Quick viewing(Text Mode)

Lifehistory Evolution in Rangeshifting Populations

Lifehistory Evolution in Rangeshifting Populations

CONCEPTS & SYNTHESIS EMPHASIZING NEW IDEAS TO STIMULATE RESEARCH IN

Ecology, 91(6), 2010, pp. 1617–1627 Ó 2010 by the Ecological Society of America

Life-history evolution in range-shifting populations

1 BENJAMIN L. PHILLIPS, GREGORY P. BROWN, AND RICHARD SHINE School of Biological Sciences, University of , New South Wales 2006

Abstract. Most evolutionary theory does not deal with populations expanding or contracting in space. Invasive , climate change, epidemics, and the breakdown of dispersal barriers, however, all create populations in this kind of spatial disequilibrium. Importantly, spatial disequilibrium can have important ecological and evolutionary outcomes. During continuous range expansion, for example, populations on the expanding front experience novel evolutionary pressures because frontal populations are assorted by dispersal ability and have a lower density of conspecifics than do core populations. These conditions favor the evolution of traits that increase rates of dispersal and reproduction. Additionally, lowered density on the expanding front eventually frees populations on the expanding edge from specialist, coevolved enemies, permitting higher investment into traits associated with dispersal and reproduction rather than defense against pathogens. As a result, the process of range expansion drives rapid life-history evolution, and this seems to occur despite ongoing serial founder events that have complex effects on at the expanding front. Traits evolving on the expanding edge are smeared across the landscape as the front moves through, leaving an ephemeral signature of range expansion in the life-history traits of a species across its newly colonized range. Recent studies suggest that such nonequilibrium processes during recent population history may have contributed to many patterns usually ascribed to evolutionary forces acting in populations at spatial equilibrium. Key words: climate change; contemporary evolution; defense allocation; ; dispersal rate; enemy escape; genotype ‘‘smearing’’; nonequilibrium processes; range shift; rapid life-history evolution; reproductive rate.

INTRODUCTION Understanding the evolutionary forces at work during range shift is of more than theoretical interest. For The vast majority of evolutionary theory is predicated example, the rate at which spread is an on a reasonable simplifying assumption: that popula- important consideration in our attempts to manage their tions are at spatial equilibrium. By this, we mean that impacts (Travis and Park 2004). Such rates can be the available to the population (or metapopula- radically modified by evolutionary shifts in life history tion) does not expand or contract in space. That (Holt et al. 2005, Phillips et al. 2008). The rates and assumption is often invalid, however. Populations move correlates of population expansion also affect our through space; ranges expand and contract, and understanding of historical and thus the dispersal barriers are surmounted (Vermeij 2005). In historical events that have generated current this review, we argue that such nonequilibrium popula- (Schoonmaker and Foster 1991, Clark et al. 1998, Davis tions experience different selective forces than do et al. 2005). Range shifts are also critical to the impact populations in spatial equilibrium, and we suggest that of climate change on biodiversity, because species that range shift can have profound impacts on the evolution can physically track climate change are likely to face a of a species’ life history. These evolutionary forces can lower impact than those that cannot (Thomas et al. cause change in life history over ecological timescales. 2004, Mustin et al. 2009). If species’ life histories evolve during climate-induced range shift, then the ramifica- tions for biodiversity conservation may be far-reaching Manuscript received 24 May 2009; revised 1 September 2009; indeed. accepted 15 September 2009; final version received 18 October 2009. Corresponding Editor: J. J. Stachowicz. Any case of continuous range shift includes two 1 E-mail: [email protected] unavoidable aspects: on the edge of the range, an 1617 1618 BENJAMIN L. PHILLIPS ET AL. Ecology, Vol. 91, No. 6

invading population (1) is assorted by dispersal ability which habitat quality is homogenous and not limiting (i.e., only the best dispersers make it to the new range (and hence, a population is expanding into new suitable edge each generation), and (2) exhibits a lower-than- habitat). In these situations, where a species is shifting normal population density. These outcomes (which we its range, some very clear predictions emerge. In will explain in more detail) have at least three particular, during range expansion we can expect consequences for life-history evolution in the vanguard dispersal rates on the advancing front to evolve upward of invading populations: first, they encourage the due to spatial selection. evolution of increased dispersal rate; second, they encourage the evolution of increased reproductive rate; Spatial selection: a runaway evolutionary process and third, they lead to the extirpation of specialist Why do we expect dispersal to evolve upward on the enemies on the invasion front. expanding front during range advance? The shift is a natural consequence of range expansion and occurs due EVOLUTION OF DISPERSAL to a process that we refer to here as ‘‘spatial selection’’ Populations spread through a combination of dis- (Fig. 1). Imagine a population where all the individuals persal and (Fisher 1937, Skellam disperse, then reproduce, and then die. If this population 1951). Traditionally, models of range advance suggested is placed into a large area of suitable vacant habitat, the waves of advance that traveled at constant speed, an population will advance its range as individuals disperse assumption at odds with many empirical observations. (Fig. 1A). Importantly, the process of dispersal effec- Theoretical advances in the last 30 years show that tively sorts individuals through space by dispersal accelerating range advance can arise from subtleties in ability. That is, individuals on the edge of the expanding how dispersal and population growth are modeled (Kot population front are at that edge simply because they et al. 1996, Shigesada and Kawasaki 1997, Hastings et dispersed farther than other individuals in the popula- al. 2005), but these models have continued to treat tion: the process of dispersal has spatially assorted the dispersal and population growth as constant through best dispersers in the population and placed them on the time. Here, we suggest that evolution often may drive expanding front (Fig. 1B). Now, because all the best changes in these critical parameters during invasion, and dispersers are in the same place at the same time, they thus accelerate range advance. YNTHESIS will tend to breed with each other (the ‘‘Olympic Village Dispersal, defined here as displacement between effect’’). Thus, if any component of dispersal ability is birth and reproduction (Ronce 2007), appears to be a

&S heritable in this , the offspring of the highly labile trait (Roff and Fairbairn 2001, Bowler individuals on the front will tend to have higher and Benton 2005, Ronce 2007). Part of this lability dispersal ability than the offspring of individuals from stems from the trait’s complexity; dispersal can be the core of the population (Fig. 1C). If we imagine this achieved actively (in which case, numerous behavioral process occurring every generation as the population traits are important; Barton´et al. 2009), passively (in ONCEPTS expands, the process of spatial assortment for dispersal

