BIOTIC RESISTANCE TO NON-INDIGENOUS PLANTS: ARE PHYLOGENETICALLY NOVEL INVADERS MORE LIKELY TO ESCAPE ENEMIES?
By
Steven Burton Hill
A thesis submitted in conformity with the requirements
for the degree of Doctor of Philosophy
Graduate Department of Ecology and Evolutionary Biology
University of Toronto
© Copyright by Steven Burton Hill (2009) Biotic resistance to non-indigenous plants: are phylogenetically novel invaders more likely to escape enemies?
Steven B. Hill
Doctor of Philosophy
Department of Ecology & Evolutionary Biology
University of Toronto
June 2009
Abstract
The degree to which biotic interactions influence invasion success may partly depend on the evolutionary relationship between invaders and native species. In particular, since host-use by enemies such as invertebrate herbivores and fungal pathogens tends to be phylogenetically conserved, exotic plants that have close native relatives in the invaded range should be more likely to interact with enemies. In this thesis, I explore this idea using a series of experiments and field surveys at nested taxonomic levels.
My results indicate that exotics from multiple plant families experience lower damage if their average phylogenetic distance from locally co-occurring native family members is higher. I then demonstrate that within the Asteraceae, foliar and capitular damage are lower on exotic compared to native species. Both damage types had a relatively large phylogenetic component, but did not decline with phylogenetic distance to native or exotic confamilials. Finally, I show that communities with versus without close relatives are unlikely to differ in resistance to the novel invader, Solidago virgaurea: biotic resistance imposed by competitors, generalist vertebrates, and specialist invertebrates
ii resulted in similar patterns of damage and mortality regardless of the presence of congeneric natives. In some cases, effects of biota were positive: growth of S. virgaurea seedlings in soils collected near congeneric natives was enhanced more than in soils from communities where congenerics were absent.
Overall, these results suggest that biotic interactions between exotic and native species can be phylogenetically structured, although trends based on distance measures tend to be weak. In some cases, damage does decline with phylogenetic distance to native species; however this trend is unlikely to be a strong force limiting invasion or structuring plant communities. These results have significant implications for current theories of invasion biology including the "Enemy Release Hypothesis" and "Darwin's Naturalization
Hypothesis", as well as for community phylogenetics.
iii Acknowledgements
I would like to acknowledge all of the people who have provided support while completing this dissertation. I am indebted to my Ph.D. advisor Dr. Peter Kotanen who has fostered the perfect environment for me to grow as a graduate student. From beginning to end, in the lab and in the field, Peter has gone above and beyond the role as an advisor. I am equally thankful to my wife, Sheri, who has been an unending source of love, understanding and emotional support. I undoubtedly owe her for the many summer days that I spent away from home while in the field at KSR. I would also like to thank all of my other family and friends that have supported my academic pursuits (for this thesis and my last). In particular, my dad (Brian), mom (Susan), step dad (Larry), and sister (Karen) have always been enthusiastic supporters of my work.
Many discussions with official and unofficial members of the Kotanen Lab, including Andrew MacDonald, James Mackay, Katya Kostyukova, and Megan Saunders- Caradonna, the Plant Insect Group (Marc Johnson, Jessamyn Manson, and Russell Dinnage), and my advisory committee (Dr. Spencer Barrett and Dr. Sasa Stefanovic) helped keep me on track and clarify fuzzy thoughts on ecology and evolution.
As with any large research project, completion would not be easy without dedicated logistical help. Katya Kostyukova provided excellent assistance whether collecting data in the field or lab for over three years. As well, Nathalie Taraban-Lagois, Gilbert Tang, Anna Simonsen, Russell Dinnage and Justin Anchetta all provided field assistance during the 2005 field season. Greenhouse help from Jutta Stein was invaluable for keeping my experimental plants alive at UTM while I prepared for fieldwork or was out of town. The KSR grounds crew (John Jensen and his staff) were always ready to plow a plot or deliver a watering tank.
Funding for the work completed in this dissertation was provided by NSERC, OGS, and the University of Toronto.
iv Table of Contents
Abstract...... ii Acknowledgements...... iv Table of Contents...... v List of Tables...... vii List of Figures...... ix CHAPTER 1 - General Introduction...... 1 CHAPTER 2 - Evidence that phylogenetically novel non-indigenous plants experience less herbivory...... 17 Abstract...... 17 Introduction...... 18 Methods...... 21 Results...... 27 Discussion...... 28 CHAPTER 3 - Enemy release from foliar enemies for exotic species in the Asteraceae: a phylogenetically controlled study...... 46 Abstract...... 46 Introduction...... 47 Methods...... 50 Results...... 54 Discussion...... 55 CHAPTER 4 - Enemy release from pre-dispersal seed predators in Asteraceae: a phylogenetically controlled study...... 76 Abstract...... 76 Introduction...... 77 Methods...... 80 Results...... 85 Discussion...... 86 CHAPTER 5 - Biotic constraints experienced by a novel invader: close relatives have little impact on herbivore damage or competition...... 106 Abstract...... 106 Introduction...... 107 Methods...... 110 Results...... 115 Discussion...... 119
v CHAPTER 6 - Positive effects of soil biota for a novel invader: implications for habitat-specific invasion success when close relatives are present...... 138 Abstract...... 138 Introduction...... 139 Methods...... 143 Results...... 148 Discussion...... 151 CHAPTER 7 - General Discussion...... 168
vi List of Tables
Table 2.1 List of exotic plants used in common garden experiments and field surveys...... 41
Table 2.2 Regression and AIC results for foliar damage and phylogenetic distance.... 43
Table 3.1 Study species belonging to Asteraceae used to measure foliar damage...... 69
Table 3.2 Multiple regression results of damage versus significant phylogenetic eigenvectors...... 70
Table 3.3 Regression results of foliar damage versus phylogenetic distance to confamilial species...... 71
Table 4.1 Study species belonging to Asteraceae used to measure proportion of 98 capitula damaged......
Table 4.2 ANOVA results for proportion of capitula damage...... 99
Table 4.3 Multiple regression results of capitular damage versus significant phylogenetic eigenvectors...... 100
Table 4.4 Regression results of capitular damage versus phylogenetic distance to confamilial species...... 101
Table 5.1 Summary of leaf area damage and proportion survival...... 129
Table 5.2 Split-plot ANOVA for habitat and vegetation removal on damage, damage type principal components, and survival for Solidago virgaurea...... 130
vii Table 5.3 Regression parameters from ANCOVA of foliar damage...... 131
Table 5.4 Correlation between different types of damage on Solidago virgaurea...... 132
Table 5.5 Summary of principal component analysis of damage types on Solidago virgaurea...... 133
Table 6.1 Split-plot factorial ANOVA for habitat and soil sterilization on shoot and root mass of Solidago virgaurea...... 162
Table 6.2 Split-plot factorial ANOVA for habitat and fungicide application on seed germination of Solidago virgaurea...... 162
viii List of Figures
Figure 2.1 Regression of herbivory and phylogenetic distance for exotic species used in common garden experiments...... 44
Figure 2.2 Regressions of herbivory and phylogenetic distance for exotic species sampled in naturally occurring populations at KSR...... 45
Figure 3.1 Difference in mean foliar damage between native and exotic Asteraceae 72
Figure 3.2 Community phylogeny for native and exotic Asteraceae with summed phylogenetic eigenvectors that predict foliar damage...... 73
Figure 3.3 Difference in mean phylogenetically-corrected foliar damage between native and exotic Asteraceae...... 74
Figure 3.4 Proportion of leaf area damaged on exotic Asteraceae versus phylogenetic distance to confamilials at KSR...... 75
Figure 4.1 Community phylogeny for native and exotic Asteraceae with summed phylogenetic eigenvectors that predict capitular damaged...... 102
Figure 4.2 Difference in mean capitular damage between native and exotic Asteraceae...... 103
Figure 4.3 Difference in mean phylogenetically-corrected capitular damage between native and exotic Asteraceae and sample time...... 104
Figure 4.4 Proportion of capitula damaged versus phylogenetic distance to confamilials at KSR...... 105
Figure 5.1 Interaction plot of foliar damage on Solidago virgaurea based on adjacent vegetation removal and habitat type...... 134
ix Figure 5.2 Foliar damage on Solidago virgaurea versus density of native Solidago spp...... 135
Figure 5.3 Bi-plot of herbivory and fungal infection principal components for Solidago virgaurea grouped by habitat type and adjacent vegetation removal...... 136
Figure 5.4 Interaction plot of Solidago virgaurea survival based on adjacent vegetation removal and habitat type...... 137
Figure 6.1 Shoot and root biomass for Solidago virgaurea when grown in sterilized and non-sterilized soils from different habitats...... 164
Figure 6.2 Proportion seed germination for Solidago virgaurea, Solidago canadensis, and Solidago nemoralis in the absence and presence of soil fungi in different habitats...... 165
x Chapter 1
General Introduction
"As species of the same genus have usually, though by no means invariably, some similarity in habits and constitution, and always in structure, the struggle will be more severe between species of the same genus, when they come into contact with each other, than between species of distinct genera."
(Darwin, 1859)
Biological invasions are the outcome of the successful transport, colonization, and subsequent spread of species in areas not previously occupied (Elton 1958). Although not necessarily human-driven, the frequency and rapidity at which novel species are being introduced has dramatically increased with the onset of modern global transport and commerce (Mack et al. 2000). Whether deliberate or accidental, this movement of species represents one of the most important conservation concerns worldwide (Mack et al. 2000; Pimentel et al. 2000). Successful invaders can alter ecosystem function (e.g.
Vitousek & Walker 1989; Mack et al. 2001), impact abundances of native species (e.g.
Braun 1950; King 1985), homogenize global community composition (Olden et al.
2006), and cost billions of dollars in economic damage (Mack et al. 2000; Pimentel et al.
2000; Colautti et al. 2006). Identifying the species that are likely to invade and those that will have the largest impact is critically important for their management (Mack et al.
2000).
1 Predicting invasions - traditional approaches
Since Elton's (1958) landmark publication, there has been considerable scientific and practical interest in predicting invasions. Attempts have typically been made from two perspectives: i) testing whether particular life history traits are correlated with invader success (Rejmanek & Richardson 1996; Williamson & Fitter 1996) and ii) testing whether invaded communities have particular characteristics that make them more susceptible (Elton 1958; Naeem et al. 2000; Fargione & Tilman 2005). The predictive ability of life history traits is mixed (Baker 1974; Williamson & Fitter 1996; Mack 2003), and in many cases this approach may only be useful among groups of closely related species (Richardson & Bond 1991). For example, characteristics of species in the pine family can explain their invasion success in South Africa (Richardson & Bond 1991); however, the utility of the same set of traits for other invaders may be limited. Other approaches such as climate matching have sometimes proven successful, but the accuracy of the results are directly proportional to how well climatic conditions in their home range reflect their true potential (Mack 1996). For example, if species have not yet reached their full geographic extent in their native range, or they occupy microclimates that are uncharacteristic of the mean climate, these unknown factors would produce false positive or negatives when attempting to predict future spread (Mack 1996). Alternative approaches have been made to predict invasion based on community susceptibility. For example, resistance to invaders can vary depending on niche availability (Mack 1996), disturbance (MacDougall & Turkington 2005), and biotic interactions (Elton 1958;
Naeem et al. 2000; Fargione & Tilman 2005). However, testing the relative roles of these factors is complicated by the idiosyncratic nature of introduction pressures (i.e. propagule
2 pressure (Lockwood et al. 2005) and the lack of historical knowledge of the ecological contexts during the very early stages of introduction (Mack et al. 2000).
Among the community-based hypotheses for invasion success, escape from biotic constraints has played a prominent role in attempts to explain why invaders are successful (Darwin 1859; Elton 1958; Strong et al. 1984; Mack 1996; Keane & Crawley
2002; Torchin et al. 2003; Cappuccino & Carpenter 2005). Invaders often are believed to lose their natural enemies such as consumers, pathogens, and parasites during invasion of new areas; loss of these enemies can translate into individual and population-level performances that are greatly enhanced compared to conspecifics in their native range
(Maron & Vila 2001; Colautti et al. 2004); this has been termed the "Enemy Release
Hypothesis" (ERH) (Elton 1958; Keane & Crawley 2002; Mitchell & Power 2003;
Torchin et al. 2003). However, biogeographic comparisons of enemy constraints may be less relevant than comparisons between invaders and the native species with which they are likely to compete (Keane & Crawley 2002). Studies that restrict this comparison to close relatives have found varied results (Colautti et al. 2004). In an example where the same species have been studied for multiple years, patterns of damage between exotic and native species can be contradictory (Agrawal & Kotanen 2003; Agrawal et al. 2005): in the first year of an experiment, exotics experienced higher foliar damage than congeneric (and one confamilial) native species; in the following year using the exact same species, exotics experienced lower damage. More importantly, in both years, differences in damage among exotic and native pairs varied significantly; for example, the exotic Plantago major L. tended to experience similar or higher damage than the
3 native Plantago rugellii Decne., whereas Cynanchum rossicum (Kleopov) Borh. tended to experience lower damage than the native Asclepias syriaca L. (both in the
Asclepiadaceae).
One reason that enemy release may fail, is if invaders are readily attacked by enemies in their invaded range; the risk of this attack may depend on an invader's relationship to native species. There is convincing evidence that host use by enemies tends to occur within specific taxonomic groups or phylogenetic lineages, particularly for invertebrate herbivores and pathogens (Novotny & Basset 2005; Odegaard et al. 2005; Weiblen et al.
2006; Gilbert & Webb 2007). When host recognition by such enemies is even moderately phylogenetically conserved, variability in an exotic species' escape from enemies should reflect their phylogenetic relationship to native members of the community being invaded
(Connor et al. 1980; Mack 1996; Cappuccino & Carpenter 2005; Dawson et al. 2009): invaders with close relatives should be more likely to share enemies and experience higher damage, than those that are distantly related to native species.
Predicting invasions - a community phylogenetic approach
If enemy escape of an invader can be predicted by its phylogenetic relationship to native species, this would tell us a great deal about how macroevolutionary dynamics can influence contemporary biotic interactions (Ehrlich & Raven 1964; Strong et al. 1984;
Becerra 2007; Johnson & Stinchcombe 2007) and mechanisms of community assembly
(Keddy 1992; Webb et al. 2002). In a practical context, it could help identify species that are more likely to succeed or become invasive because of an absence of enemy
4 constraints. In fact, phylogenetically informed hypothesis of invasion success have existed for over 150 years. Indeed, Darwin's initial assumption that close relatives are more likely to co-occur given their shared affinities for similar environments was reversed after observing that invaders in North America tended to belong to novel genera
(Darwin 1859). Both of these hypotheses showed tremendous insight, and now form the fundamental framework of community phylogenetics, the emerging discipline that links patterns and processes of community assembly with evolutionary history (Webb et al.
2002). There is even evidence that both patterns may occur simultaneously at different spatial scales (Diez et al. 2008; Proches et al. 2008): at larger scales close relatives may be more likely to co-occur when environmental conditions filter species based on trait suitability, whereas at small scales, biotic interactions such as competition among species with similar traits are assumed to limit co-occurrence (Diez et al. 2008; Proches et al.
2008).
There have been various attempts to evaluate whether phylogenetic information can predict invasion success. Approaches can be broadly categorized into one of three, non- mutually exclusive groups: i) phylogenetic and taxonomic bias among invaders, ii) increased invasion success of phylogenetically or taxonomically novel invaders, and iii) those that test the biotic mechanisms that can result in phylogenetic patterns in invasion success.
5 i) Phylogenetic and taxonomic bias in invasions
On a global scale, invasion has often been repeated by species belonging to the same taxonomic and phylogenetic groups. Among plants, the majority of successful invaders belong to relatively few families such as the Asteraceae, Brassicaceae, Fabaceae, and
Poaceae (Mack 1996; Rejmanek 1996). In part, this may simply reflect the large size of some these families; however, even after correcting for diversity, biases remain. For example, recent work utilizing a phylogenetic approach has shown that two of these groups, Asteraceae and Poaceae, tend to be over-represented as invaders in Australia, but only at the continental scale (Cadotte et al. 2009). Phylogenetic bias in invasion success is typically associated with the idea of niche conservatism, where traits that are associated with potential weediness (Baker 1974) or invasiveness (Richardson & Bond 1991) are more likely to occur among close relatives allowing them to be successful invaders.
However, these patterns could also reflect human choice for a non-random subset of species, with traits that are known to be desirable for a given region (Cadotte et al. 2009).
That phylogenetically non-random patterns exist for invaders at continental scales suggests that phylogenetically conserved traits associated with environmental and climatic conditions may be provide an effective means to predict patterns of invasion success at large scales; it is less clear how well this concept predicts invasion success at local scales, where biotic interactions directly or indirectly affect species in communities.
6 ii) Invasion success of phylogenetically novel invaders
Whether the presence of close relatives limits invasion into a community is central to
Darwin's prediction that taxonomically novel invaders are more successful (Darwin
1859). This idea has subsequently been termed Darwin's Naturalization Hypothesis, DNH
(Daehler 2001), and explicit tests of the hypothesis vary. Evidence that European grass species introduced to California tend to belong to novel genera provides support
(Rejmanek 1996); for this same group, the most invasive species tend to be less closely related to native species than those that are more benign (Strauss et al. 2006). In contrast, regional to continental surveys of naturalized and native species in Australia, Hawaii,
New Zealand, and Mediterranean Europe all reject establishment patterns expected by
DNH (Daehler 2001; Duncan & Williams 2002; Lambdon & Hulme 2006; Cadotte et al.
2009). These studies may reflect two processes operating at different spatial scales: at large spatial scales, invaders may be more likely to invade if they share traits with native species that are, for example, associated with broad climatic and environmental tolerances (environmental filtering); at small spatial scales where exotic and native species interact, niche similarities are expected to limit co-occurrence (Diez et al. 2008;
Proches et al. 2008). Since much of the work that has set out to test DNH has been based on large-scale surveys, the results may be more reflective of environmental filtering.
Indeed, Diez et al. (2008) found evidence for this among naturalized plants in New
Zealand: naturalization was positively correlated with the number of native species in an invader's respective genera, but negatively related to native congener abundance at local scales.
7 iii) Biotic mechanisms and phylogenetic patterns
Despite the number of studies that test patterns of naturalization for invaders relative to their phylogenetic or taxonomic affiliation to native species, none have tested the ecological mechanisms that would produce the predicted patterns of co-occurrence among native and exotic species using an explicitly phylogenetic approach. In fact, very little is known about whether the commonly assumed mechanistic underpinnings of DNH are responsible for patterns of exotic and native species co-occurrence in the field, particularly at local spatial scales. Consistent with Darwin's (1859) intuition, most studies that infer processes of community assembly based on community patterns assume that competition is the restrictive interaction when close relatives co-occur less often than expected by chance (Cavender-Bares et al. 2009). This assumption however, has rarely been tested, and alternative processes may be likely to limit species co-occurrence. For example, theoretical and empirical studies support the idea that shared enemy interactions among competing species can limit their co-occurrence (Holt 1977). Where this is more likely among close relatives, patterns of co-occurrence in the field will likely reflect those predicted by resource competition.
Currently, only one study has compiled a large amount of experimental data and used a phylogenetic approach to determine if competition is stronger among close relatives
(Cahill et al. 2008); their results suggested phylogenetic trends do exist, but are quite weak. Since their study evaluated competitive interactions between species using controlled experiments, the confounding effects of other environmental and biotic factors
8 were minimized. It is unknown how these results translate into patterns of coexistence and community assembly in real communities as interspecific interactions will undoubtedly be more complex than those represented by the experimental conditions used in the studies included by Cahill et al. (2008).
An alternative to the competition assumption, may be that shared enemies such as consumers, pathogens, and parasites produce patterns predicted by DNH. The degree to which close relatives share enemies is likely dictated by the conserved nature of traits among hosts and how well enemies are able to respond to those traits (Ehrlich & Raven
1964; Berenbaum 1995; Janz & Nylin 1998). In fact, evidence exists that the susceptibility of various plant hosts to foliar pathogens declines with increasing phylogenetic distance (Gilbert & Webb 2007), as well as evidence that host use by many invertebrate consumers is tightly linked to their host's phylogeny (Strong et al. 1984;
Mitter et al. 1991; Odegaard et al. 2005; Weiblen et al. 2006; Becerra 2007). To date, however, there has been no explicit phylogenetic approach used to ask the question: does escape from enemy constraints decline with the phylogenetic novelty of exotic species relative to native species in the community being invaded? I address this question in the five research chapters of this thesis in a series of nested taxonomic levels.
Goals of this Thesis
In the first chapter, I evaluate the relationship between foliar damage and phylogenetic relatedness for a broad group of plant invaders. Specifically, I investigated whether damage to exotic species declined with their phylogenetic novelty relative to native
9 species. In the second and third chapters, I focus on patterns of foliar damage and pre- dispersal seed predation on native and exotic plants in a single family, the Asteraceae. I evaluated whether damage by these enemies was lower on exotics even after controlling for phylogeny, and whether phylogenetically novel invaders experienced lower damage.
Finally, in chapters four and five, I used field and glasshouse experiments to determine if enemies in habitats with close relatives were more likely to constrain a novel invader,
Solidago virgaurea L.
The results from these chapters suggest that: (i) release from enemy constraints may only be significantly related to an invaders phylogeny when they belong to novel lineages at the family level or above; (ii) phylogeny and exotic status are strong predictors of foliar herbivory and pre-dispersal seed mortality, but this does not necessarily mean that phylogenetically novel invaders are likely to escape these constraints; and (iii) community resistance by invertebrates and pathogens may be higher when close relatives are present, however the impact of generalist herbivores, competitors, disturbance, and mutualists are likely to obscure any phylogenetic signal that may occur at the community scale. In each chapter, I discuss the implications of these results in the context of enemy escape (Mack 1996; Keane & Crawley 2002; Torchin et al. 2003) , Darwin's
Naturalization Hypothesis (Darwin 1859; Daehler 2001), and community phylogenetics
(Webb et al. 2002).
10 References
Agrawal A.A. & Kotanen P.M. (2003). Herbivores and the success of exotic plants: a
phylogenetically controlled experiment. Ecology Letters, 6, 712-715.
Agrawal A.A., Kotanen P.M., Mitchell C.E., Power A.G., Godsoe W. & Klironomos J.
(2005). Enemy release? An experiment with congeneric plant pairs and diverse
above- and belowground enemies. Ecology, 86, 2979-2989.
Baker H.G. (1974). The evolution of weeds. Annual Review of Ecology and Systematics,
5, 1-24.
Becerra J.X. (2007). The impact of herbivore-plant coevolution on plant community
structure. Proceedings of the National Academy of Sciences of the United States
of America, 104, 7483-7488.
Berenbaum M.R. (1995). The chemistry of defense - theory and practice. Proceedings of
the National Academy of Sciences of the United States of America, 92, 2-8.
Braun E.L. (1950). Deciduous forests of eastern North America. Hafner, New York, New
York, USA.
Cadotte M.W., Hamilton M.A. & Murray B.R. (2009). Phylogenetic relatedness and plant
invader success across two spatial scales. Diversity and Distributions, 15, 481-
488.
Cahill J.F., Kembel S.W., Lamb E.G. & Keddy P.A. (2008). Does phylogenetic
relatedness influence the strength of competition among vascular plants?
Perspectives in Plant Ecology Evolution and Systematics, 10, 41-50.
Cappuccino N. & Carpenter D. (2005). Invasive exotic plants suffer less herbivory than
non-invasive exotic plants. Biology Letters, 1, 435-438.
11 Cavender-Bares J., Kozak K.H., Fine P.V.A. & Kembel S.W. (2009). The merging of
community ecology and phylogenetic biology. Ecology Letters, in press.
Colautti R.I., Bailey S.A., van Overdijk C.D.A., Amundsen K. & MacIsaac H.J. (2006).
Characterised and projected costs of nonindigenous species in Canada. Biological
Invasions, 8, 45-59.
Colautti R.I., Ricciardi A., Grigorovich I.A. & MacIsaac H.J. (2004). Is invasion success
explained by the enemy release hypothesis? Ecology Letters, 7, 721-733.
Connor E.F., Faeth S.H., Simberloff D. & Opler P.A. (1980). Taxonomic isolation and
the accumulation of herbivorous insects - a comparison of introduced and native
trees. Ecological Entomology, 5, 205-211.
Daehler C.C. (2001). Darwin's naturalization hypothesis revisited. American Naturalist,
158, 324-330.
Darwin C. (1859). On the origin of species by means of natural selection. Murray,
London, UK.
Dawson W., Burslem D. & Hulme P.E. (2009). Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions, 15, 141-147.
Diez J.M., Sullivan J.J., Hulme P.E., Edwards G. & Duncan R.P. (2008). Darwin's
naturalization conundrum: dissecting taxonomic patterns of species invasions.
Ecology Letters, 11, 674-681.
Duncan R.P. & Williams P.A. (2002). Ecology - Darwin's naturalization hypothesis
challenged. Nature, 417, 608-609.
12 Ehrlich P.R. & Raven P.H. (1964). Butterflies and plants: a study in coevolution.
Evolution, 18, 586-608.
Elton C.S. (1958). The ecology of invasions by animals and plants. Methuen, London,
UK.
Fargione J.E. & Tilman D. (2005). Diversity decreases invasion via both sampling and
complementarity effects. Ecology Letters, 8, 604-611.
Gilbert G.S. & Webb C.O. (2007). Phylogenetic signal in plant pathogen-host range.
Proceedings of the National Academy of Sciences of the United States of America,
104, 4979-4983.
Holt R.D. (1977). Predation, apparent competition, and structure of prey communities.
Theoretical Population Biology, 12, 197-229.
Janz N. & Nylin S. (1998). Butterflies and plants: a phylogenetic study. Evolution, 52,
486-502.
Johnson M.T.J. & Stinchcombe J.R. (2007). An emerging synthesis between community
ecology and evolutionary biology. Trends in Ecology & Evolution, 22, 250-257.
Keane R.M. & Crawley M.J. (2002). Exotic plant invasions and the enemy release
hypothesis. Trends in Ecology & Evolution, 17, 164-170.
Keddy P.A. (1992). Assembly and response rules - 2 goals for predictive community
Ecology. Journal of Vegetation Science, 3, 157-164.
King W.B. (1985). Island birds: will the future repeat the past? In: International Council
for Bird Preservation (ed. Moors PJ). Cambridge University Press Cambridge,
UK, pp. 3-15.
13 Lambdon P.W. & Hulme P.E. (2006). How strongly do interactions with closely-related
native species influence plant invasions? Darwin's naturalization hypothesis
assessed on Mediterranean islands. Journal of Biogeography, 33, 1116-1125.
Lockwood J.L., Cassey P. & Blackburn T. (2005). The role of propagule pressure in
explaining species invasions. Trends in Ecology & Evolution, 20, 223-228.
MacDougall A.S. & Turkington R. (2005). Are invasive species the drivers or passengers
of change in degraded ecosystems? Ecology, 86, 42-55.
Mack M.C., D'Antonio C.M. & Ley R.E. (2001). Alteration of ecosystem nitrogen
dynamics by exotic plants: A case study of C-4 grasses in Hawaii. Ecological
Applications, 11, 1323-1335.
Mack R.N. (1996). Predicting the identity and fate of plant invaders: Emergent and
emerging approaches. Biological Conservation, 78, 107-121.
Mack R.N. (2003). Phylogenetic constraint, absent life forms, and preadapted alien
plants: A prescription for biological invasions. International Journal of Plant
Sciences, 164, S185-S196.
Mack R.N., Simberloff D., Lonsdale W.M., Evans H., Clout M. & Bazzaz F.A. (2000).
Biotic invasions: Causes, epidemiology, global consequences, and control.
Ecological Applications, 10, 689-710.
Maron J.L. & Vila M. (2001). When do herbivores affect plant invasion? Evidence for
the natural enemies and biotic resistance hypotheses. Oikos, 95, 361-373.
Mitchell CE & Power AG (2003) Release of invasive plants from fungal and viral
pathogens. Nature 421: 625-627
14 Mitter C., Farrell B. & Futuyma D.J. (1991). Phylogenetic studies of insect plant
interactions - insights into the genesis of diversity. Trends in Ecology &
Evolution, 6, 290-293.
Naeem S., Knops J.M.H., Tilman D., Howe K.M., Kennedy T. & Gale S. (2000). Plant
diversity increases resistance to invasion in the absence of covarying extrinsic
factors. Oikos, 91, 97-108.
Novotny V. & Basset Y. (2005). Host specificity of insect herbivores in tropical forests.
Proceedings of the Royal Society B-Biological Sciences, 272, 1083-1090.
Odegaard F., Diserud O.H. & Ostbye K. (2005). The importance of plant relatedness for
host utilization among phytophagous insects. Ecology Letters, 8, 612-617.
Olden J.D., Poff N.L. & McKinney M.L. (2006). Forecasting faunal and floral
homogenization associated with human population geography in North America.
Biological Conservation, 127, 261-271.
Pimentel D., Lach L., Zuniga R. & Morrison D. (2000). Environmental and economic
costs of nonindigenous species in the United States. Bioscience, 50, 53-65.
