Metabolic consequences of various fruit-based diets in a generalist insect species

Most insects that feed on green plants are specialists, meaning that they feed on just a narrow range of plant species. This reduces competition, especially if the host plant contains chemical deterrents that are toxic to other insects. But specialists cannot easily switch to feed on other plants, making them vulnerable to changes in the availability of the particular food type that they eat.

Generalist insects, on the other hand, are able to consume a wide range of diets. This makes them more robust to changes in food availability, but it is unclear how these insects deal with the wider range of chemical compositions of their food. Do they convert food into energy using the same chemical process, or metabolism, for all the different things they eat? Or do generalists have a specific metabolic pathway for each food type?

To answer this question, Olazcuaga, Baltenweck et al. studied the metabolism of a generalist fruit fly species. The team compared four types of fruit (blackcurrant, cherry, cranberry and strawberry) and isolated separate groups of flies so that they each ate only one type of fruit. By comparing the chemical composition of the flies with that of the fruit they ate, they were able to work out how each fruit type was metabolised. They found that the flies converted food into energy using the same process regardless of the type of fruit they ate. This lack of a specialist metabolic pathway for each fruit type meant that some chemicals were not metabolised and accumulated in the fly’s body instead. This build-up of unprocessed chemicals is likely to be harmful to the fly.

The results of Olazcuaga, Baltenweck et al. suggest that generalist insects do not actively adapt their metabolism to new food types. It’s more likely that they try different types of food as the opportunity arises, regardless of the fact that some of the food will not be converted into energy and may harm them long term. These findings are important because they give us an insight into how the chemistry of a plant can shape the physiology of the organisms that consume it, and vice-versa. These insights are a crucial step in developing sustainable agriculture practices that must consider tackle how plants are pollinated, how plant seeds are dispersed and what type of pest control to use.

One of the most salient features of living organisms is their diversity in terms of ecological requirements. Some species show extreme specialization in their environment (e.g. obligate symbionts), while others are able to use a wide range of resources and tolerate many environmental conditions. Furthermore, these relationships between organisms and their biotic and abiotic environments are dynamic at various spatial and temporal scales. As such, the concepts of niche breadth and ecological specialization are of paramount importance in ecology and evolutionary biology (Futuyma and Moreno, 1988; Sexton et al., 2017).

The dietary requirements of phytophagous insects are a textbook example of variance in niche breadth (Forister et al., 2012; Futuyma and Moreno, 1988; Jaenike, 1990). In this group, the vast majority of species are specialists of one family of plants (Forister et al., 2015) and this specialization is often accompanied by extreme morphological and/or physiological specializations to exploit a particular plant (e.g. Berenbaum, 1981; Hotti and Rischer, 2017). In contrast, some rarer species display a much wider niche breadth, allowing them to consume up to several dozen plant families (Clarke et al., 2005; Forister et al., 2015). The larger proportion of specialists than generalist species (Forister et al., 2015), as well as the fact that transitions between specialism and generalism, are bidirectional (Day et al., 2016; Janz et al., 2001; Nosil, 2002; Nylin et al., 2014), raises questions about the origin and maintenance of generalism (Forister et al., 2012; Singer, 2008), and its biochemical basis.

From a physiological point of view, dietary generalism has been much less experimentally studied than specialism, probably in part because it requires the simultaneous consideration of numerous host species and the tracking of physiological responses that may involve far more components than specialism. Additionally, the study of dietary generalism lacks the intellectual appeal of bi-species interactions leading to biochemical and genetic arm-races. Indeed, many research scrutinized the evolution of plant defenses and their corresponding detoxifying pathways in herbivores (e.g. the glucosinolate–myrosinase system of Brassicaceae countered by Spodoptera frugiperda; Winde and Wittstock, 2011). Comparatively, knowledge about how generalist species cope with a wide variety of plant biochemistry is scarce (but see Roy et al., 2016).

The core question of dietary generalism at the level of the organism is how a generalist species metabolizes different diets, which involves some basic issues that are still pending. For instance, do individuals of a generalist species have a unique biochemical response to all diets or do they differ according to their diet? In evolutionary terms, do dietary generalist species exhibit a canalized response to different dietary environments through a single, neutral, and generic use of hosts (H1; hereafter termed ‘metabolic generalism’), or do they show multiple specialized, adapted uses of different hosts (H2; hereafter termed ‘multi-host metabolic specialism’)? Broad organismal-level patterns of biochemical composition can help answer this question because individuals on different diets should be chemically similar across diets according to the former hypothesis, or mostly dissimilar according to the latter (Figure 1).

Figure 1 with 4 supplements see all

Download asset Open asset

To refine our understanding of diet generalism, we here use an untargeted metabolomics approach to characterize the metabolic impact of various fruit-based diets on the generalist insect species Drosophila suzukii (Figure 1A). The range of molecular compounds detected and quantified by current metabolomic methods may span many hundreds or thousands of molecules, ensuring a high-quality snapshot of the individuals’ physiological state (Johnson et al., 2016). Additionally, untargeted metabolomics provides the opportunity to directly compare diets and individuals that consume them, in an objective and repeatable way. The invasive Drosophila suzukii (the spotted wing Drosophila) displays an exceptional diet breadth of over 20 plant families (Asplen et al., 2015; Bellamy et al., 2013; Cini et al., 2014; Kanzawa, 1939; Kenis et al., 2016; Lee et al., 2011; Lee et al., 2015; Poyet et al., 2015). While multiple experimental studies have found that different diets have an influence on its larval development (Kenis et al., 2016; Olazcuaga et al., 2019; Poyet et al., 2015), it is widely accepted that D. suzukii is a diet generalist species.