C which case, numerous morphological traits are impor- tant), or both; dispersal also can be condition ability behaves like a runaway evolutionary process, dependent (Ferriere et al. 2000, Bowler and Benton continually selecting for increased rates of dispersal on 2005). Even an apparently simple example, a wind- the advancing front. dispersed seed, will disperse according to its mass, the Density effects interact with this process of spatial height from the ground of its release, its buoyancy, the assortment. Indeed, the only process that can counteract timing of its release, and its tendency to settle. The the Olympic Village effect is an , where low- complex phenotypic determinants of dispersal mean density populations confer lower survival or reproduc- that selection for enhanced rates of dispersal has tive rate on their members (Travis and Dytham 2002). In numerous phenotypic targets. the absence of such an effect, individuals near the When habitat is homogenous, natural selection on advancing front may actually benefit from a low-density dispersal is driven by issues such as the avoidance of environment (i.e., fewer conspecifics in their vicinity) inbreeding and (Hamilton and May 1977, and, as a consequence, leave more offspring. Those Ferriere et al. 2000, Clobert et al. 2001, Ronce 2007). offspring will tend to inherit their parent’s high dispersal Where habitat is heterogeneous (on range edges or in rates. Thus, in cases where Allee effects are weak relative dynamic , for example), dispersal re- to the Olympic Village effect, dispersal rates will evolve sponds in complex ways to a combination of natural upward on the expanding front. Several simulation selection and spatial assortment by dispersal ability. In models now show that spatial selection drives the these cases, the dispersal strategy that evolves depends evolution of increased rates of dispersal. All of these strongly upon the spatial and temporal variation in models are spatially explicit, individual-based models, habitat quality, as well as the cost of dispersal (e.g., but vary in other aspects (Travis and Dytham 2002, McPeek and Holt 1992, Travis and Dytham 1999, Hughes et al. 2007, Phillips et al. 2008; O. J. Burton, Dytham 2009). In the present review, we ignore variance J. M. J. Travis, and B. L. Phillips, unpublished in habitat quality and focus instead on a simpler case in manuscript). June 2010 LIFE-HISTORY EVOLUTION AND RANGE SHIFT 1619

Dispersal, however, is only one of the two factors determining the rate of range expansion; the other is the rate of population growth. Increases in the rate of population growth also will accelerate invasion (Holt et al. 2005). As we will explain, we expect evolution to lead to increases in the rate of population growth on the expanding population front.

EVOLUTION OF TRAITS INCREASING POPULATION GROWTH RATE Density dependence, density independence, and the r–K selection spectrum It is a venerable result of life-history theory that, in exponentially growing populations (such as may often be the case at an invasion front), variance in fitness is summarized by variance in individuals’ reproductive rates (MacArthur and Wilson 1967); high reproductive rates are favored in exponentially growing populations. This is simply because individuals that replace them- selves more rapidly than others will soon numerically dominate such a population. In contrast, in populations that are density regulated (such as may often be the case at the core of the species’ range), fitness is tied to C variance in individuals’ ability to survive and reproduce ONCEPTS at high conspecific density (MacArthur and Wilson 1967, Roff 1993); individuals that leave more offspring behind in environments of high conspecific density are favored, so selection maximizes the of &S the population. For idealized, single life-history stage, discrete-generation organisms (as are found in many ecological and evolutionary models), this notion of YNTHESIS changed selection pressures between density-regulated and non-density-regulated populations is encapsulated in the r–K selection spectrum. The application of r and K selection to real organisms (such as are found in nature) is controversial, however. FIG. 1. Spatial selection on dispersal ability. (A) A population’s initial extent is represented by the dark gray area, This controversy has involved the assignation of and its extent after a single generation of dispersal is shown by particular traits to r-selected or K-selected life histories the light gray area. (B) The population’s dispersal kernel (the (Pianka 1970), and the discovery that mortality rates distribution of per-generation dispersal distances in the that vary with life-history stage can produce complex population) defines both the spatial extent of the population following dispersal and the density of conspecifics in the newly selection scenarios that are not adequately explained by colonized area. After a dispersal event, only the best dispersers the r–K spectrum (Reznick et al. 2002). Nonetheless, r will be on the population front (area of kernel and population and K selection remains as a powerful theoretical delimited by the dashed lines). Given that these individuals are concept (Roff 1993), primarily because it neatly captures all in the same place at the same time, they will tend to breed the idea that optimal life-history strategies differ with each other as ‘‘edge parents’’ (assortative mating by dispersal ability). (C) If dispersal distance is heritable, the between density-regulated and non-density-regulated offspring of the individuals on the expanding front will have a populations: it is this idea that we refer to in our use dispersal kernel shifted to higher mean dispersal distances. of r and K selection here. The difficulty lies in applying Distances and frequencies are in arbitrary units, without values; this straightforward theoretical result to complex life tick marks are simply to show that the three panels are aligned on the same x-axis. histories, where, for example, reproductive rate depends upon survival at different ages, age at first reproduction, number of offspring, variation in number of offspring else is not equal; the number of offspring may increase, with age, and age at last reproduction, by claiming that but another trait (e.g., survival) may decrease. These a shift in a single trait, e.g., number of offspring at a practical complexities have seriously muddied the water given age, accurately measures a shift in reproductive around the theoretical concept of r and K selection. rate. All else being equal, an increase in the number of Nonetheless, life-history theory clearly shows that offspring will increase the reproductive rate, but often optimal life-history strategies differ between density- (particularly in comparative studies across species) all regulated and non-density-regulated populations. In- 1620 BENJAMIN L. PHILLIPS ET AL. Ecology, Vol. 91, No. 6

place where population growth is unfettered by density regulation. Therefore, even in species with irruptive , the front will tend to be more consistently r selective than the range core. On the expanding edge of a population, life-history strategies with high reproductive rates will be favored. Given that the expanding range edge also must be made up of the best dispersers in the population, we can see that range expansion selects for a population of highly dispersive individuals with high reproductive rates (O. J. Burton, J. M. J. Travis, and B. L. Phillips, unpublished manuscript). If we remember that the rate at which a population expands is a function of dispersal and population growth rate, then we can see that the process of range expansion causes selection on the very traits that drive that expansion. Evolution will cause accelerating range advance.

TRAIT SMEARING AND EMPIRICAL PATTERNS Given that the invasion front is the seed for all later populations that arise in the areas it passes over, genotypes on the invasion front will tend to be ‘‘smeared’’ over the landscape (Fisher 1937, Klopfstein et al. 2006, Excoffier and Ray 2008, McInerny et al. 2009). This smearing effect can even lead to maladaptive genes being widespread behind the invasion front (Travis et al. 2007, Burton and Travis 2008). Consider- YNTHESIS FIG. 2. The shift in the selective environment on an invasion front, where K is carrying capacity. As the invasion able time may be needed for natural selection to remove proceeds there will always be an area at the expanding maladapted genes following the passing of the invasion &S population front experiencing assortative mating by dispersal front, particularly if there is little or no variation at that ability, low conspecific density, serial founder events, and lowered rates of parasite transmission/immigration. By con- locus (in which case mutation or immigration has to trast, these conditions do not occur behind the invasion front. boost diversity before selection can operate; Excoffier Thus, on the expanding front, life-history traits are under a and Ray 2008). Insofar as they have a genetic basis, profoundly different selective environment when compared to phenotypes at the invasion front also will be smeared ONCEPTS the core of the range.

