Harmonia axyridis invasions: Deducing evolutionary causes and consequences
John J. Sloggett, Maastricht Science Programme, Maastricht University, P.O. Box 616, 6200 MD Maastricht, The Netherlands. Email: firstname.lastname@example.org
I consider evolutionary approaches to deducing factors that have made the ladybird beetle Harmonia axyridis such a successful invader, and the contribution that studies of this species in its native range can make. Work aiming to demonstrate which (pre)adaptations have made the species so successful often fails to compare these putative characters with those of other ladybirds. This has led to a tendency for “argument by design”-type claims on characters widely shared by non-invasive coccinellids. There is good evidence from genetic studies that evolutionary change occurred in invasive populations, contributing to their success. There is some evidence for subsequent evolutionary change after the establishment of invasive H. axyridis, primarily in the native organisms with which the ladybird interacts. I show here that there appears to have been little adaptation in H. axyridis, over about 20 generations, to the alkaloids of one North American native intraguild prey, the ladybird Coleomegilla maculata. Studies of H. axyridis in its native range are important, as they provide a snapshot of the ancestral ladybird, unobscured by subsequent evolutionary change related to its invasiveness. They provide baseline data about phenomena such as interactions with natural enemies and intraguild predation, and they also can provide pointers as to how H. axyridis might further adapt in the regions it has colonized. Harmonia axyridis represents an ideal opportunity for greater international co-operation between scientists studying this species in its native range in Asia and scientists studying it in Europe, America and Africa, where it is an invasive exotic.
Harmonia axyridis (Pallas) is an aphidophagous species of ladybird native to Asia, in China, Japan, Korea, Mongolia and eastern Russia (Iablokoff-Khnzorian 1982; Kuznetsov 1997; Brown et al. 2011a). Its relatively common occurrence means that it has long been a subject of academic study in these regions, with particular interest being shown in its color pattern polymorphism and its potential use in the biocontrol of aphid crop pests (e.g. Komai 1956; Hukusima & Ohwaki 1972; Kawai 1976; Sasaji 1980; Arefin & Ivliev 1984; Hu et al. 1989; Nakata 1995; Osawa 2000).
In the last quarter of a century the species has also become established outside its native range, in North and South America, Europe and Africa, probably in large part due to efforts to use this species as a biocontrol agent outside of its native range (Brown et al. 2011a). In many of the regions where it has established, the species has become the dominant aphidophagous species in many habitats, at least amongst the Coccinellidae (e.g. LaMana & Miller 1996; Brown & Miller 1998; Lucas et al. 2007; Brown et al. 2011b). A number of possible or known deleterious effects of the invader have been recorded, including suppression of native insect populations through competition or predation, problematic overwintering aggregations in buildings, and contamination or injury to fruit crops and their products, notably wine (Majerus et al. 2006; Koch & Galvan 2008; Kenis et al. 2010).
In large part because of its remarkable success as an exotic and its deleterious effects, for the last two decades there has been intense interest in research on this species worldwide (Koch 2003; Sloggett 2005; Roy & Wajnberg 2008). Between 1995 and 2004 the proportion of published experimental or observational papers including H. axyridis increased from 11% of all experimental and observational studies on aphidophagous coccinellids to 44% (Sloggett 2005). In 2010, 35% of papers included H. axyridis (38 of 110 papers from the Web of Science; method as for Sloggett 2005). The majority of papers focus on H. axyridis from regions where it is invasive (for 2010, 25 of the 38, 63%). Many of these papers concentrate on H. axyridis invasiveness (defined here as successful establishment and spread and ecological dominance with associated deleterious effects on native species), its causes and consequences. What is it that has made this species “the most invasive ladybird on earth” (Roy et al. 2006)? How has it been able to establish and spread in the new regions it has colonized? How does it interact with the organisms it encounters there and what are its effects on these organisms?
It is an often-repeated maxim that “nothing in biology makes sense except in the light of evolution” (Dobzhansky 1973) and this is as true of H. axyridis invasion as any other biological phenomenon. In this paper I discuss work on H. axyridis in this light. First, I consider work aiming to deduce which features of H. axyridis biology in its native range, if any, have preadapted this species to be such a successful invader. Second, I discuss whether particular characters have evolved in invasive populations that have rendered it successful. Third, I discuss the evidence that subsequent to invasion, evolutionary changes have occurred in exotic H. axyridis or in the native organisms that have recently come into contact with this species in its new ranges. I conclude by discussing the importance of continuing work on the species in its native range in answering these questions.