Proches S., Wilson J.R.U., Richardson D.M. & Rejmanek M. (2008). Searching for
phylogenetic pattern in biological invasions. Global Ecology and Biogeography,
17, 5-10.
Rejmanek M. (1996). A theory of seed plant invasiveness: The first sketch. Biological
Conservation, 78, 171-181.
Rejmanek M. & Richardson D.M. (1996). What attributes make some plant species more
invasive? Ecology, 77, 1655-1661.
15 Richardson D.M. & Bond W.J. (1991). Determinants of plant distribution - evidence
from pine invasions. American Naturalist, 137, 639-668.
Strauss S.Y., Webb C.O. & Salamin N. (2006). Exotic taxa less related to native species
are more invasive. Proceedings of the National Academy of Sciences of the United
States of America, 103, 5841-5845.
Strong D.R., Lawton J.H. & Southwood R. (1984). Insects on plants. Harvard University
Press, Cambridge, Mass., USA.
Torchin M.E., Lafferty K.D., Dobson A.P., McKenzie V.J. & Kuris A.M. (2003).
Introduced species and their missing parasites. Nature, 421, 628-630.
Vitousek P.M. & Walker L.R. (1989). Biological invasion by Myrica faya in Hawai'i:
plant demography, nitrogen-fixation, ecosystem effects. Ecological Monographs,
59, 247-265.
Webb C.O., Ackerly D.D., McPeek M.A. & Donoghue M.J. (2002). Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475-505.
Weiblen G.D., Webb C.O., Novotny V., Basset Y. & Miller S.E. (2006). Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62-
S75.
Williamson M.H. & Fitter A. (1996). The characters of successful invaders. Biological
Conservation, 78, 163-170.
16 Chapter 2
Evidence that phylogenetically novel non-indigenous plants
experience less herbivory
Steven B. Hill & Peter M. Kotanen
With kind permission from Springer Science+Business Media:
Abstract
The degree to which biotic interactions influence invasion by non-indigenous species may be partly explained by the evolutionary relationship of these invaders with natives.
Darwin's Naturalization Hypothesis controversially proposes that non-native plants are more likely to invade if they lack close relatives in their new range. A possible mechanism for this pattern is that exotics that are more closely related to natives are more likely to share their herbivores, and thus will suffer more damage than phylogenetically isolated species.
We tested this prediction using exotic plants in Ontario, Canada. We measured herbivore damage to 32 species of exotic plants in a common garden experiment, and 52 in natural populations. We estimated their phylogenetic distances from locally-occurring natives in three ways: as mean distance (age) to all native plants, mean distance to native members of the same family, and distance to the closest native species.
In the common garden, the proportion of leaves damaged and the average proportion of leaf area damaged declined with mean phylogenetic distance to native family relatives by
17 late summer. Distance to native confamilials was a better predictor of damage than distance to the closest native species, while mean distance to the entire native plant community failed to predict damage. No significant patterns were detected for plants in natural populations, likely because uncontrolled site-to-site variation concealed these phylogenetic trends.
To the extent that herbivory has negative demographic impacts, these results suggest that exotics that are more phylogenetically isolated from native confamilials should be more invasive; conversely, native communities should be more resistant to invasion if they harbour close familial relatives of potential invaders. However, the large scatter in this relationship suggests that these often are likely to be weak effects; as a result, these effects often may be difficult to detect in uncontrolled surveys of natural populations.
Introduction
Invasions by non-indigenous (exotic) plants now have affected most ecosystems around the world (Mack et al. 2000; Pimentel et al. 2000; Crall et al. 2006). Despite their pervasiveness, predicting the invasion potential of exotic species remains difficult. Some of this difficulty may stem from the complexity of interactions between potential invaders and their natural enemies and competitors in the invaded region. Numerous theories have been proposed to describe these biotic interactions (Mitchell et al. 2006), of which one of the most prominent is the Enemy Release Hypothesis (ERH) (Keane and Crawley 2002;
Torchin and Mitchell 2004). The ERH proposes that exotic species leave behind natural enemies in their native range, resulting in an advantage relative to native competitors
18 (Keane and Crawley 2002; Torchin and Mitchell 2004). Despite its apparent promise, tests of the ERH have found variable results (Colautti et al. 2004; Liu and Stiling 2006): some invaders apparently do benefit from reduced damage (e.g. Mitchell & Power 2003), but many do not (e.g., Agrawal et al. 2003; Parker & Hay 2005; Liu et al. 2007).
One reason that enemy release often fails is that exotic plants rapidly can form new associations with enemies in invaded regions, replacing those lost in the invasion process
(e.g., Parker and Hay 2005; Parker et al. 2006; Hawkes 2007). These interactions may be more likely for exotics with close native relatives in the invaded area, since shifts of enemies should be more likely among closely related hosts (Strong et al. 1984;
Lewinsohn et al. 2005). Numerous studies provide evidence that phytophagous insects are more likely to be shared as the phylogenetic distance between hosts declines (e.g.,
Novotny et al. 2002; Novotny and Basset 2005; Odegaard et al. 2005). For instance, in an extensive survey of host use by herbivorous tropical insects, Weiblen et al. (2006) found that 25% of the variance in herbivore community similarity was explained by the phylogenetic proximity of their hosts. Such trends extend to other taxa of natural enemies as well: by experimentally transferring fungal pathogens among plants, Gilbert & Webb
(2007) showed the likelihood that disease symptoms developed decreased with the phylogenetic separation of the species used as the source and target of the inoculum. If natural enemies can readily spread from natives to related exotics, then non-native plants with close relatives in the native flora might experience reduced enemy release and ultimately diminished invasiveness.
19 The idea that phylogeny might affect invasion success is not new. Darwin (1859) predicted that exotics related to native species would be better invaders, since they would be expected to thrive in environments similar to those that these natives evidently found suitable. He was surprised to find that in fact they were less frequently successful, a pattern which has been termed Darwin's Naturalization Hypothesis (Daehler 2001).
Empirical support for Darwin's Naturalization Hypothesis varies: Daehler (2001) and
Duncan and Williams (2002) rejected the idea, but other studies have had positive results
(e.g., Rejmánek 1996; Diez et al. 2008). One potential problem is that all of these studies used a taxonomic, rather than a phylogenetic approach. In one of the few studies which has used a phylogenetic analysis, Strauss et al. (2006) found that invasive exotic grasses in California are more likely to belong to lineages that have lower native species diversity, though they did not identify the mechanism responsible for this pattern. One possibility is competition: if closely related species compete more strongly, as Darwin
(1859) argued, then exotics might be competitively excluded from areas where native congeners occur. Whether the assumptions underlying this explanation are realistic is unclear, however; for example, Cahill et al. (2008) found only weak evidence that closer relatives compete more strongly. An attractive alternative is provided by shared natural enemies: if enemy release is less effective for exotics with close relatives in the invaded community, as suggested above, then the result could be the patterns observed by Strauss et al. (2006) and Darwin (1859).
Recent studies that have tested whether herbivory is greater for exotics with native relatives have produced both positive (Dawson et al. 2009) and negative (Cappuccino and
20 Carpenter 2005) results; none has attempted an explicitly phylogenetic approach. In our study, we used a common garden experiment and surveys of natural populations to determine whether the herbivore damage experienced by exotic plants is correlated with their phylogenetic relationship to native species. We predicted that the foliar damage experienced by an exotic species (i) decreases with phylogenetic distance to the closest locally-occurring native species, (ii) decreases with phylogenetic distance to native members of the same family, and (iii) decreases with phylogenetic distance to the entire native community. We provide evidence that damage declines with increasing distance to native confamilials, and possibly with distance to the closest native species. These results suggest that herbivory may contribute to the patterns suggested by Darwin's
Naturalization Hypothesis, but at the taxonomic scale of families, not genera. As well, they suggest both a reason for variation in the applicability of the Enemy Release
Hypothesis, and potential avenues for improving the precision and predictiveness of this hypothesis.
Materials and Methods
Study site
This study was conducted at the University of Toronto's Koffler Scientific Reserve
(KSR) at Jokers Hill, Regional Municipality of York, in southern Ontario, Canada (44º02'
N, 79º31' W, 300m ASL). This 350-ha site lies within the Oak Ridges Moraine, and is dominated by prominent hills with a thin organic layer over deep glacial sands.
Vegetation is a mixture of old fields supporting a diverse range of native and exotic
21 plants, hardwood (maple-beech-hemlock) forest, and conifer plantations. Further information on this site may be found at http://www.ksr.utoronto.ca/jh.html/.
Common garden experiment
We conducted a common garden experiment using 32 exotic old-field species (Table 2.1) selected on the basis of their presence in the KSR flora, taxonomic diversity, and availability. Twenty-six of our experimental species were germinated in a greenhouse from seed collected in southern Ontario and grown for 6-8 weeks before transplantation into the field. These were supplemented with six additional species transplanted from local populations; t-tests on damage measurements between transplanted and non- transplanted species confirm that there was no difference between seed-grown and transplanted individuals (p = 0.244 and p=0.205, respectively). In each of the eight 50 x
8m plowed and disked plots located in separate oldfield meadows, we planted six randomly selected plants of each species. Landscape fabric was laid over the plowed plots to reduce the density of competitors. Dominant species adjacent to the plots included Bromus inermis Leyss., Chrysanthemum leucanthemum L., Cirsium arvense (L.)
Scop., Daucus carota L., Poa compressa L., and Solidago canadensis L.
We measured herbivore damage to all plants twice during the 2005 growing season: once shortly after transplantation in late June (early summer), and once in late August (late summer). We made no effort to distinguish damage by particular species; this is appropriate since we wished to integrate net damage to plants by the entire herbivore community. On each date, the proportion of leaves damaged per plant was measured; if
22 the total number of leaves was greater than 50, this proportion was estimated from a haphazard sample of 50 leaves. We also examined photographs of 1 to 5 haphazardly chosen damaged leaves per plant to visually estimate the fraction of leaf area consumed
(to the nearest 5%). Accuracy of this method was assessed by comparing visual estimates with digital estimates of the minimum area removed for a subset of 380 leaves (r2 = 0.84 for comparison of visual and digital assessment). The proportion of leaves damaged and the proportion of area removed per damaged leaf were combined to produce an index of the fraction of leaf area damaged per plant:
proportion of leaf area damaged = proportion of leaves damaged × average
proportion of area removed from damaged leaves.
Natural population surveys
We also conducted a survey of damage in natural populations of 52 exotic species at KSR
(Table 2.1); between 2 and 30 plants from between 2 and 6 populations per species (as available) were sampled during July and August 2005. Since our aim was to be as inclusive as possible, we measured damage to species that are found in forest understory and riparian/wetland habitats, as well as in old fields. Damage was measured using the same approach as in the common gardens, except that the leaf area damaged was measured in the field instead of using photographs in the lab.
Estimating relationship
Reconstructing phylogenetic history using molecular information has been standard for over two decades (Hillis 1987); however, many comparative ecological studies still rely
23 on traditional taxonomic ranks as a surrogate for phylogenetic relationship (e.g., Agrawal et al. 2005; Diez et al. 2008). Here, we used phylogenetic approaches to quantify each exotic species' relationship to locally occurring natives (i.e., those recorded on the KSR checklist: http://www.ksr.utoronto.ca/: site accessed 8 October 2008). We chose this small geographic scale because biotic interactions such as competition and herbivory are expected to be important only if the potentially interacting species locally co-occur (Diez et al. 2008; Proches et al. 2008).
For our phylogenetic analyses, we created a composite phylogeny of all the plants present at KSR using the "Maximally Resolved Seed Plant" tree in the program PHYLOMATIC
(Webb and Donoghue 2005: http://www.phylodiversity.net/phylomatic/). The resulting tree had many polytomies at the family level, but we resolved within-family relationships for our study species using the following published phylogenies: Apiaceae (Downie et al.
2000), Asteraceae (Funk et al. 2005), Boraginaceae (Långström and Chase 2002),
Brassicaceae (Bailey et al. 2006), Caryophyllaceae (Oxelman et al. 1996; Smissen et al.
2002), Lamiaceae (Wagstaff et al. 1998; Paton et al. 2004), Orchidaceae (Freudenstein et al. 2004), Plantaginaceae (Ronsted et al. 2002), Ranunculaceae (Ro et al. 1997),
Rosaceae (Smedmark and Erickson 2002; Potter et al. 2007), Rubiaceae (Natali et al.
1995), Scrophulariaceae (Olmstead et al. 2001). Branch lengths for the resulting tree were adjusted using the "bladj" function in PHYLOCOM (Webb and Donoghue 2005: http://www.phylodiversity.net/phylocom/), which calibrates unknown node ages by linear interpolation of ages from Wikstrom et al. (2001).
24 Three measures of the phylogenetic relatedness of our exotics to the native species at
KSR were then extracted from this tree: 1) mean evolutionary distance (in years of divergence between tips of the tree) of each exotic to all native plants occurring at KSR;
2) the mean distance from each exotic to all locally co-occurring native family members
(in the absence of family members, the distance of each exotic to the next nearest lineage were used); and 3) the distance of each exotic to the closest native species. Distance to all native plants provides a description of each exotic's isolation from the entire native community (e.g., Strauss et al. 2006); distance to native confamilials recognizes that enemies are often shared at the family level (Novotny and Basset 2005; Odegaard et al.
2005; Weiblen et al. 2006). Regressing damage against each of these distance measures provides a test of the hypothesis that damage declines with this measure of phylogenetic isolation, consistent with the patterns predicted from Darwin's Naturalization Hypothesis.
Statistical analyses
Since we primarily were interested in interspecific variation in herbivory, we treated species means as data points. These were calculated by computing plot means (common garden experiment) or population means (field survey) for damage to individuals of each species, then averaging those values. This approach also allowed us to avoid pseudoreplication and to simplify our comparative analyses, which are difficult to apply to spatially replicated datasets. The proportion of leaves damaged and the average leaf area damaged were analyzed separately, since they are algebraically nonindependent.
Damage measurements were probit transformed to linearize proportional measurements; arcsin transformation produced very similar results. All phylogenetic distance
25 measurements were transformed using their natural logarithm to better meet statistical assumptions of normality and homoscedasticity. Statistical analyses were conducted using the programs R (R Development Core Team 2006) and JMP v5.0. (SAS Institute
Inc. 2002). Results are reported as mean ± standard error.
Linear regression was used to test whether damage declined with phylogenetic distance from native relatives, for both the common garden experiment and the field survey. In the case of the common garden data, June and August samples were analyzed independently, since they effectively tested different hypotheses: June samples acted as a control to test for unwanted and unexpected initial differences among newly-transplanted plants, while
August samples provided useful data since they recorded cumulative damage over the summer. Since the plants sampled in the field survey had not been transplanted, an initial sampling was unnecessary. We compared the predictive power of our three competing measures of phylogenetic distance (distance to nearest native species, native confamilials, and native community) by using Akaike's Information Criteria (Burnham and Anderson
2002): for cases where at least one competing model was statistically significant, the combination of high Akaike weights and high strength of fit (r2) identifies the simplest and most powerful model.
To determine whether the inclusion of multiple species from some families biased any significant results, we ran regressions of damage vs. distance using all possible combinations of species, subject to the constraint that only one species was used to represent each family. This resulted in 432 and 84480 models for the common garden
26 experiment and the survey respectively. The value of the regression slope when all species were included compared to the distribution of slopes from models lacking within- family replication allowed us to determine if our results were biased by family-level phylogenetic nonindependence.
Results
At the June sampling of the common garden there were no significant effects of phylogenetic distance on damage (Table 2.2), indicating no unwanted initial differences.
By the August sampling, however, the proportion of leaves damaged (mean ± SE: 0.44 ±
0.05) and the leaf area damaged (mean ± SE: 0.05 ± 0.003) both declined significantly with mean phylogenetic distance from native confamilials, respectively explaining 12% and 18% of variation (Fig 2.1, Table 2.2). As well, there was a marginally significant negative relationship between leaf area damaged and distance to the closest native species
(p = 0.083, r2 = 0.07; Table 2.2). Models including only one species per family were consistent with these results: 95% quantiles for slopes of these reduced models did not include 0, but did include the slopes estimated in our significant regressions; this indicates our significant results were not an artefact of including multiple species per family. In contrast, there was no statistical relationship between either damage measurement and mean distance to all locally-occurring natives. Mean distance to native family members was a much stronger predictor of damage than the other distance measures, as indicated by lower AIC and higher Akiake weights (w) and r2 values (Table
2.2).
27 In contrast with the common garden results, damage to exotics in natural populations was not significantly related to any of the phylogenetic distance measures (Table 2.2; Fig.
2.2). In all cases, adjusted r2 values were very low or slightly negative (Table 2.2), indicating that essentially none of the variation in damage could be explained by phylogenetic distance to native relatives.
Discussion
Does damage decline with phylogenetic distance between natives and exotics?
We predicted that herbivore damage to exotics would decline with phylogenetic distance from native relatives. Few tests of related hypotheses have been performed. Working in
Tanzania, Dawson et al. (2009) found herbivory in introduced plants increased with both the presence of native congeners and the diversity of native confamilials, though this was apparently unrelated to invasiveness. In contrast, also working in Ontario, Cappuccino and Carpenter (2005) found that herbivore damage to exotic plants was unrelated to the number of native congeners or confamilial native genera. Neither of these studies employed a phylogenetic approach, but using methods based on phylogenetic distance rather than the number of congeneric or confamilial native species, we found that both the proportion of leaves damaged and the average proportion of leaf area damaged declined with phylogenetic distance from confamilials (Fig. 2.1). As well, the leaf area damaged showed a nearly significant decline with phylogenetic distance from the nearest native species (Fig. 2.1). Remarkably, we were able to detect these patterns despite both phylogenetic uncertainty and the high level of noise resulting from interspecific variation in damage; nonetheless, the strength of these relationships was low (r2 < 0.2), suggesting
28 the effects of native relatives often may be small relative to other factors affecting both herbivore damage and invasiveness.
In contrast with our common garden results, we found no significant phylogenetic trends in damage to wild populations (Fig. 2.2). The most plausible explanation is that uncontrolled variation in damage among natural populations was greater than any phylogenetic trend, and therefore masked the effects of phylogenetic distance. This does not necessarily mean phylogenetic influences are unimportant in natural populations; only that they operate in addition to other sources of background variation. Herbivore communities are highly spatially variable and often differ among habitats (e.g., DeWalt et al. 2004). We sampled scores of natural populations in a wide range of habitats; site-to- site variation in background levels of herbivore damage thus likely varied considerably.
In contrast, experimental plants in our common gardens all shared the same environment, minimizing unwanted site effects. This reduction of environmental noise is one of the principal advantages of common gardens, since it can allow the detection of subtle trends that may be difficult to demonstrate in field surveys. Interestingly, the only other study we are aware of that shows an effect of phylogenetic isolation on herbivory (Dawson et al. 2009) was conducted in a botanical garden: effectively a common-garden approach.
Generally speaking, closer relatives share more of the herbivore community (Novotny et al. 2002; Novotny and Basset 2005; Odegaard et al. 2005; Weiblen et al. 2006); therefore, we expected that distance to the nearest relative (often a congeneric) would be the strongest predictor of damage. The stronger predictive value of family relatedness
29 than distance to the nearest native relative was unexpected. There are several possible explanations for our result. First, many characteristics of plants that are important to herbivores are shared at the family level (chemistry, morphology, etc.); perhaps the single closest relative is less important than the suite of confamilials sharing such characteristics. There is some evidence that family-level characters can be more important in determining herbivore communities than variation among species within a family; for example, Cappuccino and Carpenter (2005) provide evidence that, although invasive exotic plants tend to be chemically more unusual than non-invasive exotic species, phytochemical uniqueness may only be weakly correlated with phylogeny below the family level. Second, deeper phylogenetic nodes where ecological or functional similarities are conserved may represent greater species diversity than a single nearest neighbour; thus, relatives at the family level may harbour a greater diversity of insects potentially able to colonize a new invader than does a single close relative. While
Cappuccino and Carpenter (2005) found damage to exotics to be unrelated to the diversity of native relatives, Dawson et al. (2009) did find that herbivore damage increased with the diversity of native confamilials, consistent with this prediction.
Finally, the stronger effect of family relationship may occur because estimates of topology and branch lengths below the family level are less accurate than that above the family level. This is likely, both since finer relationships are often more difficult to resolve, and because we were forced to use an interpolation procedure to estimate node ages.
30 Taxonomy often represents poly- or paraphyletic groupings of taxa, and also fails to account for taxon age, while a reliable phylogeny minimizes these problems; thus, phylogeny should be more powerful in detecting factors correlated with relationship.
Nonetheless, it is important to realize that phylogenies are hypotheses about evolutionary history, and should be treated accordingly when incorporated into comparative analyses
(Webb et al. 2002). As well, fully resolved phylogenies are still not available for most plant communities, forcing some degree of reliance on taxonomic information: in our case, we still used taxonomy to inform our choice of node ages to include as predictors.
Enemy Release and Darwin's Naturalization Hypothesis.
The Enemy Release Hypothesis states that escape from enemies may contribute to the success of exotic species (Keane and Crawley 2002; Torchin and Mitchell 2004; Liu and
Stiling 2006). Tests of this hypothesis have produced variable results: exotic plants often do experience reduced levels of damage in new areas but there are numerous exceptions to this rule (Liu & Stiling 2006). Our results suggest one source of this variation: escape from enemies may depend on the phylogenetic relationship between an invader and members of the invaded community. An exotic plant closely related to natives in the invaded community thus may experience more damage than a phylogenetic outlier. The effects we found were modest, and even substantial damage does not necessarily translate to reduced invasiveness or demographic performance (Maron and Vilá 2001; Liu et al.
2007; Dawson et al. 2009); nonetheless, in at least some cases, this biotic resistance might reduce the abundance or spread of an exotic.
31 If natural enemies are more likely to attack exotics with close native relatives, this could provide a mechanism for Darwin's Naturalization Hypothesis. Although we did not have the data to explicitly test Darwin's Naturalization Hypothesis, evidence in the literature is mixed (Rejmanek 1996; Daehler 2001; Duncan and Williams 2002; Diez et al. 2008); however, this variation may partly reflect the influence of opposing ecological forces operating at different spatial scales (Proches et al. 2008). At regional scales, exotics and close native relatives may positively co-occur because of a requirement for similar physical environments, as Darwin predicted; at local scales, negative biotic interactions such as competition and herbivory may produce negative co-occurrence, as Darwin actually reported (Proches et al. 2008). Diez et al. (2008) found exactly this pattern: at a regional scale, the abundance of exotic plants near Auckland, New Zealand was positively correlated with the abundance of native congeners, while at a local (within- habitat) scale this pattern was reversed. Thus, community-scale studies (such as ours) may be required to elucidate the contribution of biotic factors to the patterns discussed by
Darwin.
Links with community phylogenetics
Darwin's Naturalization Hypothesis represents a precursor to community phylogenetics: the recent theory that communities are in part structured by patterns of evolutionarily conserved traits among component taxa (Webb et al. 2002). Many such studies assume that competition and environmental filtering are the dominant forces structuring communities (e.g., Webb 2000; Cavender-Bares et al. 2004; Cavender-Bares et al. 2006;
Webb et al. 2006; Swenson et al. 2007). For example, where close relatives do not co-
32 occur (i.e., a community is "phylogenetically overdispersed"), the mechanism is assumed to involve limiting similarity resulting in competitive exclusion. However, experimental evidence suggests that strong competition among close relatives may not in fact be a leading determinant of community structure (Cahill et al. 2008). Instead, a negative association among relatives could be explained by apparent competition (Holt 1977; Holt and Lawton 1994): the suppression of one species by the enemies of another. Where natural enemies are shared among close relatives, this also could result in phylogenetic overdispersion; our experimental results suggest this may be a plausible alternative. Thus, a century and a half after it was originally proposed, Darwin's Naturalization Hypothesis may lead to a better understanding of the role that phylogeny plays in structuring natural communities, and the mechanisms involved in the success or failure of exotic species.
Acknowledgements
This research was supported by NSERC Research and Equipment Grants (PMK), an
NSERC PGS-D (SBH), and the Koffler Scientific Reserve at Jokers Hill. Thanks to
Kateryna Kostyukova for her continuous help, to Nathalie Taraban-Lagois, Gilbert Tang,
James McKay, and Andrew MacDonald for their support, discussions, and field assistance, and to two anonymous reviewers for valuable suggestions on an earlier version of this manuscript. This is a publication of the Koffler Scientific Reserve. All of the experiments conducted in this study comply with the current laws of Canada.
References
Agrawal AA & Kotanen PM (2003) Herbivores and the success of exotic plants: a
33 phylogenetically controlled experiment. Ecol Lett 6: 712-715
Agrawal AA, Kotanen PM, Mitchell CE, Power AG, Godsoe W & Klironomos J (2005)
Enemy release? An experiment with congeneric plant pairs and diverse above- and
belowground enemies. Ecology 86: 2979-2989
Bailey CD, Koch MA, Mayer M, Mummenhoff K, O'Kane Jr SL, Warwick SI, Windham
MD & Al-Shehbaz IA (2006) Toward a global phylogeny of the Brassicaceae. Mol
Biol Evol 23: 2142-2160
Burnham KD & Anderson DA (2002) Model Selection and Multi-Model Inference, 2
edn. Springer, New York
Cahill JF, Kembel SW, Lamb EG & Keddy PA (2008) Does phylogenetic relatedness
influence the strength of competition among vascular plants? Plant Ecol Evol Syst
10: 41-50
Cappuccino N & Carpenter D (2005) Invasive exotic plants suffer less herbivory than
non-invasive exotic plants. Biol Lett 1: 435-438
Cavender-Bares J, Ackerly DD, Baum DA & Bazzaz FA (2004) Phylogenetic
overdispersion in Floridian oak communities. Am Nat 163: 823-843
Cavender-Bares J, Keen A & Miles B (2006) Phylogenetic structure of Floridian plant
communities depends on taxonomic and spatial scale. Ecology 87: S109-S122
Colautti RI, Ricciardi A, Grigorovich IA & MacIsaac HJ (2004) Is invasion success
explained by the enemy release hypothesis? Ecol Lett 7: 721-733
Crall AW, Meyerson LA, Stohlgren TJ, Jarnevich CS, Newman GJ & Graham J (2006)
Show me the numbers: what data currently exist for non-native species in the USA?
Front Ecol Environ 4: 414-418
34 Daehler CC (2001) Darwin's naturalization hypothesis revisited. Am Nat 158: 324-330
Darwin C (1859) On the origin of species by means of natural selection. Murray, London,
UK
Dawson W, Burslem DFRP & Hulme PE (2009) Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions 15: 141-147.
Dewalt SJ, Denslow JS & Ickes K (2004) Natural-enemy release facilitates habitat
expansion of the invasive tropical shrub Clidemia hirta. Ecology 85: 471-483
Diez JM, Sullivan JJ, Hulme PE, Edwards G & Duncan RP (2008) Darwin's
naturalization conundrum: dissecting taxonomic patterns of species invasions. Ecol
Lett 11: 674-681
Downie SR, Katz-Downie DS & Watson MF (2000) A phylogeny of the flowering plant
family Apiaceae based on chloroplast DNA rpl16 and rpoCl intron sequences:
towards a suprageneric classification of subfamily Apioideae. Am J Bot 87: 273-
292
Duncan RP & Williams PA (2002) Ecology - Darwin's naturalization hypothesis
challenged. Nature 417: 608-609
Freudenstein JV, van den Berg C, Goldman DH, Kores PJ, Molvray M & Chase MW
(2004) An expanded plastid DNA phylogeny of Orchidaceae and analysis of
jackknife branch support strategy. Am J Bot 91: 149-157
Funk VA, Bayer RJ, Kelley S, Chan R, Watson L, Gemeinholzer B, Schilling E, Panero
JL, Baldwin BG, Garcia-Jacas N, Susanna A & Jansen RK (2005) Everywhere but
Antarctica: using a supertree to understand the diversity and distribution of the
35 Compositae. Biol Skr 55: 343-374
Gilbert GS & Webb CO (2007) Phylogenetic signal in plant pathogen-host range. P Natl
Acad Sci USA 104: 4979-4983
Hawkes CV (2007) Are invaders moving targets? The generality and persistence of
advantages in size, reproduction, and enemy release in invasive plant species with
time since introduction. Am Nat 170: 832-843
Hillis DM (1987) Molecular versus morphological approaches to systematics. Annu Rev
Ecol Syst 18: 23-42
Holt RD (1977) Predation, apparent competition, and structure of prey communities.
Theor Popul Biol 12: 197-229
Holt RD & Lawton JH (1994) The ecological consequences of shared natural enemies.
Annu Rev Ecol Syst, 25: 495-520
Keane RM & Crawley MJ (2002) Exotic plant invasions and the enemy release
hypothesis. Trends Ecol Evol 17: 164-170
Långström E & Chase MW (2002) Tribes of Boraginoideae (Boraginaceae) and
placement of Antiphytum, Echiochilon, Ogastemma, and Sericostoma: a
phylogenetic analysis based on atpB plastid DNA sequence data. Plant Syst Evol
234: 137-153
Lewinsohn TM, Novotny V & Basset Y (2005) Insects on plants: diversity of herbivore
assemblages revisited. Annu Rev Ecol Syst 36: 597-620
Liu H & Stiling P (2006) Testing the enemy release hypothesis: a review and meta-
analysis. Biol Inv 8: 1535-1545
Liu H, Stiling P & Pemberton RW (2007) Does enemy release matter for invasive plants?