C across the landscape. Evolution occurring on the invasion front thus can leave an ephemeral signature deed, a shift from density-dependent to density-inde- of range advance in the variation of life-history traits pendent regulation on an invasion front, and the across the invasion history (Travis and Dytham 2002, selection that follows from this shift, may help to Phillips et al. 2008). explain the acceleration of species invasions. The signature of range advance is ephemeral because By definition, on an expanding population front, the natural selection behind the expanding front will act to density of conspecifics grades from something near optimize the phenotype for the new conditions of (local) carrying capacity to zero (Fig. 2). One fundamental spatial equilibrium. Thus, trait values that evolve on the implication of this density gradient is that individuals on expanding front will be less fit, following colonization of the newly colonized edge of the range probably will an area, than trait values that evolve in the population experience exponential population growth. Individuals core. Fitness is, however, relative to local phenotypes, so farther back into the core of the range, however, will the rate at which trait values decay back to equilibrium probably experience density-regulated population values will depend upon the rate at which mutation and growth (Brook and Bradshaw 2006). Thus, an invasion immigration can provide the variance on which selection front can be thought of as a steep gradient between r- can act. Nonetheless, some time after the population selective and K-selective environments. Critically, this stops expanding, the signature of range advance will shift in environments is experienced within a single have completely decayed. species over a relatively narrow . Thus, more Because traits are smeared across the landscape correctly, invasion front individuals face an r-selective during invasion, comparisons between frontal and environment relative to the individuals within the core of successively older populations of an ongoing invasion the range. As the population expands through space, the should yield clines in life-history traits that reflect the invasion front always represents an environment with differing selective forces operating on the invasion front low conspecific density and thus consistently provides a vs. at spatial equilibrium (behind the front). June 2010 LIFE-HISTORY EVOLUTION AND RANGE SHIFT 1621

SELECTION ON TRAITS AFFECTING DISPERSAL Traits associated with reproductive rate AND REPRODUCTIVE RATES Reproductive rate will be increased by an increase in Unless we can measure intergenerational dispersal or fecundity or a decrease in the age at first reproduction the rate of population growth (which is always difficult, (Cole 1954, Lewontin 1965). In perennial species, and usually impossible), we will be forced to examine reproductive rate also can be increased by a lengthening these life-history parameters using data on the distribu- of the reproductive period, or by a decrease in the peak tions of phenotypic traits. In such cases (e.g., where we reproductive age, although a decrease in the age at first measure wing length as a surrogate for dispersal, or reproduction remains the most efficient way of increas- fecundity as a surrogate for reproductive rate), we are ing the reproductive rate (Lewontin 1965). Generally forced to assume that all else is equal. This assumption is though, having more offspring, and/or having them risky when comparisons are made across species, earlier (if all else remains the same) will increase the because such comparisons are muddied by differences individual’s reproductive rate. Intriguingly, these life- in habitat and geographic range as well as greater time history traits often differ between native populations to a common ancestor, so major restructuring of life and invasive populations within a single species (‘‘home- history is possible. The ‘‘all else is equal assumption’’ is, and-away comparisons’’). In trees (genus Pinus), however, presumably less risky within a single species for example, invasive populations have shorter genera- and over the short timescales within which evolution can tion times and smaller (more dispersive) seeds (Rejma- act on life-history traits during an invasion. With these nek and Richardson 1996). In a common garden, purple caveats, we now turn to empirical observations of shifts loosestrife from its invasive range in in traits associated with dispersal and reproductive rates grows more rapidly (and so presumably reaches during range expansion. maturity earlier) than does loosestrife from its native range (Blossey and Notzold 1995). In fact, with a few

Traits associated with dispersal notable exceptions (Willis et al. 2000, van Kleunen and C Empirical results demonstrating evolved increase in Schmid 2003, Maron et al. 2004), plants from invaded ONCEPTS dispersal on expanding range fronts are beginning to areas tend to grow faster, grow larger, and/or set more accumulate. The process of spatial assortment by seed, than do conspecifics from the native range (Keane dispersal ability is nicely illustrated by Glanville fritillary and Crawley 2002, Bossdorf et al. 2005, Colautti et al. butterflies on the A˚land islands, where recently colo- 2006). &S nized environments are dominated by highly dispersive Such patterns often have been interpreted in the light individuals (Hanski et al. 2002, 2004). Spatial assort- of the ‘‘enemy release’’ hypothesis; that is, invaders ment by dispersal ability also occurs in Australian cane escape their coevolved enemies (e.g., predators, para- YNTHESIS toads, where the first individuals to colonize an area sites), and so can invest more into growth and tend to be the individuals with the longest legs (and thus, reproduction instead of defense. Although ‘‘enemy the greatest dispersal ability; Phillips et al. 2006). Under release’’ is usually attributed to accidents of transport spatial selection, we expect evolved shifts in dispersal and introduction to new lands (often, an invader’s across a species’ invasive range. Increasingly, such shifts natural enemies are not transported with it), non- are being documented. In North America, for example, equilibrial models suggest that the same phenomenon lodgepole expanded their range northward follow- can also arise within the invaded range as an outcome of ing the last glaciation; the northernmost populations are rapid range expansion. Similarly, many of the differ- less than 100 years old. Recently colonized populations ences observed between populations in native vs. have lighter seeds with a larger wing size : seed mass invaded ranges may have evolved in the invaded range ratio, indicating increased dispersal ability on the during the process of range expansion, rather than invasion front (Cwynar and MacDonald 1987). In reflecting the nonrandom nature of successful establish- Australia, cane toads on an accelerating invasion front ment (‘‘propagule bias,’’ Colautti et al. 2006, Blackburn show similarly clear evidence of increased dispersal et al. 2009), or evolution occurring at spatial equilibrium ability, via changes in morphology (leg length) as well as (Blossey and Notzold 1995). In support of this behavior (frontal toads move more often, move farther possibility, several recent studies have reported evolved when they do move, and follow straighter paths: shifts in life history within the invaded range. For (Phillips et al. 2006, 2008, Alford et al. 2009). In Britain, example, when grown in a common garden, Chinese crickets expanding their range show evolved increases in tallow trees from their invaded range grow faster and wing length and flight stamina on the expanding front mature earlier than do those from their native range. (Simmons and Thomas 2004). Also in Britain, butterflies That pattern is potentially consistent with enemy release expanding northward because of climate change exhibit (optimality) or propagule bias. Within their invaded evolved shifts toward larger wings and flight muscles range, however, tallow trees from recently colonized (Hughes et al. 2007). As we look further into the areas also grow faster and mature earlier in a common possibility of rapid evolution of dispersal-relevant traits garden than do trees from older populations in the during range advance, many more examples are likely to invaded range (Siemann and Rogers 2001). That second emerge. pattern cannot be attributed to propagule bias, and 1622 BENJAMIN L. PHILLIPS ET AL. Ecology, Vol. 91, No. 6