By directly comparing the metabolomes of fruits and flies that consume them, we investigated the following question: Do generalist individuals have a single physiological response to various diets (‘metabolic generalism’) or do they exhibit specific responses to these various diets (‘multi-host metabolic specialism’)? We first verified how biochemically different were the selected major host plant species used by D. suzukii. Then, we examined whether generalist individuals use different host plants in a similar way, qualitatively and quantitatively. Finally, we attempted to discriminate individuals of a generalist species according to their diet and identify diet-specific metabolites in generalist individuals.

The metabolomes of four major crop fruits (blackcurrant, cherry, cranberry, and strawberry) and their corresponding flies (that consumed them both at the larval and early adult stages) were characterized simultaneously using ultra-high-performance liquid chromatography coupled to mass spectrometry (UHPLC-MS; Figure 1A). Our untargeted approach enabled the detection and relative quantification of a set of 11,470 unique ions that were found in at least one fruit or fly sample. While the simultaneous processing of fruit and flies’ samples provided a single integral dataset enabling direct comparison between diets and consumers, some analyses required the fruit-related and fly-related data to be separated, resulting in a fruit dataset and a fly dataset. Although the integral, fruit, and fly datasets are based on ions characterized by UHPLC-MS, some ions will be referred to as ‘compounds’ or ‘metabolites’ later in the manuscript, depending on the context. Importantly, we took care to standardize the fly and integral dataset using control flies (i.e. flies that developed on an artificial diet only) to avoid including ions derived from the artificial component of our fruit diets (see Methods section for more details). Controlled and raw fly datasets yielded similar results (see Appendix 1). We here present the results of the analyses of the controlled fly dataset.

Host use

Whether individuals of a generalist species process all types of diet in a similar, generic way (‘metabolic generalism’) or show specific pathways to metabolize unique compounds of different diets (‘multi-host metabolic specialism’) entails largely non-overlapping expectations for the both qualitative and quantitative biochemical composition of individuals relatively to their diets (Figure 1).

The first approach is based on the fact that metabolic generalism is expected to result in particular patterns of overlap between ions when all fruits and flies are considered together (Figure 1B). If D. suzukii is a metabolic generalist, flies should all share a large part of their ions, fly ions specific to a particular diet should be rare or absent, and/or ions specific to single fruit-fly pair (i.e. a pair of fruit and flies grown over the same fruit) should also be rare (Figure 1C, left panel). On the contrary, multi-host metabolic specialism would be detected by any given fruit-fly pair sharing large amounts of ions but not with other pairs, and/or large amounts of ions found only in flies that are specific to each diet (Figure 1C, right panel). To test these predictions, we computed intersection sizes between all fruits’ and flies’ ions from a presence/absence (qualitative) dataset derived from our quantitative integral dataset. Comparison of the overall intersections of ions present in the different fruits and the flies grown over these fruits showed that most ions are shared between large groups of fruits and flies (i.e. with higher degrees of intersection; Figure 2A). The largest intersection includes 3691 ions shared by all fruits and all flies (32%). The second largest intersection comprises 1103 ions shared among all flies (and absent in fruits; 10%). The fourth largest intersection contains 412 ions shared among all fruits (and no flies; 4%). High-order degree characterizes all other largest intersections, except for a set of 283 ions uniquely found in blackcurrant fruits. Figure 2B focuses on the sets of ions found uniquely in fruits, flies, or fruit-fly pairs (i.e. the intersections of degree 1 and 2). While ions found uniquely in each fruit are not uncommon (>100 ions for all fruits), ions found uniquely in flies raised on a particular diet are rare (two in cherry-fed flies, seven in cranberry-fed flies, and none in strawberry-fed and blackcurrant-fed flies). Ions that are distinctive of a particular fruit-fly pair are also seldom, ranging from 69 in the blackcurrant fruit-fly pair to only seven in the strawberry fruit-fly pair.

Figure 2

Download asset Open asset

Similarity and dissimilarity in the use of different host plants could stem from the metabolic fate of common and unique fruit metabolites when ingested by flies. Differently-fed flies metabolizing compounds that are shared between fruits to produce metabolites shared between flies would be indicative of ‘metabolic generalism’ (Figure 1D). On the opposite, differently-fed flies metabolizing compounds private to a given fruit to produce new metabolites (mostly not shared between flies) would be indicative of ‘multi-host metabolic specialism.’ We thus inspected the intersection degrees of all fruits and the intersection degrees of all flies, together with their relationships (Figure 3). First, the intersection of all fruits (left column) is largely formed from ions common to all fruits (48.6% of all ions present in fruits). Ions present in lower degree intersections are found in approximately equal proportions to each other (intersections of degree 3, 2, and 1 represent 20.1%, 14.8%, and 16.5% respectively). Approximately 10% of all ions found in this study are not detected in fruits, but only in flies (yellow stack, left column). Second, the intersection degrees of flies show marked differences from those of the fruits (Figure 3, right column). These include a higher proportion of metabolites common to all flies (76.3% of all ions present in flies), smaller proportions of lower degree intersections (intersections of degrees 3, 2, and 1 represent 10.6%, 4%, and 9% of all ions present in flies, respectively), and a large proportion of fruit metabolites that are absent in flies (19.5% of all detected ions). Third, the inspection of the relationships between fruit and flies’ metabolomes shows that most common fruit ions are the largest, but not unique, part of the most common fly ions. It is important to note that 92% of ions that were present in flies but not in any fruit (i.e. derived from de novo produced metabolites; yellow stack, left column) are ions common to all flies regardless of their diet (Figure 3, yellow flow from 0 to 4). In contrast to the most common fruit ions, ions unique to a single fruit or shared by a few fruits (fruit intersection degrees 1–3) follow roughly the same distribution of 1/3 being shared by all flies, 1/3 being shared by a few flies, and 1/3 being shared by no flies. This includes ions unique to a given fruit, 39.4% of which are found in all flies. Fly ions that are specific to a given diet come from fruit ions with various intersection degrees. In particular, 27.1% are common to all fruits, 21.5% are unique to a single fruit and only 1.1% are produced de novo (nine ions).