IS H. AXYRIDIS PREDAPTED FOR INVASION?
An important aim of invasion biology is to identify biological characteristics that can allow us to deduce which species are likely to become invasive (Williamson 1996; Kolar & Lodge 2002). This is particularly important in the case of biological control agents such as predatory ladybirds, which are often intentionally introduced to new regions (Roy et al. 2011a). An assumption of such work is that these characteristics are already possessed in its native range; that is, that the species is preadapted to become an invader.
There have been a large number of studies that have claimed to throw light on which characteristics of H. axyridis in its native range make it such a successful invader outside of it. Numerous different characters of the species have been suggested to be of importance (Table 1). However, even though different characteristics are undoubtedly required to facilitate successful establishment, spread or ecological dominance, it is unlikely that all of the factors on this diverse list could have played major roles in H. axyridis invasions. In fact, there is a strong possibility of argument by design; that is, that researchers can invent arguments to support a role for any characteristic of H. axyridis being linked to its invasiveness, due to the characteristic's mere occurrence in the species.
Table 1. Factors suggested as predisposing Harmonia axyridis towards invasiveness†
|Life history features||Large body size, high fecundity, short developmental time, an extra larval instar||Michaud 2002; Félix & Soares 2004; Labrie et al. 2006; Hemptinne et al. 2011|
|Generalist behavior||Habitat generalist, extremely generalist aphid diet, polyphagy extending beyond aphid prey||Tedders & Schaefer 1994; Majerus et al. 2006; Roy et al. 2006; Berkvens et al. 2008, 2010a|
|Winter survival||Overwintering behavior, cold hardiness||Labrie et al. 2008; Berkvens et al. 2010b|
|Dispersal ability||High rate of spread||Brown et al. 2011a; Hemptinne et al. 2011|
|Low susceptibility to natural enemies||High immunity, strong chemical defense, powerful physical defense such as larval spines||Cottrell & Shapiro-Ilan 2008; Ware and Majerus 2008; Nedvěd et al. 2010|
|Highly cannibalistic||Better able to withstand low prey availability or poor prey quality||Burgio et al. 2002; Michaud 2003|
|Strong intraguild predator||Frequent victor in aggressive interactions with other species, ability to develop on intraguild prey||Cottrell 2004, 2005; Ware & Majerus 2008; Pell et al. 2008|
|Color pattern polymorphism||Direct relationship to generalism through ability of different morphs to tolerate different (micro)climates, indirect effects of differential diet suitability for the morphs||Majerus et al. 2006; Soares et al. 2008|
To avoid this in these studies, researchers need to address two issues. The first of these is to deduce what characteristics are unique to H. axyridis (or are shared with other invasive ladybirds, such as Coccinella septempunctata L. within the Coccinellidae. This requires a good knowledge not only of H. axyridis but of other coccinellids too. I have previously argued that our knowledge of coccinellids is too narrowly based, being in large part restricted to a few very well studied generalist species, including H. axyridis (Sloggett 2005). Were this not the case we would have more species with which to compare H. axyridis, to deduce which of its characters are unique. This call has largely been ignored, however: indeed the observation that so much work is focused on H. axyridis has even been used as a rationale for further work focused on this species (Koch et al. 2006a; McCornack et al. 2007).