36 evidence from a comparison of insect herbivore damage among invasive, non-
invasive and native congeners. Biol Inv 9: 773-781
Mack RN, Simberloff D, Lonsdale WM, Evans H, Clout M & Bazzaz FA (2000) Biotic
invasions: Causes, epidemiology, global consequences, and control. Ecol Appl 10:
689-710
Maron JL & Vila M (2001) When do herbivores affect plant invasion? Evidence for the
natural enemies and biotic resistance hypotheses. Oikos 95: 361-373
Mitchell CE, Agrawal AA, Bever JD, Gilbert GS, Hufbauer RA, Klironomos JN, Maron
JL, Morris WF, Parker IM, Power AG, Seabloom EW, Torchin ME & Vazquez DP
(2006) Biotic interactions and plant invasions. Ecol Lett 9: 726-740
Mitchell CE & Power AG (2003) Release of invasive plants from fungal and viral
pathogens. Nature 421: 625-627
Morton JK & Venn JM (1990) A checklist of the flora of Ontario: vascular plants.
University of Waterloo Biology Series Waterloo, Ontario, Canada
Natali A, Manen J-F & Ehrendorfer F (1995) Phylogeny of the Rubiaceae-Rubioideae, in
particular the tribe Rubieae: evidence from a non-coding chloroplast DNA
sequence. Ann Missouri Bot Garden 82: 428-439
Novotny V & Basset Y (2005) Host specificity of insect herbivores in tropical forests. P
Roy Soc B-Biol Sci 272: 1083-1090
Novotny V, Basset Y, Miller SE, Weiblen GD, Bremer B, Cizek L & Drozd P (2002)
Low host specificity of herbivorous insects in a tropical forest. Nature 416: 841-844
Odegaard F, Diserud OH & Ostbye K (2005) The importance of plant relatedness for host
utilization among phytophagous insects. Ecol Lett 8: 612-617
37 Olmstead RA, dePamphilis CW, Wolfe AD, Young ND, Ellisons WJ & Reeves PA
(2001) Disintegration of the Scrophulariaceae. Am J Bot 88: 348-361
Oxelman B, Lidén M & Berglund D (1996) Chloroplast rps16 intron phylogeny of the
tribe Sileneae (Caryophyllaceae) . Plant Syst Evol 206: 393-410
Parker JD, Burkepile DE & Hay ME (2006) Opposing effects of native and exotic
herbivores on plant invasions. Science 311: 1459-1461
Parker JD & Hay ME (2005) Biotic resistance to plant invasions? Native herbivores
prefer non-native plants. Ecol Lett 8: 959-967
Paton AJ, Springate D, Suddee S, Otieno D, Grayer RJ, Harley MM, Willis F, Simmonds
MSJ, Powell MP & Savolainen V (2004) Phylogeny and evolution of basils and
allies (Ocimeae, Labiatae) based on three plastid DNA regions. Mol Phylogenet
Evol 31: 277-299
Pimentel D, Lach L, Zuniga R & Morrison D (2000) Environmental and economic costs
of nonindigenous species in the United States. Bioscience 50: 53-65
Potter D, Erickson T, Evans RC, Oh S, Smedmark JEE, Morgan DR, Kerr M, Robertson
KR, Arsenault M, Dickinson TA & Campbell CS (2007) Phylogeny and
classification of Rosaceae. Plant Syst Evol 266: 5-43
Proches S, Wilson JRU, Richardson DM & Rejmanek M (2008) Searching for
phylogenetic pattern in biological invasions. Global Ecol Biogeogr 17: 5-10
Rejmanek M (1996) A theory of seed plant invasiveness: the first sketch. Biol Conserv,
78: 171-181
R Development Core Team (2006) R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria ISBN 3-
38 900051-07-0, URL http://wwwR-projectorg
Ro K-E, Keener CS & McPheron BA (1997) Molecular phylogenetic study of the
Ranunculaceae: utility of the nuclear 26S ribosomal DNA in inferring intrafamilial
relationships. Mol Phylogenet Evol 8: 117-127
Ronsted N, Chase MW, Albach DC & Bello MA (2002) Phylogenetic relationships
within Plantago (Plantaginaceae): evidence from nuclear ribosomal ITS and plastid
trnL-F sequence data. Bot J Linn Soc 139: 323-338
SAS Institute Inc (2002) JMP version 5.0 In SAS Institute. Cary, North Carolina, USA
Smedmark JEE & Erickson T (2002) Phylogenetic relationships of Geum (Rosaceae) and
relatives inferred from the nrITS and trnL-trnF regions. Syst Bot 27: 303-317
Smissen RD, Clement JC, Garnock-Jones PJ & Chambers GK (2002) Subfamilial
relationships within Caryophyllaceae as inferred from 5' ndhF sequences. Am J Bot
89: 1336-1341
Strauss SY, Webb CO, Salamin, N (2006) Exotic taxa less related to native species are
more invasive. PNAS 15: 5841-5845
Strong DR, Lawton JH & Southwood R (1984) Insects on plants. Harvard University
Press, Cambridge, Mass, USA
Swenson NG, Enquist BJ, Thompson J & Zimmerman JK (2007) The influence of spatial
and size scale on phylogenetic relatedness in tropical forest communities. Ecology
88: 1770-1780
Torchin ME & Mitchell CE (2004) Parasites, pathogens, and invasions by plants and
animals. Front Ecol Environ 2: 183-190
Wagstaff SJ, Hickerson L, Spangler R, Reeves PA & Olmsted RA (1998) Phylogeny in
39 Labiatae s l inferred from cpDNA sequences. Plant Syst Evol 209: 265-274
Webb CO (2000) Exploring the phylogenetic structure of ecological communities: an
example for rain forest trees. Am Nat 156: 145-155
Webb CO, Ackerly DD, McPeek MA & Donoghue MJ (2002) Phylogenies and
community ecology. Annu Rev Ecol Syst 33: 475-505
Webb CO & Donoghue MJ (2005) Phylomatic: tree assembly for applied phylogenetics.
Mol Ecol Notes 5: 181-183
Webb CO, Gilbert GS & Donoghue MJ (2006) Phylodiversity-dependent seedling
mortality, size structure, and disease in a Bornean rain forest. Ecology 87: S123-
S131
Weiblen GD, Webb CO, Novotny V, Basset Y & Miller SE (2006) Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology 87: S62-
S75
Wikstrom N, Savolainen V & Chase MW (2001) Evolution of the angiosperms:
calibrating the family tree. P Roy Soc Lond B Bio 268: 2211-2220
40 Table 2.1 List of exotic plants sampled in common gardens and natural populations.
Bolded families have native members at the study site (the Koffler Scientific Reserve).
Nomenclature follows Morton and Venn (1990)
Family Species Common Garden Natural Populations Amaranthaceae Amaranthus retroflexus L. X Apiaceae Daucus carota L. X X Asclepiadaceae Cynanchum rossicum (Kleopov) Borh. X Asteraceae Arctium minus (Hill) Bernh. X X Asteraceae Centaurea jaceae L. X Asteraceae Cirsium arvense (L.) Scop. X X Asteraceae Cirsium vulgare (Savi) Tenore X Asteraceae Cichorium intybus L. X X Asteraceae Chrysanthemum leucanthemum L. X X Asteraceae Hieracium aurantiacum L. X Asteraceae Hieracium caespitosum Dumort. X Asteraceae Inula helenium L. X Asteraceae Sonchus arvensis L. X Asteraceae Tragopogon dubius Scop. X Asteraceae Tragopogon pratensis L. X Asteraceae Tussilago farfara L. X Boraginaceae Cynoglossum officinalis L. X Boraginaceae Echium vulgaris L. X X Brassicaceae Alliaria petiolata (M. Bieb.) Cav. & Gr. X Brassicaceae Capsella bursa-pastoris (L.) Medikus X Brassicaceae Erysimum cheiranthoides L. X Brassicaceae Hesperis matronalis L. X Brassicaceae Lepidium campestre (L.) R.Br. X Brassicaceae Sinapis arvensis L. X Brassicaceae Thlaspi arvense L. X Campanulaceae Campanula rapunculoides L. X Caryophyllaceae Dianthus armeria L. X Caryophyllaceae Saponaria officinalis L. X Caryophyllaceae Silene latifolia Poiret X X Caryophyllaceae Silene noctiflora L. X Caryophyllaceae Silene vulgaris (Moench) Garcke X X Chenopodiaceae Chenopodium album L. X X Clusiaceae Hypericum perforatum L. X X Convolvulaceae Convolvulus arvensis L. X 41 Dipsacaceae Dipsacus sylvestris Hudson X Euphorbiaceae Euphorbia cyparissias L. X X Fabaceae Lotus corniculatus L. X X Fabaceae Medicago lupulina L. X Fabaceae Medicago sativa L. X Fabaceae Melilotus alba Medikus X X Fabaceae Melilotus officinalis (L.) Pallas X Fabaceae Trifolium pratense L. X X Fabaceae Vicia cracca L. X Fabaceae Vicia tetrasperma (L.) Moench X Lamiaceae Glechoma hederaceae L. X Lamiaceae Leonurus cardiaca L. X Lamiaceae Mentha X piperita L. X Lamiaceae Nepeta cataria L. X X Lythraceae Lythrum salicaria L. X Orchidaceae Epipactis helleborine (L.) Crantz X Papaveraceae Chelidonium majus L. X X Plantaginaceae Plantago lanceolata L. X Plantaginaceae Plantago major L. X X Polygonaceae Rumex crispus L. X X Ranunculaceae Ranunculus acris L. X X Rubiaceae Galium verum L. X X Rosaceae Potentilla recta L. X X ScrophulariaceaeLinaria vulgaris Hill X X ScrophulariaceaeVerbascum thapsus L. X X Solanaceae Solanum dulcamara L. X X Total species 32 52
42 Table 2.2 Regression and AIC results for herbivory (proportion of leaves damaged and
leaf area damaged) vs. phylogenetic distance to native species occurring at KSR. Damage
variables were probit transformed and phylogenetic distance measures were natural
logarithm transformed before analysis. Headers represent degrees of freedom (df) used to
test the significance of the regression slope (F, p), and Akaike's information criteria
(AIC), Akaike weights (w), and variance explained (r2) to determine the best
phylogenetic distance model
Proportion of Leaves Damaged Leaf Area Damaged Phylogenetic Data df F p AIC w r2 F p AIC w r2 distance All native 1,30 0.030 0.865 74.700 0.30 0.00 0.041 0.840 48.706 0.26 0.00 Common taxa Garden Native family 1,30 0.544 0.467 74.156 0.40 0.00 1.302 0.263 47.391 0.49 0.01 (June) taxa Closest native 1,30 0.000 1.000 74.731 0.30 0.00 0.004 0.953 48.746 0.25 0.00
All native 1,30 0.138 0.713 99.203 0.06 0.00 0.955 0.336 36.121 0.03 0.00 Common taxa Garden Native family 1,30 5.299 0.028 94.145 0.78 0.12 7.957 0.008 29.597 0.86 0.18 (August) taxa Closest native 1,30 1.939 0.174 97.346 0.16 0.03 3.214 0.083 33.867 0.10 0.07
All native 1,50 2.232 0.142 22.715 0.59 0.00 0.003 0.956 81.079 0.31 0.00 taxa Survey Native family 1,50 0.003 0.958 24.983 0.19 0.00 0.000 0.997 81.082 0.31 0.00 taxa Closest native 1,50 0.341 0.562 24.633 0.22 0.00 0.440 0.510 80.626 0.39 0.00
43 Figure 2.1 Regressions of herbivory (proportion of leaves damaged and proportion of leaf area damaged) against phylogenetic distance to all native taxa, native family members, and closest native species at KSR during August 2005. Points represent means for each species included in the common garden experiment. Intact lines indicate significance at p < 0.05, broken line indicates significance at p < 0.1
44 Figure 2.2 Regressions of herbivory (proportion of leaves damaged and proportion of leaf area damaged) against phylogenetic distance to all native taxa, native family members, and closest native species at KSR from surveys of naturally occurring populations between July and August 2005. Points represent means for each species sampled in the field
45 Chapter 3
Release from foliar enemies for exotic species in the Asteraceae:
a phylogenetically controlled study
Abstract
The shared evolutionary history of exotic and native species may provide a framework for predicting whether biotic interactions between them are likely to occur. In this study, I used exotic and native plants in the Asteraceae to evaluate if i) exotics experience less herbivore damage than natives when phylogenetic relationships are controlled, and ii) whether phylogenetically novel exotics experience lower damage that those that are more closely related to other locally-occurring members of the same family. My results show that exotics experience lower damage than native species both with and without using phylogenetic correction methods. Phylogenetically structured variation in herbivory tended to be associated with tribes; therefore, phylogenetically corrected patterns in damage suggest that exotics on average experience lower damage than natives in the same tribe. Contrary to my prediction that phylogenetically novel exotics would experience lower damage, exotics that had higher distances to closest native family members tended to experience slightly higher damage. These results suggest that phylogenetically novel exotics may be more naive to native enemies, though the pattern was relatively weak. Future tests of the role of enemies and other biotic interactions that influence invasion success may benefit from including the phylogenetic relationship of exotic and co-occurring native species.
46 Introduction
Predicting whether direct or indirect biotic interactions will occur among exotic and native species may be important for better understanding and managing biological invasions (Mitchell et al. 2006). How an invader interacts with competitors (Naeem et al.
2000; Fargione & Tilman 2005), their susceptibility to novel enemies (consumers, herbivores, pathogens, and parasites) (Agrawal & Kotanen 2003; Agrawal et al. 2005;
Parker & Hay 2005; Parker & Gilbert 2007), and their facilitation by mutualists
(Richardson et al. 2000) can have important consequences for invasion success (Levine et al. 2004). In the absence of prior information about how invaders interact with these species, shared evolutionary history may provide a framework for evaluating the direction and strength of these interactions (Connor et al. 1980; Cappuccino & Carpenter
2005; Dawson et al. 2009). Evolutionary relatedness among co-occurring species has been used to infer whether processes such as competition or environmental filtering are likely to structure communities (Webb et al. 2002; Cavender-Bares et al. 2004; Kembel
& Hubbell 2006; Cavender-Bares et al. 2009; Vamosi et al. 2009). Similar approaches have been used to evaluate the degree of phylogenetic similarity of exotic species relative to native species (Strauss et al. 2006; Proches et al. 2008; Cadotte et al. 2009). One of the emerging patterns is that exotics that are phylogenetically novel relative to native communities may be more successful as invaders (Cappuccino & Carpenter 2005; Strauss et al. 2006).
The idea that phylogeny affects invasion success is not new. Observations made by de
Candolle (1855) and Darwin (1859) resulted in the proposition that invaders will tend to
47 be less closely related to native species; subsequently this idea has been named Darwin's
Naturalization Hypothesis (DNH) (Daehler 2001). Darwin (1859) suggested that this pattern would result from close relatives having similar niches, and thus competing more intensely. Tests of this hypothesis vary among studies: Daehler (2001) and Duncan and
Williams (2002) found evidence against the idea, but others have found support
(Rejmanek & Richardson 1996; Diez et al. 2008). These contradictions may echo the scale dependency of the processes responsible for the patterns expected by DNH (Proches et al. 2008). At local scales, competition can result in limited co-occurrence of close relatives; whereas at larger scales, exotics may be more successful if they are closely related to natives, as they are more likely to share traits that facilitate survival given the environmental conditions. A recent test of the competition-evolutionary relatedness prediction using plants, however, suggests that a phylogenetic signal may exist for some groups, but the trends tend to be weak (Cahill et al. 2008). Even though competition was proposed to be the underlying mechanism of DNH, other interactions such as herbivory resulting from shared host-use may result in similar patterns of community assembly and invasion success (Mack 1996; Cavender-Bares et al. 2009).
Escape from enemies may play an important role in invasions. The Enemy Release
Hypothesis predicts that exotic species leave behind specialist enemies in their native range, giving them an advantage over co-occurring natives competitors (Keane &
Crawley 2002; Torchin et al. 2003). Despite its prominence, empirical tests of the ERH show varied results (Colautti et al. 2004; Liu & Stiling 2006): many exotics do experience reduced damage in their invaded range, but compared to co-occurring natives,
48 many do not (Agrawal & Kotanen 2003; Colautti et al. 2004; Liu & Stiling 2006). In part, this can result from exotics rapidly accumulating new enemies in their invaded range (Hawkes 2007; Parker & Gilbert 2007). These interactions may be more likely for exotics with close native relatives in the invaded area, since shifts of enemies should be more likely among closely related hosts (Strong et al. 1984; Lewinsohn et al. 2005;
Gossner et al. 2009). Numerous studies provide evidence that phytophagous insects are more likely to be shared as the phylogenetic distance between hosts declines (e.g.,
Novotny et al. 2002; Novotny and Basset 2005; Odegaard et al. 2005). There is also evidence that exotic species that are less closely related to native species experience less damage by natural enemies (Connor et al. 1980; Dawson et al. 2009; Hill & Kotanen
2009). Together, this evidence suggests that phylogenetically novel exotics may be more likely to escape enemies (however, see Cappuccino & Carpenter 2005). Thus, evaluating the ERH in a phylogenetic context may help reconcile the conflicting tests of the hypothesis, by accounting for differences in damage experienced by exotic species of different relatedness to native species.
Here I evaluate whether herbivory on exotic plants is influenced by their phylogenetic relationship to co-occurring species among plants in the Asteraceae. I hypothesized that that i) exotic species experience lower herbivory than natives, ii) that this is true when phylogenetic relationships are accounted for, and iii) that phylogenetically novel exotics experience lower herbivory than those that are less isolated, consistent with DNH. I tested this idea by measuring herbivore damage to native and exotic members of the
Asteraceae, and mapping these data onto a phylogeny. My results suggest that exotic
49 Asteraceae experience lower damage than native species even when phylogenetic relationships are controlled; however, for the exotic species, damage did not decline with phylogenetic novelty relative to natives.
Methods
Study site and natural history
This study was conducted at the University of Toronto's Koffler Scientific Reserve
(KSR) at Jokers Hill, Regional Municipality of York, in southern Ontario, Canada (44º02'
N, 79º31' W, 300m ASL). This 350-ha site lies within the Oak Ridges Moraine, and is dominated by prominent hills with a thin organic layer over deep glacial sands.
Vegetation is a mixture of old fields supporting a diverse range of native and exotic plants, hardwood (maple-beech-hemlock) forest, and conifer plantations. Further information on this site may be found at http://www.ksr.utoronto.ca.
I chose to work with Asteraceae as they are well represented at KSR: there are 77 species
(~12% of the local flora), of which 44 are native, the remaining 33 exotic. Species are present in all major habitats found on site, including old-field meadows, forest understories, wet meadows, and ponds. Their phylogeny is well understood (Funk et al.
2005), with exotics represented among many of the different lineages locally present.
50 Herbivory survey
During May to November 2008, damage was measured on 21 native and 14 exotic species (Table 3.1); species were chosen to represent as many lineages within Asteraceae as possible, and based on their abundance in the field. For each species, five to ten individuals were randomly sampled from one to three locations. Native and exotic plants were sampled in all respective habitat types (e.g. oldfield, forest understory, and wetland) to attempt to avoid biases associated with habitat-related damage on native and exotic plants (e.g. Parker & Gilbert 2007). For individual plants, the total number of leaves, the number of damaged leaves, and the percent area of damage to each leaf were recorded.
Percent leaf area damage was estimated visually to the nearest 5%. The mean proportion of leaf area damaged and the proportion of leaves damaged were multiplied to give an index of plant level damage; the product represents an estimate of the total proportion of foliar area lost to herbivory. For all analyses, damage for individual species was pooled based on means for each population as I was interested in among-species variation and its relationship to phylogeny.
Community phylogeny
The backbone of the community phylogeny for this family was drawn from Funk et al.
(2005). Where polytomies existed, I supplemented the topology based on the following studies: Erigeron (Noyes 2000), Hieracium (Gaskin & Wilson 2007), Lactuceae
(Whitton et al. 1995), Solidago (Semple & Cook 2006). Node ages for the resulting tree were then estimated using the function "bladj" in PHYLOCOM (Webb et al. 2008), with node age calibration points from (Kim et al. 2005). 51 Statistical Analyses
To evaluate differences in damage between native and exotic species I used one-tailed t- tests, as I expected a priori that exotics would experience less damage. To evaluate the phylogenetic component of herbivory, I used an eigenvector decomposition method
(Diniz-Filho & de Sant'Ana 1998; Desdevises et al. 2003). I extracted the 34 eigenvectors from a principal co-ordinate analysis of the phylogenetic distance matrix from the
Asteraceae community phylogeny. These eigenvectors extract unique (orthogonal) predictors of phylogenetic structure in a trait (in this case herbivory) based on distances between species. The eigenvectors predict increasingly finer-scale patterns of phylogenetically structured variation, such that the first eigenvector detects the broadest pattern, and the last eigenvector detects the finest-scale pattern (Diniz-Filho & de
Sant'Ana 1998). In general, traits that have broad phylogenetic structure will be associated with deeper splits within the phylogeny, whereas fine phylogenetic structures will be associated with more recent splits (closer to the tips of the phylogeny). For my analysis, each eigenvector was then correlated with the proportion of leaf area damage; those that were significant at p ≤ 0.10 were retained as predictors of phylogenetically structured patterns of foliar damage. I then calculated the residuals from a multiple regression of foliar damage versus the significant phylogenetic eigenvectors; these residuals were used to represent non-phylogenetically structured foliar damage. One- tailed t-tests using these residuals were then used to test the hypothesis that exotic species experienced lower herbivory than natives after correcting for their phylogenetic relationship. I also used phylogenetic independent contrasts (PICs) and a one-tailed paired t-test to test the hypothesis that exotic species experienced lower damage than
52 native species (Felsenstein 1985); PICs were calculated using the function "aot" in the statistical package PHYLOCOM (Webb et al. 2008), t-tests were calculated in Excel.
To determine if phylogenetically novel exotic species experienced lower herbivory, I calculated four descriptors of novelty relative to the rest of the Asteraceae family present at KSR: i) mean distance to all taxa (native and exotic), ii) mean distance to closest relative, iii) mean distance to all native taxa, and iv) mean distance to closest native taxa.
Distances in years among individual taxa were based on the tip-to-tip measurements given the branch lengths in the Asteraceae community phylogeny, and were calculated from the phylogenetic distance matrix using the package "ape" in R (R development core team 2009). The predictors, although superficially similar, test different hypotheses. The first two evaluate the degree to which interactions with both native and exotic relatives explain damage; this may be the case if interactions occur with enemies of other related exotic species. The third tests whether interactions with herbivores depends on the mean distance to only native species within the family as a whole. The fourth tests if it is the distance to the closest native relative that matters; this is expected to be true if herbivore interactions are most likely to be shared among closest relatives. I used each predictor in a separate linear regression with proportion leaf area damage to determine if any were significant, and if so which explained the most variability in damage.
53 Results
Overall, the proportion of leaf area damaged was lower for exotics (mean ± SE: 0.038 ±
0.02) than for natives (0.073 ± 0.04) (Fig. 3.1). Without considering phylogenetic relatedness among individuals, this supports the hypothesis that exotics tend to experience lower damage than native species (p = 0.022).
When the phylogenetic distance matrix from the community phylogeny was decomposed using a principal coordinate analysis, two eigenvectors were significantly correlated with herbivory and explained 32% of variation in damage among species (Table 3.2). In general, phylogenetically structured patterns of damage were associated with deeper nodes within the family lineage that tend to represent taxonomic tribes (Funk et al. 2005).
The most apparent trends were associated with low damage in the lineage containing
Antennaria neglecta, Gnaphalium obtusifolium, and Anaphalis margaritacea
(Gnaphalieae) and relatively high damage in the lineage containing Senecio aureus and
Tussilago farfara (Senecioneae) (Fig. 3.2). There was also a weak phylogenetic signal in damage associated with the Astereae group (Solidago spp., Erigeron spp., Conyza sp.,
Euthamia sp., and Aster spp.) (Fig. 3.2).
When the residuals of the multiple regression of damage on the phylogenetic eigenvectors were used to represent phylogenetically corrected damage, exotic species still experienced lower damage than native species (p = 0.005, Fig. 3.3). Phylogenetically independent contrasts also supported this pattern: the mean difference for contrasts of
54 native and exotic species was 0.046 ± 0.025 (t = 1.843, df = 3, p = 0.0695) without considering branch lengths; when contrasts were standardized by branch lengths, the result became even stronger (t = 2.975, df = 3, p = 0.020).
Contrary to my prediction, phylogenetically novel exotic species tended to experience higher damage than those that were closely related to natives (Table 3.3, Fig. 3.4).
Statistically, however, these trends were weak for all phylogenetic distance predictors
(Table 3.3); of the four measures, distance to the closest native member of the Asteraceae was close to significance (p = 0.092). I suspected that these trends were influenced by high damage on Cirsium arvense and Arctium minus (Table 3.1, Fig. 3.2), which in part may have resulted from biocontrol insects (Mason & Huber 2002). When the analyses were run without damage estimates for these two species, there was no significant relationship between damage and any of phylogenetic distance measures, including the distance to closest native relative (0.195 ≤ p ≤ 0.818 for all cases).
Discussion
In this study, I show that exotic species in the Asteraceae experience lower herbivory than co-occurring native confamilials; this is also true when phylogenetic relationships are accounted for. However, contrary to my prediction, phylogenetically novel exotics tended to experience higher rates of herbivory than those that were more closely related to natives.
55 Implications for Enemy-Release and DNH
The Enemy Release Hypothesis predicts that exotic species benefit from leaving behind co-evolved enemies in their native range (Keane & Crawley 2002; Torchin et al. 2003;
Colautti et al. 2004). Tests of the hypothesis have produced variable results (Maron &
Vila 2001; Agrawal & Kotanen 2003; Colautti et al. 2004; Liu & Stiling 2006; Parker &
Gilbert 2007). Indeed, many species do tend to experience lower damage in their introduced range; however, when compared to co-occurring natives, exotics can experience similar levels of damage, or even higher rates of damage (Agrawal & Kotanen
2003; Parker et al. 2006). Comparing foliar damage on a broad suite of locally co- occurring Asteraceae, my data are consistent with the observations from other studies that damage on exotics can be variable, and even greater than on native species; for example, exotic species such as Arctium minus, Cirsium arvense, and Tussilago farfara experienced higher damage than the majority of native species, and native species such as
Gnaphalium obtusifolium, Anaphalis margaritacea, and Antennaria neglecta experienced damage that was similar or lower than the majority of exotic species. That exotics experience lower herbivory before and after phylogenetic correlations were taken into account, however, suggests that exotic species tend to experience lower herbivory when compared to natives. Since many of the exotic species can be found in the same habitats as native species, these reductions in herbivory may give them a competitive advantage, potentially increasing their impact as an invader (Maron & Vila 2001; Keane & Crawley
2002; Colautti et al. 2004; Liu & Stiling 2006).
56 Some of the measured damage was phylogenetically structured (r2 = 0.32). Where this was the case, damage tended to be similar within phylogenetically broad groups associated with taxonomic tribes (Funk et al. 2005). For these groups, this indicates that native and exotic species generally had similar levels or damage within a given lineage, though exotics tended to have slightly lower damage: for example the exotic
Chrysanthemum leucanthemum experienced similar rates of damage to native species in the same lineage (Erigeron sp., Conyza sp., and Solidago spp.) (Table 3.1). Similarly, damage on the native Prenanthes altissima was only slightly higher than damage on exotic species in the same lineage such as Sonchus arvensis, Sonchus asper, and Crepis capillaris (Table 3.1). In some cases, however, this pattern did not hold. For example, damage on the native Eupatorium spp. tended to be very high, but the closely related
Galinsoga quadriradiata experienced the lowest damage among all exotic species (Table
3.1).
Contrary to expectation, exotics that were less closely related to native species experienced higher rates of herbivory, though the pattern was weak (p = 0.092). This trend is consistent with other work that has summarized patterns of herbivory on exotic species across a number of studies (Ricciardi & Ward 2006), but is in contrast to patterns that I observed using exotic species from multiple plant families at KSR (Hill & Kotanen
2009). Ricciardi & Ward (2006) found that phylogenetically novel exotics may be more susceptible to enemies, as herbivory tended to be higher on exotic plants belonging to novel rather than native genera (Parker et al. 2006; Ricciardi & Ward 2006). My previous work however, found that invaders less related to native family members, or that
57 belonged to novel families, experienced lower damage (Hill & Kotanen 2009). The opposite result of Hill & Kotanen (2009) compared to patterns observed in this chapter and Ricciardi & Ward (2006) may be reconciled if enemy host use tends to be associated with particular plant families (Bernays & Chapman 1994; Futuyma & Mitter 1996;
Odegaard et al. 2005; Weiblen et al. 2006), but the ability of enemies to extend to, or shift among novel hosts within families is less predictable (Janz & Nylin 1998;
Lewinsohn et al. 2005). Such irregularity in host-use has been shown to vary among lineages within families; for example, Frenzel & Brandl (2001) found that phytophagous enemy communities on hosts in the Brassicaceae tended to be more similar (i.e. higher generalism) than within the tribe Cynaroideae (Asteraceae). My data may reflect this relatively restricted use of hosts within tribes of the Asteraceae, as phylogenetically structured herbivory occurred at about this taxonomic level.