Figure 3 with 1 supplement see all

Download asset Open asset

Finally, we visually investigated the multi-dimensional clustering of fruits’ and flies’ metabolomes using a PCA using the integral dataset (Figure 1F). If diets are the main influence on flies’ metabolomes, flies should not cluster together and rather localize close to their corresponding fruit, especially on the second dimension (Figure 1G, right panel). On the contrary, if flies similarly process different diets, flies’ metabolomes should cluster together, away from the fruits (Figure 1G, left panel). Results of the PCA show that while different fruits are easily discriminated, differently-fed flies all clustered into a single group (Figure 3—figure supplement 1).

Altogether, our analyses underscore that the metabolomes of flies fed on different fruits are mostly identical, especially from a qualitative standpoint, and unanimously point to metabolic generalism, over multi-host metabolic specialism, as a mechanism promoting diet breadth.

Metabolomic discrimination of diets in flies

Given the overall qualitative similarity of flies that developed on different diets, we then investigated whether quantitative hallmarks of diet consumption could be detected in flies. A PCA focusing on the fly dataset demonstrated that flies’ diets can be readily inferred (Figure 4). While the first dimension of the PCA (20% of variance explained) discriminates between biological replicates of our experiment (i.e. generations of flies), the second and third dimensions of the PCA (18.3% and 10.5% of variance explained, respectively) discriminate fly metabolomes according to their diet.

Figure 4 with 7 supplements see all

Download asset Open asset

To further detect which ions were indicative of the development of particular fruits, we then computed diet-specific binomial linear models with Elastic Net regularization. We specifically aimed at producing two lists of diet-specific ions based on their quantitative levels in flies’ metabolomes and on two sets of parameters designed (i) to optimally classify flies according to their diet (‘large’ list) and (ii) to be as short as possible to allow manual curation of the identified ions (‘compact’ list, see Materials & methods for details). This statistical approach yielded a ‘large’ list consisting of 169 fly ions associated with a blackcurrant diet, 103 with a cherry diet, 121 with a cranberry diet, and 40 with a strawberry diet. All four ‘large’ models enabled a perfect classification of flies according to their diet (ratio of deviance explained >98%). As expected, these ions are part of the top set of ions driving the explained variance in the dimensions of the previous PCA, as shown by their location in the ion space (Figure 4—figure supplement 1). The ‘compact’ set of Elastic Net parameters attained a similar deviance ratio and yielded a list of 62 fly ions associated with a blackcurrant diet, 40 with a cherry diet, 49 with a cranberry diet, and 15 with a strawberry diet. All fly ions in the ‘compact’ list were present in the ‘large’ list. As shown in Figure 4—figure supplement 2, the ‘large’ list of fly ions can perfectly class fly samples according to their diet. It also suggests that most of these fly ions are present in similar quantities in their respective fruit, with few exceptions. Consistent with these results, we verified that ions of the ‘large’ list for a given flies’ diet were significantly more present in flies of the said diet than in flies developed on other diets and also more present in the said fruit than on other fruits (Figure 4—figure supplement 3). Depending on the diet, quantities of these distinctive ions found in flies could be correlated to the quantity found in the fruit (e.g. blackcurrant; Figure 4—figure supplement 4A, B) or not (e.g. cherry, cranberry, strawberry; Figure 4—figure supplement 4C, D).

Diet-associated ions were found to be part of intersections of every degree, both in flies and in fruits (Figure 4—figure supplement 5). In flies, large parts of these ions were specific to the flies that consumed the diet, especially for blackcurrant-fed flies (degree 1: 76%, 31%, 33%, and 25% in blackcurrant, cherry, cranberry, and strawberry-fed flies, respectively). However, ions contained in intersections of larger degrees were also observed, including ions found in all flies (degree 4: 16%, 48%, 31%, and 53% in blackcurrant, cherry, cranberry, and strawberry-fed flies, respectively; Figure 4—figure supplements 5 and 6). While these diet-distinctive metabolites did not always originate from fruit-specific metabolites and could be detected in other flies and other fruits (Figure 4—figure supplement 5), their quantities were lower there than in the focal fruit and flies who consumed it. It is worth noting that few diet-specific ions were produced de novo by flies (fruits intersection of degree zero: 0%, 5%, 5%, and 3% ions in blackcurrant, cherry, cranberry, and strawberry-fed flies, respectively), consistent with the metabolic generalism hypothesis.

Altogether, we found that quantitative levels of ions are critical to correctly classify fly metabolomes according to diet. Consistent with this finding, we observed that qualitative and quantitative datasets did not have the same resolving power when it came to distinguishing flies based on their diet (Figure 4—figure supplement 7). While randomly picking ~300 ions would lead to a 75% classification accuracy for flies’ diet when using their quantification levels, a similar performance was attained with a sampling of ~1000 ions when using only their presence. Regardless, flies’ diet could be readily inferred based on a relatively small subset of ions.