The dangers of such an exclusive approach are manifest. It is, for example, often claimed that H. axyridis is highly polyphagous, with a diet encompassing both non-homopteran prey (e.g. Tedders & Schaefer 1994; Koch 2003) and plant material (e.g. Berkvens et al. 2008, 2010a). This polyphagy, it is argued, may have played a role in its invasiveness, facilitating enhanced survival or even reproduction when aphid prey are scarce (e.g. Berkvens et al. 2008), or posing a threat to certain non-homopteran extraguild prey invertebrates such as butterflies (e.g. Koch et al. 2003, 2006b). It is true that H. axyridis consumes a wide diversity of food extending well beyond aphid prey, but comparison with other aphidophagous coccinellids suggests that it is hardly unique in that. For example, Niijima et al. (1986) list three other species of unrelated aphidophagous ladybird that perform just as well as H. axyridis on a non-aphid diet of drone honeybee powder in the laboratory, while further species exhibit similarly good larval development and/or adult longevity on the diet. There are also numerous papers documenting consumption of plant food or non-homopteran prey by diverse aphidophagous ladybirds in the field (e.g. Forbes 1883; Putman 1964; Hemptinne & Desprets 1986; Pemberton & Vandenberg 1993; Cook & Webb 1995; Ricci & Ponti 2005; Davidson & Evans 2010) and researchers are becoming increasingly aware that this non-homopteran food plays an integral role in the ecology of many species considered aphidophagous (Evans 2009; Lundgren 2009). Given that polyphagy is a widespread attribute, it seems unlikely that this has made H. axyridis such a uniquely successful colonizer of new regions. Similarly, those claiming that invasive H. axyridis represent a threat to non-homopteran extraguild insects like butterflies must explain why this threat is more significant than that historically posed by native coccinellids (e.g. see Warren & Tadic 1967; Cook & Webb 1995; Phoofolo et al. 2001).
In the example of polyphagy, there is already sufficient information available in the literature to develop an informed argument about the importance of the character in question. This is not, however, always the case. For example, our knowledge of intraguild predation between ladybirds derives in large part from H. axyridis itself, either as a potential predator or prey (review in Pell et al. 2008). By contrast, detailed studies of intraguild predation between pairs of coevolved ladybird species (i.e. species sharing a long evolutionary history together) remain rather limited, with extensive work on a single system restricted to the model studies of Adalia bipunctata (L.) and C. septempunctata carried out by Dixon and colleagues (Agarwala & Dixon 1992; Hemptinne et al. 2000a,b). While it is certainly true to say that H. axyridis displays a greater propensity for intraguild predation than many other ladybird species, even now there is some indication that some other species, like the highly polyphagous Coleomegilla maculata (De Geer) (Michaud & Jyoti 2007), may also be habitual intraguild predators.
A greater number of studies focused on other species (Sloggett 2005) would certainly provide a large part of the solution to the argument-by-design problem. However, direct experimental or meta-analytical comparisons of H. axyridis with other species are a more immediate means of deducing which characteristics of H. axyridis are sufficiently unique to have played a likely role in this species' invasiveness. These analyses should include a number of other species. Use of a single comparator species makes the formulation of general conclusions about unique characters impossible, as it can not be deduced which of the two species being compared is unusual. Recently, there have been two studies that have used a broad comparative approach to good effect. In experiments tests using Cladocera and ants and several ladybird species, Nedvěd et al. (2010) showed that H. axyridis chemical defenses are among both the most toxic and unpalatable. Their results provide the most convincing support yet for the frequent assertion that H. axyridis is better defended against natural enemies than other ladybirds. In another comparative study, Hemptinne et al. (2011) examined dispersal ability in North American invasive species, and reproductive rate and body size in a total of 30 ladybird species, to suggest that the invasiveness of H. axyridis and C. septempunctata is related to their large body size. On the basis of their comparisons, they argue that this gives these two species a very high reproductive rate and dispersal ability.
The second and more difficult issue for those studying H. axyridis invasiveness is to go beyond the correlative link between an unusual character and its occurrence in H. axyridis, to demonstrating that the character provides a clear benefit to the ladybird during colonization or spread, or in bestowing ecological dominance. This work needs to move beyond mere documentation of particular characteristics, to examining their role in a real world ecological context. The area where the most such progress has been made is probably intraguild predation. Studies have now moved beyond mere documentation of H. axyridis as a victor in intraguild interactions against native species in the laboratory, to addressing the conditions necessary for intraguild predation by H. axyridis to occur (e.g. Musser & Shelton 2003; Wells et al. 2010; Ingels & De Clercq 2011), its costs and benefits under realistic ecological scenarios (e.g. Ware et al. 2009; Sloggett et al. 2009a), its likelihood and occurrence in the field (e.g. Cottrell & Yeargan 1998a,b; Sato & Dixon 2004; Jansen & Hautier 2008; Hautier et al. 2011) and how it may interact with or mediate other intraguild interactions such as competition for aphid prey (e.g. Alhmedi et al. 2010). A similar approach needs to be applied in other areas. For example, although the study of Nedvěd et al. (2010) showed that H. axyridis is exceptionally well chemically defended, it is unclear to what extent this provides additional natural protection against diverse natural enemies (Sloggett et al. 2011) and, more importantly, whether such natural enemies are responsible for regulating ladybird populations, including those of H. axyridis (Ceryngier & Hodek 1996; Roy et al. 2011b).