The phylogenetically structured patterns that were observed in this study are unlikely to produce patterns of invasion success predicted by DNH (Rejmanek & Richardson 1996;
Daehler 2001; Duncan & Williams 2002; Diez et al. 2008): phylogenetically structured damaged tended to be conserved among deeper phylogenetic lineages within the family associated with taxonomic tribes, and exotics that were less closely related to natives tended to experience higher damage. Similar to (Ricciardi & Ward 2006), my work here suggests that within Asteraceae, phylogenetically novel invaders may be slightly more susceptible to enemies; in part this may result if these invaders are evolutionarily naive, lacking the necessary defenses to avoid or tolerate damage from the enemies they encountered (Hokkanen & Pimentel 1989; Verhoeven et al. 2009). Evaluating, for other
58 plant families, the degree to which phylogenetic novelty renders an invader more or less susceptible to natural enemies will allow an assessment of the generality of the trends that were observed among exotic and native plants in the Asteraceae.
Since this study was a survey on naturally occurring plants, I could not control other confounding factors that may be responsible for producing the observed patterns of herbivory among plants. For instance, some species may experience more damage because they have been targeted for biocontrol. This may be the case within the
Asteraceae for this region. In particular, Cirsium arvense, in the sub-family lineage
Cardueae (Funk et al. 2005), has often been targeted because of its invasiveness in agricultural habitats (Mason & Huber 2002). Biocontrol organisms that have been intended for use on C. arvense can colonize other species in the same tribe such as
Arctium minus, and these two species experienced some of the highest rates of damage among the exotic species that I sampled (Mason & Huber 2002). Removing these two species from the analyses resulted in non-significant trends between foliar damage and all phylogenetic distances. Therefore, the positive results that were observed when damage was regressed against the distance to its closest native relative may have been inflated by biocontrol insects using these species. If this was the case, and damage was caused by biocontrol agents, then patterns of damage on the remaining exotic species still showed no pattern with their phylogenetic distance to locally co-occurring native and exotic species. Thus, there still would be no evidence to support the prediction of reductions in damage with increasing phylogenetic distance to native and exotic species.
59 Implications for Community Phylogenetics
Darwin's Naturalization Hypothesis is a precursor to the contemporary field of community phylogenetics; the idea that community structure can be predicted by the phylogenetic relationship among co-occurring taxa (Webb et al. 2002). Currently, most studies use community phylogenetics to infer mechanisms underlying patterns of community structure (Vamosi et al. 2009): if close relatives are over-represented, then community assembly is assumed to result from environmental conditions filtering species traits (Webb et al. 2002); if close relatives are evenly distributed or under-represented, then community assembly is assumed to result from competition (Webb et al. 2002;
Vamosi et al. 2009). Biotic interactions other than competition, however, could be responsible for patterns that are observed in these types of studies (Cavender-Bares et al.
2009). For example, among close relatives, facilitation and apparent competition (Holt
1977) can result in opposing patterns of phylogenetic clustering and evenness, respectively (Vamosi et al. 2009). If apparent competition is strong enough to result in biotic resistance in natural communities, results from this study suggest that slightly higher damage by natural enemies on phylogenetically novel exotics may result in patterns of within-family phylogenetic clustering; where this occurs among exotics belong to native genera, patterns in the field would mimic environmental filtering, or facilitation by mutualists (Webb et al. 2002; Vamosi et al. 2009).
References
Agrawal A.A. & Kotanen P.M. (2003). Herbivores and the success of exotic plants: a
phylogenetically controlled experiment. Ecology Letters, 6, 712-715. 60 Agrawal A.A., Kotanen P.M., Mitchell C.E., Power A.G., Godsoe W. & Klironomos J.
(2005). Enemy release? An experiment with congeneric plant pairs and diverse
above- and belowground enemies. Ecology, 86, 2979-2989.
Bernays E.A. & Chapman R.F. (1994). Host plant selection by phytophagous insects.
Chapman & Hall, New York.
Cadotte M.W., Hamilton M.A. & Murray B.R. (2009). Phylogenetic relatedness and plant
invader success across two spatial scales. Diversity and Distributions, 15, 481-
488.
Cahill J.F., Kembel S.W., Lamb E.G. & Keddy P.A. (2008). Does phylogenetic
relatedness influence the strength of competition among vascular plants?
Perspectives in Plant Ecology Evolution and Systematics, 10, 41-50. de Candolle, A.L.P. (1855). Géographie Botanique raissoné. V. Masson, Paris.
Cappuccino N. & Carpenter D. (2005). Invasive exotic plants suffer less herbivory than
non-invasive exotic plants. Biology Letters, 1, 435-438.
Cavender-Bares J., Ackerly D.D., Baum D.A. & Bazzaz F.A. (2004). Phylogenetic
overdispersion in Floridian oak communities. American Naturalist, 163, 823-843.
Cavender-Bares J., Kozak K.H., Fine P.V.A. & Kembel S.W. (2009). The merging of
community ecology and phylogenetic biology. Ecology Letters, in press.
Colautti R.I., Ricciardi A., Grigorovich I.A. & MacIsaac H.J. (2004). Is invasion success
explained by the enemy release hypothesis? Ecology Letters, 7, 721-733.
61 Connor E.F., Faeth S.H., Simberloff D. & Opler P.A. (1980). Taxonomic isolation and
the accumulation of herbivorous insects - a comparison of introduced and native
trees. Ecological Entomology, 5, 205-211.
Daehler C.C. (2001). Darwin's naturalization hypothesis revisited. American Naturalist,
158, 324-330.
Darwin C. (1859). On the origin of species by means of natural selection. Murray,
London, UK
Dawson W., Burslem D. & Hulme P.E. (2009). Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions, 15, 141-147.
Desdevises Y., Legendre P., Azouzi L. & Morand S. (2003). Quantifying
phylogenetically structured environmental variation. Evolution, 57, 2647-2652.
Diez J.M., Sullivan J.J., Hulme P.E., Edwards G. & Duncan R.P. (2008). Darwin's
naturalization conundrum: dissecting taxonomic patterns of species invasions.
Ecology Letters, 11, 674-681.
Diniz-Filho J.A.F. & de Sant'Ana C.E.R. (1998). An eigenvector method for estimating
phylogenetic inertia. Evolution, 52, 1247-1262.
Duncan R.P. & Williams P.A. (2002). Ecology - Darwin's naturalization hypothesis
challenged. Nature, 417, 608-609.
Fargione J.E. & Tilman D. (2005). Diversity decreases invasion via both sampling and
complementarity effects. Ecology Letters, 8, 604-611.
62 Felsenstein J. (1985). Phylogenies and the comparative method. American Naturalist,
125, 1-15.
Frenzel M. & Brandl R. (2001). Hosts as habitats: faunal similarity of phytophagous
insects between host plants. Ecological Entomology, 26, 594-601.
Funk V.A., Bayer R.J., Kelley S., Chan R., Watson L., Gemeinholzer B., Schilling E.,
Panero J.L., Baldwin B.G., Garcia-Jacas N., Susanna A. & Jansen R.K. (2005).
Everywhere but Antarctica: using a supertree to understand the diversity and
distribution of the Compositae. Biologiske Skrifter, 55, 343-374.
Futuyma D.J. & Mitter C. (1996). Insect-plant interactions: the evolution of component
communities. Philosophical Transactions of the Royal Society B-Biological
Sciences, 351, 1361-1366.
Gaskin J.F. & Wilson L.M. (2007). Phylogenetic relationships among native and
naturalized Hieracium (Asteraceae) in Canada and the United States based on
plastid DNA sequences. Systematic Botany, 32, 478-485.
Gossner M.M., Chao A., Bailey R.I. & Prinzing A. (2009). Native fauna on exotic trees:
phylogenetic conservatism and geographic contingency in two lineages of
phytophages on two lineages of trees. American Naturalist, 173, 599-614.
Hawkes C.V. (2007). Are invaders moving targets? The generality and persistence of
advantages in size, reproduction, and enemy release in invasive plant species with
time since introduction. American Naturalist, 170, 832-843.
Hill S.B. & Kotanen P.M. (2009). Evidence that phylogenetically novel non-indigenous
plants experience less herbivory. Oecologia, in-press.
63 Hokkanen H.M.T. & Pimentel D. (1989). New associations in biological control: theory
and practice. Canadian Entomologist, 121, 829-840.
Holt R.D. (1977). Predation, apparent competition, and structure of prey communities.
Theoretical Population Biology, 12, 197-229.
Janz N. & Nylin S. (1998). Butterflies and plants: A phylogenetic study. Evolution, 52,
486-502.
Keane R.M. & Crawley M.J. (2002). Exotic plant invasions and the enemy release
hypothesis. Trends in Ecology & Evolution, 17, 164-170.
Kembel S.W. & Hubbell S.P. (2006). The phylogenetic structure of a neotropical forest
tree community. Ecology, 87, S86-S99.
Kim K.J., Choi K.S. & Jansen R.K. (2005). Two cholorplast DNA inversions originated
simultaneously during the early evolution of the sunflower family (Asteraceae).
Molecular Biology and Evolution, 22, 1783-1792.
Levine J.M., Adler P.B. & Yelenik S.G. (2004). A meta-analysis of biotic resistance to
exotic plant invasions. Ecology Letters, 7, 975-989.
Lewinsohn T.M., Novotny V. & Basset Y. (2005). Insects on plants: diversity of
herbivore assemblages revisited. Annual Review of Ecology and Systematics, 36,
597-620.
Liu H. & Stiling P. (2006). Testing the enemy release hypothesis: a review and meta-
analysis. Biological Invasions, 8, 1535-1545.
64 Mack R.N. (1996). Biotic barriers to plant naturalization. In: Proceedings 9th
International Symposium of Biological Control of Weeds. Stellenbosch University
of Cape Town, South Africa, pp. 39-46.
Maron J.L. & Vila M. (2001). When do herbivores affect plant invasion? Evidence for
the natural enemies and biotic resistance hypotheses. Oikos, 95, 361-373.
Mason P.G. & Huber J.T. (2002). Biological control programmes in Canada 1981-2000.
CABI Publishing, Ottawa, ON, Canada.
Mitchell C.E., Agrawal A.A., Bever J.D., Gilbert G.S., Hufbauer R.A., Klironomos J.N.,
Maron J.L., Morris W.F., Parker I.M., Power A.G., Seabloom E.W., Torchin M.E.
& Vazquez D.P. (2006). Biotic interactions and plant invasions. Ecology Letters,
9, 726-740.
Mitchell C.E. & Power A.G. (2003). Release of invasive plants from fungal and viral
pathogens. Nature, 421, 625-627.
Naeem S., Knops J.M.H., Tilman D., Howe K.M., Kennedy T. & Gale S. (2000). Plant
diversity increases resistance to invasion in the absence of covarying extrinsic
factors. Oikos, 91, 97-108.
Novotny, V. & Basset, Y. (2005). Host specificity of insect herbivores in tropical forests.
Proceedings of the Royal Society B, 272, 1083-1090.
Novotny, V., Basset, Y., Miller, S.E., Weiblen, G.D., Bremer, B., Cizek, L, & Drozd, P.
(2002). Low host specificity of herbivorous insects in a tropical forest. Nature,
416, 841-844.
65 Noyes R.D. (2000). Biogeographical and evolutionary insights on Erigeron and allies
(Asteraceae) from ITS sequence data. Plant Systematics and Evolution, 220, 93-
114.
Odegaard F., Diserud O.H. & Ostbye K. (2005). The importance of plant relatedness for
host utilization among phytophagous insects. Ecology Letters, 8, 612-617.
Parker I.M. & Gilbert G.S. (2007). When there is no escape: The effects of natural
enemies on native, invasive, and noninvasive plants. Ecology, 88, 1210-1224.
Parker J.D., Burkepile D.E. & Hay M.E. (2006). Opposing effects of native and exotic
herbivores on plant invasions. Science, 311, 1459-1461.
Parker J.D. & Hay M.E. (2005). Biotic resistance to plant invasions? Native herbivores
prefer non-native plants. Ecology Letters, 8, 959-967.
Proches S., Wilson J.R.U., Richardson D.M. & Rejmanek M. (2008). Searching for
phylogenetic pattern in biological invasions. Global Ecology and Biogeography,
17, 5-10.
R Development Core Team (2006) R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria ISBN 3-
900051-07-0, URL http://wwwR-projectorg
Rejmanek M. & Richardson D.M. (1996). What attributes make some plant species more
invasive? Ecology, 77, 1655-1661.
Ricciardi A. & Ward J.M. (2006). Comment on "Opposing effects of native and exotic
herbivores on plant invasions". Science, 313.
66 Richardson D.M., Allsopp N., D'Antonio C., Milton S.J. & Rejmanek M. (2000). Plant
invasions - the role of mutualisms. Biological Reviews, 75, 65-93.
Semple J.C. & Cook R.E. (2006). Solidago. In: Flora of North America (ed. Committee
FNAE). Oxford University Press Oxford, pp. 107-166.
Strauss S.Y., Webb C.O. & Salamin N. (2006). Exotic taxa less related to native species
are more invasive. Proceedings of the National Academy of Sciences of the United
States of America, 103, 5841-5845.
Strong D.R., Lawton J.H. & Southwood R. (1984). Insects on plants. Harvard University
Press, Cambridge, Mass., USA.
Torchin M.E., Lafferty K.D., Dobson A.P., McKenzie V.J. & Kuris A.M. (2003).
Introduced species and their missing parasites. Nature, 421, 628-630.
Vamosi S.M., Heard S.B., Vamosi J.C. & Webb C.O. (2009). Emerging patterns in the
comparative analysis of phylogenetic community structure. Molecular Ecology,
18, 572-592.
Verhoeven K.J.F., Biere A., Harvey J.A. & van der Putten W.H. (2009). Plant invaders
and their novel natural enemies: who is naive? Ecology Letters, 12, 107-117.
Webb C.O., Ackerly D.D. & Kembel S.W. (2008). Phylocom: software for the analysis
of phylogenetic community structure and trait evolution. Bioinformatics, 24,
2098-2100.
Webb C.O., Ackerly D.D., McPeek M.A. & Donoghue M.J. (2002). Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475-505.
67 Weiblen G.D., Webb C.O., Novotny V., Basset Y. & Miller S.E. (2006). Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62-
S75.
Whitton J., Wallace R.S. & Jansen R.K. (1995). Phylogenetic-relationships and patterns
of character change in the tribe Lactuceae (Asteraceae) based on chrloroplast
DNA restriction site variation. Canadian Journal of Botany-Revue Canadienne
De Botanique, 73, 1058-1073.
68 Table 3.1 Study species belonging to the Asteraceae present at KSR used for this study, their origin, mean proportion of foliar damage, and SE of proportion foliar damage. For
SE proportion damage, NA indicates one population of a species was sampled
SE Proportion Species Origin Proportion Damage Damage Anaphalis margaritacea (L.) Benth. & Hook. Native 0.025 0.014 Antennaria neglecta Greene Native 0.008 0.003 Arctium minus (Hill) Bernh. Exotic 0.101 0.032 Aster cordifolius L. Native 0.210 0.032 Aster novae-angliae L. Native 0.078 0.017 Bidens cernua L. Native 0.209 NA Bidens frondosa L. Native 0.179 0.007 Centaurea jacea L. Exotic 0.031 0.009 Chrysanthemum leucanthemum L. Exotic 0.026 0.009 Cichorium intybus L. Exotic 0.066 NA Cirsium arvense (L.) Scop. Exotic 0.088 0.037 Conyza canadensis (L.) Cron. Native 0.031 0.003 Crepis capillaris (L.) Wallr. Exotic 0.033 0.015 Erigeron annuus (L.) Pers. Native 0.056 0.026 Erigeron philadelphicus L. Native 0.040 0.009 Erigeron strigosus Muhl. Native 0.033 0.018 Eupatorium maculatum L. Native 0.147 0.059 Eupatorium perfoliatum L. Native 0.422 0.023 Euthamia graminifolia (L.) Salisb. Native 0.078 0.007 Galinsoga quadriradiata Ruiz & Pavon Exotic 0.009 NA Gnaphalium obtusifolium L. Native 0.026 NA Hieracium caespitosum L. Exotic 0.020 0.005 Hieracium pilosella L. Exotic 0.017 0.009 Inula helenium L. Exotic 0.043 0.007 Lactuca serriola L. Exotic 0.009 0.001 Prenanthes altissima L. Native 0.069 0.020 Rudbeckia hirta L. Native 0.059 0.020 Senecio aureus L. Native 0.074 0.024 Solidago caesia L. Native 0.053 0.010 Solidago canadensis L. Native 0.209 0.047 Solidago flexicaulis L. Native 0.104 0.028 Sonchus arvensis L. Exotic 0.062 0.037 Sonchus asper (L.) Hill Exotic 0.058 0.026 Tragopogon dubius Scop. Exotic 0.014 0.004 Tussilago farfara L. Exotic 0.079 0.035
69 Table 3.2 Multiple regression results of damage versus the eigenvectors from the
Principle Coordinate Analysis of the community phylogeny that were significantly related to herbivory. The sum of the eigenvector loadings are shown in Fig. 3.2.
Residuals from the multiple regression of damage on these eigenvectors were used as the phylogenetically corrected damage estimate, shown in Fig. 3.3. Numbers in square brackets for eigenvectors represent the rank among the 34 eigenvectors that were used to partition phylogenetic distances; lower numbers represent eigenvectors that explain broader patterns of phylogenetically structured variation. Degrees of freedom for each t- test was 32
Parameter Estimate SE t p Intercept -1.581 0.069 -22.776 <0.001 eigenvector [6] -0.020 0.008 -2.482 0.016 eigenvector [25] 0.097 0.028 3.417 0.002
70 Table 3.3 Regressions results of foliar damage versus different measures of phylogenetic uniqueness (mean distance to all (i) Asteraceae, (ii) closest relative, (iii) all native
Asteraceae, and (iv) closest native Asteraceae. Distance to closest native Asteraceae is marginally significant and explains the most variation in damage (15%)
Model df slope se F p r2 All Asteraceae 1,12 0.747 0.541 1.909 0.192 0.07 Closest Asteraceae 1,12 0.309 0.189 2.676 0.128 0.11 All native Asteraceae 1,12 0.437 0.436 1.008 0.335 <0.01 Closest native Asteraceae 1,12 0.301 0.165 3.353 0.092 0.15
71 Figure 3.1 Difference in mean ± SE damage between native and exotic Asteraceae without phylogenetic correction. The one-tailed Welch's t-test (equal variance was not assumed, so corrected degrees of freedom are used) supports the hypothesis that exotic species experience lower damage than native species
72 Figure 3.2 Community phylogeny of Asteraceae sampled from species present at KSR.
Bars on right show the sum of the resulting eigenvector scores that best explain the phylogenetic component of herbivory. Two eigenvectors had a significant relationship with herbivory and together explained 32% of the variation in foliar damage (Table 3.2)
73 Figure 3.3 Difference in mean ± SE damage between native and exotics when using the residuals from the multiple regression from the significant eigenvectors (Table 3.2). The one-tailed Welch's t-test (equal variance was not assumed, so corrected degrees of freedom are used) supports the hypothesis that exotic species experience lower damage than native species
74 Figure 3.4 Proportion of leaf area damage versus phylogenetic novelty (distance to all
Asteraceae, closest Asteraceae, all native Asteraceae, and closest native Asteraceae) for exotic species only. Damage tends to increase with phylogenetic novelty, but was marginally significant only when distance to closest native relative was used as the predictor (see statistics in Table 3.3). Analyses were performed on Z-score transformed damage data, and ln transformed phylogenetic distances to better meet assumptions of normality and homoscedasticity
75 Chapter 4
Enemy release from pre-dispersal seed predators in Asteraceae:
a phylogenetically controlled study
Abstract
Exotic species more closely related to native species may be more susceptible to native enemies if host use by enemies is phylogenetically conserved. Where this is the case, the use of phylogenies that include co-occurring native and exotic species may identify which exotic species are more likely to interact with enemies. In this study, I measured damage caused by pre-dispersal seed predators on native and exotic plants in the
Asteraceae. Damage was then mapped onto a community phylogeny. I tested the predictions that exotic plants experience lower damage than native species after controlling for evolutionary history, and that phylogenetically novel exotic species would experience lower damage. Consistent with my first prediction, exotic plants experienced lower damage than native plants. However, there was no support for the phylogenetic novelty prediction. Despite this result, phylogenetically structured damage by pre- dispersal seed predators tended to occur among nodes within the family lineage associated with taxonomic tribes. This suggests that exotics do suffer less damage than related natives, but damage experienced by exotics may depend on the phylogenetic scale of comparison with native species.
76 Introduction
The Enemy Release Hypothesis (ERH) predicts that exotic species leave specialist enemies behind in their native range, giving them an advantage over co-occurring native species in their invaded range (Keane & Crawley 2002; Torchin et al. 2003). Tests of this hypothesis have found varied results (Maron & Vila 2001; Colautti et al. 2004; Liu &
Stiling 2006): exotic species do tend to experience reduced damage in their invaded range
(Maron & Vila 2001; Colautti et al. 2004; Liu & Stiling 2006), but some exotics experience similar or higher damage than native competitors (Agrawal & Kotanen 2003).
This variation may exist, in part, because some exotics can rapidly accumulate enemies in their invaded range (Hawkes 2007); this is expected to be more common when host-use by enemies is phylogenetically conserved (Odegaard et al. 2005; Weiblen et al. 2006;
Gilbert & Webb 2007). Such interactions with enemies, therefore, may be more likely for exotics that have close relatives in their introduced range (Strong et al. 1984; Lewinsohn et al. 2005; Dawson et al. 2009; Gossner et al. 2009).
The idea that phylogenic relationship between exotic and native species can influence invasion success is not new. Both de Candolle (1855) and Darwin (1859) observed that
North American exotic species often belonged to novel genera, resulting in the proposition that invasion by close relatives is limited by competition (Darwin 1859).
Subsequently this idea was named Darwin's Naturalization Hypothesis (Daehler 2001).
Tests of this hypothesis vary: Daehler (2001) and Duncan and Williams (2002) found evidence against the idea, while other studies have found support (Rejmanek &
Richardson 1996; Diez et al. 2008). Phylogenetic approaches have also revealed that
77 introduced grasses in California tend to be more invasive if they are less closely related to native grasses (Strauss et al. 2006). An emerging pattern from these studies is that phylogenetically novel exotic species may be better invaders.
Although other biotic interactions can potentially result in community-level patterns consistent with DNH, competition is usually assumed to be the mechanism (Darwin
1859; Webb et al. 2002). Experimental evidence that compares competitive interactions among species based on their phylogenetic relatedness suggests that trends tend to be weak: Cahill et al. (2008) found that there were only marginal relationships between competition and phylogenetic distance among a broad group of plants. Alternatively, biotic interactions such as herbivory may cause patterns similar to those predicted by
DNH through mechanisms such as apparent competition (Holt 1977; Holt & Lawton
1994; Mack 1996). As outlined above, host use by enemies (invertebrate and pathogens) tends to be phylogenetically conserved (Odegaard et al. 2005; Weiblen et al. 2006;
Gilbert & Webb 2007). Evidence that being closely related to native species affects the susceptibility of an invader to herbivory, however, is mixed (Cappuccino & Carpenter
2005; Dawson et al. 2009; Hill & Kotanen 2009): for instance, Cappuccino & Carpenter
(2005) found weak support for the prediction that taxonomic isolation for invaders in eastern North America resulted in lower foliar damage. In contrast, Dawson et al. (2009) found that taxonomically isolated trees introduced to Tanzania tend to escape damage from foliar herbivores. My previous work is consistent with Dawson et al. (2009), and has shown that phylogenetically novel plants experience less damage than those that that are more closely related to co-occurring native family members (Hill & Kotanen 2009).
78 Many studies evaluate enemy escape on plants using damage by folivorous insects
(Agrawal & Kotanen 2003; Agrawal et al. 2005; Hill & Kotanen 2009). However, evaluating the phylogenetic component of enemy escape may require measuring the impacts of enemies that tend to have more specialized host-use (Frenzel & Brandl 2001;
Joshi & Vrieling 2005). Insects that consume seeds while still attached to the maternal plant (predispersal seed predators) may offer a guild of such enemies (Fenner & Lee
2001). This group of enemies includes specialized flies, beetles, and Lepidoptera; furthermore, the host fidelity of many predispersal seed predators often results in their use as biocontrol agents for invasive exotic species (Louda & O'Brien 2002). As well, damage by folivorous insects may only have weak effects on an invaders fitness or vigour
(Crawley 1989; Parker & Gilbert 2007), whereas predispersal seed predators can cause dramatic declines in seed production (Hawthorn & Hayne 1978); in some cases, these reductions can negatively affect population growth (e.g. Louda & O'Brien 2002).
In another study I examined patterns of foliar damage on native and exotic plants in the
Asteraceae (Chapter 3). In this previous work, I found that foliar damage on exotic plants was lower than that on native species before and after controlling for phylogenetic relationships (Chapter 3). However, there was no support for the prediction that novel invaders experienced less damage (Connor et al. 1980; Dawson et al. 2009; Hill &
Kotanen 2009). Here I investigated patterns of damage caused by pre-dispersal seed predators for native and exotic species in the same family. Specifically I tested the hypotheses that i) pre-dispersal damage to capitula is lower for exotic species than native
79 species, ii) that this is also true when adjusted for phylogenetic relationship within the family, and iii) that phylogenetically novel exotic species experience lower rates of damage than those more closely related to the native flora. Similar to patterns observed in the previous chapter, I show that the capitular damage is phylogenetically structured among broad groups of native and exotic plants within the family. Also, exotics experience lower rates of damage when phylogenetic corrections are used, but phylogenetically novel exotics did not experience lower damage.
Methods
Study site and natural history
This study was conducted at the University of Toronto's Koffler Scientific Reserve
(KSR) at Jokers Hill, Regional Municipality of York, in southern Ontario, Canada (44º02'
N, 79º31' W, 300m ASL). This 350-ha site lies within the Oak Ridges Moraine, and is dominated by prominent hills with a thin organic layer over deep glacial sands.
Vegetation is a mixture of old fields supporting a diverse range of native and exotic plants, hardwood (maple-beech-hemlock) forest, and conifer plantations. Further information on this site may be found at http://www.ksr.utoronto.ca.
I chose to work with native and exotic plants in the Asteraceae as this group is well represented at KSR: there are 77 species in the family (~12% of the local flora), 44 are native and the remaining 33 are exotic (Table 4.1). Members in this group are present in all major habitats found on site, including old-field meadows, forest understories, and wetlands. Their phylogeny is well understood (Funk et al. 2005), with exotic species
80 represented among many of the lineages that are present at KSR. World-wide, numerous members of this family have also become important invasive exotic species (e.g. Cadotte et al. 2009).
Survey of Capitulum Damage
During May to November 2008, I collected capitula for 21 native and 14 exotic species in the Asteraceae (Table 4.1); these species were chosen based on their abundance in the field and to represent as many lineages within the group as possible. Native and exotic plants were sampled in all respective habitat types (e.g. oldfield, forest understory, and wetland) to attempt to avoid biases associated with habitat-related damage on native and exotic plants (e.g. Parker & Gilbert 2007). For each species, capitula were collected from two to ten randomly sampled individuals in one to four locations separated by at least 200 m at KSR. To allow time for colonization by seed predators, I collected the capitula for each species two to four weeks after first noting that flowering had begun. In the lab, capitula from each location were mixed together, and between 50 to 100 were randomly selected to count; the resulting measure of damage represented the population-level of damage by predispersal seed predators. Capitula were counted as damaged if larvae of any kind (typically belonging to Diptera, Lepidoptera, or Chrysomelidae) were found amongst the developing seeds or buried in the receptacle. I also counted damage where larvae were absent, but damage had clearly been caused by these larvae. This included bore holes in seeds with missing endosperm, bore holes in the receptacle, and missing seeds but the presence of insect frass. For statistical analyses, estimates of damage for each species were then calculated as the proportion of damaged capitula to total capitula
81 summed across populations. These proportions were transformed by calculating the associated Z-score for the proportion of damaged capitula; when there were no capitula damaged, 0.001 was added to the zero-score to avoid undefined numbers (Crawley 2007).
Z-score transformations are similar to probit transformations, and linearize the response variable, which is typically more appropriate for linear models (Crawley 2007).
Community phylogeny
The backbone of the Asteraceae community phylogeny was drawn from Funk et al.