Identification and fate of metabolites associated with fruit diets

To gain a better understanding of the distinctive features of flies and their diet, we tried to identify diet-specific fly ions of the ‘compact’ list together with the 50 major ions (i.e. those having the largest peak area) for each investigated fruit. Manual curation and detailed mass spectra analysis allowed us to identify 71 metabolites, 26 of which were confirmed with commercial standards. Identified metabolites belonged to major plant metabolite families, including anthocyanins, flavonoids, organic and hydroxycinnamic acids, and terpenes (Supplementary file 1). A two-dimensional clustering analysis of metabolites associated with diet illustrates that even a limited set of these compounds was sufficient to infer flies’ diet (Figure 5). A survey of the published literature showed that many of the metabolites identified here in specific fruits (and flies that were grown on them) had already been detected and identified in the same fruits (Figure 5, Supplementary file 1), thereby supporting their correct identification.

Figure 5 with 1 supplement see all

Download asset Open asset

Depending on fruits and specific metabolites, relative amounts of fruit-derived metabolites detected in flies did not necessarily mirror their relative abundance in fruits (Figure 5—figure supplement 1). Indeed, some major fruit metabolites, such as the anthocyanins delphinidin in blackcurrant, cyanidin in cherry, or pelargonidin in strawberry, were detected in flies, albeit with a relatively low relative abundance (fly/fruit ratio between 0.01 and 0.001). Conversely, other less abundant fruit metabolites, such as terpenes, were relatively abundant in flies (fly/fruit ratio >0.5). Despite the relatively low number of identified metabolites, our quantification results demonstrate that only some fruit metabolites are metabolized (or excreted), while others tend to accumulate in flies.

As expected from previous studies, organic acids tended to be shared by all fruits, but with a particular balance depending on fruits: malic, quinic, and citric acids were always present in fruits (and flies) but with elevated levels in cherry, cranberry, and blackcurrant fruits (and flies that consumed them), respectively. Another chemical family ubiquitous in plants are flavonoids, with many potential physiological effects for consumer organisms, such as neurobehavioral modulation of activity (Bugel and Tanguay, 2018), feeding, and oviposition (Simmonds, 2001; Treutter, 2006). All fruits harbored major flavonoids such as quercetin, kaempferol, or myricetin, with specific substituent groups: quercetin arabinoside and rhamnoside in cranberry, quercetin rutinoside in cherry, quercetin glucoside in blackcurrant and quercetin glucuronide in strawberry. Those substituents have been shown to modulate the bioactivity of flavonoids (Bugel and Tanguay, 2018) and might mediate differential developmental responses to fruit diet in D. suzukii (Olazcuaga et al., 2019). While these substituents can undergo profound modifications by the gut microbiome (e.g. hydrolysis), we found no large-scale trace of microbiota influence in the present study, with flavonoids being essentially carried ‘as is’ in flies’ metabolomes. As an interesting subclass of flavonoids, major anthocyanins characteristic of the investigated fruits (and of their color) were readily found in flies that consumed these fruits. Strawberry-fed flies displayed elevated levels of pelargonidin glucoside, the main anthocyanin of strawberry fruits (da Silva et al., 2007). Similarly, cranberry- and cherry-fed flies presented high levels of cyanidin arabinoside, and cyanidin glucoside and rutinoside, the main anthocyanins found in cranberry and cherry, respectively (Acero et al., 2019; Seeram et al., 2004). Finally, fruits and flies that consumed them also displayed specific chemical signatures associated with terpenoids, considered as the largest class of plant secondary metabolites (Dudareva et al., 2004). Our results show that blackcurrant-fed flies can easily be discriminated by their levels in some linalool derivatives. While this volatile, floral-scented monoterpene is found in many plant families (Knudsen et al., 2006; Pichersky and Lewinsohn, 2011), specific linalool derivatives, e.g., carboxylinalool glycoside and linalooloxide glycoside, were predictive of the flies’ blackcurrant diet. Similarly, flies that developed on strawberry can be differentiated based on their elevated content in nerolidol, a sesquiterpene found in the floral scent of many plant families (Knudsen et al., 2006) and in fruits including strawberry (Aharoni et al., 2004).

While most of these signatures involve a passive accumulation of plant secondary metabolites, we also uncovered traces of the metabolization of fruit compounds. Indeed, we found elevated levels of hippuric acid in all flies, and particularly in cranberry-fed flies. Hippuric acid originates from the metabolization of various phenolic compounds (such as flavonoids and chlorogenic acids) by the gut microbiome and is a biomarker of fruit (and especially berries) consumption in various species including humans (De Simone et al., 2021; Ulaszewska et al., 2020). This result highlights that the metabolization process of major fruit compounds by D. suzukii may be relatively universal, and lack specificity, as shown by the accumulation of monoterpenes in blackberry-fed flies. However, it is noteworthy that some compounds belonging to other families of toxic plant metabolites do not seem to affect D. suzukii. Both cherry puree and cherry-fed flies were found to contain amygdalin, which belongs to the cyanogenic glucoside family. The toxicity of cyanogenic glucosides has been widely documented in the context of plant-insect interactions (Zagrobelny et al., 2004). However, cherry has been identified as an excellent host for D. suzukii, in both oviposition preference and larval performance experiments (Olazcuaga et al., 2019).

The direct comparison of metabolomes of fruits and of individuals that were grown on them sheds some light on basic issues pertaining to diet generalism (Forister et al., 2012). Indeed, we found that consumption by generalist individuals of dissimilar diets resulted in a diet-unspecific qualitative response to host biochemistry, i.e., ‘metabolic generalism.’ This tolerant reaction to diet also leads to the passive accumulation of diet-specific metabolites in individuals. As a result, while individuals were mostly similar across diets, detecting their particular diet was straightforward.