The majority of studies of H. axyridis invasiveness use ladybirds from invasive populations. This is an additional problem for the researcher who wishes to find any pre-existing predisposition to invasiveness in this species, because of possible subsequent evolutionary change (Whitney & Gabler 2008). Such change might have occurred in invasive populations early in their colonization of other regions, making them successful; alternatively, evolutionary changes subsequent to invasion might act to obscure those characteristics that made the species so successful in the first place.
Evolutionary change at or subsequent to introduction can be divided up according to its timing. Evolutionary changes occurring at the time of colonization and establishment can facilitate these processes (Sakai et al. 2001; Bossdorf et al. 2005) and thus the overall invasiveness of a species. It should, however be noted that genotypic or phenotypic changes occurring at this time may not necessarily be adaptive, but can arise through factors such as founder effects or genetic drift: consequently, such changes may not be indicative of a role in invasion or even adaptive (Keller & Taylor 2008).
Other subsequent evolution can be seen as a consequence of invasion: this can include processes such as hybridization with native species (which does not occur in H. axyridis), selected responses to environmental gradients, such as temperature, and coevolutionary changes with native members of the invaded community (Sakai et al. 2001; Lee 2002).
Unlike the factors predisposing a species to invasiveness, studies of these factors largely rely on intraspecific, interpopulational or interindividual comparisons. They include comparisons of native and invasive populations, and population studies of genetic and phenotypic architecture, including studies of phenotypic plasticity and genetic variation in particular traits and thus their potential for change.
EVOLUTIONARY CHANGES IN H. AXYRIDIS FACILITATING INVASIVENESS
Suspected examples exist of species where evolutionary change occurring subsequent to their arrival or introduction outside of their native range has facilitated their colonization and/or establishment (Sakai et al. 2001; Lee 2002; Bossdorf et al. 2005; Hufbauer 2008). Multiple biocontrol releases of H. axyridis in new regions were originally carried out, providing the possibility of genetic mixing between differentiated populations of origin; additionally these releases were often from captive cultures, which themselves might have undergone selection prior to release. Thus, H. axyridis is a prime candidate for such evolutionary change (Brown et al. 2011a). Great strides have already been made in studying the genetic and phenotypic architecture of H. axyridis in a series of studies by Estoup, Lombaert and coworkers (Lombaert et al. 2008, 2010, 2011; Facon et al. 2011a,b; Turgeon et al. 2011).
Using microsatellites and Bayesian analysis, Lombaert et al. (2010) were able to compare different invasion scenarios. Their analyses indicate that the invasive eastern North American population acted as a bridgehead for further invasions of South America, Europe and Africa. This suggests that an evolutionary shift facilitating invasiveness might have occurred in H. axyridis in eastern North America (Lombaert et al. 2010). It appears that the North American population originated from a mixture of native eastern and western populations (Lombaert et al. 2011), which are genetically distinct (Blekhman 2008; Blekhman et al. 2010; Lombaert et al. 2011). This suggests that the shift in North America may have come about due to new genomic combinations resulting from population admixture (Brown et al. 2011a; Lombaert et al. 2011). The eastern North American population appears to have undergone a population bottleneck, with associated purging of deleterious recessive alleles: consequently even inbred individuals display high fitness (Facon et al. 2011a). There is also some evidence that invasive populations display greater phenotypic plasticity than the biocontrol cultures from which they are descended (Lombaert et al. 2008). Invasion in Europe may also have been assisted by further genetic mixing between ladybirds from eastern North America and ladybirds originating from European biocontrol cultures (Facon et al. 2011b; Turgeon et al. 2011).