(2005). Where polytomies existed, I supplemented the topology based on the following studies: Erigeron (Noyes 2000), Hieracium (Gaskin & Wilson 2007), Lactuceae
(Whitton et al. 1995), Solidago (Semple & Cook 2006). Node ages for the resulting tree were then estimated using the function "bladj" in PHYLOCOM v.4.0.1 (Webb &
Donoghue 2005), with node age calibration points from a phylogeny based on rbcL and ndhF sequence data (Kim et al. 2005) (see Fig. 4.1).
Statistical Analyses
I evaluated whether damage covaried with receptacle width or average seed size using separate least square regressions. To determine if there was a difference between species sampled at different times I categorized them into early flowering plants (May-August), and late flowering plants (September - November); this was based on observations in the field, and reflected the general flowering phenology of the plants. I used a t-test to determine if damage differed between these two phenological groups.
82 As in the previous chapter, to evaluate the phylogenetic component of herbivory, I used an eigenvector decomposition method (Diniz-Filho & de Sant'Ana 1998; Desdevises et al. 2003). Using a principle co-ordinate analysis (PCoA) of the phylogenetic distance matrix from the Asteraceae community phylogeny, 34 eigenvectors were extracted to predict phylogenetically structured damage. The PCoA eigenvectors represent unique
(orthogonal) predictors of phylogenetic structure in a trait (in this case capitular damage) based on phylogenetic distances between species. The order of the eigenvectors predicts increasingly finer-scale patterns of phylogenetically structured variation, such that the first eigenvector detects the broadest pattern, and the last eigenvector detects the finest- scale pattern (Diniz-Filho & de Sant'Ana 1998). In general, traits that have broad phylogenetic structure are associated with deeper splits within the phylogeny, whereas fine phylogenetic structures are associated with recent splits, closer to the tips of the phylogeny. Each eigenvector was then compared to the proportion of capitula damaged using a linear regression, and only those that were significant at p ≤ 0.10 were retained as phylogenetic predictors of damage. These eigenvectors were then used in a multiple regression of proportion of capitula damaged versus the significant phylogenetic eigenvectors to remove the phylogenetic component of herbivory. The residuals from this analysis were used in the one-tailed t-tests outlined below.
To test the hypothesis that exotic species experienced lower rates of pre-dispersal seed predation, for both the non-corrected and phylogenetically-corrected data I used a two- factor analysis of variance including terms for origin (native versus exotic) and flowering time (since preliminary analysis suggested it should be included as a covariate). Terms
83 that were non-significant (p > 0.05) were removed from the statistical model, until a minimum adequate model was identified to explain the data (Crawley 2007). If the reduced model resulted in only the species' origin factor (native or exotic) being significant, I used a one-tailed t-test to determine if exotic species experienced lower damage than native species.
To determine if phylogenetically novel invaders experienced lower capitular damage than those more closely related, four predictors of phylogenetic novelty from the rest of the
Asteraceae present at KSR were calculated: i) mean distance to all native and exotic taxa, ii) mean distance to closest native or exotic relative, iii) mean distance to all native taxa, and iv) mean distance to closest native taxa. Distances among individual taxa were based on the tip-to-tip measurements given the branch lengths in the Aster community phylogeny. The four predictors, although superficially similar, test different hypotheses.
The first two evaluate the degree to which interactions among native and exotic relatives matter; this may be the case if herbivore interactions have developed on other exotic species within the family. The third tests whether interactions with herbivores depends on the mean distance to only native species within the family as a whole, the fourth tests if it is the distance to the closest native relative that matters.
All analyses were completed using packages ape, stats, and vegan in R (R
Development Core Team 2006) or JMP v5.0 (SAS 2002).
84 Results
Covariates
The proportion of capitula damaged was not significantly related to either receptacle size
(F1,33 = 0.568, p = 0.457) or seed size (F1,33 = 0.451, p = 0.506). However, damage was marginally higher for species sampled earlier in the season (May - August) versus late
season (September - October) (t32.1 = 1.654, p = 0.108), therefore it was included in the statistical models along with species origin.
Native versus Exotic
Capitulum damage was similar for native and exotic species when no phylogenetic correction was made (p = 0.457, Table 4.2, Fig. 4.2). However, time of sampling remained marginally significant (p = 0.070, Table 4.2), with higher damage during the early season compared to the late season plants. The lack of an interaction between origin and sampling time (p = 0.215, Table 4.2) indicates that changes in damage levels over the season were similar for native and exotic species. These results changed when the phylogenetic component of capitular damage was removed (Table 4.2, Fig. 4.3). Native plants tended to experience higher damage than exotics (p = 0.061, Table 4.2, Fig. 4.3), season became non-significant (p = 0.233, Table 4.2), and the interaction between origin and sampling time remained non-significant (p = 0.408, Table 4.2). When the non- significant effects of the interaction and sampling time were removed, a one-tailed t-test to determine if mean damage for exotics was lower than for natives provided stronger support for my prediction (t = -1.7902, df = 20.578, p = 0.044). This result was further supported using the five available phylogenetically independent contrasts of the
85 difference between damage on native versus exotic plants (mean difference ± SE: 0.16 ±
0.052, t = 3.073, df = 4, p = 0.019).
Phylogenetic novelty
The prediction of decreased capitular damage for phylogenetically novel exotic species was not supported. Damage was not statistically related to any of the four phylogenetic novelty measures for exotic species (Table 4.3, Fig. 4.4). As in chapter 3, I suspected that high rates of capitular damage on Cirsium arvense and Arctium minus were, in part, the result of biocontrol insects (Mason & Huber 2002). When data for these two species were removed from the phylogenetic distance regressions, none had significant slopes (0.554 ≤ p ≤ 0.907 for all cases).
Discussion
Whether exotic species are likely to interact with enemies may influence their success and invasiveness in new environments (Maron & Vila 2001; Levine et al. 2004; Fagan et al. 2005; Hawkes 2007). In this study I predicted that exotic species would experience lower damage than natives, and that this would be true after correcting for phylogeny. I did not find support for the first prediction in the uncorrected dataset; however, after correcting for phylogenetic relationship among the study species, I found that exotic species experienced significantly lower damage. I also tested the hypothesis that among exotic species, those that are phylogenetically novel would be more likely to escape enemies. My data did not support this prediction, as exotic species consistently
86 experienced rates of damage that were independent of their evolutionary novelty relative to locally co-occurring native and exotic confamilials.
The lack of an effect of origin for non-corrected damage indicates that the exotic
Asteraceae in this study have been colonized by pre-dispersal seed predators since their introduction to North America. In some cases, the effects of these seed predators resulted in capitular damage that was higher than native species. Indeed, Arctium minus had the highest proportion (89%) of damaged capitula, even though it is an exotic; and native species such as Bidens cernua, Eupatorium perfoliatum, and Gnaphalium obtusifolium had no evidence of damaged capitula. That exotics experience lower damage when corrected for phylogenetic relationship suggests that, among exotic and native species belonging to the same lineage, exotics will tend to experience lower than expected rates of predispersal seed predation, as also found by Fenner & Lee's (2001) study on
Asteraceae in New Zealand.
The apparent lack of relationship between damage and phylogenetic distance of invaders to co-occurring family members suggests that the degree of evolutionary divergence among these plants had a minimal effect on their susceptibility to attack. However, since a large component of variability in damage was explained by the phylogenetic eigenvectors (r2 = 0.59), processes that result in similar levels of capitular damage do seem to be conserved among sub-lineages within the family. Similar to phylogenetically structured patterns of foliar damage observed on native and exotic Asteraceae (Chapter
3), capitular damage tended to be conserved across lineages within the family that are
87 represented by taxonomic tribes. That rates of damage were generally similar among these groups within the family, but that distances among invaders and other species were not significant, suggests that patterns of host-use may be idiosyncratic within these sub- family lineages and difficult to detect with the distance measures used here (Janz & Nylin
1998; Lewinsohn et al. 2005).
In part, this also might help explain apparent inconsistencies between the lack of a phylogenetic distance pattern in damage below the family-level in this study, versus decreases in damage among families when exotics were less closely related to native family members (Chapter 2). The observations here are consistent with those from the previous chapter, and therefore a similar mechanism may explain these patterns: although patterns of host-use by enemies tend to be shared within families (Odegaard et al. 2005;
Weiblen et al. 2006) and shared host-use can decline with increasing genetic distance
(Brandle & Brandl 2006; Gilbert & Webb 2007), the strength of phylogenetically conserved host-use can vary among families (Frenzel & Brandl 2001). Therefore, one might expect that patterns of damage below the family would be difficult to detect with phylogenetic distances that measure evolutionary novelty based on closest relative or all relatives, as were used in this study. Future work that contrasts patterns of host-use on exotic and native species, both within and among families should reveal whether the patterns that were observed within the Asteraceae are consistent with other plant families.
88 ERH, DNH, & Community Phylogenetics
Results from this study suggest that release from enemies may at best be considered a relative process: exotics may experience lower damage than expected based on their evolutionary history with native species, rather than an absolute reduction relative to levels of damage experienced by all co-occurring native plants. This has important implications for contemporary ideas in invasion biology and community assembly such as the "Enemy Release Hypothesis", "Darwin's Naturalization Hypothesis", and community phylogenetics.
Overall, the idea that exotic species escape enemies in their introduced range relative to native species is supported here. If niche space is also conserved among these close relatives (Silvertown 2004), reduced herbivory on exotics could give them a competitive advantage over native species with which they are more likely to compete (Keane &
Crawley 2002). However, this only occurred when phylogenetic relatedness was controlled. These results suggest that future tests the ERH may benefit from including an explicit phylogenetic component that evaluates to which lineage an invader belongs relative to native species. Incorporating a phylogenetic framework into tests of the ERH may also help explain the varied results among cases studies that are observed in the literature (Maron & Vila 2001; Colautti et al. 2004; Liu & Stiling 2006).
The initial prediction that damage would be negatively related to phylogenetic novelty was rejected for all four phylogenetic distance measures. This implies that pre-dispersal seed predators are unlikely to produce within-family patterns of co-occurrence predicted
89 by "Darwin's Naturalization Hypothesis". Since the proportion of capitula damaged among invaders was independent of their phylogenetic novelty, but had a strong phylogenetic signal associated with taxonomic tribes, factors within these lineages may be important for the levels of damage that were observed on exotics. Identifying the characters responsible for the phylogenetic signal and how well they are conserved among native and exotic species belonging to these groups should help predict susceptibility to enemies, and put caveats on where DNH may not operate among close relatives.
Community phylogenetics predicts that patterns of co-occurrence among species will reflect their evolutionary history (Webb et al. 2002). Communities are expected to be phylogenetically clustered when structured by environmental conditions, and to be phylogenetically even or over-dispersed if structured by competitive interactions (Webb et al. 2002). Similar to DNH, these patterns are predicated on the idea that competition among close relatives is more intense as a result of niche conservatism (Vamosi et al.
2009). However, biological interactions other than competition can be associated with phylogenetic relatedness (Weiblen et al. 2006; Gilbert & Webb 2007), and presumably result in similar co-occurrence patterns among communities. For example, direct or indirect facilitation among close relatives could result in phylogenetic clustering
(Richardson et al. 2000), and interactions with shared enemies could produce patterns of phylogenetic evenness, or over-dispersion if close relatives are more likely to share enemies (Mitchell et al. 2006). Results from this study indicate that there is a strong phylogenetic component to capitular damage among native and exotic plants in the
90 Asteraceae (r2 = 0.59), but damage is not necessarily related to phylogenetic novelty of exotic species based on the distance measures we used. The neutral effect of phylogenetic distance on damage then presumably has little effect on how communities are structured, and is unlikely to yield patterns of co-occurrence that mimic competition.
References
Agrawal A.A. & Kotanen P.M. (2003). Herbivores and the success of exotic plants: a
phylogenetically controlled experiment. Ecology Letters, 6, 712-715.
Agrawal A.A., Kotanen P.M., Mitchell C.E., Power A.G., Godsoe W. & Klironomos J.
(2005). Enemy release? An experiment with congeneric plant pairs and diverse
above- and belowground enemies. Ecology, 86, 2979-2989.
Brandle M. & Brandl R. (2006). Is the composition of phytophagous insects and parasitic
fungi among trees predictable? Oikos, 113, 296-304.
Cadotte M.W., Hamilton M.A. & Murray B.R. (2009). Phylogenetic relatedness and plant
invader success across two spatial scales. Diversity and Distributions, 15, 481-
488.
Cahill J.F., Kembel S.W., Lamb E.G. & Keddy P.A. (2008). Does phylogenetic
relatedness influence the strength of competition among vascular plants?
Perspectives in Plant Ecology Evolution and Systematics, 10, 41-50. de Candolle, A.L.P. (1855). Géographie Botanique raissoné. V. Masson, Paris.
Cappuccino N. & Carpenter D. (2005). Invasive exotic plants suffer less herbivory than
non-invasive exotic plants. Biology Letters, 1, 435-438.
91 Colautti R.I., Ricciardi A., Grigorovich I.A. & MacIsaac H.J. (2004). Is invasion success
explained by the enemy release hypothesis? Ecology Letters, 7, 721-733.
Connor E.F., Faeth S.H., Simberloff D. & Opler P.A. (1980). Taxonomic isolation and
the accumulation of herbivorous insects - a comparison of introduced and native
trees. Ecological Entomology, 5, 205-211.
Crawley M.J. (1989). Insect herbivores and plant population dynamics. Annual Review of
Ecology and Systematics, 34, 531-564.
Crawley M.J. (2007). The R book. John Wiley, New York.
Daehler C.C. (2001). Darwin's naturalization hypothesis revisited. American Naturalist,
158, 324-330.
Darwin C. (1859). On the origin of species by means of natural selection. Murray,
London, UK.
Dawson W., Burslem D. & Hulme P.E. (2009). Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions, 15, 141-147.
Desdevises Y., Legendre P., Azouzi L. & Morand S. (2003). Quantifying
phylogenetically structured environmental variation. Evolution, 57, 2647-2652.
Diez J.M., Sullivan J.J., Hulme P.E., Edwards G. & Duncan R.P. (2008). Darwin's
naturalization conundrum: dissecting taxonomic patterns of species invasions.
Ecology Letters, 11, 674-681.
Diniz-Filho J.A.F. & de Sant'Ana C.E.R. (1998). An eigenvector method for estimating
phylogenetic inertia. Evolution, 52, 1247-1262.
92 Duncan R.P. & Williams P.A. (2002). Ecology - Darwin's naturalization hypothesis
challenged. Nature, 417, 608-609
Fagan W.F., Lewis M.L., Neubert M.G., Aumann C., Apple J.L. & Bishop J.G. (2005).
When can herbivores slow or reverse the spread of an invading plant? A test case
from Mount St. Helens. American Naturalist, 166, 669-685.
Fenner M. & Lee W.G. (2001). Lack of pre-dispersal seed predators in introduced
Asteraceae in New Zealand. New Zealand Journal of Ecology, 25, 95-99.
Frenzel M. & Brandl R. (2001). Hosts as habitats: faunal similarity of phytophagous
insects between host plants. Ecological Entomology, 26, 594-601.
Funk V.A., Bayer R.J., Kelley S., Chan R., Watson L., Gemeinholzer B., Schilling E.,
Panero J.L., Baldwin B.G., Garcia-Jacas N., Susanna A. & Jansen R.K. (2005).
Everywhere but Antarctica: using a supertree to understand the diversity and
distribution of the Compositae. Biologiske Skrifter, 55, 343-374.
Gaskin J.F. & Wilson L.M. (2007). Phylogenetic relationships among native and
naturalized Hieracium (Asteraceae) in Canada and the United States based on
plastid DNA sequences. Systematic Botany, 32, 478-485.
Gilbert G.S. & Webb C.O. (2007). Phylogenetic signal in plant pathogen-host range.
Proceedings of the National Academy of Sciences of the United States of America,
104, 4979-4983.
Gossner M.M., Chao A., Bailey R.I. & Prinzing A. (2009). Native Fauna on Exotic
Trees: Phylogenetic Conservatism and Geographic Contingency in Two Lineages
of Phytophages on Two Lineages of Trees. American Naturalist, 173, 599-614.
93 Hawkes C.V. (2007). Are invaders moving targets? The generality and persistence of
advantages in size, reproduction, and enemy release in invasive plant species with
time since introduction. American Naturalist, 170, 832-843.
Hawthorn W.D. & Hayne P.D. (1978). Seed production and predispersal seed predation
in the biennial composite species, Arctium minus (Hill) Bernh. and A. lappa L.
Oecologia, 34, 283-295.
Hill S.B. & Kotanen P.M. (2009). Evidence that phylogenetically novel non-indigenous
plants experience less herbivory. Oecologia, in-press.
Holt R.D. (1977). Predation, Apparent Competition, and Structure of Prey Communities.
Theoretical Population Biology, 12, 197-229.
Holt R.D. & Lawton J.H. (1994). The Ecological Consequences of Shared Natural
Enemies. Annual Review of Ecology and Systematics, 25, 495-520.
Janz N. & Nylin S. (1998). Butterflies and plants: A phylogenetic study. Evolution, 52,
486-502.
Joshi J. & Vrieling K. (2005). The enemy release and EICA hypothesis revisited:
incorporating the fundamental difference between specialist and generalist
herbivores. Ecology Letters, 8, 704-714.
Keane R.M. & Crawley M.J. (2002). Exotic plant invasions and the enemy release
hypothesis. Trends in Ecology & Evolution, 17, 164-170.
Kim K.J., Choi K.S. & Jansen R.K. (2005). Two cholorplast DNA inversions originated
simultaneously during the early evolution of the sunflower family (Asteraceae).
Molecular Biology and Evolution, 22, 1783-1792.
94 Levine J.M., Adler P.B. & Yelenik S.G. (2004). A meta-analysis of biotic resistance to
exotic plant invasions. Ecology Letters, 7, 975-989.
Lewinsohn T.M., Novotny V. & Basset Y. (2005). Insects on plants: diversity of
herbivore assemblages revisited. Annual Review of Ecology and Systematics, 36,
597-620.
Liu H. & Stiling P. (2006). Testing the enemy release hypothesis: a review and meta-
analysis. Biological Invasions, 8, 1535-1545.
Louda S.M. & O'Brien C.W. (2002). Unexpected ecological effects of distributing the
exotic weevil, Larinus planus (F.), for the biological control of Canada thistle.
Conservation Biology, 16, 717-727.
Mack R.N. (1996). Predicting the identity and fate of plant invaders: Emergent and
emerging approaches. Biological Conservation, 78, 107-121.
Maron J.L. & Vila M. (2001). When do herbivores affect plant invasion? Evidence for
the natural enemies and biotic resistance hypotheses. Oikos, 95, 361-373.
Mason P.G. & Huber J.T. (2002). Biological control programmes in Canada 1981-2000.
CABI Publishing, Ottawa, ON, Canada.
Mitchell C.E., Agrawal A.A., Bever J.D., Gilbert G.S., Hufbauer R.A., Klironomos J.N.,
Maron J.L., Morris W.F., Parker I.M., Power A.G., Seabloom E.W., Torchin M.E.
& Vazquez D.P. (2006). Biotic interactions and plant invasions. Ecology Letters,
9, 726-740.
Noyes R.D. (2000). Biogeographical and evolutionary insights on Erigeron and allies
(Asteraceae) from ITS sequence data. Plant Systematics and Evolution, 220, 93-
114.
95 Odegaard F., Diserud O.H. & Ostbye K. (2005). The importance of plant relatedness for
host utilization among phytophagous insects. Ecology Letters, 8, 612-617.
Parker I.M. & Gilbert G.S. (2007). When there is no escape: The effects of natural
enemies on native, invasive, and noninvasive plants. Ecology, 88, 1210-1224.
R Development Core Team (2006) R: A language and environment for statistical
computing. R Foundation for Statistical Computing, Vienna, Austria ISBN 3-
900051-07-0, URL http://wwwR-projectorg
Rejmanek M. & Richardson D.M. (1996). What attributes make some plant species more
invasive? Ecology, 77, 1655-1661.
Richardson D.M., Allsopp N., D'Antonio C., Milton S.J. & Rejmanek M. (2000). Plant
invasions - the role of mutualisms. Biological Reviews, 75, 65-93.
SAS. (2002). JMP. In. SAS Institute Cary, North Carolina, USA.
Semple J.C. & Cook R.E. (2006). Solidago. In: Flora of North America (ed. Committee
FNAE). Oxford University Press Oxford, pp. 107-166.
Silvertown J. (2004). Plant coexistence and the niche. Trends in Ecology & Evolution, 19,
605-611.
Strauss S.Y., Webb C.O. & Salamin N. (2006). Exotic taxa less related to native species
are more invasive. Proceedings of the National Academy of Sciences of the United
States of America, 103, 5841-5845.
Strong D.R., Lawton J.H. & Southwood R. (1984). Insects on plants. Harvard University
Press, Cambridge, Mass., USA.
Torchin M.E., Lafferty K.D., Dobson A.P., McKenzie V.J. & Kuris A.M. (2003).
Introduced species and their missing parasites. Nature, 421, 628-630.
96 Vamosi S.M., Heard S.B., Vamosi J.C. & Webb C.O. (2009). Emerging patterns in the
comparative analysis of phylogenetic community structure. Molecular Ecology,
18, 572-592.
Webb C.O., Ackerly D.D., McPeek M.A. & Donoghue M.J. (2002). Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475-505.
Webb C.O. & Donoghue M.J. (2005). Phylomatic: tree assembly for applied
phylogenetics. Molecular Ecology Notes, 5, 181-183.
Weiblen G.D., Webb C.O., Novotny V., Basset Y. & Miller S.E. (2006). Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62-
S75.
Whitton J., Wallace R.S. & Jansen R.K. (1995). Phylogenetic-relationships and patterns
of character change in the tribe Lactuceae (Asteraceae) based on chrloroplast
DNA restriction site variation. Canadian Journal of Botany-Revue Canadienne
De Botanique, 73, 1058-1073.
97 Table 4.1 List of native and exotic species in the Asteraceae present at KSR, used for surveys of pre-dispersal seed predation. Where species were sampled from only one location, standard errors of the mean proportion of capitula damaged are reported as NA
Mean Proportion SE Proportion Species Origin Capitula Capitula damaged damaged Achillea millefolium L. Native 0.826 0.078 Anaphalis margaritacea (L.) Benth. & Hook. Native 0.029 0.029 Antennaria neglecta Greene Native 0.685 0.057 Arctium minus (Hill) Bernh. Exotic 0.890 0.030 Aster cordifolius L. Native 0.211 0.086 Aster novae-angliae L. Native 0.315 0.023 Bidens cernua L. Native 0.000 NA Bidens frondosa L. Native 0.020 0.000 Centaurea jacea L. Exotic 0.112 0.068 Chrysanthemum leucanthemum L. Exotic 0.480 0.080 Cichorium intybus L. Exotic 0.000 NA Cirsium arvense (L.) Scop. Exotic 0.169 0.028 Conyza canadensis (L.) Cron. Native 0.067 0.027 Crepis capillaris (L.) Wallr. Exotic 0.020 NA Erigeron annuus (L.) Pers. Native 0.146 0.146 Erigeron strigosus Muhl. Native 0.219 0.054 Eupatorium maculatum L. Native 0.070 0.070 Eupatorium perfoliatum L. Native 0.000 0.000 Euthamia graminifolia (L.) Salisb. Native 0.176 0.033 Galinsoga quadriradiata Ruiz & Pavon Exotic 0.000 NA Gnaphalium obtusifolium L. Native 0.000 NA Hieracium caespitosum L. Exotic 0.050 0.050 Inula helenium L. Exotic 0.075 0.025 Lactuca serriola L. Exotic 0.110 0.090 Prenanthes altissima L. Native 0.153 0.134 Rudbeckia hirta L. Native 0.477 0.150 Senecio aureus L. Native 0.153 NA Solidago caesia L. Native 0.173 0.077 Solidago canadensis L. Native 0.100 0.060 Solidago flexicaulis L. Native 0.201 0.013 Solidago nemoralis Ait. Native 0.113 0.035 Sonchus arvensis L. Exotic 0.214 0.039 Sonchus asper (L.) Hill Exotic 0.026 0.026 Taraxacum officinalis Weber Exotic 0.039 0.025 Tragopogon dubius Scop. Exotic 0.008 0.008
98 Table 4.2 Proportion of capitula damaged, both non-corrected, and phylogenetically corrected, compared using a two-way ANOVA including terms for origin (native or exotic) and sampling time (May-August), and (September-October). Proportion of damage was transformed using associated Z-scores for p + 0.001, where p is proportion of capitular damaged. A one-tailed t-test was also used to evaluate the effect of origin in the corrected example, as the minimum adequate model included only this term
No Phylogenetic Correction Phylogenetic Correction Model Parameters df F p F p Origin 1,31 0.5508 0.464 3.795 0.061† Season 1,31 3.5106 0.07 1.546 0.223 Origin X Season 1,31 1.6054 0.215 0.704 0.408 †corresponding one tailed t-test = -1.7, df = 20.1, p = 0.04
99 Table 4.3 Multiple regression of the relationship between Z-transformed proportion of capitula damage and significant phylogenetic eigenvectors. The residuals from this multiple regression were used as the phylogenetically corrected estimates of damage
(Table 4.2). The overall variance explained by the five phylogenetic eigenvectors was r2
= 0.59. Numbers in square brackets for eigenvectors represent the rank among the 34 eigenvectors that were used to partition phylogenetic distances; lower numbers represent eigenvectors that explain broader patterns of phylogenetically structured variation
Term Estimate SE t p Intercept -1.297 0.117 -11.042 <0.001 eigenvector [2] 0.0185 0.006 2.985 0.006 eigenvector [3] -0.031 0.009 -3.629 0.001 eigenvector [8] 0.054 0.017 3.134 0.004 eigenvector [19] -0.095 0.031 -3.028 0.005 eigenvector [20] 0.118 0.034 3.492 0.002
100 Table 4.4 Regression models for Z-score transformed proportion of capitula damaged, versus four measures of phylogenetic novelty among Asteraceae: mean distance to all
Asteraceae regardless of origin (All Asteraceae), closest Asteraceae regardless of origin
(closest Asteraceae), mean distance to native Asteraceae (All native Asteraceae), and closest native Asteraceae
Phylogenetic Distance df slope SE F p All Asteraceae 1,12 2.662 1.724 2.386 0.148 Closest Asteraceae 1,12 -0.072 0.667 0.011 0.916 All native Asteraceae 1,12 1.551 1.365 1.290 0.278 Closest native Asteraceae 1,12 0.577 0.532 1.177 0.299
101 Figure 4.1 Community phylogeny for species in the Asteraceae that were used in this study. Bars to the right of species names are the summed eigenvector loadings from the five eigenvectors that were used to explain the phylogenetic component of capitular damage
102 Figure 4.2 Z-score transformed mean proportion of capitulum damage (± SE) between exotic and native species
103 Figure 4.3 Phylogenetically-corrected Z-transformed proportion capitulum damage for native and exotic species in the Asteraceae grouped by sampling time at KSR. Exotics experience lower pre-dispersal seed predation than native species (see Table 4.2 for statistics)
104 Figure 4.4 Z-transformed proportion of capitula damage and phylogenetic novelty of exotic Asteraceae at KSR, based on mean distance to all Asteraceae, distance to closest
Asteraceae, mean distance to all native Asteraceae, and mean distance to closest native
Asteraceae
105 Chapter 5
Biotic constraints experienced by a novel invader:
close relatives have little impact on herbivore damage and competition
Abstract
The success of non-indigenous (exotic) species may be influenced by biotic interactions that occur during the initial stages of invasion. For example, natural enemies that are likely to be shared among close relatives may make habitats with native relatives more resistant to invasion. In this study, I predicted that habitats with close relatives would be more resistant to the novel exotic Solidago virgaurea. I planted S. virgaurea into habitats with close relatives (Solidago cf canadensis, and mixed Solidago sites) and without close relatives (grass dominated sites). Habitat type was also crossed with a disturbance treatment, where I removed adjacent vegetation. Overall, community type did not affect levels of damage or plant survival. Removal of adjacent competitors resulted in a significant decline in damage, and significant increase in survival. There was also a significant interaction between community type and adjacent competitor removals, as decreases in damage and increases in survival were largest for plants in grass habitats.
When adjacent vegetation was removed, damage tended to have a higher invertebrate component than in the intact treatment. Damage caused by a rust fungus tended to be higher in Solidago canadensis habitats. Overall, the absence of evidence for more severe herbivory and competition in communities dominated by native congenerics suggests that the presence of close relatives is unlikely to make communities more resistant to invasion by the study species.
106 Introduction
Almost all ecosystems on Earth have been affected by non-indigenous (exotic) species
(Mack et al. 2000; Pimentel et al. 2000; Crall et al. 2006). Despite their ubiquitous nature, predicting which exotic species will succeed and become invasive remains difficult. In part, this may be the consequence of the lack of data regarding invasion dynamics during the initial stages of the invasion process (Mack et al. 2000). During this stage, invaders may go undetected as populations are small, making them relatively inconspicuous and innocuous (Elton 1958; Williamson 1996). However, biotic interactions with the existing flora and fauna during early invasion may be a critical factor influencing the long-term success of an invader, and its subsequent invasiveness
(Levine et al. 2004; Mitchell et al. 2006). Therefore, understanding the impact that biotic interactions have on exotic species during the very early stage of invasion is critical for managing biological invasions.