The biochemical composition of stone fruits, berries, and pseudocarps has been explored in numerous studies, highlighting the major contributions of sugars, organic acids, and secondary metabolites to the chemistry of fruits. Our untargeted overview of fruit composition enabled us to confirm that these fruits differ widely in their biochemical composition, and most importantly, that phylogenetic differences correspond to different biochemical landscapes for frugivorous species in our case. It is important to note, however, that the phylogenetic classification of the studied fruit species does not represent their proximity in terms of chemical composition. This strengthens the view that the taxonomic scale may not be the best scale to evaluate diet breadth and the adaptive challenges posed by generalism. For herbivores, it is also important to consider which plant organs are consumed (roots, leaves, fruits), as their respective composition may offer different dietary challenges and different proximities between plant species.

The use of diverse hosts had a variety of biochemical consequences on individuals of a generalist species, all consistent with the ‘metabolic generalism’ hypothesis. Indeed, diet-specific metabolites were rare, most fruit metabolites were incorporated in all individuals regardless of diet and overall metabolite pattern indicated a proximity across all insect samples. On the one hand, individuals showed a remarkably fixed qualitative chemical response to diet, coherent with a canalized response. This pattern mirrored the metabolic homeostasis found in some generalist insects reared on various protein:carbohydrates (P:C) ratios (Silva-Soares et al., 2017; Watanabe et al., 2019; Young et al., 2018). On the other hand, different diets had a significant impact on the quantity of a multitude of ions, resulting in the straightforward identification of insects’ diets based on their metabolome. Our study thus highlights that the physiological response to various diets, while generic in structure, was plastic when relative metabolite amounts were considered. This biochemical response could be expected from studies underlining the transcriptional plasticity in generalist species from all major insect clades (Birnbaum and Abbot, 2020; Celorio-Mancera et al., 2016; Celorio-Mancera et al., 2012; Celorio-Mancera et al., 2013; Christodoulides et al., 2017; Ragland et al., 2015; Simon et al., 2015).

The simultaneous observation of a single general qualitative response and diverse quantitative responses to diets is coherent with the metabolic generalism hypothesis and may help reconcile paradoxical notions about the consequences of generalism at the organismal level. On the one hand, individuals from generalist species do not always differ in response to various diets (e.g. Leptinotarsa decemlineata, Forister et al., 2007; or Lymantria dispar, Barbosa, 1978; Lazarevic et al., 1998). It is a defining characteristic of generalist species that they can tolerate (i.e. display relatively constant developmental responses under) a wider range of dietary circumstances when compared to specialist species. It makes sense, therefore, that the structural backbone of the metabolome of generalist individuals remains largely unaffected by diet, as uncovered here. On the other hand, the diverse quantitative responses of individuals’ metabolomes to diet highlight the passive metabolic response of generalist individuals to the idiosyncrasies of their dietary environment. In particular, we observed that distinctive fruit secondary metabolites, relating to their specific flavor (flavonoids), color (anthocyanins), or scent (terpenoids), were accumulated in corresponding individuals, and that very few de novo diet-specific individuals’ metabolites were produced. This may be a common feature in generalist frugivores as they have to face the exceptional diversity of secondary metabolites found in fruits (Kessler and Kalske, 2018; Wetzel et al., 2020; Whitehead et al., 2022).

Our study thus suggests that diet generalization could stem from reduced attention to hosts’ biochemical peculiarities and a neutral, passive stance towards potential hosts. While most of the ecological and evolutionary studies of host-plant interactions have focused on intricate plant-insect relations relying on strong selective forces, a passive and generic metabolic response could help species extend their diet breadth, at least transiently on evolutionary timescales. In this case, diet generalism would originate from a general tolerance rather than from the accumulation of independent genetic features associated with the exploitation of different resources (Forister et al., 2012; but see Calla et al., 2017; Rane et al., 2019), and the resulting plastic quantitative response to the diet may not necessarily be adaptive.

Immediate costs of weakened defenses against specific plant compounds can be observed in generalist species (Ali and Agrawal, 2012). Here, the passive accumulation of terpenoids in blackcurrant-fed flies indicates that D. suzukii may not have evolved efficient monoterpene metabolization pathways. In addition to their attractant or deterrent properties (Binder and Robbins, 1997), some terpenes have been shown to exhibit toxic properties and to be subjected to detoxification pathways in insects (Blomquist et al., 2021; Scalerandi et al., 2018). Indeed, in D. suzukii, a blackcurrant diet causes a relatively high larval mortality and leads to the extinction of the population in only a couple of generations (Olazcuaga et al., 2019; Olazcuaga et al., 2021). However, other fruits harboring potentially harmful compounds, such as amygdalin in cherry, can be excellent hosts for D. suzukii, including elevated oviposition preference and larval performance (Olazcuaga et al., 2019). Therefore, metabolic generalism may not exclude some adaptations to cope with specific classes of toxic compounds, potentially reflecting the ancestral niche of generalist herbivores.

Metabolic generalism could also procure both immediate and longer-term benefits. For instance, the accumulation of flavonoids increases the fitness of some phytophagous species (e.g. Polyommatus icarus; Simmonds, 2003). While chemical sequestration is an active process in specialist species, passive mechanisms may be common in generalist species (Petschenka and Agrawal, 2016). Here, the probably passive ‘sequestration’ of flavonoids by generalist individuals might be useful against predators or parasitoids. On larger ecological scales, metabolic generalism could also entail benefits, e.g., by enabling a geographical expansion triggering enemy-release or by reducing intra-guild competition. When successive generations of herbivores must exploit different hosts seasonally (e.g. in frugivores), an immediate reward of metabolic generalism could stem from resource allocation alternatives, e.g., a homeostatic immune response.