This work provides a very clear picture of the invasion genetics of H. axyridis and of genetic factors that might have assisted in its colonization and establishment in new regions. Clearly if initial invasive populations of H. axyridis were small or if matings between close relatives occur regularly on the invasion front, then lowered inbreeding depression is highly beneficial (Facon et al. 2011a). Similarly higher phenotypic plasticity may allow the ladybird to colonize a wider diversity of habitats or regions (Lombaert et al. 2008). The exact phenotypic adaptations responsible for colonization and establishment remain a matter for debate, however, in spite of some intriguing findings (Lombaert et al. 2010; Brown et al. 2011a). Most interestingly, Facon et al. (2011a) found that the generation time of invasive populations was on average 6.3 days shorter than that of native populations. As they point out, this would lead to an increased population growth rate in invasive populations of the species. It is especially intriguing when considered together with the pre-existing high reproductive rate and dispersal ability consequent on body size, postulated by Hemptinne et al. (2011). Together these studies suggest that it was both pre-existing and simultaneous life history adaptations that made H. axyridis such a successful invader, providing the ladybird with the potential for massive population growth in a short time. This is consistent with invasion theory (Sakai et al. 2001), although clearly further work is necessary to confirm this scenario for H. axyridis. Such work should examine trade offs and costs related to these life history traits, which could potentially limit their beneficial effects. These could be both physiological (for example, trade-offs between developmental rate or body size and other characters such as chemical defense: see Holloway et al. 1993) or more ecological (for example, a limited ability to exploit certain aphid species or populations with a large body size: see Sloggett 2008). Estimates of potential population growth could also be linked to life table data and data on the historical spread of H. axyridis, and could be used to examine H. axyridis competitiveness with native coccinellid species.
EVOLUTIONARY CONSEQUENCES OF INVASION
After initial establishment, further evolutionary change occurs in response to both abiotic and biotic aspects of the new environment in which an invasive species finds itself. In the case of the biotic environment, changes are coevolutionary, involving not just the invader but the organisms with which it interacts. Thus far there is little published work on how H. axyridis is adapting to the abiotic environment in the new regions in which it lives, with most of the focus on coevolution with other organisms, notably natural enemies and intraguild prey.
Coevolution with natural enemies
A number of parasitoids, parasites and pathogens have been recorded parasitizing H. axyridis in its new range (Roy et al. 2011b) and two examples of apparently new interactions have elicited particular interest. The first is a laboubenialean fungus Hesperomyces virescens Thaxter, which parasitizes a wide range of coccinellids globally (Riddick et al. 2009). This was first recorded on North American H. axyridis in 2002, some 14 years after the ladybird was first recorded as established there (Garcés & Williams 2004; Riddick & Schaefer 2005). It is now extremely abundant on H. axyridis in North America (e.g. Harwood et al. 2006; Nalepa & Weir 2007; Riddick 2010) and has more recently also established on H. axyridis in Europe, over a decade after the ladybird's first arrival there (Steenberg & Harding 2010; Roy et al. 2011b). The second is a sexually transmitted parasitic mite of ladybirds, Coccipolipus hippodamiae (McDaniel & Morrill), which was recorded naturally occurring in North America in 2007 (19 years after first recording of H. axyridis establishment) and Europe in 2009 (over a decade after first H. axyridis establishment) (Rhule et al. 2010; Riddick 2010).
A problem with both these and other studies of natural enemies is the relative paucity of information on the natural enemies of H. axyridis in its native range. It is difficult to claim with a high degree of certainty that any interaction is completely new. Because H. virescens attack is very obvious, due to the prominent yellow thalli occurring on the external cuticle of the ladybird (e.g. see Riddick & Schaefer 2005; Harwood et al. 2006), it seems unlikely that H. virescens infection would have passed unnoticed on ladybirds either in the native range of H. axyridis or, if it had occurred on them much earlier, in their colonization of North America and Europe. Thus, in this case, a new association with the host does seem likely. However, this is not the case for Coccipolipus mites, which generally occur underneath the elytra of ladybirds and are more difficult to detect (see Webberley et al. 2004; Riddick 2010). This is likely also to be true for other parasitoids, parasites or pathogens, which often require dissection or rearing from hosts, or even molecular approaches in the laboratory to be detected.