instead must reflect processes at work in the invaded case, traits associated with dispersal and reproduction range; presumably, evolution occurring on the invasion may be able to increase at the expense of traits that are front. Similarly, in their invasive Australian range, cane not tied to fitness in the invasion vanguard (e.g., traits toads from recently colonized areas grow faster than do associated with defense against enemies). toads from older, long-established populations, even when raised under identical conditions (Phillips 2009). ESCAPE FROM ENEMIES DURING INVASION Glanville Fritillaries in recently colonized environments Introduced species often have fewer parasites and tend not only to be highly dispersive, but also to invest pathogens in their invasive range than in their native more in reproduction (Hanski et al. 2006, Saastamoinen range (Mitchell and Power 2003, Torchin et al. 2003, 2007). Few studies to date have looked for such changes Perkins et al. 2008). This phenomenon is driven in life history traits of introduced species across their primarily by founder events: typically only a few invasive range (Gaston 2009). We expect a rapid individuals are transported to the introduced range increase in such studies, and hence in insights into this and these few individuals are, by chance, unlikely to process, in the years ahead. Importantly, such studies harbor all the parasites and pathogens that attack the will need to pay careful attention to the invasion history species in its native range. Thus, as a general rule, of the organism in its invaded range. It may be that individual invaders will need to invest less in defenses many of the exceptions (species that do not show a shift against natural, coevolved enemies than had been true in in life history in home-and-away comparisons) may, their native range (Muller-Scharer et al. 2004). This given our new understanding, simply reflect samples well-understood pattern is often invoked to explain the taken from an inappropriate part of the invaded range. success of invasive species (Blossey and Notzold 1995, Keane and Crawley 2002, Colautti et al. 2004). Less TRADE-OFFS appreciated, however, is the ongoing nature of the Both models and data suggest that, during continuous founder effect leading to escape from enemies. Follow- range expansion, selection drives the evolution of ing colonization, as a population expands through increased dispersal and reproductive rate. Whether or space, the vanguard population undergoes a founder not individuals in invasive populations can evolve to event every generation (few individuals make up the become more highly dispersive, rapidly growing, and expanding edge; Klopfstein et al. 2006). Thus, in each YNTHESIS highly fecund (relative to conspecifics in older popula- generation, as the population expands, there is a high tions) depends on the nature of trade-offs between traits probability that parasites and pathogens will not be &S (O. J. Burton, J. M. J. Travis, and B. L. Phillips, present on the invasion front (Fig. 2, unless parasites unpublished manuscript). If dispersal trades off directly disperse independently from, and farther than, their against the traits determining reproductive rates (as it hosts; Hilker et al. 2005). When parasite transmission is does in many theoretical models to date), then both of density dependent, this serial founder effect is com- these attributes cannot increase at the same time. Given pounded by low transmission rates at low host densities ONCEPTS on the invasion front. Thus, over multiple generations of C the complexity of dispersal and reproductive rate, however, it seems unlikely that these two attributes expansion, even high-prevalence pathogens and para- directly trade off against each other. The complex sites will be left behind (Phillips et al. 2010). Thus, phenotypic determinants of dispersal and reproductive invasive species not only start with fewer coevolved rate provide many selective targets, generating a vast enemies, but also, once range advance begins, on the number of possible trade-off relationships. As a simple expanding front they may lose even the few coevolved example, if dispersal is determined entirely by three parasites that they brought with them from the native traits (e.g., seed mass, wing size, and release height), and range. Similarly, spreading their range reproductive rate is determined entirely by three traits through climatic shift may lose their coevolved enemies (e.g., growth rate, size at maturity, and seed number), on the expanding population front. then there are nine possible trade-offs between traits The recently discovered phenomenon of ongoing determining dispersal and population growth. What are enemy escape during invasion has now been demon- the chances that all of these trade-offs are negative? If strated in two different spatially explicit models (Moor- some of them are neutral, or even positive (e.g., perhaps croft et al. 2006, Phillips et al. 2010). Recently the effect there is no relationship between seed mass and size at has been empirically demonstrated in butterflies expand- maturity), then both dispersal and reproductive rates ing northward due to climate change in Britain may increase simultaneously under the selective forces (Menendez et al. 2008), and in introduced cane toads experienced at the expanding range edge. in Australia (Phillips et al. 2010). In these cases, parasite Alternatively, dispersal and reproductive traits may richness and/or prevalence was lower on the expanding trade off against other traits unrelated to dispersal and front. reproduction. In particular, dispersal and reproductive Invasive populations do eventually accumulate ene- rates may trade off against traits that are important in mies in their introduced range (e.g., Carroll and Boyd equilibrium populations; traits that increase fitness in 1992, Perkins et al. 2008). Sometimes these enemies times and places of high conspecific density. In such a make instantaneous shifts to the invasive species, but June 2010 LIFE-HISTORY EVOLUTION AND RANGE SHIFT 1623 more often some time elapses before enemies can host- a random sample of the native population Colautti et al. switch effectively. Given this common lag time in host 2006, Blackburn et al. 2009), second as a consequence of switching, vanguard populations of invaders also may range advance (evolution in nonequilibrium popula- be relatively free of non-coevolved enemies. Vanguard tions), and then last as a consequence of standard populations may have arrived so recently that local evolutionary processes (at spatial equilibrium). Differ- enemies have had no time to effect host-switching. Thus, ences between native and invaded ranges can be due to on the invasion front, an invader will only experience any one or all of these processes, so such comparisons impact from the most generalist of enemies: those that necessarily are fraught with ambiguity. Nonetheless, can host-switch instantaneously. One clear implication these home-and-away studies make up the bulk of our of these patterns is that, on an expanding population information to date. What changes in defense do we front, resources normally invested in defense against observe between native and introduced ranges? coevolved or specialized enemies can now be redirected to dispersal and reproduction. Plants We can arbitrarily divide plant chemical defenses into THE EVOLUTION OF DEFENSES AGAINST ENEMIES two categories: quantitative and qualitative (Stamp If vanguard populations of a range-expanding species 2003). This division is artificial and will not capture are free of coevolved and specialized enemies, and are the full variation in plant defenses, but is a useful under strong selection for increased rates of dispersal didactic distinction. Qualitative defenses are toxins (e.g., and reproduction, we expect to see evolved shifts in alkaloids and glycosides); they tend to be cheap to defenses against enemies. General defenses will still be produce and effective against broad suites of enemies important, as the invader will encounter generalist (Muller-Scharer et al. 2004). Qualitative toxins tend to enemies as it invades new areas (Muller-Scharer et al. be less effective against specialized, coevolved, enemies