The passive accumulation of plant metabolites, such as anthocyanins, may also illustrate a greater opportunity for diet generalism due to plant domestication, especially for frugivorous species (Whitehead et al., 2017). Indeed, anthocyanins are known to trade-off with plant defense systems, and the domestication of fruit crops, by acting on anthocyanins accumulation, may have lowered plant defenses and thereby allowed generalist species to exploit niches that were ancestrally too challenging. For instance, it has been demonstrated that the domestication of cranberry cultivars, by selecting upon anthocyanin-related traits such as ripening and fruit color intensity, resulted in lowered plant defenses responses related to the jasmonic acid pathway (Rodriguez-Saona et al., 2011). In strawberries, it is known that biochemical insect resistance traits correlate negatively with important agronomical traits such as yield (Hancock et al., 2008). On the contrary, blackcurrant domestication mainly focused on plant pathogens and frost resistance for much of its history (Brennan, 2008), and this could explain the high levels of detrimental terpenoids in this species. Overall, biochemical changes induced by plant domestication are expected to lead to (i) increased diet generalism and (ii) increased levels of non-lethal plant secondary metabolites in herbivores, especially in frugivores.

Finally, by showing that fruit-specific metabolites were readily found in metabolomes, our study suggests that metabolomic data might help track the diet of phytophagous insects in the wild (Alberdi et al., 2019; Nielsen et al., 2018). This could complement high-throughput DNA sequencing-based diet reconstruction, which relies on the presence of foreign DNA material. Prerequisites for achieving this goal include an increased ability to correctly identify plant secondary metabolites and laboratory experiments to study the effect of various diets on the herbivores’ metabolomes during development.

In the future, the investigation of different spatial and temporal scales in resource use could help to shed light on the equilibrium between diet generalism and specialization (Devictor et al., 2010), and on the origins and limits of diet generalism (Bolnick et al., 2003; Roughgarden, 1972; Sexton et al., 2017). To date, the interaction of phytophagous insect species such as leaf-chewing and mining insects with their host plants has drawn much of the research attention. Compared to more general plant tissues, fruits are specialized organs interacting with multiple species, with detrimental or beneficial impacts to plant fitness, and this may have shaped a greater complexity of the chemical landscape in fruits (Cipollini and Levey, 1997). If the response to insect pests tends to decrease the attractiveness or palatability of fruits for more desirable seed dispersers, such as vertebrate frugivores, anti-insect defenses may perturb the reproductive potential of host plants (Andersen, 1988; but see Wilson et al., 2012). Understanding the ecological constraints exerted on both the chemistry of fruits and the physiology of their consumers, impacting the evolution of niche breadth itself, will require a detailed knowledge of molecular- to community-level interactions across multiple trophic levels.

All data & code generated during this study have been deposited in the INRAE dataverse. The shiny application enabling the exploration and analysis of our complete dataset (PCA, GLM/Elastic Net and associated visualizations) is available here.

1. 1. Foucaud J
   2. Olazcuaga L
   3. Baltenweck R
   4. Leménager N
   5. Maia-Grondard A
   6. Claudel P
   7. Hugueney P

   (2022) Dataverse INRAE

   DATA and CODE for "Metabolic consequences of various fruit-based diets in a generalist insect species".

   https://doi.org/10.57745/G4D3PG

1. 1. Acero N
   2. Gradillas A
   3. Beltran M
   4. García A
   5. Muñoz Mingarro D

   (2019) Comparison of phenolic compounds profile and antioxidant properties of different sweet cherry (Prunus avium L.) varieties

   Food Chemistry 279:260–271.

   https://doi.org/10.1016/j.foodchem.2018.12.008

   • PubMed
   • Google Scholar
2. 1. Acharjee A

   (2013) Comparison of regularized regression methods for ~omics data

   Journal of Postgenomics Drug & Biomarker Development 03:1000126.

   https://doi.org/10.4172/2153-0769.1000126

   • Google Scholar
3. 1. Aharoni A
   2. Giri AP
   3. Verstappen FWA
   4. Bertea CM
   5. Sevenier R
   6. Sun Z
   7. Jongsma MA
   8. Schwab W
   9. Bouwmeester HJ

   (2004) Gain and loss of fruit flavor compounds produced by wild and cultivated strawberry species

   The Plant Cell 16:3110–3131.

   https://doi.org/10.1105/tpc.104.023895

   • PubMed
   • Google Scholar
4. 1. Alberdi A
   2. Aizpurua O
   3. Bohmann K
   4. Gopalakrishnan S
   5. Lynggaard C
   6. Nielsen M
   7. Gilbert MTP

   (2019) Promises and pitfalls of using high-throughput sequencing for diet analysis

   Molecular Ecology Resources 19:327–348.

   https://doi.org/10.1111/1755-0998.12960

   • PubMed
   • Google Scholar
5. 1. Ali JG
   2. Agrawal AA

   (2012) Specialist versus generalist insect herbivores and plant defense

   Trends in Plant Science 17:293–302.

   https://doi.org/10.1016/j.tplants.2012.02.006

   • PubMed
   • Google Scholar
6. 1. Andersen AN

   (1988) Insect seed predators may cause far greater losses than they appear to

   Oikos 52:337.

