, Johnson MT (2011) Plant traits that predict resistance to herbivores. Functional Ecology 25: 358–367.
  63. 63. Turesson G (1922) The genotypical response of the plant species to the habitat. Hereditas 3: 211–350.
  64. 64. Morgenstern EK (1996) Geographic Variation in Forest Trees. Vancouver: UBC Press.
  65. 65. Albrectsen BR, Witzell J, Robinson KM, Wulff S, Luquez VMC, et al. (2010) Large scale geographic clines of parasite damage to Populus tremula L. Ecography 33: 483–493.
  66. 66. Hemming JDC, Lindroth RL (1995) Intraspecific variation in aspen phytochemistry: effects on performance of gypsy moths and forest tent caterpillars. Oecologia 103: 79–88.
  67. 67. Smith DS, Bailey JK, Shuster SM, Whitham TG (2010) A geographic mosaic of trophic interactions and selection: trees, aphids and birds. Journal of Evolutionary Biology 24: 422–429.
  68. 68. Wimp GM, Wooley S, Bangert RK, Young WP, Martinsen GD, et al. (2007) Plant genetics predicts intra-annual variation in phytochemistry and arthropod community structure. Molecular Ecology 16: 5057–5069.
  69. 69. Delaplanque A, editor. (1998) Les insects associés aux peupliers. Brussels: Éditions Memor.
  70. 70. Coulianos C, Holmåsen I (1991) Galler: En fälthandbok om gallbildningar på vilda och odlade växter. Stockholm: Interpublishing.
  71. 71. Abramoff MD, Magelhaes PJ, Ram SJ (2004) Image Processing with ImageJ. Biophotonics International 11: 36–42.
  72. 72. Ossipova S, Ossipov V, Haukioja E, Loponen J, Pihlaja K (2001) Proanthocyanidins of mountain birch leaves: quantification and properties. Phytochemical Analysis 12: 128–133.
  73. 73. R Development Core Team (2006) R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna.
  74. 74. Hadfield J (2010) MCMC methods for Multi–response Generalised Linear Mixed Models: The MCMCglmm R Package. Journal of Statistical Software 33: 1–22.
  75. 75. Lynch M, Walsh B (1997) Genetics and analysis of quantitative traits. Sunderland: Sinauer Associates Inc..
  76. 76. Oksanen J, Blanchet FG, Kindt R, Legendre P, Minchin PR, et al. (2011) vegan: Community Ecology Package. R package version 2.0-2.
  77. 77. Anderson MJ (2001) A new method for non-parametric multivariate analysis of variance. Austral Ecology 26: 32–46.
  78. 78. López-de-Heredia U, Sierra-de-Grado R, Pando V, Cristóbal MD, Martínez-Zurimendi PMartínTA (2004) Comparison of isozyme and morphological markers to assess the within population variation in small populations of European aspen (Populus tremula L.) in Spain. Silvae Genetica 53: 227–233.
  79. 79. Barnes BV (1969) Natural Variation and Delineation of Clones of Populus tremuloides and P. grandidentata in Northern Lower Michigan. Silvae Genetica 18: 130–142.
  80. 80. Wu RL, Bradshaw HD, Stettler RF (1997) Molecular genetics of growth and development in Populus (Salicaceae). V. Mapping quantitative trait loci affecting leaf variation. American Journal of Botany 84: 143–153.
  81. 81. Marron N, Dillen SY, Ceulemans R (2007) Evaluation of leaf traits for indirect selection of high yielding poplar hybrids. Environmental and Experimental Botany 61: 103–116.
  82. 82. Wu RL, Stettler RF (1994) Quantitative genetics of growth and development in Populus. I. A three-generation comparison of tree architecture during the first 2 years of growth. Theoretical and Applied Genetics 89: 1046–1054.
  83. 83. Frewen BE, Chen THH, Howe GT, Davis J, Rohde A, et al. (2000) Quantitative trait loci and candidate gene mapping of bud set and bud flush in Populus. Genetics 154: 837–845.
  84. 84. Rae AM, Robinson KM, Street NR, Taylor G (2004) Structural, morphological and biochemical traits linked to fast growth in short-rotation coppice poplar for biomass. Canadian Journal of Forest Research 34: 1488–1498.
  85. 85. Strong DR, Lawton JH, Southwood TRE (1984) Insects on plants: community patterns and mechanisms. Oxford: Blackwell.
  86. 86. Fukui A (2001) Indirect interactions mediated by leaf shelters in animal–plant communities. Population Ecology 43: 31–40.
  87. 87. Peeters PJ (2002) Correlations between leaf structural traits and the densities of herbivorous insect guilds. Biological Journal of the Linnean Society 77: 43–65.
  88. 88. Chang SX, Robison DJ (2003) Nondestructive and rapid estimation of hardwood foliar nitrogen status using the SPAD-502 chlorophyll meter. Forest Ecology and Management 181: 331–338.
  89. 89. Donaldson JR, Lindroth RL (2007) Genetics, environment, and their interaction determine the efficacy of chemical defense in trembling aspen. Ecology 88: 729–739.
  90. 90. Winkel-Shirley B (2001) Flavonoid biosynthesis. A colorful model for genetics, biochemistry, cell biology, and biotechnology. Plant Physiology 126: 485–493.
  91. 91. Tsai C-J, Harding SA, Tschaplinski TJ, Lindroth RL, Yuan Y (2006) Genome-wide analysis of the structural genes regulating defense phenylpropanoid metabolism in Populus. New Phytologist 172: 47–62.
  92. 92. Chen F, Liu C-J, Tschaplinski TJ, Zhao N (2009) Genomics of Secondary Metabolism in Populus: Interactions with Biotic and Abiotic Environments. Critical Reviews in Plant Science 28: 375–392.
  93. 93. Donaldson JR, Lindroth RL (2007) Genetics, environment, and their interaction determine the efficacy of chemical defense in trembling aspen. Ecology 88: 729–739.
  94. 94. Campos RI, Vasconcelos HL, Ribeiro SP, Neves FS, Soares JP (2006) Relationship between tree size and insect assemblages associated with Anadenanthera macrocarpa. Ecography 29: 442–450.
  95. 95. Santos JC, Fernandes GW (2010) Mediation of herbivore attack and induced resistance by plant vigor and ontogeny. Acta Oecologica 36: 617–625.
  96. 96. Donaldson JR, Stevens MT, Barnhill HR, Lindroth RL (2006) Age-Related shifts in leaf chemistry of clonal aspen (Populus tremuloides). Journal of Chemical Ecology 32: 1415–1429.
  97. 97. Barton KE, Koricheva J (2010) The ontogeny of plant defense and herbivory: characterising general patterns using meta-analysis. American Naturalist 175: 481–493.
  98. 98. Boege K, Marquis RJ (2005) Facing herbivory as you grow up: the ontogeny of resistance in plants. Trends in Ecology and Evolution 20: 441–448.