2004). In contrast, specific defenses (those tailored to (which can evolve numerous coping mechanisms, and C deal with coevolved and specialized enemies) will no may even use the qualitative toxins to identify their host; ONCEPTS longer be under strong maximizing selection. Thus, Muller-Scharer et al. 2004). The other major type of evolved shifts in defenses may occur as a consequence of plant chemical defenses, quantitative defenses, act by trade-offs between defense capacity vs. traits associated decreasing the plant’s nutritional value: either by with dispersal and reproduction. Strong directional decreasing digestibility (e.g., via high lignin content), selection on dispersal and reproductive rates may create or nutrient availability (e.g., via high tannin content; &S strong directional selection against enemy defenses if Stamp 2003). As a gross generalization, quantitative any trade-offs exist, and the release from coevolved defenses are costly, because they are composed of large YNTHESIS enemies in the population vanguard allows lowered molecules that need to be at high concentrations to be investment in these defenses to evolve. effective (Stamp 2003). Similarly, quantitative toxins Dispersal and reproductive rates are complex traits, tend to be constitutive defenses (i.e., they are always and the same can be said of traits associated with present), whereas qualitative defenses can be either defense against enemies. In motile animals, these constitutive or inducible following attack (Strauss et al. defenses will be manifest as predator avoidance/escape 2002). Quantitative defenses are useful against specialist strategies (Edmunds 1974), and in the immune system. because they constrain the nutrient uptake In plants and sessile animals, these defenses will be (and thus growth and reproduction) of the . manifest as physical and chemical defenses. Very few Thus, as a plant expands its range and leaves specialist studies examine shifts in these traits within the invasive enemies behind, we expect to see decreased quantitative range, but numerous studies have examined shifts in defenses, and may or may not see increased qualitative defenses between native and invasive ranges (home-and- defense (Muller-Scharer et al. 2004, Lankau 2007). We away comparisons). Although these studies are often may also expect a shift toward inducible defenses (which unclear about exactly where invasive range samples elicit costs only after being induced), as opposed to came from in relation to the spread history, we note constitutive defenses (which incur a constant cost). that, more often than not, these studies find changed Presumably, all defenses are costly, although these investment in defenses in the invaded range (Keane and costs may range from undetectably small to quite Crawley 2002, Lee and Klasing 2004). Again, we point obvious, depending on the defensive trait being exam- out that these changes can occur during range expansion ined. Importantly, these costs are often expressed via and be left behind as an ephemeral signature of that direct trade-offs with growth and reproduction (Strauss range expansion. Evolutionary change occurring as a et al. 2006). Thus, on the expanding range edge of range- consequence of mechanisms acting in equilibrium shifting plants, we might expect evolved increases in populations (i.e., natural selection at spatial equilibri- dispersal and reproductive rates, which, in turn, drive um) is then overlaid on this trait landscape. Thus, decreases in costly defenses (usually quantitative defens- differences between invaded and native ranges occur es directed against specialist enemies). Lowered quanti- first as a consequence of founder and establishment tative defenses, in turn, may encourage greater biases (‘‘propagule bias’’: introduced populations are not investment in less costly, inducible defenses (usually 1624 BENJAMIN L. PHILLIPS ET AL. Ecology, Vol. 91, No. 6

qualitative defenses, to maintain protection against and that innate responses (although costly when generalist enemies). Empirical results comparing inva- activated) are cheaper, in the long term, than humoral sive and native populations support this idea (Muller- immunity, particularly when serious immune challenges Scharer et al. 2004). First, plants from invasive are rare. Given that humoral immunity is also mediated populations often show higher growth rates than plants by the innate arm of the immune system, and that from the native range (Keane and Crawley 2002, Joshi humoral immunity is the derived state, presumably and Vrieling 2005). Second, these higher growth rates having evolved to deal with an increased evolutionary often are associated with reduced quantitative defenses discrepancy between long-lived organisms and their (Siemann and Rogers 2001, Muller-Scharer et al. 2004). short-lived pathogens, an argument could be mounted Last, plants from introduced populations tend to be for the down-regulation of humoral immunity in better adapted to generalist herbivores than are those response to life-history evolution and enemy release from native populations (Schierenbeck et al. 1994, Joshi during range expansion. Similarly, innate immunity and Vrieling 2005, Engelkes et al. 2008, Cano et al. might be sufficient to deal with nonspecialist enemies 2009). Unfortunately, to our knowledge, there are no (although systemic inflammation would still be costly, studies examining defensive traits across an invasive and may be down-regulated independently of the rest of range. Given that defensive traits are likely to evolve the innate system), reducing selection for the mainte- during range advance, this is a serious gap in our nance of humoral immunity in vanguard populations. understanding of the evolutionary dynamics of range These predictions are in partial opposition to those of shift. Lee and Klasing (2004). Although improving steadily, our understanding of Animals plant and animal defenses and their various costs (as Animals can exhibit behavioral or morphological measured by trade-offs with dispersal and reproductive to avoid or escape enemies, and these traits rate) is poor. Hence, it is difficult to predict the are likely to be under selection during range advance. evolution of defensive traits during range expansion. Because these traits vary from species to species, we Nonetheless, when taken on a case-by-case basis, we confine ourselves to discussion of a defense trait expect negative selection on an expanding population common to most animals: the immune system. The edge against defensive traits that carry reproductive or YNTHESIS immune system is somewhat more complex than plant dispersal costs, and that are tailored toward specialist defenses, but some parallels can be found. Lee and enemies. To our knowledge, no studies have compared &S Klasing (2004) suggest that when animals are confront- aspects of the immune system over the course of an ed with fewer, less specialized enemies, costly innate invasion. However, Møller and Cassey (2004) show an immune defenses (i.e., systemic inflammation) should be increase in T-cell-mediated immunity associated with down-regulated. This rapid-response immune defense is invasion success in , and Lee et al. (2005) show not necessary, they argue, when facing attack by a non- lowered inflammatory response in an invasive species ONCEPTS