   https://doi.org/10.2307/3565207

   • Google Scholar
7. 1. Asplen MK
   2. Anfora G
   3. Biondi A
   4. Choi DS
   5. Chu D
   6. Daane KM
   7. Gibert P
   8. Gutierrez AP
   9. Hoelmer KA
   10. Hutchison WD
   11. Isaacs R
   12. Jiang ZL
   13. Kárpáti Z
   14. Kimura MT
   15. Pascual M
   16. Philips CR
   17. Plantamp C
   18. Ponti L
   19. Vétek G
   20. Vogt H
   21. Walton VM
   22. Yu Y
   23. Zappalà L
   24. Desneux N

   (2015) Invasion biology of spotted wing Drosophila (Drosophila suzukii): a global perspective and future priorities

   Journal of Pest Science 88:469–494.

   https://doi.org/10.1007/s10340-015-0681-z

   • Google Scholar
8. 1. Backhaus B
   2. Sulkowski E
   3. Schlote FW

   (1984)

   A semi-synthetic, general-purpose medium for Drosophila melanogaster

   Drosophila Information Service 60:210–212.

   • Google Scholar
9. 1. Barbosa P

   (1978) Host plant exploitation by the gypsy moth, Lymantria dispar

   Entomologia Experimentalis et Applicata 24:228–237.

   https://doi.org/10.1111/j.1570-7458.1978.tb02777.x

   • Google Scholar
10. 1. Bellamy DE
    2. Sisterson MS
    3. Walse SS

    (2013) Quantifying host potentials: indexing postharvest fresh fruits for spotted wing Drosophila, Drosophila suzukii

    PLOS ONE 8:e61227.

    https://doi.org/10.1371/journal.pone.0061227

    • PubMed
    • Google Scholar
11. 1. Berenbaum M

    (1981) Patterns of furanocoumarin distribution and insect herbivory in the Umbelliferae: plant chemistry and community structure

    Ecology 62:1254–1266.

    https://doi.org/10.2307/1937290

    • Google Scholar
12. 1. Binder BF
    2. Robbins JC

    (1997) Effect of terpenoids and related compounds on the oviposition behavior of the European corn borer, Ostrinia nubilalis (Lepidoptera: Pyralidae)

    Journal of Agricultural and Food Chemistry 45:980–984.

    https://doi.org/10.1021/jf960400z

    • Google Scholar
13. 1. Birnbaum SSL
    2. Abbot P

    (2020) Gene expression and diet breadth in plant-feeding insects: summarizing trends

    Trends in Ecology & Evolution 35:259–277.

    https://doi.org/10.1016/j.tree.2019.10.014

    • PubMed
    • Google Scholar
14. 1. Blomquist GJ
    2. Tittiger C
    3. MacLean M
    4. Keeling CI

    (2021) Cytochromes P450: terpene detoxification and pheromone production in bark beetles

    Current Opinion in Insect Science 43:97–102.

    https://doi.org/10.1016/j.cois.2020.11.010

    • PubMed
    • Google Scholar
15. 1. Bolnick DI
    2. Svanbäck R
    3. Fordyce JA
    4. Yang LH
    5. Davis JM
    6. Hulsey CD
    7. Forister ML

    (2003) The ecology of individuals: incidence and implications of individual specialization

    The American Naturalist 161:1–28.

    https://doi.org/10.1086/343878

    • PubMed
    • Google Scholar
16. 1. Bowers MD

    (1983) The role of iridoid glycosides in host-plant specificity of checkerspot butterflies

    Journal of Chemical Ecology 9:475–493.

    https://doi.org/10.1007/BF00990220

    • PubMed
    • Google Scholar
17. Book

    1. Brennan RM

    (2008) Currants and gooseberries

    In: Hancock JF, editors. Temperate Fruit Crop Breeding: Germplasm to Genomics. Springer Netherlands. pp. 177–196.

    https://doi.org/10.1007/978-1-4020-6907-9_6

    • Google Scholar
18. 1. Bugel SM
    2. Tanguay RL

    (2018) Multidimensional chemobehavior analysis of flavonoids and neuroactive compounds in zebrafish

    Toxicology and Applied Pharmacology 344:23–34.

    https://doi.org/10.1016/j.taap.2018.02.019

    • PubMed
    • Google Scholar
19. 1. Calla B
    2. Noble K
    3. Johnson RM
    4. Walden KKO
    5. Schuler MA
    6. Robertson HM
    7. Berenbaum MR

    (2017) Cytochrome P450 diversification and hostplant utilization patterns in specialist and generalist moths: birth, death and adaptation

    Molecular Ecology 26:6021–6035.

    https://doi.org/10.1111/mec.14348

    • PubMed
    • Google Scholar
20. 1. Celorio-Mancera MP
    2. Heckel DG
    3. Vogel H

    (2012) Transcriptional analysis of physiological pathways in a generalist herbivore: responses to different host plants and plant structures by the cotton bollworm, Helicoverpa armigera

    Entomologia Experimentalis et Applicata 144:123–133.

    https://doi.org/10.1111/j.1570-7458.2012.01249.x

    • Google Scholar
21. 1. Celorio-Mancera MP
    2. Wheat CW
    3. Vogel H
    4. Söderlind L
    5. Janz N
    6. Nylin S

    (2013) Mechanisms of macroevolution: polyphagous plasticity in butterfly larvae revealed by RNA-seq

    Molecular Ecology 22:4884–4895.

    https://doi.org/10.1111/mec.12440

    • PubMed
    • Google Scholar
22. 1. Celorio-Mancera MP
    2. Wheat CW
    3. Huss M
    4. Vezzi F
    5. Neethiraj R
    6. Reimegård J
    7. Nylin S
    8. Janz N