Assuming that both H. virescens and C. hippodamiae association with invasive H. axyridis do represent new host associations, it is also not yet clear what the relative contributions of host and parasite evolution to the formation of these associations is. The long initial absence of the H. virescens on H. axyridis in the North America, coupled to the fact that it already parasitized other ladybird species there (Thaxter 1931), suggests that a physiological change in the fungus might have underlain its host shift to H. axyridis. This is much less likely for C. hippodamiae, as artificial transfers from European hosts were achieved in the laboratory, even before the association was observed in the wild; furthermore, mite performance on H. axyridis was similar to that on a pre-existing host A. bipunctata (Rhule et al. 2010).
An intriguing possibility is that the ladybird might have become more vulnerable to such parasites due to genetic trade-offs between defensive and competitive capabilities. The evolution of increased competitive ability (EICA) hypothesis suggests that reallocation of resources from defense to population growth could occur in invaders as a response to an absence of natural enemies (Blossey & Nötzold 1995; Roy et al. 2011c). In the initial stages of invasion this would be beneficial; however at later stages it might render the invader more vulnerable to potential native enemies. It has already been suggested that reallocation of resources to population growth might have occurred through an enhanced developmental rate (Facon et al. 2011a; see previous section).
The evidence that H. axyridis defense against natural enemies is reduced in invasive populations varies with respect to parasites/pathogens and predators. Two studies exist in which the effects of pathogens or parasites have been tested on both native and invasive H. axyridis at the same time. In a study of the effects of the entomopathogenic fungus, Beauveria bassiana (Balsamo) Vuillemin on native Japanese and invasive British H. axyridis, Roy et al. (2008) found a non-significant trend (P ≈ 0.07) for lower mortality in British over Japanese H. axyridis. Fewer Japanese H. axyridis also oviposited after infection. In another study, Koyama & Majerus (2008) examined parasitism of H. axyridis by the hymenopteran parasitoid of ladybirds, Dinocampus coccinellae (Schrank), for which H. axyridis forms a marginally suitable host (Hoogendoorn & Heimpel 2002; Firlej et al. 2007; Berkvens et al. 2010c). This study not only used native Japanese and invasive British ladybirds, but also D. coccinellae from both regions as well. Wasp origin did not affect likelihood of development in H. axyridis, but more wasps developed in Japanese H. axyridis than British ones (Koyama & Majerus 2008). The results of both studies, with D. coccinellae (Koyama & Majerus 2008) and B. bassiana (Roy et al. 2008), suggest that the invasive population was better protected than the native one: this is the opposite result from that predicted by the EICA hypothesis. However, in contrast to these experiments, Rieder et al. (2008) found some evidence that eggs of invasive North American H. axyridis were less repellent and toxic than those of native Japanese H. axyridis to Coccinella septempunctata brucki Mulsant larvae. The quantity of alkaloid in H. axyridis eggs is known to affect their repellency and toxicity to C. septempunctata larval intraguild predators (Kajita et al. 2010); thus it seems possible that North American H. axyridis do invest less in chemical defense than native conspecifics. This is consistent with the EICA hypothesis. It is probable that other defenses such as host immunity are more important than defensive chemicals in defending H. axyridis against pathogens and parasites, such as B. bassiana and D. coccinellae, possibly explaining the contradiction between studies. If enhanced development time has been bought through lower investment in chemical defense in invasive populations, this would not be apparent through studies of such pathogenic or parasitic natural enemies.
Coevolution with intraguild prey
While some light has been thrown on coevolution between H. axyridis and its natural enemies, less has been thrown on coevolution with its prey. This is surprising, given the original biocontrol role of this species and the postulated deleterious effects it is having through intraguild predation where it is invasive. The very limited work of relevance available relates to intraguild prey.