The Enemy Release Hypothesis (ERH) is a prominent idea that is often invoked to explain the success of exotic species (Keane & Crawley 2002; Maron & Vila 2001;
Torchin et al. 2003; Colautti et al. 2004). This hypothesis predicts that exotic species lose specialist enemies when introduced to new areas, giving them a competitive advantage over native species (Keane & Crawley 2002; Torchin et al. 2003). Support for the ERH depends on the type of 'release' that is studied: biogeographical comparisons for the same species in its invaded and native ranges tend to support the idea (Keane & Crawley 2002;
Torchin et al. 2003; Colautti et al. 2004); however, results from studies that compare enemy escape between invading species and native competitors are equivocal (Agrawal
107 & Kotanen 2003; Colautti et al. 2004; Parker & Gilbert 2007). In some cases, invaders can experience higher rates of damage than closely related native species (Agrawal et al.
2005). Since enemies (invertebrate herbivores and pathogens) are often shared among close relatives (Odegaard et al. 2005; Weiblen et al. 2006; Gilbert & Webb 2007), some of the interspecific variability in escape could be associated with an increased likelihood of interacting with enemies associated with close relatives (Strong et al. 1984; Lewinsohn et al. 2005; Dawson et al. 2009); a phenomenon that may be less likely when close relatives are absent.
Other biotic interactions that are conserved among close relatives may also play an important role in invasion success (Darwin 1859; Mack 1996; Webb et al. 2002). Indeed, the idea that competition is more intense among close relatives and influences community assembly was predicted by Darwin (1859), and is a fundamental principle of community phylogenetics (Webb et al. 2002). A recent test of this prediction using a meta-analysis of phytometer studies found only a weak correlation between competitive interactions and phylogenetic distance (Cahill et al. 2008). It is unknown how these results translate into patterns of coexistence in real communities as interspecific interactions will undoubtedly be more complex than those represented by the experimental conditions used in the studies included by Cahill et al. (2008).
That exotic species experience biotic constraints in their invaded range is consistent with the idea that natural communities are inherently resistant to invasion (i.e. biotic resistance) (Elton 1958; Mack 1996; Levine et al. 2004). However, the degree to which
108 biotic interactions act to resist invasion by novel species can be strongly influenced by disturbance (Levine et al. 2004). Disturbance frees up space and resources, and potentially interacts with naturally occurring biotic constraints; thus communities that, after disturbance, maintain their effectiveness at repelling invaders should be less likely to be colonized by exotic species. Disturbance is expected to reduce the effectiveness of natural barriers that are localized such as competition, but may have less influence on constraints that are more mobile such as natural enemies (Herzig & Root 1996). As suggested above, community assembly can be influenced by the presence of close relatives (Webb et al. 2002). Therefore, in communities where close relatives are present, disturbance is expected to have more of an impact on biotic resistance imposed by enemies than when those relatives are absent.
Here I tested the prediction that communities with close relatives (congeneric native species) compared to those without would be more resistant to invasion by a novel exotic,
Solidago virgaurea L. I also predicted that this effect would be reduced when intact communities were disturbed by removing adjacent competitors. Specifically, I tested the hypotheses that S. virgaurea would experience higher damage and lower survival (i) in communities with close relatives, (ii) in intact versus disturbed communities, and (iii) that there would be a community by disturbance interaction, such that effects would be strongest in communities where close relatives are present. I found support only for the second hypothesis: S. virgaurea experienced higher damage and lower survival in intact versus disturbed treatments. There was no statistical difference in damage or survival among community types, and although the interaction between community type and
109 disturbance was significant, the strongest effects were observed in grass communities, not communities where close relatives were present.
Materials and Methods
Study site
This study was conducted at the University of Toronto's Koffler Scientific Reserve
(KSR) at Jokers Hill, Regional Municipality of York, in southern Ontario, Canada (44º02'
N, 79º31' W, 300m ASL). This 350-ha site lies within the Oak Ridges Moraine, and is dominated by prominent hills with a thin organic layer over deep glacial sands.
Vegetation is a mixture of old fields supporting a diverse range of native and exotic plants, hardwood (maple-beech-hemlock) forest, and conifer plantations. Further information on this site may be found at http://www.ksr.utoronto.ca.
Study Species
The genus Solidago is a North American group in the subtribe Solidagininae, family
Asteraceae (Beck et al. 2004). The only non-North American goldenrod is Solidago virgaurea L., which is a perennial native to Europe and Asia (Jobin et al. 1996). In its native range S. virgaurea grows in many habitats including rocky outcrops, disturbed areas, and old field meadows (Davis et al. 2000). The species is not yet naturalized in
North America, but can be readily purchased as an ornamental or for herbal remedies at horticultural stores. North American Solidago species such as Solidago gigantea Ait. and
Solidago altissima L. are successful invaders throughout much of Europe (Jakobs et al.
2004), and can co-occur regionally with the native S. virgaurea (Jobin et al. 1996).
110 Therefore, I anticipated that S. virgaurea would have the ecological amplitude to grow along side North American Solidago species at this study site. The two representative species that I studied were Solidago cf canadensis L. (including Solidago canadensis and
Solidago altissima, hereafter S. canadensis) and Solidago nemoralis Ait. At KSR, S. canadensis is dominant in mesic to seasonally wet old-field meadows, whereas S. nemoralis tends to grow in slightly drier sites, usually mixed with S. canadensis.
Experimental Design
In spring 2006, I established 15 study locations in five different meadows for each of the three habitat types dominated by grasses (primarily Bromus inermis and Poa pratensis), dominated by S. canadensis, or mixed S. canadensis and S. nemoralis. Within each study location, I planted 20 individual S. virgaurea seedlings into plots where the adjacent vegetation was randomly chosen to be left intact or removed. Experimental plants were spaced at least two metres apart. Vegetation removal included digging up all the plant biomass in an approximately 1m diameter circle around the focal plant. I replaced as much of the organic soil horizon as possible to the plot. Plots were weeded on a weekly basis throughout the growing season during the experiment.
During June, July, August 2006, and June 2007, I determined whether each plant survived and if so, its amount of damage. At each sample time, damage was measured in two ways: the proportion of leaves with damage, and the proportion of leaf area damaged grouped into one of twenty 5% damage classes. These damage measures were then multiplied to produce an index of the fraction of total plant damage:
111 Dtotal = Dl × Da
Where Dtotal is the fraction of total plant damage, Dl is the proportion of leaves damaged, € and Da is the mean proportion of leaf area damage for all leaves on a given plant. The
€ index varies between 0 and 1, where 0 indicates€ that leaves had no damage, and 1
€ indicates that all leaf area had been consumed. For the statistical analyses (see below) I
calculated the mean damage from the four sample times as an integrated index of overall
herbivore pressure.
I also evaluated the relative types of damage on each plant. This included scoring the
proportion of damage that was contributed by vertebrates, invertebrates, fungi, and
unknown discoloration or senescence which may include pathogen damage. Plants where
the leaves had been clipped at the base of the petiole were assumed to be damaged by
vertebrates; this type of damage was typically combined with fecal pellets from voles,
mice, or rabbits. Leaves that had been clipped, but remained below the plant were
assumed to have been caused by slugs or snails. Missing leaf tissue either at or within the
leaf margin was assumed to be caused by insects. Any damage in the form of
discoloration or senesced tissue that could not be categorized was scored as unknown
(and is presented as such in the analyses).
After completing the experiment, I measured the density of native Solidago species
occurring within two metres of each focal plant; in the biomass removal treatment, I
avoided the disturbed areas. Only S. canadensis and Solidago nemoralis were present
near the plots; other close relatives present in the surrounding meadows included
112 Solidago rugosa and Euthamia graminifolia, as well as other species in the Asteraceae.
For statistical analyses, I used the natural logarithm of the combined density for S. canadensis and S. nemoralis, to better meet statistical assumptions. As mentioned above, the mean damage index for an individual plant across all sample period that it survived was used as an index of the damage that each plant experienced.
Statistical Analysis
Herbivory
I used split-plot factorial analysis of variance to analyze the damage data among habitats and vegetation removal treatments (Kirk 1982); the model included terms for habitat type
(n = 3), meadow nested in habitat type (n = 5), vegetation removal (n = 2), and the habitat type by vegetation removal interaction. Habitat type, vegetation removal, and their interaction were fixed effects, block nested within habitat was a random effect. In this statistical model, the F-ratio for habitat type uses the mean square error of the block nested in habitat as the denominator, and the interaction of block nested in habitat with vegetation removal is used in the denominator of the F-ratio for vegetation removal, and the habitat by vegetation removal interaction (Kirk 1982). Although I had replication for the treatment represented by meadow (nested in habitat type) crossed with vegetation removal, I used the mean damage within each site by vegetation removal treatment for statistical analyses. This was done to achieve a more accurate measure of habitat effects, and maintain a balanced experimental design. Damage measurements were also Z-score transformed to linearize the data and better meet statistical assumptions (Crawley 2007).
113 To determine if the density of native Solidago species influenced rates of damage, I used an analysis of covariance model; terms included estimates for vegetation removal treatment, density of native Solidago species adjacent to experimental plants, and the interaction between the two.
Damage type
Since damage type for each plant was measured as the percentage contribution of vertebrate, invertebrate, fungi, or unknown (i.e. discoloured or senesced), the measures were highly correlated with one another (Table 5.4). To remove this correlation, I treated damage type as a multivariate response using principal component analysis. Two principal components were chosen based on their correlation with the different damage types: the first for vertebrate and invertebrate damage, and the third for fungal damage
(Table 4). Since the principal component axes were orthogonal, I used them as separate measures of their respective damage types. The same statistical model outlined above was used with the resulting eigenvectors to determine if damage by vertebrates, invertebrates, and fungi differed among habitat and vegetation removal treatments.
Mortality
To determine if survival of S. virgaurea differed among habitat and disturbance treatments, I used the proportion of plants surviving in each meadow x vegetation removal combination at the end of the experiment (June 2007). The same split-plot factorial analysis of variance was used as outlined above. Proportion survival
114 measurements were Z-score transformed to linearize the data and better meet the assumptions of the statistical test.
All statistical analyses were done in JMP v.5 (SAS 2002).
Results
Density of native Solidago species
The density of native Solidago species (S. canadensis and S. nemoralis) was significantly different among community types (ANOVA, p < 0.05). Communities with only S. canadensis had the highest density (of 33.92 stems/m2), whereas grass dominated communities had the lowest (1.04 stems/m2), and Solidago mixed communities were intermediate (8.17 stems/m2).
Foliar Damage
Among adjacent vegetation removal treatments within community types, proportion leaf area damaged experienced by S. virgaurea ranged from 0.24 ± 0.05 to 0.68 ± 0.08 (Table
5.1). Damage was similar among habitat types (Table 5.2, Fig. 5.1). Over all communities, damage was significantly higher when adjacent vegetation was not disturbed (Table 5.2, Fig. 5.1). However, this effect was not consistent across communities as the interaction term was significant (p = 0.003) (Fig. 5.1). Damage decreased the most between intact and disturbed treatments in grass communities where this effect was significant (Fig. 5.1). Damage also declined in both Solidago habitats, but not significantly (Fig. 5.1). The nature of this statistical interaction resulted in individual
115 treatment means that had relatively complicated patterns of statistical significance (Fig.
5.1). Patterns that are important for testing the hypotheses included: damage being significantly higher in intact communities for grass and S. canadensis compared to
Solidago mixed communities provided partial support for the hypothesis that damage would be higher in intact communities, and no significant differences in damage among communities when they were disturbed, which rejected the hypothesis that biotic constraints would be higher when close relatives were present (Fig. 5.1)
When damage was regressed against the mean density of native Solidago spp. adjacent to experimental plants I found opposite trends depending on whether the adjacent vegetation had been removed (Fig. 5.2). Overall, damage was different depending on adjacent vegetation removal (F1,26 = 16.445, p < 0.001), with no significant relationship of native
Solidago density (F1,26 = 0.632, p = 0.434), but the interaction between vegetation removal and Solidago density was significant (F1,26 = 5.0717, p = 0.033). Slopes for the relationship between damage and native Solidago density were significant, both when adjacent vegetation was left intact (t = -2.205, df = 32, p = 0.037) and when it was removed (t = 2.252, df = 32, p = 0.033): damage declined as Solidago density increased when adjacent vegetation was left intact, whereas damage increased as Solidago density increased when adjacent vegetation was removed (Fig. 5.2).
116 Foliar Damage Type
The principal component analysis resulted in two axes explaining the different types of foliar damage, which together explained 87% of the variance in damage type (Fig. 5.3).
The first represented a gradient of vertebrate vs. invertebrate damage and explained 83% of the variance in damage type: lower eigenvector loadings were associated with vertebrate damage, whereas higher values were associated with invertebrate damage
(Table 4). The second axis was used to represent damage caused by fungi, and explained
4% of the overall variability in damage type. For this axis, eigenvector loadings were positively correlated with fungal damage (Table 4).
For the first principal component axis (herbivory vs. invertebrate damage), there was no statistical effect of habitat type, indicating that vertebrate and invertebrate damage was similar regardless of community composition (Table 5.2, Fig. 5.3). For the same axis, over all communities, eigenvector scores were statistically different between intact and disturbed treatments (Table 5.2, Fig. 5.3); the resulting pattern indicated that vertebrate damage tended to be higher in intact communities, and invertebrate damage increased in importance when communities were disturbed (Fig. 5.3). However, the interaction of community type and vegetation removal was significant (Fig. 5.3): the interaction was primarily the effect of disturbance altering the type of damage more in grass habitats than in both the Solidago habitats (Fig. 5.3). Among the habitat x treatment combinations for axis 1, the mean eigenvector score for the intact grass treatment was significantly lower
117 (p < 0.05) than all but the intact Solidago canadensis habitat. This reflected the high rate of vertebrate herbivory when grass and Solidago canadensis communities were intact.
Eigenvector scores for the second principal component axis were significantly different among community types, but not between adjacent plant removal treatments, or among individual habitat and vegetation removal treatments (Table 5.2). The significant habitat effect was the result of a higher mean eigenvector scores in S. canadensis-dominated habitats (Fig. 5.3); the mean for which was higher than in grass-dominated and Solidago mixed communities (p < 0.05). Since the second principal component axis was associated with damage caused by rust fungus, the difference in eigenvector scores reflects a generally higher rate of this type of damage in Solidago canadensis habitats.
Survival
The proportion of plants that survived from June 2006 to June 2007 was similar among habitats (Table 5.2, Fig. 5.4). Within habitats, the removal of adjacent vegetation resulted in significantly higher survival (Table 5.2). However, the effect on survival of removing adjacent vegetation was different among habitats (Table 5.2): increases in survival were highest in grass habitats, intermediate in S. canadensis habitats, and lowest in Solidago mixed habitats (Fig. 5.4). However, only plants in grass habitats realized a significant increase in survival (p < 0.05). Despite this higher survival differential, the mean survival of plants among the disturbed communities was not statistically different (p > 0.05; Fig.
5.4).
118 Discussion
The success of exotic species in part depends on biotic interactions with co-occurring native species in the introduced range (Darwin 1859; Mack 1996; Rejmanek 1996;
Daehler 2001). Using S. virgaurea to represent a novel invader, I found that intact communities, regardless of composition, provided a higher biotic constraint: damage was higher and survival lower when adjacent vegetation was left intact versus when it was removed. The reduction in damage and increase in survival between intact versus disturbed grass communities were significant, but this was not the case for communities where close relatives were present. This significant effect suggests that biotic resistance was reduced the most when grass was the dominant competitor. However, these trends resulted in damage and survival that were similar among community types in the disturbed treatment, suggesting that biotic constraints are likely to be equally ineffective regardless of the presence of close relatives in disturbed communities.
Biotic constraints and enemy escape
Overall, natural biotic constraints were strongest when communities were intact (i.e. damage was higher and survival lower); this was likely the combined impact of herbivory by small mammals and competition from pre-existing vegetation. Although I did not positively identify the agent of herbivory, meadow voles (Microtus pennsylvanicus) are a common generalist herbivore at KSR and were the likely source. Previous studies have shown that meadow voles can have large community-wide impacts on vegetation composition, especially during peak abundance years, and in grassland and edge habitats
(Manson et al. 2001; Kauffman & Maron 2006). Given that these herbivores have a broad
119 diet (Belanger & Bergeron 1987) and their foraging behaviour may be influenced more by the presence of intact vegetation cover than vegetation composition (Moenting &
Morris 2006), it is not surprising that the presence or absence of close relatives in intact communities had little impact on damage or survival of S. virgaurea. The large reduction in damage on plants between intact and disturbed treatments in the grass community is likely the result of changes in the behaviour of these herbivores, as small disturbances can alter the foraging pattern of voles (Moenting & Morris 2006). Recent work has suggested that biotic constraints experienced during invasion can be regulated more strongly by generalist than specialist enemies (Parker & Hay 2005); reducing the impact of generalists, then may result in communities being easier to invade. These results are consistent with this idea, as biotic constraints imposed by the generalist herbivores were much more effective when communities were intact versus disturbed.
When communities were left intact, the negative impact of vertebrate herbivores was very likely compounded by competition from resident vegetation. Competition may directly reduce the success of S. virgaurea; as well, plants suffering from interspecific competition may be less able to tolerate herbivore damage (Strauss & Agrawal 1999). By the end of the 2006 growing season, surrounding vegetation in the grass and S. canadensis communities had 100% ground cover (personal observation). Although I did not directly measure the impact that percent cover had on S. virgaurea, the experimental plants in the intact treatments were often overtopped by undisturbed vegetation in grass and S. canadensis communities; this would have resulted in significant light competition and very likely reduced the ability of S. virgaurea to survive. In other experiments, the
120 removal of the resident vegetation has been shown to significantly improve establishment and vigour of individual invaders (Levine et al. 2004). Among these studies, however, the degree to which competition repels invaders was less clear (Levine et al. 2004). The results here indicate that the combined effect of herbivory and competition may be effective at repelling S. virgaurea: in two S. canadensis meadows and one grass meadow there were no surviving plants (out of 10 individuals initially planted) where resident vegetation was left intact, despite between 4 to 6 plants surviving when resident vegetation was removed in the same sites.
The relatively low rates of damage caused by invertebrates across habitats and vegetation removal treatments suggest that biotic resistance relating to this group of enemies is unlikely to play a role in habitat specific invasion success of S. virgaurea. Of the invertebrate herbivores observed on S. virgaurea, Microrhopala vittata and Trirhabda canadensis are known to specialize on Solidago spp. (Maddox & Root 1987; Carson &
Root 1999; Stireman et al. 2005). Experimental work has shown that both of these beetles can reduce the abundance of their native hosts, particularly in outbreak years (Carson &
Root 1999). Establishment of S. virgaurea may then be reduced by damage caused by these beetles when they are at high abundance, but since they were observed on plants in all communities, it is unlikely that their impact will be limited to plants in Solidago communities. In part, colonization of S. virgaurea by herbivores when native hosts were absent may be the result of this experiment being conducted on a relatively small spatial scale: in some cases, study sites were less than 300m apart and small patches of S. canadensis were present in all of the grass communities. The mobility of M. vittata and T.
121 canadensis would easily overcome such distances (Herzig & Root 1996). The spillover of these herbivores from resident Solidago spp. may form a considerable biotic constraint, even in communities where the native hosts are locally absent or at low abundance.
Evaluating the importance of interactions such as these would require a spatially- structured experimental design to determine how proximity and density of native hosts influences colonization of novel exotic species (e.g., MacKay & Kotanen 2008).
Damage caused by fungal infection was the result of an unidentified species of rust, which caused only a minor percentage of overall damage. Despite the low rate of damage, differences did occur among communities: infection was highest on plants in S. canadensis communities. Since S. canadensis seemed to be the host species for the rust, the rate at which the rust infected S. virgaurea related directly to the proximity of host plants. Depending on the impacts of the rust, the higher rates of infection in S. canadensis habitats may offer a potential biotic constraint.
Implications for DNH
"Darwin's Naturalization Hypothesis" predicts that invaders are likely to be less successful where close relatives are present (Darwin 1859). Although it usually is assumed that the mechanism for this pattern is increased competition among close relatives, shared enemies can also negatively affect the co-occurrence of related species
(Holt 1977; Mitchell et al. 2006). Results from this study suggest that the presence or absence of close relatives is unlikely to systematically affect biotic constraints such that
S. virgaurea would be more successful when close relatives were absent. The lack of a
122 herbivore effect seems to be a result of the combined influence of high rates of damage by generalist herbivores when communities were intact, and in part, damage caused by
Solidago spp. specialists in all communities when the adjacent vegetation was removed.
The lack of a competition effect indicates both grasses and native Solidago spp. have functionally similar effects. The only interaction that may result in community-specific biotic constraint was higher damage by an unidentified rust fungus in S. canadensis communities. However, damage was slight, and I anticipate that the community-specific nature of this interaction would be short lived given the dispersal ability of fungal spores.
Together, these results suggest that patterns predicted by DNH would not occur, at least at the scale of this experiment (~200 ha).
References
Agrawal A.A. & Kotanen P.M. (2003). Herbivores and the success of exotic plants: a
phylogenetically controlled experiment. Ecology Letters, 6, 712-715.
Agrawal A.A., Kotanen P.M., Mitchell C.E., Power A.G., Godsoe W. & Klironomos J.
(2005). Enemy release? An experiment with congeneric plant pairs and diverse
above- and belowground enemies. Ecology, 86, 2979-2989.
Beck J.B., Nesom G.L., Calie P.J., Baird G.I., Small R.L. & Schilling E. (2004). Is
subtribe Solidagininae (Asteraceae) monophyletic? Taxon, 53, 691-698.
Belanger J. & Bergeron J.M. (1987). Can alkaloids and tannins in plants influence food
choice of meadow voles (Microtus pennsylvanicus). Canadian Journal of
Zoology-Revue Canadienne De Zoologie, 65, 375-378.
123 Cahill J.F., Kembel S.W., Lamb E.G. & Keddy P.A. (2008). Does phylogenetic
relatedness influence the strength of competition among vascular plants?
Perspectives in Plant Ecology Evolution and Systematics, 10, 41-50.
Carson W.P. & Root R.B. (1999). Top-down effects of insect herbivores during early
succession: influence on biomass and plant dominance. Oecologia, 121, 260-272.
Colautti R.I., Ricciardi A., Grigorovich I.A. & MacIsaac H.J. (2004). Is invasion success
explained by the enemy release hypothesis? Ecology Letters, 7, 721-733.
Crall A.W., Meyerson L.A., Stohlgren T.J., Jarnevich C.S., Newman G.J. & Graham J.
(2006). Show me the numbers: what data currently exist for non-native species in
the USA? Frontiers in Ecology and the Environment, 4, 414-418.
Crawley M.J. (2007). The R book. John Wiley, New York.
Daehler C.C. (2001). Darwin's naturalization hypothesis revisited. American Naturalist,
158, 324-330.
Darwin C. (1859). On the origin of species by means of natural selection. Murray,
London, UK.
Davis M.A., Grime J.P. & Thompson K. (2000). Fluctuating resources in plant
communities: a general theory of invasibility. Journal of Ecology, 88, 528-534.
Dawson W., Burslem D. & Hulme P.E. (2009). Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions, 15, 141-147.
Elton C.S. (1958). The ecology of invasions by animals and plants. Methuen, London,
UK.
124 Gilbert G.S. & Webb C.O. (2007). Phylogenetic signal in plant pathogen-host range.
Proceedings of the National Academy of Sciences of the United States of America,
104, 4979-4983.
Herzig A.L. & Root R.B. (1996). Colonization of host patches following long-distance
dispersal by a goldenrod beetle, Trirhabda virgata. Ecological Entomology, 21,
344-351.
Holt R.D. (1977). Predation, apparent competition, and structure of prey communities.
Theoretical Population Biology, 12, 197-229.
Jakobs G., Weber E. & Edwards P.J. (2004). Introduced plants of the invasive Solidago
gigantea (Asteraceae) are larger and grow denser than conspecifics in the native
range. Diversity and Distributions, 10, 11-19.
Jobin A., Schaffner U. & Nentwig W. (1996). The structure of the phytophagous insect
fauna on the introduced weed Solidago altissima in Switzerland. Entomologia
Experimentalis Et Applicata, 79, 33-42.
Kauffman M.J. & Maron J.L. (2006). Consumers limit the abundance and dynamics of a
perennial shrub with a seed bank. American Naturalist, 168, 454-470.
Keane R.M. & Crawley M.J. (2002). Exotic plant invasions and the enemy release
hypothesis. Trends in Ecology & Evolution, 17, 164-170.
Kirk R.E. (1982). Experimental design: Procedures for the behavioural sciences. Second
ed. Brooks and Cole, Monterey.
Levine J.M., Adler P.B. & Yelenik S.G. (2004). A meta-analysis of biotic resistance to
exotic plant invasions. Ecology Letters, 7, 975-989.
125 Lewinsohn T.M., Novotny V. & Basset Y. (2005). Insects on plants: diversity of
herbivore assemblages revisited. Annual Review of Ecology and Systematics, 36,
597-620.
Mack R.N. (1996). Predicting the identity and fate of plant invaders: Emergent and
emerging approaches. Biological Conservation, 78, 107-121.
Mack R.N., Simberloff D., Lonsdale W.M., Evans H., Clout M. & Bazzaz F.A. (2000).
Biotic invasions: Causes, epidemiology, global consequences, and control.
Ecological Applications, 10, 689-710.
MacKay J. & Kotanen P.M. (2008). Local escape of an invasive plant, common ragweed
(Ambrosia artemisiifolia L.), from above-ground and below-ground enemies in its
native area. Journal of Ecology, 96, 1152-1161.
Maddox G.D. & Root R.B. (1987). Resistance to 16 diverse species of herbivorous
Insects within a population of goldenrod, Solidago altissima - genetic variation
and Heritability. Oecologia, 72, 8-14.
Manson R.H., Ostfeld R.S. & Canham C.D. (2001). Long-term effects of rodent
herbivores on tree invasion dynamics along forest-field edges. Ecology, 82, 3320-
3329.
Maron J.L. & Vila M. (2001). When do herbivores affect plant invasion? Evidence for
the natural enemies and biotic resistance hypotheses. Oikos, 95, 361-373.
Mitchell C.E., Agrawal A.A., Bever J.D., Gilbert G.S., Hufbauer R.A., Klironomos J.N.,
Maron J.L., Morris W.F., Parker I.M., Power A.G., Seabloom E.W., Torchin M.E.
& Vazquez D.P. (2006). Biotic interactions and plant invasions. Ecology Letters,
9, 726-740.
126 Moenting A.E. & Morris D.W. (2006). Disturbance and habitat use: is edge more
important than area? Oikos, 115, 23-32.
Odegaard F., Diserud O.H. & Ostbye K. (2005). The importance of plant relatedness for
host utilization among phytophagous insects. Ecology Letters, 8, 612-617.
Parker I.M. & Gilbert G.S. (2007). When there is no escape: The effects of natural
enemies on native, invasive, and noninvasive plants. Ecology, 88, 1210-1224.
Parker J.D. & Hay M.E. (2005). Biotic resistance to plant invasions? Native herbivores
prefer non-native plants. Ecology Letters, 8, 959-967.
Pimentel D., Lach L., Zuniga R. & Morrison D. (2000). Environmental and economic
costs of nonindigenous species in the United States. Bioscience, 50, 53-65.
Rejmanek M. (1996). A theory of seed plant invasiveness: The first sketch. Biological
Conservation, 78, 171-181.
SAS (2002). JMP. SAS Institute. Cary, North Carolina, USA.
Stireman J.O., Nason J.D. & Heard S.B. (2005). Host-associated genetic differentiation in
phytophagous insects: General phenomenon or isolated exceptions? Evidence
from a goldenrod-insect community. Evolution, 59, 2573-2587.
Strauss S.Y. & Agrawal A.A. (1999). The ecology and evolution of plant tolerance to
herbivory. Trends in Ecology & Evolution, 14, 179-185.
Strong D.R., Lawton J.H. & Southwood R. (1984). Insects on plants. Harvard University
Press, Cambridge, Mass., USA.
Torchin M.E., Lafferty K.D., Dobson A.P., McKenzie V.J. & Kuris A.M. (2003).
Introduced species and their missing parasites. Nature, 421, 628-630.
127 Webb C.O., Ackerly D.D., McPeek M.A. & Donoghue M.J. (2002). Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475-505.
Weiblen G.D., Webb C.O., Novotny V., Basset Y. & Miller S.E. (2006). Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62-
S75.
Williamson M. (1996). Biological Invasions. Chapman and Hall, London, UK.
128 Table 5.1 Summary of proportion survival and leaf area damage for each community
(grass = grass, scan = S. canadensis, smix = mixed Solidago) and treatment (intact = adjacent vegetation left intact, open = adjacent vegetation removed) type. Proportion survival is based on n = 50 plants for each community treatment combination. Proportion leaf area damage (±SE) is mean based on average damage for plants from each community * treatment combination.