C coevolved pathogen, and may, in fact, cause damaging when compared with a noninvasive congener. Martin et overreaction to a pathogen that is incapable of infecting al. (2007) show that immune systems are geared toward its novel host anyway. When facing non-coevolved low-cost generic defenses in short-lived, high-growth pathogens, a heightened humoral response (i.e., that species of Peromyscus mouse more than in long-lived, driven primarily by antibodies) may be adequate. slow-growing congeners. Given the clear importance of Humoral responses are slower but (they argue) less shifts in the immune system for understanding the costly, and can ‘‘remember’’ pathogens (such that a process of biological control, as well as the long-term repeat exposure will generate a more rapid response). outcomes of range shift on native species shifting their Therefore, Lee and Klasing (2004) suggest that species range through climate change, this avenue of study with a low systemic inflammatory response but a high merits further investigation. humoral response will make better invaders. Inverte- brates (which lack humoral immunity) can become CONCLUSION better invaders simply by down-regulating their innate One clear implication from our emerging understand- responses. ing of evolution on expanding range edges is that we As with our discussion of plant defenses, these need to be very careful about how we design studies generalities are a gross simplification of reality. Indeed, investigating the evolution of invasive species. Many the distinction between innate and humoral immunity is studies have compared traits between native and somewhat artificial (much like the distinction between invaded ranges (home-and-away comparisons), but quantitative and qualitative defenses in plants), but patterns at this level are a consequence of three remains a useful didactic device (Murphy et al. 2008). evolutionary phases. Additionally, the costs of the immune system are poorly First, the introduction and establishment phase will understood. For example, it could equally be argued determine the genotypes available in the invaded range that maintenance and activation of the humoral system (Blackburn et al. 2009). Where multiple introductions is costly (massive cell proliferation and maintenance), occur, a population with heightened genetic June 2010 LIFE-HISTORY EVOLUTION AND RANGE SHIFT 1625 diversity may be the starting point (Kolbe et al. 2004). northward since the last glaciation showed clear patterns Alternatively, if there is a single introduction, lowered associated with evolution under spatial disequilibrium. genetic diversity may be the starting point. In either At the recently expanding population edge, seed case, the founder event of introduction and the morphology indicated heightened dispersal rates relative demographic process of establishment means that the to trees from older populations. If we were not aware genotypes and phenotypes available in the introduced that this species had recently shifted its range, we may range are unlikely to be representative of those in the have come up with an (erroneous) explanation for this invaded range (Blackburn and Duncan 2001, Simons pattern based on evolution at spatial equilibrium. This 2003, Colautti et al. 2006, Blackburn et al. 2009). Thus, example shows the critical importance of understanding differences between invaded and native ranges can population history before investigating the evolution of emerge instantaneously at this juncture. life-history traits using population comparisons. The second evolutionary phase, spatial disequilibri- An interesting outcome of evolution on expanding um, occurs following introduction, during range expan- range edges is that life-history traits on expanding range sion. This is the phase we have talked about extensively edges are selected toward values that increase the rate of in this paper, defined by two processes: spatial spread. Thus, lags in spread rate and accelerated spread assortment by dispersal ability and lowered conspecific (different sides of the same coin) are explicable through density on the invasion front. These two unavoidable rapid evolution. Evolutionarily accelerated range shift consequences of range expansion lead to predictable may be the most appropriate null expectation, rather evolutionary pressures in the vanguard population, and than constant rates of range expansion. Models of range this vanguard population smears these evolved traits spread that do not incorporate evolved shifts in life across the landscape, leaving an ephemeral signature of history are simpler, but must be carefully parameterized range expansion in the organism’s traits across its to give accelerated spread, and do not generate the many invasive range. testable hypotheses emerging from the framework that C

The last evolutionary phase occurring in species we present here. In particular, our prediction that both ONCEPTS introductions is that with which we are most familiar: dispersal and traits associated with reproductive rate are evolution by natural selection in populations at spatial maximized during range advance begs the obvious equilibrium. Many of the studies documenting shifts in question of trade-offs and thus suggests selective forces life history in invasive species have made evolutionary behind the evolution of defense mechanisms in range- arguments based on evolution at spatial equilibrium. shifting taxa. Previous hypotheses erected to explain &S For example, the ‘‘evolution of increased competitive shifts in life history between invaded and native ranges ability’’ hypothesis argues that invasive species will have focus purely on equilibrium theory, and so cannot point YNTHESIS fewer enemies in their introduced range, so the balance to selective forces driving these shifts; instead, they rely of any trade-off between competitive ability and defense on optimality arguments. will be tipped in favor of competitive ability (Blossey Evolution on expanding range edges is only now being and Notzold 1995). This may be true, but if the species understood as a potent force in the evolution of life becomes invasive and spreads its range following histories. Hence, there is massive scope for empirical and introduction, the process of range expansion drives theoretical work to investigate this process. Although we selection for increased reproductive rate on the invasion can make reasonably clear predictions about the shifts in front, and this process will leave an ephemeral signature life history that should occur during range expansion, through the invasive population. Rates of population the interaction between these selective processes and the growth and dispersal are likely to be highest on the effects of serial foundering and Allee effects will be a expanding edge of the invasive range and lowest at the challenging future problem. Acknowledging the critical core. Ideally, we should not attempt to test equilibrium role of rapid life-history evolution during range shift not hypotheses without first understanding the effect of only will clarify observed empirical patterns (such as nonequilibrium dynamics on the traits we are examin- shifts in life history between invaded and native ranges), ing. Comparison of life-history traits between invasive but also will clarify the process of range expansion. At a and native populations (‘‘home-and-away’’) is a com- time of high concern over invasive species and the mon research design, but rarely is attention paid to the biodiversity impacts of climate change, a focus on the location of the samples within the invasive range. We evolution of life histories during range advance may suggest that an understanding of variation within the have important outcomes. invasive range is critical before sensible comparisons can ACKNOWLEDGMENTS be made with the native range. We thank Justin Travis, Stuart Baird, Mark Urban, Olivia In species that are shifting (or that have recently Burton, and Arnaud Estoup for stimulating discussion around shifted) their range as a consequence of climate change, these issues. Much of this discussion was had around a small we expect to see two distinct evolutionary phases (rather workshop in Darwin funded by the Environmental Futures Network. Tim Blackburn, Justin Travis, and an anonymous than the three of invasive species): spatial disequilibri- referee provided valuable comments on a previous draft. um, followed by spatial equilibrium. In 1987, Cwynar Additional support was provided by the Australian Research and MacDonald showed that lodgepole pines expanding Council. 1626 BENJAMIN L. PHILLIPS ET AL. Ecology, Vol. 91, No. 6