It is to be expected that native species at risk from intraguild predation by invasive H. axyridis should evolve strategies for avoidance of this species, as apparently occurs in other species in the native range of H. axyridis (Agarwala et al. 2003; Sato et al. 2003). Such behavior has only been observed in the syrphid Episyrphus balteatus (De Geer) and parasitoid Aphidius ervi Halliday, which avoid areas contaminated with chemical cues from H. axyridis for oviposition (Almohamad et al. 2010; Meisner et al. 2011). By contrast, avoidance tactics, in this case dropping by larvae to avoid intraguild predation, appear to be missing in North American A. bipunctata, although, as suggested by the authors, this could also be due to the species' arboreal habitat (Sato et al. 2005). Clearly further work on the evolution of avoidance strategies by native species in areas where H. axyridis is invasive is necessary. Continuing monitoring for the evolution of such behaviors over time would be particularly valuable, although possibly difficult to achieve in practice.
Because invasive H. axyridis are already successful intraguild predators on native species, there appear to be limited selective pressures on this ladybird related to intraguild predation. Moser et al. (2010) suggested that satiated invasive North American H. axyridis larvae avoided chemical cues from larvae of the North American native ladybird C. maculata. However, they neither provided a clear adaptive explanation for this behavior nor did they consider it in the context of the short coevolutionary history between the species. In fact in intraguild interactions between the two species, C. maculata larvae are typically the losers (Cottrell & Yeargan 1998c; Moser & Obrycki 2009): this suggests that putative avoidance of C. maculata by H. axyridis would be non-adaptive or maladaptive.
Although invasive H. axyridis are both successful and habitual intraguild predators of native ladybirds, there is one aspect of H. axyridis intraguild predation on other ladybirds where this species' performance is poor. This is in its resistance to the toxic effects of intraguild prey chemical defenses. Ladybirds display considerable interspecific variation in the alkaloids that they use for chemical defense and communities of ladybirds typically comprise species with a diversity of alkaloid defenses (Daloze et al. 1995; Sloggett et al. 2009b). In its native range H. axyridis is well adapted to resist any toxic effects of the alkaloids of ladybird intraguild prey it encounters there (e.g. Yasuda & Ohnuma 1999; Sato et al. 2008; Sloggett & Davis 2010). However, where it is invasive, it encounters ladybird prey with novel types of alkaloids that it does not encounter, or rarely encounters, in its native range: although these prey are consumed, their alkaloids can have deleterious toxic effects on the predator (Sloggett et al. 2009a, 2010; Sloggett & Davis 2010). These effects clearly reduce the value of intraguild predation to H. axyridis although under certain circumstances, such as shortage of aphid prey, such intraguild predation can still be beneficial (Sloggett et al. 2010, 2011). It is therefore to be expected that better resistance to the alkaloids of these novel prey should evolve, just as alkaloid resistance has evolved in the native range of H. axyridis.
I have been able to address whether this has occurred for the alkaloids of one particular North American native prey ladybird, C. maculata. Although H. axyridis consumes eggs of this species in the field (Cottrell & Yeargan 1998a,b) and can even develop on a diet of these eggs (Cottrell & Yeargan 1998c; Cottrell 2004), the eggs, which contain the alkaloid myrrhine (Sloggett et al. 2009b), are partially toxic to the predator: they cause increased mortality and slowed development relative to intraguild prey defended by alkaloids from the native range of H. axyridis (Sloggett et al. 2009a). By an extremely fortunate chance, two experiments on the effects of intraguild predation of eggs of C. maculata, both using native C. maculata and invasive H. axyridis from the same area (Lexington, KY, USA), were carried out approximately a decade apart (Cottrell & Yeargan 1998c; Sloggett et al. 2009a). Assuming bivoltinism in H. axyridis (LaMana & Miller 1996; Koch & Hutchison 2003), this translates into about 20 generations of H. axyridis. In both cases, as part of wider studies, first instar H. axyridis were fed an exclusive diet of eggs of C. maculata and mortality within the first instar monitored. These results can consequently be compared.