Adjacent Proportion Proportion Leaf Area Habitat Vegetation Survival Damage (±SE) grass Intact 0.18 0.68 (0.08) grass Open 0.70 0.24 (0.04) scan Intact 0.14 0.53 (0.10) scan Open 0.40 0.36 (0.05) smix Intact 0.34 0.35 (0.04) smix Open 0.48 0.24 (0.05)
129 Table 5.2 Results of split-plot analysis of variance for damage, damage type (PC1 and
PC3), and survival of Solidago virgaurea. Habitat includes grass, Solidago canadensis, and Solidago mixed habitats. Adjacent vegetation includes vegetation being left intact or removed around the focal plant. HxV is the interaction between habitat type and vegetation removal. Damage and survival measures were Z-score transformed. PC1 and
PC3 were eigenvector loadings from a principal component analysis on damage type
(vertebrate, invertebrate, fungi, and unknown)
Damage PC1 PC3 Survival Source df F p F p F p F p Habitat 2 2.212 0.152 1.797 0.208 7.012 0.010 2.621 0.114 Adjacent Vegetation 1 54.671 0.007 16.553 0.002 0.103 0.754 29.465 0.000 HxV 2 9.602 0.003 4.594 0.033 0.386 0.688 3.715 0.056 note: all F values tested with df = 12 in the denominator
130 Table 5.3 Regression parameters from analysis of covariance for foliar damage (Z-score transformed) compared using terms for adjacent vegetation removal treatment (intact or removed), density of native Solidago species (natural logarithm transformed), and the interaction between the two
Coefficient Treatment Estimate SE t p Intercept Intact 0.196 0.131 1.491 0.148 Slope Intact -0.098 0.044 -2.205 0.037 Intercept Open -0.645 0.182 -4.618 <0.001 Slope Open 0.037 0.060 2.252 0.033
131 Table 5.4 Correlation between different types of damage on Solidago virgaurea
Invertebrate Fungi Unknown Vertebrate -0.8917 -0.4641 -0.7622 Invertebrate 0.3359 0.4197 Fungi 0.3589
132 Table 5.5 Relationship of principal component eigenvectors (PC1, PC2, PC3) and type of damage (vertebrate, invertebrate, fungi). RSS is residual sum of squares between proportion damage and the eigenvector loading for each principal component; r is the correlation coefficient, and r2 is the amount of variance explained. PC1 best explained variance in vertebrate and invertebrate damage, PC3 best explained fungal damage.
PC1 PC2 PC3 RSS r r2 RSS r r2 RSS r r2 Vertebrate 0.011 -0.999 0.996 2.780 -0.035 -0.034 2.783 0.003 -0.036 Invertebrate 0.265 0.911 0.824 1.300 -0.406 0.135 1.555 -0.034 -0.035 Fungi 0.241 0.464 0.188 0.301 0.147 -0.013 0.073 0.873 0.753
133 a
ab
bc bc
c c
Figure 5.1 Interaction plot of foliar damage (Z-score transformed) plotted against adjacent vegetation removal treatment, grouped by habitat type. Symbols are mean proportion foliar damage (Z-score transformed) for each habitat by vegetation removal treatment combination. Letters next to each symbol represent Tukey HSD test groupings, different letters indicate significant differences (p < 0.05)
134 Figure 5.2 Relationship between foliar damage (Z-score transformed) and native
Solidago density (natural logarithm transformed). Symbols represent mean damage at each site when adjacent vegetation was intact (filled symbols) and disturbed (open symbols). Lines show significant relationships (p < 0.05) between damage and Solidago density for each treatment type
135 c
ab c b
b d b a b d d d
Figure 5.3 Principal component axes that best explained vertebrate, invertebrate, and fungal damage on Solidago virgaurea. Axis 1 is a gradient of vertebrate damage (on the left) to invertebrate damage (on the right). Axis 2 is an increasing gradient of fungal damage (primarily rust fungus). Symbols represent mean eigenvector loadings where adjacent vegetation was left intact (filled symbols) and was disturbed (open symbols).
Lines through symbols are standard errors for eigenvector loadings along respective axes.
Letters (a & b) on horizontal SE lines represent mean groupings, different letters indicate significant differences along the herbivory principal component (p < 0.05); letters (c & d) on vertical SE lines represent significant differences (p < 0.05) along the fungal infection principal component
136 a
ab
abc abc
bc
c
Figure 5.4 Survival of Solidago virgaurea in different habitats when adjacent vegetation was left intact and was disturbed. Symbols represent Z-score transformed mean proportion survival in the different habitat and vegetation removal treatments. The vertical line through each symbol represents one standard error. Letters next to symbols indicate groupings of mean survival: different letters indicates means are significantly different (p < 0.05)
137 Chapter 6
Positive effects of soil biota for a novel invader:
implications for habitat- specific invasion success
when close relatives are present
Abstract
The success or failure of a non-indigenous species may reflect habitat-specific biotic interactions in its introduced range. Specifically, effects of biotic interactions on an invader might be stronger in habitats containing close relatives, which may act as reservoirs for host-specific associations. Here I evaluate the effects of soil biota on seedling growth and seed mortality of a novel exotic in habitats differing in the abundance of close native relatives. Growth of the European Solidago virgaurea was evaluated in a greenhouse experiment which I exposed seedlings to soils collected in
North American old field habitats dominated by one or more native Solidago species, or grass species. As well, seed mortality was evaluated in a field experiment in which germination of fungicide-protected seeds was compared with controls in these same habitats. Contrary to the expectation, root and shoot biomass were higher in soil inoculated with biota from habitats with native Solidago. Non-sterile soils generally improve root and shoot growth; this effect tended to be greater for plants inoculated with soil biota collected near close relatives. Seed mortality tended to be reduced by fungicide treatment, but was so great that it was difficult to determine if habitat effects were present. This contrasts with two native Solidago species, in which germination rates were much higher: for these species, germination was highest in grass-dominated habitats, and
138 was improved by fungicide treatment. Overall, these results suggest that soil biota may enhance the growth of S. virgaurea in habitats dominated by native congenerics. In contrast, fungal pathogens may reduce germination for native Solidago spp. (and possibly
S. virgaurea), but this does not seem to depend on whether relatives are present. Success of seedlings of exotics such as S. virgaurea thus may differ among habitats because of biotic facilitation by mutualists.
Introduction
Non-indigenous (exotic) species can leave behind the majority of species with which they interacted in their native range, only to form new interactions with species upon introduction to new areas (Richardson et al. 2000; Colautti et al. 2004; Mitchell et al.
2006). The outcome of these interactions with potential enemies, mutualists and competitors can have important consequences for their invasion success (Maron & Vila
2001; Levine et al. 2004; Mitchell et al. 2006). Therefore, predicting whether invaders are likely to interact with native species is critical for better understanding and managing biological invasions.
For instance, one of the more prominent ideas that predicts why exotic species succeed is that they leave behind enemies in their native range (Keane & Crawley 2002; Mitchell &
Power 2003; Torchin et al. 2003; Colautti et al. 2004). This 'Enemy Release Hypothesis'
(ERH) proposes that this loss of enemies provides invaders with a competitive advantage over native species (Maron & Vila 2001; Keane & Crawley 2002; Torchin et al. 2003).
However, the importance of ERH in explaining enemy reductions and subsequent
139 invasion success remains controversial, as exotics can accumulate enemies in their new range (Agrawal & Kotanen 2003; Colautti et al. 2004; Hawkes 2007; Parker et al. 2006;
Parker & Gilbert 2007; Brandle et al. 2008). This can result in exotic species experiencing similar or even higher damage than native species (e.g. Agrawal & Kotanen
2003; Parker & Hay 2005).
Host-use by enemies can be phylogenetically conserved (Odegaard et al. 2005; Brandle
& Brandl 2006; Weiblen et al. 2006). Where these phylogenetically conserved patterns of host-use extend to novel invaders, exotic plants that are more closely related to native communities may be more likely to accumulate enemies and experience higher rates of damage. Evidence to support this pattern is mixed: Cappuccino & Carpenter (2005) found no relationship between taxonomic isolation and damage to introduced plants in eastern North America, whereas Dawson et al. (2009) found that taxonomically isolated exotic trees in Africa were more likely to experience lower herbivory that those more closely related to native species. Some of these inconsistencies may depend on the phylogenetic scale of comparisons that are made between exotic and native species
(Chapters 3 & 4, Hill & Kotanen 2009): for example, foliar damage on a broad taxonomic sample of exotic species declined with their phylogenetic novelty among locally co-occurring native species (Hill & Kotanen 2009), whereas comparisons limited to within one family, the Asteraceae, found no significant relationships between foliar and capitular damage and phylogenetic novelty (Chapter 3 & 4).
140 A phylogenetic framework also can be used to predict whether competitive and mutualistic interactions are likely to be realized by exotic species (Mitchell et al. 2006).
In fact, the idea that invaders are limited by competition forms the basis of what is now known as "Darwin's Naturalization Hypothesis" (Daehler 2001): DNH assumes that competition is more intense among close relatives, resulting in exotic species being less likely to become naturalized when they are closely related to native species (Darwin
1859). Although there is some evidence to support competition as a mechanism, experimental work has shown that the relationship between competitive ability and phylogenetic distance is quite weak (Cahill et al. 2008). In contrast, interactions with mutualistic species may be critical for invasion success of some exotic species
(Richardson et al. 2000). As arbuscular mycorrhizae are thought to be relatively generalized in their host use, the likelihood of forming novel mycorrhizal associations may only be weakly associated with an invader's phylogenetic relationship with native species (Richardson et al. 2000); however, mycorrhizae can greatly vary in their impact on plant vigor (Bever 2002; Klironomos 2003). Therefore, interacting with the appropriate mycorrhizae may be more likely when invaders are more closely related to native species (e.g. Maherali & Klironomos 2007).
In this study, I was interested in determining if habitat-specific soil biota, including both mutualists and pathogens, affected seedling growth and seed mortality of an exotic species, and whether these interactions were further influenced by the presence of congeneric native relatives. Previous work that has investigated the role of soil biota on plant invasion dynamics has typically used species that are naturalized in their new range
141 (Klironomos 2002). Consequently, many of these species (both the invader and the native soil biota) may have had time to adapt to one another (Muller-Scharer et al. 2004; Parker
& Gilbert 2004); this obscures the understanding of how phylogenetic relationship among invading and native species may influence the critical early stages of the invasion process. To avoid this, I used a species that has not yet colonized North America, but is widely available at seed supply stores: Solidago virgaurea L.
I ran two experiments to evaluate whether habitat interactions are influenced by the presence of close relatives: the first tested how total soil biota, including both mutualists and pathogens, from different habitats affected seedling growth. The second investigated how soil fungi from different habitats affected seed mortality of the novel exotic, along with two native congeneric species. The habitats from which the soils were collected represented plant communities with and without native congenerics of S. virgaurea. Net effects of soil biota were expected to be positive if mutualists are more important than pathogens, and negative if the reverse is true. Therefore, if pathogens predominate, I expected soil sterilization treatments to improve growth; if mutualists dominate, performance should be best in non-sterile treatments. If these microbes are shared with relatives, both effects should be stronger in soils from native Solidago-dominated communities. I expected that these patterns should be true for both seedling growth and seed survival when exposed to sterilized and non-sterilized soils.
142 Materials and Methods
Study Site
This study was conducted at the University of Toronto's Koffler Scientific Reserve
(KSR) at Jokers Hill, Regional Municipality of York, in southern Ontario, Canada (44º02'
N, 79º31' W, 300m ASL). This 350-ha site lies within the Oak Ridges Moraine, and is dominated by prominent hills with a thin organic layer over deep glacial sands.
Vegetation is a mixture of old fields supporting a diverse range of native and exotic plants, hardwood (maple-beech-hemlock) forest, and conifer plantations. Further information on this site may be found at http://www.ksr.utoronto.ca/jh.html.
Study Species
The genus Solidago is primarily a North American group in the subtribe Solidagininae, family Asteraceae (Beck et al. 2004). The only non-North American goldenrod is
Solidago virgaurea L., which is a perennial native to Europe and Asia. In its native range
S. virgaurea grows in many habitats including rocky outcrops, disturbed areas, and old fields (Davis et al. 2000). The species is not yet naturalized in North America, but can be readily purchased as an ornamental or for herbal remedies at horticultural stores. North
American Solidago species such as Solidago gigantea Ait. and Solidago altissima L. are successful invaders throughout much of Europe (Jakobs et al. 2004), and can co-occur with the native S. virgaurea (Jobin et al. 1996). Therefore, I anticipated that S. virgaurea would have the ecological amplitude to grow along side North American Solidago species at this study site. The two representative species that I studied were Solidago cf canadensis L. (including Solidago canadensis and Solidago altissima; hereafter S.
143 canadensis) and Solidago nemoralis Ait. At KSR, S. canadensis is dominant in mesic to seasonally wet old-field meadows, whereas S. nemoralis tends to grow in slightly drier sites, usually mixed with S. canadensis.
Experimental Design
Seedling Experiment
Soils were collected from five meadows representing each of three old-field habitats at
KSR; habitat types included grass (dominated by Bromus inermis Leyss. and Poa pratensis L.), S. canadensis, and mixed Solidago (including S. nemoralis and S. canadensis). In general, edaphic conditions were similar across habitat types; however, grass and S. canadensis habitats were found on mesic to dry old-field sites, where as mixed Solidago habitats tended to be on drier exposed soils. In the last decade, S. canadensis has aggressively invaded many grass-dominated old-fields at this site, including the meadows used in this study.
Solidago virgaurea seeds were purchased from a local seed supply store in southern
Ontario. A maximum of 20 seeds were placed on 8cm diameter Whatman filter paper in
9cm diameter petri dishes and were germinated using an environmental chamber set to a
12 hour dark-light cycle, with overnight low temperature set to 10oC and daytime high temperature set to 20oC. After germinating, the root length of each seedling was measured to test whether initial (pre-treatment) seedling size was a predictor of final size.
One seedling was then planted per 70 mm pot with the habitat and sterilization treatments outlined below.
144 When the seedlings were ready for planting, soil collected at KSR was brought back to the lab, where it was kept in refrigerators at 4oC for 2 days. These soils were then either sterilized by autoclaving the soil twice (two consecutive 20 min cycles) (e.g. MacKay &
Kotanen 2008), or was left intact. In each pot, approximately 100g of a common growth medium (field-collected sand mixed with potting soil and sterilized) was then inoculated with 5 g of either the sterilized soil or the non-sterilized soil from each meadow and thoroughly mixed. I replicated each habitat and treatment combination five times for a total of 150 plants. Seedlings were grown under ambient conditions in the greenhouse at the University of Toronto Mississauga between January 11 and March 24, 2008. Using soils collected from the field, I expected that this approach emulated the initial effect of soil biota upon colonization, rather than soil-feedback, which measures the effects of soil biota accumulation in successive generations (Klironomos 2002).
Seed Mortality Experiment
The same S. virgaurea seed stock as above was used for the seed mortality experiment.
Seeds for the native Solidago species were collected from 50 individuals across five locations at KSR during the fall 2007. For each species, seeds were thoroughly mixed before use. In the lab, experimental units were created by randomly choosing 20-50 seeds of a single species and placing them between two layers of 0.5mm nylon mesh (small enough to contain the seeds of all three species), held together using 35mm photography slide mounts (Blaney & Kotanen 2001a). Parafilm was wrapped around the edges of the slide mounts to prevent seeds from escaping. I created 150 of these units, and each was
145 randomly assigned to one of six habitat (grass, Solidago canadensis, mixed Solidago) x treatment (fungicide, control) combinations; where each habitat type was represented by five different meadows. Before planting, seeds in the fungicide-addition treatment were dipped in a solution of Maestro 75 DF (10g/L with active ingredient 75% by weight of
Captan; Zenecan Agro, Stoney Creek, ON, Canada). Captan is a non-systemic heterocyclic nitrogen fungicide effective against a broad spectrum of Oomycetes,
Ascomycetes, and Basidiomycetes (Sharvelle 1961; Torgeson 1969; Neergaard 1977), and has been used to successfully reduce fungal seed mortality for various experiments conducted at KSR (Blaney & Kotanen 2001b; Schafer & Kotanen 2003; O'Hanlon-
Manners & Kotanen 2004; MacKay & Kotanen 2008). Slide mounts with seeds not exposed to fungicide were dipped in water. Slide mounts were then planted at a depth of
1 - 5cm in their respective sites. At one-week intervals during the subsequent fortnight, approximately 100 mL of the same solution of fungicide was applied to soils surrounding where the slide mounts were buried for the respective treatment.
Slide mounts were collected the following spring on June 11, 2008. In the lab, 20 seeds were randomly chosen from each slide mount and placed on 8cm diameter Whatman filter paper, on top of moist sterilized sand, in 9 cm diameter petri dishes. Petri dishes were then randomized and placed on benches in the glass house at University of Toronto at Mississauga. Water was added to each petri dish every second or third day. To minimize environmental effects of the glass house on germination, petri dishes were randomly reassigned a location on the bench following weekly surveys. Surveys of germination success were made over five weeks until July 20, 2008. The cumulative
146 number of seeds that germinated from the 20 that were sub-sampled was used for statistical analyses. A subset of the non-germinated seeds were dissected to determine if they were viable; otherwise, they were assumed to be dead.
Statistical Analyses
For the seedling and seed experiments, I used mean biomass and proportion germination success, respectively, to determine if there were effects i) across habitats, ii) relating to soil biota, and iii) that differed across habitats relating to soil biota. I used the means of habitat x treatment replicates (n = 30) to create a balanced analysis, and more precise estimate of plant biomass. Results are very similar when the data were not pooled in this manner, unless otherwise indicated. For both experiments, I used traditional mixed effect analysis of variance in the program JMP v. 5 (SAS 2002). I used a split-plot factorial analysis of variance model (Kirk 1982) where habitat type, soil biota treatment, and their interaction were considered fixed effects, and meadow nested in habitat type was considered a random factor.
For the seedling growth experiment, I also evaluated whether seedling survival was influenced by treatment type; the associated Z-score for the proportion survival for each treatment type was used to linearize the data and was analyzed using the same statistical model explained above.
147 Results
Seedling experiment
Seedling mortality and covariates
Of the 150 plants grown for the seedling experiment 62 survived the 8-week growth period. At least one representative for each site and treatment combination survived,
which allowed the experiment to be analyzed with a balanced design. Habitat type (F2,12,
= 0.257), treatment (F1,12 = 1.500), and their interaction (F2,12 = 0.132) were not statistically related to seedling mortality (p > 0.05 for all cases). Shoot (F1,28 = 1.863. p =
0.183) and root (F1,28 = 0.982, p = 0.330) biomass were not statistically related to initial root length, so initial root length was not included as a covariate in these final analyses.
Shoot biomass
There was a marginally significant difference among habitat types (p = 0.061; Table 6.1).
Shoot mass was lowest for plants grown in soils from grass habitats (mean and standard error: 0.23 ± 0.05 g), followed by Solidago canadensis habitats (0.41 ± 0.05 g), and mixed Solidago habitats (0.50 ± 0.05 g) (Fig. 6.1a). Plants grown in sterilized soil tended to be smaller than in non-sterilized soil (0.32 g compared to 0.43 g), but this result was also only marginally significant (p = 0.089). Despite the interaction between habitat type and soil sterilization being non-significant, differences in biomass between sterile and non-sterile soils did tend to different among communities (Fig. 6.1a). Increases in growth when in the presence of soil biota were lowest in grass habitats (0.01 g), intermediate in
Solidago canadensis habitats (0.11 g), and the highest in mixed Solidago habitats (0.21 g). When the differences in shoot growth between non-sterile versus sterile soil for both
148 Solidago habitats were compared to that in grass habitats using a t-test, the differences in growth between community types were still non-significant (t = 1.317, df = 13, p =
0.211).
Root biomass
Root biomass differed significantly among habitat types (p = 0.056), and soil sterilization treatment (p = 0.010, Table 6.1). Overall, root biomass was highest in soils from Solidago habitats (0.25 ± 0.03 g, for both habitats), and lowest in the grass habitat (0.13 ± 0.03 g)
(Fig. 6.1b). Plants grown in non-sterilized soils produced 1.7 times more root biomass
(0.27g compared to 0.16g). There was no significant interaction between habitat and soil sterilization (p = 0.233). However, effects of soil sterilization did tend to be different across habitat type (Fig. 6.1b). Increases in root biomass were very similar in non- sterilized Solidago soils (0.14g and 0.17g for Solidago canadensis and mixed sites, respectively), compared to only a 0.02 g increase in soils from grass habitats. When differences in root growth between non-sterilized soils and sterilized soils for both
Solidago communities were compared to that in grass habitats using a t-test, there was a marginally significant difference (t= 1.845, df = 13, p = 0.088): increases in root growth tended to be higher in soils from Solidago communities.
Seed experiment
There were large differences in the overall germination success of the three species used:
S. canadensis (proportion = 0.32) and S. nemoralis (0.24) had much higher germination than the novel exotic S. virgaurea (0.01). The rates of germination for S. canadensis and
149 S. virgaurea seeds used in the experiment were lower than the proportions observed in laboratory controls (0.53 and 0.12, respectively); unfortunately, no comparison was made with S. nemoralis as the seed stock went missing during the experiment. Among seeds that did not germinate, a subset was dissected and the endosperm was found to be liquefied or rotted in all cases.
Native Solidago spp.
Germination success of S. canadensis seeds was affected by habitat type (p = 0.072;
Table 6.2). Among the three habitat types, germination success was significantly higher
(contrasts, p = 0.026) in grass habitats (0.44 ± 0.05) than in Solidago canadensis and mixed Solidago habitats (0.25 ± 0.05, and 0.29 ± 0.05, respectively) (Fig. 6.2). Fungicide treatment also significantly improved germination success (p = 0.002). The proportion of seeds treated with fungicide that germinated was more than twice as large as those not treated (0.48 ± 0.06 compared to 0.17 ± 0.06, respectively). The interaction between habitat type and fungicide treatment was not significant (p > 0.926), indicating that declines in germination success in unsterilized soils were consistent among habitats (Fig
2).
Patterns of germination success for S. nemoralis seeds were similar to those of S. canadensis (Fig. 6.2). Habitat type significantly affected germination (p = 0.002; Table
6.2). Again, germination was higher in grass habitats (0.40 ± 0.04) compared to the combined mean in Solidago canadensis (0.15 ± 0.04) and mixed Solidago (0.13 ± 0.05) habitats (linear contrasts, p < 0.001) (Fig. 6.2). Fungicide treatment also significantly
150 increased germination success (p = 0.002; Table 6.2): when fungicide was applied, germination success was 0.29 ± 0.02 compared to 0.16 ± 0.02 (Fig. 6.2). The interaction between habitat type and fungicide application was not significant (p = 0.161), indicating that the impact of fungi on germination success was consistent across habitat types (Fig.
6.2).
Solidago virgaurea
There was no difference in germination success across habitats for S. virgaurea (p =
0.156, Table 6.2). Treating seeds with fungicide had a marginally significant impact on germination success (p = 0.062; Table 6.2); success was slightly higher when fungicide was applied (0.016 ± 0.006) compared to when it was not (<0.001 ± 0.006). The interaction between habitat type and fungicide application was not significant (p = 0.184), indicating that fungal effects were statistically similar across habitat types. Unfortunately, the extremely low germination success of S. virgaurea made it difficult to have confidence in these results; only 12 seeds out of 1475 germinated for the experiment.
Despite this, the only seeds that did survive were those that were treated with fungicide.
Controls using non-experimental seeds suggested that at least 12% of seeds should have germinated, so low success in the experiment was not completely the result of a bad seed stock.
Discussion
I found that the net effects of soil biota on the exotic S. virgaurea were opposite when measured as seedling growth and seed mortality: seedlings tended to grow better when
151 exposed to soil biota collected from the field, but unprotected seeds showed higher rates of mortality. These effects varied only slightly across habitats: soil biota from habitats with close native relatives tended to have stronger positive effects on growth than soils from grass communities; but, this interaction was not significant in the ANOVA models and only marginally significant for root biomass when the data for Solidago habitats were compared to grass habitats using a t-test. Seed mortality was so high that it is difficult to put confidence in habitat-related patterns for S. virgaurea, but the few seeds that survived had all been treated with fungicide, and fungicide greatly improved germination in the native Solidago species. As well, for the native Solidago species, germination was higher for seeds in grass habitats than in habitats with conspecifics.
Opposing effects of soil biota
Higher overall growth (for shoot and root biomass) of plants in non-sterilized soils indicates a significant influence of mutualists, and suggests that the net effect of the soil community (pathogens + mutualists) was positive. Consistent with this, root growth was more productive in non-sterilized soil inoculated with biota from the Solidago habitats vs. biota from grass habitats, suggesting that mutualists might be more prevalent in sites where congenerics are present. If this is the case, the most likely candidates are arbuscular mycorrhizae (AMF). Though not highly host specific (Richardson et al. 2000),
AMF can have considerably different effects on host species (Klironomos 2003). Since growth differences tended to be better in Solidago sites, appropriate AMF may have been more abundant where the native congeneric Solidago were present. These benefits could have been compounded by the soil a higher diversity of suitable AMF (Maherali &
152 Klironomos 2007). In contrast, sterilization had no effect on plants grown in soil inoculated with biota from grass habitats, suggesting an equal influence of pathogens and mutualists, or their absence. If these interactions persist throughout the seasonal development of S. virgaurea in the field, growth would tend to be higher when close relatives are present. Contrary to the prediction that habitats with close relatives would be more resistant to invasion, their presence may influence soil biota such that S. virgaurea would be facilitated (Richardson et al. 2000).
Both native Solidago species experienced greatly improved germination when protected with fungicide, indicating the presence of fungal seed pathogens. In contrast, the extremely low germination success of S. virgaurea compared to native Solidago species even when treated with fungicide suggests that either i) they were dormant ii) seeds were dead before planting, iii) seed pathogens resistant to Captan are important for this species. The first option is unlikely, as a number (N ~ 20 for each species) of non- germinated seeds were dissected, and in all cases the embryo showed signs of infection by pathogens; the endosperm was liquefied or had rotted. The second possibility is consistent with lab controls, which indicated that initial germinability was low.
Nonetheless, at least 12% of 60 lab control seeds germinated, indicating approximately
177 of the 1475 were expected to germinate in the absence of additional mortality, rather
2 than the 12 that were observed ( χ c = 47.03, df = 1, p < 0.001) However, it is certainly possible that some pathogens may have been unaffected by the fungicide treatment, as suggested in option iii.€ This could have been a resistant fungus, or a non-fungal pathogen, likely a bacterium (Baskin & Baskin 2006). This also may explain why germination was
153 lower for seeds of both S. canadensis and S. nemoralis in habitats where conspecific or congeneric natives were present, even after treatment with fungicide. If so, this pathogen may have had more severe effects on S. virgaurea than on the native Solidago spp., as might be the case if the natives had evolved resistance, but the naive exotic had not
(Hokkanen & Pimentel 1989; Verhoeven et al. 2009).
Implications for ERH and DNH
The reduced impact of enemies for exotic species is often used to explain their success when introduced to new areas (Colautti et al. 2004; Keane and Crawley 2002; Torchin et al. 2003). The results from this study provide mixed support for predictions of the ERH: there was little evidence that soil pathogens affected seedling growth, but seeds may have been very susceptible to an unidentified soil pathogen. If so, contrary to escaping enemies, long-term invasion success would likely rely on overcoming the large negative effects of pathogens that caused seed mortality. To the degree that mutualistic interactions do benefit seeds that successfully germinate, they might be more likely to form in the presence of close relatives.
Among the different communities, the effects of AMF may differ enough to facilitate S. virgaurea colonization when close relatives are present, a pattern that would oppose predictions of DNH. However, long-term success will depend on the cumulative effect of all biotic interactions from seed to reproducing adult (Mitchell et al. 2006). Contrary to this observed pattern of better growth when close relatives were present, in a separate herbivory experiment (Chapter 5), S. virgaurea survival to maturity in the field was
154 slightly higher in grass habitats, though not significant. Furthermore, S. virgaurea experienced slightly higher damage from invertebrates when native Solidago were present. Therefore, benefits that may result from interactions with AMF in the field are likely to be equalized by negative effects from herbivores.
Implications for Community Phylogenetics
Recent work has shown that community structure, or the likelihood of plants co- occurring may depend on their phylogenetic relationship (Vamosi et al. 2009). There is also evidence that the phylogenetic relatedness of exotic species to natives can influence their invasiveness (Strauss et al. 2006). Similar to DNH, one of the fundamental assumptions of the approaches used by these studies is species that are more closely related likely compete more intensely (Vamosi et al. 2009). Therefore, the species composition of a community is expected to reflect the underlying mechanisms structuring it: communities are predicted to be composed of distant relatives when competition is important (phylogenetically over-dispersed), and close relative when environmental filtering is important (phylogenetically clustered) (Webb et al. 2002). These patterns, however may also reflect biotic interactions other than competition: if close relatives are more likely to share enemies, communities may become phylogenetically over-dispersed; in contrast, if close relatives are more likely to share mutualists, communities may become phylogenetically clustered. If S. virgaurea has better growth when exposed to soil biota from sites with congeneric natives, this suggests that shared mutualists could produce phylogenetically clustered patterns of community structure. The dynamics of these types of interactions should be considered as viable alternatives to the idea that
155 abiotic conditions are responsible for filtering species' traits (Vamosi et al. 2009), particularly when mechanisms of community assembly are inferred from the patterns of species co-occurrence in the field.