LITERATURE CITED Excoffier, L., and N. Ray. 2008. Surfing during population expansions promotes genetic revolutions and structuration. Alford, R. A., G. P. Brown, L. Schwarzkopf, B. L. Phillips, and Trends in Ecology and Evolution 23:347–351. R. Shine. 2009. Comparisons through time and space suggest Ferriere, R., J. R. Belthoff, I. Olivieri, and S. Krackow. 2000. rapid evolution of dispersal behaviour in an invasive species. Evolving dispersal: where to go next? Trends in Ecology and Research 36:23–28. Evolution 15:5–7. Barton´, K. A., B. L. Phillips, J. M. Morales, and J. M. J. Travis. Fisher, R. A. 1937. The wave advance of advantageous genes. 2009. The evolution of an ‘intelligent’ dispersal strategy: Annals of Eugenics 7:355–369. biased, correlated random walks on patchy landscapes. Oikos Gaston, K. J. 2009. Geographic range limits: achieving 118:309–319. synthesis. Proceedings of the Royal Society B 276:1395–1406. Blackburn, T. M., and R. P. Duncan. 2001. Establishment Hamilton, W. D., and R. M. May. 1977. Dispersal in stable patterns of exotic birds are constrained by non-random . Nature 269:578–581. patterns in introduction. Journal of Biogeography 28:927– Hanski, I., C. J. Breuker, K. Scho¨ps, R. Setchfield, and M. 939. Nieminen. 2002. Population history and life history influence Blackburn, T. M., J. L. Lockwood, and P. Cassey. 2009. Avian the migration rate of female Glanville fritillary butterflies. invasions. Oxford University Press, Oxford, UK. Oikos 98:87–97. Blossey, B., and R. Notzold. 1995. Evolution of increased Hanski, I., C. Era¨lahti, M. Kankare, O. Ovaskainen, and H. competitive ability in invasive non-indigenous plants: a Sire´n. 2004. Variation in migration propensity among hypothesis. Journal of Ecology 83:887–889. individuals maintained by landscape structure. Ecology Bossdorf, O., H. Auge, L. Lafuma, W. E. Rogers, E. Siemann, Letters 7:958–966. and D. Prati. 2005. Phenotypic and genetic differentiation Hanski, I., M. Saastamoinen, and O. Ovaskainen. 2006. between native and introduced plant populations. Oecologia Dispersal-related life-history trade-offs in a butterfly meta- 144:1–11. population. Journal of Animal Ecology 75:91–100. Bowler, D. E., and T. G. Benton. 2005. Causes and Hastings, A., et al. 2005. The spatial spread of invasions: new consequences of animal dispersal strategies: relating individ- developments in theory and evidence. Ecology Letters 8:91– ual behaviour to spatial dynamics. Biological Reviews 80: 101. 205–225. Hilker, F. M., M. A. Lewis, H. Seno, M. Langlais, and H. Brook, B. W., and C. J. A. Bradshaw. 2006. Strength of Malchow. 2005. Pathogens can slow down or reverse evidence for density dependence in time series of invasion fronts of their hosts. Biological Invasions 7:817–832. 1198 species. Ecology 87:1445–1451. Holt, R. D., T. H. Keitt, M. A. Lewis, B. A. Maurer, and M. L. Burton, O. J., and J. M. J. Travis. 2008. Landscape structure Taper. 2005. Theoretical models of species’ borders: single and boundary effects determine the fate of mutations species approaches. Oikos 108:18–27. Hughes, C. L., C. Dytham, and J. K. Hill. 2007. Modelling and

YNTHESIS occurring during range expansions. Heredity 101:329–340. Cano, L., J. Escarre, K. Vrieling, and F. X. Sans. 2009. analysing evolution of dispersal in populations at expanding Palatability to a generalist herbivore, defense and growth of range boundaries. Ecological Entomology 32:437–445.

&S invasive and native Senecio species: testing the evolution of Joshi, J., and K. Vrieling. 2005. The enemy release and EICA increased competitive ability hypothesis. Oecologia 159:95– hypothesis revisited: incorporating the fundamental differ- 106. ence between specialist and generalist herbivores. Ecology Carroll, S. P., and C. Boyd. 1992. Host race radiation in the Letters 8:704–714. soapberry bug: natural history, with the history. Evolution Keane, R. M., and M. J. Crawley. 2002. Exotic plant invasions 46:1052–1069. and the enemy release hypothesis. Trends in Ecology and ONCEPTS Clark, J. S., et al. 1998. Reid’s paradox of rapid plant Evolution 17:164–170. C migration. BioScience 48:13–24. Klopfstein, S., M. Currat, and L. Excoffier. 2006. The fate of Clobert, J., E. Danchin, A. A. Dhondt, and J. D. Nichols, mutations surfing on the wave of range expansion. Molecular editors. 2001. Dispersal. Oxford University Press, Oxford, Biology and Evolution 23:482–490. UK. Kolbe, J. J., R. E. Glor, L. R. Schettino, A. C. Lara, A. Larson, and J. B. Losos. 2004. Genetic variation increases during Colautti, R. L., I. A. Grigorovich, and H. J. MacIsaac. 2006. biological invasion by a Cuban lizard. Nature 431:177–181. Propagule pressure: a null model for biological invasions. Kot, M., M. A. Lewis, and P. van den Driessche. 1996. Biological Invasions 8:1023–1037. Dispersal data and the spread of invading organisms. Colautti, R. L., A. Ricciardi, I. A. Grigorovich, and H. J. Ecology 77:2027–2042. Macisaac. 2004. Is invasion success explained by the enemy Lankau, R. A. 2007. Specialist and generalist herbivores exert release hypothesis? Ecology Letters 7:721–733. opposing selection on a chemical defense. New Phytologist Cole, L. C. 1954. The population consequences of life-history 175:176–184. phenomena. Quarterly Review of Biology 29:103–137. Lee, K. A., and K. C. Klasing. 2004. A role for immunology in Cwynar, L. C., and G. M. MacDonald. 1987. Geographical invasion biology. Trends in Ecology and Evolution 19:523– variation of lodgepole pine in relation to population history. 529. American Naturalist 129:463. Lee, K. A., L. B. Martin, and M. Wikelski. 2005. Responding Davis,M.B.,R.G.Shaw,andJ.R.Etterson.2005. to inflammatory challenges is less costly for a successful avian Evolutionary responses to changing climate. Ecology 86: invader, the (Passer domesticus), than its less- 1704–1714. invasive congener. Oecologia 145:244–251. Dytham, C. 2009. Evolved dispersal strategies at range margins. Lewontin, R. C. 1965. Selection for colonizing ability. Pages Proceedings of the Royal Society Biological Sciences B 276: 79–94 in H. Baker and G. Stebbins, editors. The genetics of 1407–1413. colonizing species. Academic Press, New York, New York, Edmunds, M. 1974. Defence in animals. Longman, New York, USA. New York, USA. MacArthur, R. H., and E. O. Wilson. 1967. The theory of Engelkes, T., E. Morrien, K. J. F. Verhoeven, T. M. Bezemer, island biogeography. Princeton University Press, Princeton, A. Biere, J. A. Harvey, L. M. McIntyre, W. L. M. Tamis, and New Jersey, USA. W. H. van der Putten. 2008. Successful range-expanding Maron, J. L., M. Vila, R. Bommarco, S. Elmendorf, and P. plants experience less above-ground and below-ground Beardsley. 2004. Rapid evolution of an invasive plant. enemy impact. Nature 456:946–948. Ecological Monographs 74:261–280. June 2010 LIFE-HISTORY EVOLUTION AND RANGE SHIFT 1627