In the earlier results from the late 1990s, Cottrell and Yeargan (1998c) recorded 44.8% mortality in a total sample of 29 first instar larvae: thus 12 of the 29 larvae died in the first instar. In the later results, from 2007, mortality was eight of 15 larvae (53.3%) (Sloggett et al. 2009a). Although the results are rather too small-scale for a formal statistical analysis, mortality rates in both cases are within 10% of each other, and, in any case, mortality is marginally higher in the later results. There is thus no support for an increase in H. axyridis resistance to the myrrhine defenses of C. maculata. This finding is all the more striking as studies, also carried out in Lexington, show that H. axyridis does consume C. maculata eggs in the wild (Cottrell & Yeargan 1998a,b) and C. maculata is a common species in the area, as well as elsewhere in North America. Chemical resistance appears to result from metabolism of the prey alkaloids involved (Sloggett & Davis 2010) and is relatively alkaloid-specific (Sloggett et al. 2009a, 2010, 2011). The results here indicate that it is probably difficult to evolve, and consequently takes a long time to appear, even after predation of a particular species has provided the selective pressure for its evolution.
ROLE OF STUDIES OF H. AXYRIDIS IN THE NATIVE RANGE
As pointed out in the introduction, the majority of studies of H. axyridis are now focused on invasive populations outside of the native range of the species. However, studies on invasive populations alone are not sufficient to understand the success of this species in colonizing new geographic regions and even dominating the aphidophagous fauna there. Studies from the native range of the species have much to contribute because the species in its native range represents the ancestral state before invasion. Harmonia axyridis has now existed as an invader for close on a quarter of a century and it should not necessarily be assumed that invasive populations are identical in all respects to those that originally colonized new regions. Even was this not the case, it is impractical to monitor evolutionary changes as they occur within the invasive range over time. Such work will probably remain limited to the odd fortuitous comparison, such as that on intraguild predation of C. maculata described in the previous section.
Geographic comparisons are easier to make because they are instantaneous and do not require long periods of time. There is a continuing promising role for studies that compare invasive populations with native ones to deduce what evolutionary changes have occurred from the native range to the invasive one. These studies are as important for work aiming to demonstrate pre-existing predispositions in the species towards invasiveness as they are for work looking at evolutionary changes occurring during or after colonization and establishment. Such work should ideally extend across a multiplicity of native and invasive populations to fully document variation in individual traits. Recent studies of H. axyridis showing that the species may be comprised of genetically distinct eastern and western subspecies that have mixed in the invasive range (see section “Evolutionary changes in H. axyridis facilitating invasiveness”) serve to emphasize just how important population variation in the native range might be in understanding the invasiveness of H. axyridis.
There are a number of research areas that have been better studied in the invasive than the native range of H. axyridis, and further studies are needed in the native range to provide important baseline information for findings made in the invasive range. One example of this is work on natural enemies, where more information from the native range, especially from the field, is required. The same is true for intraguild predation. While studies on intraguild predation in the invasive range have become ever more complex and realistic, studies in the native range have moved little beyond staged laboratory trials. In some cases, we have interesting information on intraguild predation between H. axyridis and ladybird species from its native range that is entirely limited to the laboratory, with no supporting studies on co-occurrence of the species in the field or natural intraguild predation occurring between them (e.g. Ware et al. 2008; Ware & Majerus 2008). This makes interpretation of the laboratory results very difficult.
Lastly, studies in the native range may not only tell us about the past of H. axyridis, they may tell us about the future. By looking at the native range we can predict what traits will evolve in the invasive range in the future and deduce something about the “evolvability” of particular traits. The example of resistance to C. maculata alkaloids by invasive H. axyridis intraguild predators serves to make this point. There is good reason to think that resistance to C. maculata toxic alkaloids will evolve in North America, because this is what has already happened in response to intraguild prey alkaloids in the native range. The fact that it has not happened yet suggests that such traits evolve slowly, but we would still expect them ultimately to evolve. This insight would not be there but for our knowledge of alkaloid resistance in the native range.
There is no doubt that in the last decades, the invasiveness of H. axyridis has propelled it from a species of regional academic and biocontrol interest to a globally researched model species, of importance not only in biocontrol, but as a pest to humans and threat to biodiversity in its own right. Although almost twice as much of this research occurs where H. axyridis is invasive, more than ever now there is an overriding requirement for research on or incorporating native populations of this species. At a time when international collaborative ventures between research groups are not only encouraged, but often supported by substantial funding, Harmonia axyridis, the truly global ladybird, holds great promise for truly global research.
I thank Andrew Davis and Markus Wagner for their assistance in work on datasets relating to research on H. axyridis, and Helen Roy and Ilja Zeilstra for comments on earlier drafts of the manuscript.