References
Agrawal A.A. & Kotanen P.M. (2003). Herbivores and the success of exotic plants: a
phylogenetically controlled experiment. Ecology Letters, 6, 712-715.
Baskin C.C. & Baskin J.M. (2006). The natural history of soil seed banks of arable land.
Weed Science, 54, 549-557.
Beck J.B., Nesom G.L., Calie P.J., Baird G.I., Small R.L. & Schilling E. (2004). Is
subtribe Solidagininae (Asteraceae) monophyletic? Taxon, 53, 691-698.
Bever J.D. (2002). Negative feedback with a mutualism: host-specific growth of
mycorrhizal fungi reduces plant benefit. Proceedings of the Royal Society B-
Biological Sciences, 263, 2595-2601.
Blaney C.S. & Kotanen P.M. (2001a). Effects of fungal pathogens on seeds of native and
exotic plants: a test using congeneric pairs. Journal of Applied Ecology, 38, 1104-
1113.
Blaney C.S. & Kotanen P.M. (2001b). Post-dispersal losses to seed predators: an
experimental comparison of native and exotic old field plants. Canadian Journal
of Botany-Revue Canadienne De Botanique, 79, 284-292.
Brandle M. & Brandl R. (2006). Is the composition of phytophagous insects and parasitic
fungi among trees predictable? Oikos, 113, 296-304.
156 Brandle M., Kuhn I., Klotz S., Belle C. & Brandl R. (2008). Species richness of
herbivores on exotic host plants increases with time since introduction of the host.
Diversity and Distributions, 14, 905-912.
Cahill J.F., Kembel S.W., Lamb E.G. & Keddy P.A. (2008). Does phylogenetic
relatedness influence the strength of competition among vascular plants?
Perspectives in Plant Ecology Evolution and Systematics, 10, 41-50.
Cappuccino N. & Carpenter D. (2005). Invasive exotic plants suffer less herbivory than
non-invasive exotic plants. Biology Letters, 1, 435-438.
Colautti R.I., Ricciardi A., Grigorovich I.A. & MacIsaac H.J. (2004). Is invasion success
explained by the enemy release hypothesis? Ecology Letters, 7, 721-733.
Daehler C.C. (2001). Darwin's naturalization hypothesis revisited. American Naturalist,
158, 324-330.
Darwin C. (1859). On the origin of species by means of natural selection. Murray,
London, UK.
Davis M.A., Grime J.P. & Thompson K. (2000). Fluctuating resources in plant
communities: a general theory of invasibility. Journal of Ecology, 88, 528-534.
Dawson W., Burslem D. & Hulme P.E. (2009). Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions, 15, 141-147.
Hawkes C.V. (2007). Are invaders moving targets? The generality and persistence of
advantages in size, reproduction, and enemy release in invasive plant species with
time since introduction. American Naturalist, 170, 832-843.
157 Hill S.B. & Kotanen P.M. (2009). Evidence that phylogenetically novel non-indigenous
plants experience less herbivory. Oecologia, in-press.
Hokkanen H.M.T. & Pimentel D. (1989). New associations in biological control: theory
and practice. Canadian Entomologist, 121, 829-840.
Jakobs G., Weber E. & Edwards P.J. (2004). Introduced plants of the invasive Solidago
gigantea (Asteraceae) are larger and grow denser than conspecifics in the native
range. Diversity and Distributions, 10, 11-19.
Jobin A., Schaffner U. & Nentwig W. (1996). The structure of the phytophagous insect
fauna on the introduced weed Solidago altissima in Switzerland. Entomologia
Experimentalis Et Applicata, 79, 33-42.
Keane R.M. & Crawley M.J. (2002). Exotic plant invasions and the enemy release
hypothesis. Trends in Ecology & Evolution, 17, 164-170.
Kirk R.E. (1982). Experimental design: Procedures for the behavioural sciences. Second
ed. Brooks and Cole, Monterey.
Klironomos J.N. (2002). Feedback with soil biota contributes to plant rarity and
invasiveness in communities. Nature, 417, 67-70.
Klironomos J.N. (2003). Variation in plant response to native and exotic arbuscular
mycorrhizal fungi. Ecology, 84, 2292-2301.
Levine J.M., Adler P.B. & Yelenik S.G. (2004). A meta-analysis of biotic resistance to
exotic plant invasions. Ecology Letters, 7, 975-989.
MacKay J. & Kotanen P.M. (2008). Local escape of an invasive plant, common ragweed
(Ambrosia artemisiifolia L.), from above-ground and below-ground enemies in its
native area. Journal of Ecology, 96, 1152-1161.
158 Maherali H. & Klironomos J.N. (2007). Influence of phylogeny on fungal community
assembly and ecosystem function. Science, 316, 1746-1748.
Maron J.L. & Vila M. (2001). When do herbivores affect plant invasion? Evidence for
the natural enemies and biotic resistance hypotheses. Oikos, 95, 361-373.
Mitchell C.E., Agrawal A.A., Bever J.D., Gilbert G.S., Hufbauer R.A., Klironomos J.N.,
Maron J.L., Morris W.F., Parker I.M., Power A.G., Seabloom E.W., Torchin M.E.
& Vazquez D.P. (2006). Biotic interactions and plant invasions. Ecology Letters,
9, 726-740.
Mitchell C.E. & Power A.G. (2003). Release of invasive plants from fungal and viral
pathogens. Nature, 421, 625-627.
Muller-Scharer H., Schaffner U. & Steinger T. (2004). Evolution in invasive plants:
implications for biological control. Trends in Ecology & Evolution, 19, 417-422.
Neergaard P. (1977). Seed Pathology. Halsted Press, New York.
O'Hanlon-Manners D.L. & Kotanen P.M. (2004). Evidence that fungal pathogens inhibit
recruitment of a shade-intolerant tree, white birch (Betula papyrifera), in
understory habitats. Oecologia, 140, 650-653.
Odegaard F., Diserud O.H. & Ostbye K. (2005). The importance of plant relatedness for
host utilization among phytophagous insects. Ecology Letters, 8, 612-617.
Parker I.M. & Gilbert G.S. (2004). The evolutionary ecology of novel plant-pathogen
interactions. Annual Review of Ecology Evolution and Systematics, 35, 675-700.
Parker I.M. & Gilbert G.S. (2007). When there is no escape: The effects of natural
enemies on native, invasive, and noninvasive plants. Ecology, 88, 1210-1224.
159 Parker J.D., Burkepile D.E. & Hay M.E. (2006). Opposing effects of native and exotic
herbivores on plant invasions. Science, 311, 1459-1461.
Parker J.D. & Hay M.E. (2005). Biotic resistance to plant invasions? Native herbivores
prefer non-native plants. Ecology Letters, 8, 959-967.
Richardson D.M., Allsopp N., D'Antonio C., Milton S.J. & Rejmanek M. (2000). Plant
invasions - the role of mutualisms. Biological Reviews, 75, 65-93.
SAS (2002). JMP version 5.0. In. SAS Institute Cary, North Carolina, USA.
Schafer M. & Kotanen P.M. (2003). The influence of soil moisture on losses of buried
seeds to fungi. Acta Oecologica-International Journal of Ecology, 24, 255-263.
Sharvelle E.G. (1961). The nature and uses of modern fungicides. Burgess Publishing,
Minneapolis.
Strauss S.Y., Webb C.O. & Salamin N. (2006). Exotic taxa less related to native species
are more invasive. Proceedings of the National Academy of Sciences of the United
States of America, 103, 5841-5845.
Torchin M.E., Lafferty K.D., Dobson A.P., McKenzie V.J. & Kuris A.M. (2003).
Introduced species and their missing parasites. Nature, 421, 628-630.
Torgeson D.C. (1969). Fungicides: an advanced treatise. Chemistry and Physiology.
Academic Press, New York.
Vamosi S.M., Heard S.B., Vamosi J.C. & Webb C.O. (2009). Emerging patterns in the
comparative analysis of phylogenetic community structure. Molecular Ecology,
18, 572-592.
Verhoeven K.J.F., Biere A., Harvey J.A. & van der Putten W.H. (2009). Plant invaders
and their novel natural enemies: who is naive? Ecology Letters, 12, 107-117.
160 Webb C.O., Ackerly D.D., McPeek M.A. & Donoghue M.J. (2002). Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475-505.
Weiblen G.D., Webb C.O., Novotny V., Basset Y. & Miller S.E. (2006). Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62-
S75.
161 Table 6.1 Split-plot factorial analysis of variance results for differences in shoot and root growth, and their ratio for Solidago virgaurea grown in soils that were inoculated with sterile or intact soil biota from different habitat types (grass, Solidago canadensis, and mixed Solidago). Since the experimental design was a split plot factorial (Kirk 1982), mean squares estimated for the random effect were used in the denominator of the F-ratio for habitat type. The remaining fixed effects were tested against the error term (in this case, the interaction between meadow nested in habitat type, and treatment)
Shoot Mass Root Mass Source df F p F p Habitat 2 3.565 0.061 3.702 0.056 Sterilization Treatment 1 3.420 0.089 9.390 0.010 H x S 2 1.100 0.364 1.647 0.233 For all F-ratios denominator df = 12
162 Table 6.2 Split-plot factorial analysis of variance for the proportion of seeds germinated for Solidago canadensis, Solidago nemoralis, and Solidago virgaurea across different habitat types (grass, Solidago canadensis, and mixed Solidago) when either treated with fungicide, or left untreated. See caption in Table 6.1 for design structure
Solidago Solidago Solidago canadensis nemoralis virgaurea Source df F p F p F p Habitat 2 3.290 0.072 11.001 0.002 2.157 0.156 Fungicide Treatment 1 15.680 0.002 15.623† 0.002 4.302† 0.062 H x F 2 0.078 0.926 2.164† 0.161 1.983† 0.184 note: F-ratio denominator df = 12, except where highlighted by †, df = 11
163 Figure 6.1 Shoot biomass (a) and root biomass (b) of Solidago virgaurea when exposed to sterilized or non-sterilized soils from habitats dominated by grass, Solidago canadensis
(Scan), and mixed Solidago (Smix). Symbols represent the mean ± SE values for each habitat and treatment combination
164 Figure 6.2 Germination success of Solidago canadensis, Solidago nemoralis, and Solidago virgaurea seeds when fungi were present and absent in soils from habitats dominated by grass, Solidago canadensis (Scan), and mixed Solidago (Smix). Symbols represent the mean ± SE proportion of seeds that germinated for each habitat and treatment combination
165 Chapter 7
General Discussion
In this thesis I have investigated the degree to which an invader's phylogenetic relationship to native species influences its escape from biotic constraints (herbivores and pathogens) using a nested series of taxonomic levels. The second chapter revealed that escape from herbivores was more likely for invaders that are less closely related to native family members. The third and fourth chapters investigated whether a similar phylogenetic pattern exists within a group of closely related plants in the Asteraceae.
Although exotic species experienced less damage than native species, there was no evidence that phylogenetically novel exotics experienced increased release. Finally, chapters five and six revealed that close relatives can influence the type of biotic interactions that the novel exotic, Solidago virgaurea, experiences. The combined effect of vertebrate herbivores and competition in intact communities provided high biotic resistance, regardless of close relatives being present. When habitats were disturbed, the presence of congeneric relatives may result in slightly increased damage, but any negative impact of this interaction may be countered by the positive effect of below- ground mutualistic interactions with mycorrhizae. Together these chapters represent a comprehensive examination of how the biotic interactions and ecological consequences experienced by exotic species can be influenced by their phylogenetic relationship to native species in the invaded range.
166 Implications for the "Enemy Release Hypothesis"
The degree to which exotic species escape biotic constraints in their invaded range has been central to theories of invasion biology (Keane & Crawley 2002; Torchin et al. 2003;
Colautti et al. 2004). To date, tests of this hypothesis have generally neglected the fact that many biotic interactions, including those with enemies, can be phylogenetically conserved (Strong et al. 1984; Williamson & Fitter 1996; Odegaard et al. 2005; Weiblen et al. 2006; Gilbert & Webb 2007). Where this is the case, invaders that have close relatives in the invaded range are expected to be more likely to interact with enemies and experience higher rates of damage (Mack 1996; Rejmanek 1996). The few studies that test the prediction of reduced damage on phylogenetically isolated plants show different trends. Dawson et al. (2009) found that damage declines with taxonomic isolation of exotic trees from native tree species in Africa. Cappuccino & Carpenter (2005) found no evidence of a relationship between foliar damage on exotic plants and the number of native confamilial plants in North America. Ricciardi and Ward (2006) even suggest that invaders belonging to novel genera tend to be more susceptible to native enemies than those represented by native genera. Using a phylogenetically informed approach, results from chapter two of this thesis are consistent with Dawson et al. (2009): damage on phylogenetically novel invaders tended to be lower. This was the case when phylogenetic novelty was measured as mean distance to family members, but not to an exotic's closest native relative.
There was also support for enemy release among invaders in the Asteraceae. When the proportion of foliar damage and the proportion of capitula damaged were measured on
167 native and exotic species in this family, phylogenetic correction methods revealed that invaders had lower damage than native species. In addition, most of the phylogenetically structured variability in these two measures was associated with sub-family lineages represented by taxonomic tribes. Together, this suggests that enemy escape is likely a relative process, where invaders tend to experience lower damage relative to native species belonging to that lineage. However, there was no general effect of phylogenetic distance. I expect that this pattern will vary across families, however, as the strength of shared host-use can differ among families (Frenzel & Brandl 2001).
On an individual level, the presence of close relatives at the habitat scale had little impact on rates of damage, and any trends that likely exist were only detected when habitats were disturbed (Chapters 5 & 6). The combined effects of generalist herbivores (e.g. meadow voles) and competition resulted in higher damage and mortality when habitats were left intact; in this case, the presence of close relatives had no detectable impact on potential invasion success, as rates of mortality were similar. Furthermore, subtle increases in damage with the density of close relatives when habitats had been disturbed may be countered by positive interactions with below-ground mutualists. These results suggest that the successful invasion of the novel invader S. virgaurea would be influenced by the net effect of both positive and negative biotic interactions (Mitchell et al. 2006).
168 Implications for "Darwin's Naturalization Hypothesis"
Coexistence may be limited where the net outcome of biotic interactions between close relatives tends to be negative. This idea is central to the prediction made by Darwin
(1859) that floras are more likely to gain taxonomically novel invaders than those that are closely related to native species. Although competition was assumed to be the mechanism that limited co-occurrence (Darwin 1859), shared enemies could produce a similar pattern
(Mack 1996; Dawson et al. 2009). Results from this thesis suggest that phylogenetically structured damage among exotic species is unlikely to act as a mechanism that would produce patterns of invasion success as strong as those predicted by DNH.
That foliar damage on invaders was better predicted by phylogenetic distances to family members, and not their closest native relative suggests that escaping biotic constraints may be more likely for exotics belonging to novel or anciently diverged families (Chapter
2). To the extent that these biotic constraints are strong enough to limit invasion success
(Levine et al. 2004), then invaders that belong to these novel lineages may be more successful (Cappuccino & Carpenter 2005; Strauss et al. 2006). Similar patterns in damage, however, were not observed for phylogenetically novel exotic plants within the
Asteraceae (Chapters 3 & 4). Although exotic species experienced lower foliar damage and losses to pre-dispersal seed predators than native species, damage was not negatively correlated with phylogenetic distance from native or exotic species in this family. Results from the S. virgaurea studies also failed to find clear support for a mechanism of DNH: overall, damage and survival of plants in the field were similar across habitats with and without close relatives. Laboratory experiments even suggest that S. virgaurea may
169 benefit from mutualistic interactions when congeners are present; though weak, such interactions could result in patterns of invasion success that oppose those predicted by
DNH (Richardson et al. 2000; Mitchell et al. 2006).
Tests of DNH have typically been performed using lists of successful invaders at large spatial scales. In part, this explains why the majority tends to reject the hypothesis
(Daehler 2001; Duncan & Williams 2002; Lambdon & Hulme 2006; Diez et al. 2008;
Cadotte et al. 2009): relying on these data may actually test the prediction that species are more likely to co-occur when environmental conditions filter the appropriate set of traits for survival (Diez et al. 2008; Proches et al. 2008). Where species lists are used to test this hypothesis, patterns of presence and absence at finer scales may be a more appropriate (e.g. Diez et al. 2008; Cadotte et al. 2009). The frequent assumption that non- random phylogenetic patterns at small-scales are the outcome of competitive interactions, however, is questionable (Cahill et al. 2008); based on the results from this thesis, a similar conclusion may be reached for interactions with shared enemies: among the tests I used, the strongest support for a relationship between damage and phylogenetic novelty was for exotics that were less closely related to family members (Chapter 2). This relationship, however, was relatively weak, only occurred when experimental plants were grown in a common garden experiment, and was not replicated in natural populations of a larger set of exotic species at the same site.
170 Implications for Community Phylogenetics
Darwin's (1859) insight that biotic interactions among close relatives can influence community assembly forms an important component of the underlying framework of community phylogenetics (Webb et al. 2002). Hence, the same general predictions of phylogenetically conserved biotic interactions apply to the community assembly of natural communities and those that are invaded: co-occurrence among close relatives is expected to be higher than expected when environmental filtering is important for community assembly, and lower than expected when negative biotic interactions are conserved among close relatives (Webb et al. 2002; Vamosi et al. 2009). Predictions under the community phylogenetic framework, again, tend assume that competition is the dominant mechanism limiting co-occurrence (Webb et al. 2002). However, resource competition may not necessarily be strong enough among close relatives to influence patterns of co-occurrence in the field (Cahill et al. 2008). The compelling evidence that enemies and their associated interactions are conserved among closely related hosts suggests that these biotic interactions should be included as viable alternatives (Odegaard et al. 2005; Weiblen et al. 2006; Becerra 2007; Gilbert & Webb 2007). Results from this thesis suggest that damage experienced by invaders was affected by their phylogenetic relationship with native species (chapters 2, 3, & 4), but at broad phylogenetic scales, not among closest relatives. That these trends tended to be weak, however, suggests that shared enemy interactions among relatives are unlikely to be consistent or strong enough to produce community-scale patterns of phylogenetic overdispersion, the pattern predicted if close rather than distance relatives are more likely to share negative biotic interactions (Webb et al. 2002; Cavender-Bares et al. 2009).
171 Practical Implications
Being able to predict exotic species' likelihood of escape from biotic constraints based on their phylogenetic relationship to native species could be an important tool for managing biological invasions. Despite the data from this thesis suggesting that the relationship between damage and an exotic's phylogenetic novelty among native species being relatively weak, there were strong phylogenetic patterns in the foliar damage (r2 = 0.32) and capitular damage (r2 = 0.59) among native and exotic Asteraceae. Therefore, consistent with traditional predictive approaches in invasion biology, using a phylogenetic approach might help identify potential invaders and determine which communities are susceptible (Baker 1974; Mack 1996; Rejmanek 1996). My results suggest that potential invaders that are less closely related to family members, or represent novel families are generally more likely to escape enemies, though this pattern was weak, limiting the utility of phylogenetic distance to family members as a predictive tool. The large amount of variability in damage that was phylogenetically structured among native and exotic Asteraceae hints at the potential for phylogenetic approaches to predict rates of damage for exotic species; using estimates of evolutionary distance between an exotic and native species within a family, however, is unlikely to be as effective. Since the strength of phylogenetically conserved host-use can vary among families (Frenzel & Brandl 2001), the degree to which the patterns I observed within the
Asteraceae are applicable to other families will likely depend on the phylogenetic structure of characters that are important for host use, and how well enemies can track those characters (Futuyma & Mitter 1996; Janz & Nylin 1998; Lewinsohn et al. 2005;
Agrawal 2007).
172 Equally beneficial to predicting which species have the potential to invade, is being able to determine which communities are likely to be invaded (Mack 1996). If agents of herbivory are important, communities with relatives of invaders might be harder to invade. Similar to diversity-based predictions of community resistance (Elton 1958;
Naeem et al. 2000; Fargione & Tilman 2005), communities that harbour a higher phylogenetic diversity may be more resistant to invasion if they are more likely to contain close relatives. This will likely depend on the net effect of biotic interactions that occur between community members and the potential invader (Mitchell et al. 2006). The results from the Solidago study suggest that above-ground enemies and below-ground mutualists may interact to influence habitat-specific patterns of invasion associated with the presence of close relatives.
Phylogenetically conserved host-use of enemies could also help inform biocontrol programs. If the presence of close relatives is likely to influence rates of herbivory on exotic species, then natural communities could be supplemented with native relatives of troublesome species. Where invasive species have resulted in declines of native species, approaches such as this could have three major benefits: increased abundance of affected species, increased competition for phylogenetically conserved niche space, and decreased release from shared enemy interactions. Utilizing native species as biocontrol also offers a more ecologically ethical approach than introducing non-native biocontrol agents from the invader's native range, as well as a potentially cheaper one (Mack et al. 2000).
173 Future Directions
As the tree of life becomes more complete and resolved, ecologists will likely be able to use this information to predict many of the biotic interactions that are fundamental to understanding invasion dynamics and community assembly (Webb et al. 2002). Results from this thesis suggest that indeed, enemy interactions between exotic and native species have a phylogenetic component. This should be considered in future tests of the Enemy
Release Hypothesis (Keane & Crawley 2002; Torchin et al. 2003; Colautti et al. 2004) and Darwin's Naturalization Hypothesis (Strauss et al. 2006). Furthermore, inferring environmental filtering or competitive exclusion as mechanisms of community assembly based on nonrandom phylogenetic patterns of species co-occurrence should be weighted against alternative hypotheses such as shared consumers, pathogens, parasites, and mutualists (Webb et al. 2002; Vamosi et al. 2009). The work here also suggests that phylogenetically conserved host-use does occur in some cases, but does not always correlate with measures of phylogenetic novelty: interactions tend to be weak, and are potentially confounded by phylogenetic scale, convergences in ecologically important phenotypes, and the spatial scale at which shared interactions are realized. Future work on invasion biology will benefit by considering these factors as well as the phylogenetic relationship of exotic species relative to native communities.
References
Agrawal A.A. (2007). Macroevolution of plant defense strategies. Trends in Ecology &
Evolution, 22, 103-109.
174 Baker H.G. (1974). The evolution of weeds. Annual Review of Ecology and Systematics,
5, 1-24.
Becerra J.X. (2007). The impact of herbivore-plant coevolution on plant community
structure. Proceedings of the National Academy of Sciences of the United States
of America, 104, 7483-7488.
Cadotte M.W., Hamilton M.A. & Murray B.R. (2009). Phylogenetic relatedness and plant
invader success across two spatial scales. Diversity and Distributions, 15, 481-
488.
Cahill J.F., Kembel S.W., Lamb E.G. & Keddy P.A. (2008). Does phylogenetic
relatedness influence the strength of competition among vascular plants?
Perspectives in Plant Ecology Evolution and Systematics, 10, 41-50.
Cappuccino N. & Carpenter D. (2005). Invasive exotic plants suffer less herbivory than
non-invasive exotic plants. Biology Letters, 1, 435-438.
Cavender-Bares J., Kozak K.H., Fine P.V.A. & Kembel S.W. (2009). The merging of
community ecology and phylogenetic biology. Ecology Letters, in press.
Colautti R.I., Ricciardi A., Grigorovich I.A. & MacIsaac H.J. (2004). Is invasion success
explained by the enemy release hypothesis? Ecology Letters, 7, 721-733.
Daehler C.C. (2001). Darwin's naturalization hypothesis revisited. American Naturalist,
158, 324-330.
Darwin C. (1859). On the origin of species by means of natural selection. Murray,
London, UK.
175 Dawson W., Burslem D. & Hulme P.E. (2009). Herbivory is related to taxonomic
isolation, but not to invasiveness of tropical alien plants. Diversity and
Distributions, 15, 141-147.
Diez J.M., Sullivan J.J., Hulme P.E., Edwards G. & Duncan R.P. (2008). Darwin's
naturalization conundrum: dissecting taxonomic patterns of species invasions.
Ecology Letters, 11, 674-681.
Duncan R.P. & Williams P.A. (2002). Ecology - Darwin's naturalization hypothesis
challenged. Nature, 417, 608-609.
Elton C.S. (1958). The ecology of invasions by animals and plants. Methuen, London,
UK.
Fargione J.E. & Tilman D. (2005). Diversity decreases invasion via both sampling and
complementarity effects. Ecology Letters, 8, 604-611.
Frenzel M. & Brandl R. (2001). Hosts as habitats: faunal similarity of phytophagous
insects between host plants. Ecological Entomology, 26, 594-601.
Futuyma D.J. & Mitter C. (1996). Insect-plant interactions: The evolution of component
communities. Philosophical Transactions of the Royal Society B-Biological
Sciences, 351, 1361-1366.
Gilbert G.S. & Webb C.O. (2007). Phylogenetic signal in plant pathogen-host range.
Proceedings of the National Academy of Sciences of the United States of America,
104, 4979-4983.
Janz N. & Nylin S. (1998). Butterflies and plants: A phylogenetic study. Evolution, 52,
486-502.
176 Keane R.M. & Crawley M.J. (2002). Exotic plant invasions and the enemy release
hypothesis. Trends in Ecology & Evolution, 17, 164-170.
Lambdon P.W. & Hulme P.E. (2006). How strongly do interactions with closely-related
native species influence plant invasions? Darwin's naturalization hypothesis
assessed on Mediterranean islands. Journal of Biogeography, 33, 1116-1125.
Levine J.M., Adler P.B. & Yelenik S.G. (2004). A meta-analysis of biotic resistance to
exotic plant invasions. Ecology Letters, 7, 975-989.
Lewinsohn T.M., Novotny V. & Basset Y. (2005). Insects on plants: diversity of
herbivore assemblages revisited. Annual Review of Ecology and Systematics, 36,
597-620.
Mack R.N. (1996). Predicting the identity and fate of plant invaders: Emergent and
emerging approaches. Biological Conservation, 78, 107-121.
Mack R.N., Simberloff D., Lonsdale W.M., Evans H., Clout M. & Bazzaz F.A. (2000).
Biotic invasions: Causes, epidemiology, global consequences, and control.
Ecological Applications, 10, 689-710.
Mitchell C.E., Agrawal A.A., Bever J.D., Gilbert G.S., Hufbauer R.A., Klironomos J.N.,
Maron J.L., Morris W.F., Parker I.M., Power A.G., Seabloom E.W., Torchin M.E.
& Vazquez D.P. (2006). Biotic interactions and plant invasions. Ecology Letters,
9, 726-740.
Naeem S., Knops J.M.H., Tilman D., Howe K.M., Kennedy T. & Gale S. (2000). Plant
diversity increases resistance to invasion in the absence of covarying extrinsic
factors. Oikos, 91, 97-108.
177 Odegaard F., Diserud O.H. & Ostbye K. (2005). The importance of plant relatedness for
host utilization among phytophagous insects. Ecology Letters, 8, 612-617.
Proches S., Wilson J.R.U., Richardson D.M. & Rejmanek M. (2008). Searching for
phylogenetic pattern in biological invasions. Global Ecology and Biogeography,
17, 5-10.
Rejmanek M. (1996). A theory of seed plant invasiveness: The first sketch. Biological
Conservation, 78, 171-181.
Ricciardi A. & Ward J.M. (2006). Comment on "Opposing effects of native and exotic
herbivores on plant invasions". Science, 313.
Richardson D.M., Allsopp N., D'Antonio C., Milton S.J. & Rejmanek M. (2000). Plant
invasions - the role of mutualisms. Biological Reviews, 75, 65-93.
Strauss S.Y., Webb C.O. & Salamin N. (2006). Exotic taxa less related to native species
are more invasive. Proceedings of the National Academy of Sciences of the United
States of America, 103, 5841-5845.
Strong D.R., Lawton J.H. & Southwood R. (1984). Insects on plants. Harvard University
Press, Cambridge, Mass., USA.
Torchin M.E., Lafferty K.D., Dobson A.P., McKenzie V.J. & Kuris A.M. (2003).
Introduced species and their missing parasites. Nature, 421, 628-630.
Vamosi S.M., Heard S.B., Vamosi J.C. & Webb C.O. (2009). Emerging patterns in the
comparative analysis of phylogenetic community structure. Molecular Ecology,
18, 572-592.
Webb C.O., Ackerly D.D., McPeek M.A. & Donoghue M.J. (2002). Phylogenies and
community ecology. Annual Review of Ecology and Systematics, 33, 475-505.
178 Weiblen G.D., Webb C.O., Novotny V., Basset Y. & Miller S.E. (2006). Phylogenetic
dispersion of host use in a tropical insect herbivore community. Ecology, 87, S62-
S75.
Williamson M.H. & Fitter A. (1996). The characters of successful invaders. Biological
Conservation, 78, 163-170.
179