Martin, L. B., Z. M. Weil, and R. J. Nelson. 2007. Immune Ronce, O. 2007. How does it feel to be like a rolling stone? Ten defense and reproductive pace of life in Peromyscus mice. questions about dispersal evolution. Annual Review of Ecology 88:2516–2528. Ecology, Evolution, and Systematics 38:231–253. McInerny, G. J., J. R. G. Turner, H. Y. Wong, J. M. J. Travis, Saastamoinen, M. 2007. Mobility and lifetime fecundity in new and T. G. Benton. 2009. How range shifts induced by climate versus old populations of the Glanville fritillary butterfly. change affect neutral evolution. Proceedings of the Royal Oecologia 153:569–578. Society B 276:1527–1534. Schierenbeck, K. A., R. N. Mack, and R. R. Sharitz. 1994. McPeek, M. A., and R. D. Holt. 1992. The evolution of Effects of herbivory on growth and allocation in dispersal in spatially and temporally varying environments. native and introduced species of Lonicera. Ecology 75:1661– American Naturalist 140:1010–1027. 1672. Menendez, R., A. Gonzalez-Megias, O. T. Lewis, M. R. Shaw, Schoonmaker, P. K., and D. R. Foster. 1991. Some implica- and C. D. Thomas. 2008. Escape from natural enemies tions of paleoecology for contemporary ecology. Botanical during climate-driven range expansion: a case study. Eco- Review 57:204–245. logical Entomology 33:413–421. Shigesada, N., and K. Kawasaki. 1997. Biological invasions: Mitchell, C. E., and A. G. Power. 2003. Release of invasive theory and practice. Oxford University Press, Oxford, UK. plants from fungal and viral pathogens. Nature 421:625–627. Siemann, E., and W. E. Rogers. 2001. Genetic differences in Møller, A. P., and P. Cassey. 2004. On the relationship between growth of an invasive tree species. Ecology Letters 4:514– T-cell mediated immunity in species and the establish- 518. ment success of introduced populations. Journal of Animal Simmons, A. D., and C. D. Thomas. 2004. Changes in dispersal Ecology 73:1035–1042. during species’ range expansions. American Naturalist 164: Moorcroft, P. R., S. W. Pacala, and M. A. Lewis. 2006. 378–395. Potential role of natural enemies during tree range expan- Simons, A. M. 2003. Invasive aliens and sampling bias. Ecology sions following climate change. Journal of Theoretical Letters 6:278–280. Biology 241:601–616. Skellam, J. G. 1951. Random dispersal in theoretical popula- Muller-Scharer, H., U. Schaffner, and T. Steinger. 2004. tions. Biometrika 38:196–218. Evolution in invasive plants: implications for biological Stamp, N. 2003. Out of the quagmire of plant defense control. Trends in Ecology and Evolution 19:417–422. hypotheses. Quarterly Review of Biology 78:23–55. Murphy, K., P. Travers, and M. Walport. 2008. Janeway’s Strauss, S. Y., J. A. Lau, and S. P. Carroll. 2006. Evolutionary C

Immunobiology. Seventh edition. Garland Science, New ONCEPTS responses of natives to introduced species: what do intro- York, New York, USA. ductions tell us about natural communities? Ecology Letters Mustin, K., T. G. Benton, C. Dytham, and J. M. J. Travis. 9:357–371. 2009. The dynamics of climate-induced range shifting; Strauss, S. Y., J. A. Rudgers, J. A. Lau, and R. E. Irwin. 2002. perspectives from simulation modelling. Oikos 118:131–137. Direct and ecological costs of to herbivory. Trends Perkins, S. E., et al. 2008. Invasion biology and parasitic in Ecology and Evolution 17:278–285.

infections. Pages 179–204 in R. S. Ostfeld, F. Keesing, and &S V. T. Eviner, editors. Infectious disease ecology: the effects of Thomas, C. D., et al. 2004. risk from climate change. on disease and disease on ecosystems. Princeton Nature 427:145–148. Torchin, M. E., K. D. Lafferty, A. P. Dobson, V. J. McKenzie, University Press, Princeton, New Jersey, USA. YNTHESIS Phillips, B. L. 2009. The evolution of growth rates on an and A. M. Kuris. 2003. Introduced species and their missing expanding range edge. Biology Letters 5:802–804. parasites. Nature 421:628–630. Phillips, B. L., G. P. Brown, J. M. J. Travis, and R. Shine. 2008. Travis, J. M. J., and C. Dytham. 1999. Habitat persistence, Reid’s paradox revisited: the evolution of dispersal in range- habitat availability and the evolution of dispersal. Proceed- shifting populations. American Naturalist 172:S34–S48. ings of the Royal Society of London B 266:723–728. Phillips, B. L., G. P. Brown, J. K. Webb, and R. Shine. 2006. Travis, J. M. J., and C. Dytham. 2002. Dispersal evolution Invasion and the evolution of speed in toads. Nature 439:803. during invasions. Research 4:1119– Phillips, B. L., C. Kelehear, L. Pizzato, G. P. Brown, D. 1129. Barton´, and R. Shine. 2010. Parasites and pathogens lag Travis, J. M. J., T. Mu¨nkemu¨ller, O. J. Burton, A. Best, C. behind their host during periods of host range-advance. Dytham, and K. Johst. 2007. Deleterious mutations can surf Ecology 91:872–881. to high densities on the wave front of an expanding Pianka, E. R. 1970. On r- and K-selection. American Naturalist population. Molecular Biology and Evolution 24:2334–2343. 104:592–597. Travis, J. M. J., and K. J. Park. 2004. Spatial structure and the Rejmanek, M., and D. M. Richardson. 1996. What attributes control of invasive alien species. Animal Conservation 7:321– make some plant species more invasive? Ecology 77:1655– 330. 1661. van Kleunen, M., and B. Schmid. 2003. No evidence for an Reznick, D., M. J. Bryant, and F. Bashey. 2002. r- and K- evolutionary increased competitive ability in an invasive selection revisited: the role of population regulation in life- plant. Ecology 84:2816–2823. history evolution. Ecology 83:1509–1520. Vermeij, G. 2005. Invasion as expectation: a historical fact of Roff, D. A. 1993. The evolution of life histories: theory and life. Pages 315–339 in D. F. Sax, J. J. Stachowicz, and S. D. analysis. Springer, New York New York, USA. Gaines, editors. Species invasions: insights into ecology, Roff, D. A., and D. J. Fairbairn. 2001. The genetic basis of evolution, and biogeography. Sinauer Associates, Sunder- dispersal and migration, and its consequences for the land, Massachusetts, USA. evolution of correlated traits. Pages 191–202 in J. Clobert, Willis, A. J., J. Memmott, and R. I. Forrester. 2000. Is there E. Danchin, A. A. Dhondt, and J. D. Nichols, editors. evidence for post-invasion evolution of increased size among Dispersal. Oxford University Press, Oxford, UK. invasive plant species? Ecology Letters 3:275–283.

Top View