    (2016) Evolutionary history of host use, rather than plant phylogeny, determines gene expression in a generalist butterfly

    BMC Evolutionary Biology 16:59.

    https://doi.org/10.1186/s12862-016-0627-y

    • PubMed
    • Google Scholar
23. Software

    1. Chiquet J
    2. Rigaill G
    3. Sundqvist M

    (2020) Aricode: efficient computations of standard clustering comparison measures, version version 1.0.0

    R Package.

    https://CRAN.R-project.org/package=aricode
24. 1. Christodoulides N
    2. Van Dam AR
    3. Peterson DA
    4. Frandsen RJN
    5. Mortensen UH
    6. Petersen B
    7. Rasmussen S
    8. Normark BB
    9. Hardy NB

    (2017) Gene expression plasticity across hosts of an invasive scale insect species

    PLOS ONE 12:e0176956.

    https://doi.org/10.1371/journal.pone.0176956

    • PubMed
    • Google Scholar
25. 1. Cini A
    2. Anfora G
    3. Escudero-Colomar LA
    4. Grassi A
    5. Santosuosso U
    6. Seljak G
    7. Papini A

    (2014) Tracking the invasion of the alien fruit pest Drosophila suzukii in Europe

    Journal of Pest Science 87:559–566.

    https://doi.org/10.1007/s10340-014-0617-z

    • Google Scholar
26. 1. Cipollini ML
    2. Levey DJ

    (1997) Secondary metabolites of fleshy vertebrate-dispersed fruits: adaptive hypotheses and implications for seed dispersal

    The American Naturalist 150:346–372.

    https://doi.org/10.1086/286069

    • PubMed
    • Google Scholar
27. 1. Clarke AR
    2. Armstrong KF
    3. Carmichael AE
    4. Milne JR
    5. Raghu S
    6. Roderick GK
    7. Yeates DK

    (2005) Invasive phytophagous pests arising through a recent tropical evolutionary radiation: the Bactrocera dorsalis complex of fruit flies

    Annual Review of Entomology 50:293–319.

    https://doi.org/10.1146/annurev.ento.50.071803.130428

    • PubMed
    • Google Scholar
28. 1. Conway JR
    2. Lex A
    3. Gehlenborg N

    (2017) UpSetR: an R package for the visualization of intersecting sets and their properties

    Bioinformatics 33:2938–2940.

    https://doi.org/10.1093/bioinformatics/btx364

    • PubMed
    • Google Scholar
29. Software

    1. Cory J
    2. Read QD

    (2020) Ggalluvial: alluvial plots in « ggplot2 », version version 0.12.3

    R Package.

    http://corybrunson.github.io/ggalluvial/
30. 1. da Silva FL
    2. Escribano-Bailón MT
    3. Pérez Alonso JJ
    4. Rivas-Gonzalo JC
    5. Santos-Buelga C

    (2007) Anthocyanin pigments in strawberry

    LWT - Food Science and Technology 40:374–382.

    https://doi.org/10.1016/j.lwt.2005.09.018

    • Google Scholar
31. 1. Day EH
    2. Hua X
    3. Bromham L

    (2016) Is specialization an evolutionary dead end? testing for differences in speciation, extinction and trait transition rates across diverse phylogenies of specialists and generalists

    Journal of Evolutionary Biology 29:1257–1267.

    https://doi.org/10.1111/jeb.12867

    • PubMed
    • Google Scholar
32. 1. De Simone G
    2. Balducci C
    3. Forloni G
    4. Pastorelli R
    5. Brunelli L

    (2021) Hippuric acid: could became a barometer for frailty and geriatric syndromes?

    Ageing Research Reviews 72:101466.

    https://doi.org/10.1016/j.arr.2021.101466

    • PubMed
    • Google Scholar
33. 1. Devictor V
    2. Clavel J
    3. Julliard R
    4. Lavergne S
    5. Mouillot D
    6. Thuiller W
    7. Venail P
    8. Villéger S
    9. Mouquet N

    (2010) Defining and measuring ecological specialization

    Journal of Applied Ecology 47:15–25.

    https://doi.org/10.1111/j.1365-2664.2009.01744.x

    • Google Scholar
34. 1. Dudareva N
    2. Pichersky E
    3. Gershenzon J

    (2004) Biochemistry of plant volatiles

    Plant Physiology 135:1893–1902.

    https://doi.org/10.1104/pp.104.049981

    • PubMed
    • Google Scholar
35. 1. Forister ML
    2. Ehmer AG
    3. Futuyma DJ

    (2007) The genetic architecture of a niche: variation and covariation in host use traits in the Colorado potato beetle

    Journal of Evolutionary Biology 20:985–996.

    https://doi.org/10.1111/j.1420-9101.2007.01310.x

    • PubMed
    • Google Scholar
36. 1. Forister ML
    2. Dyer LA
    3. Singer MS
    4. Stireman JO
    5. Lill JT

    (2012) Revisiting the evolution of ecological specialization, with emphasis on insect-plant interactions

    Ecology 93:981–991.

    https://doi.org/10.1890/11-0650.1

    • PubMed
    • Google Scholar
37. 1. Forister ML
    2. Novotny V
    3. Panorska AK
    4. Baje L
    5. Basset Y
    6. Butterill PT
    7. Cizek L
    8. Coley PD
    9. Dem F
    10. Diniz IR
    11. Drozd P
    12. Fox M
    13. Glassmire AE
    14. Hazen R
    15. Hrcek J
    16. Jahner JP
    17. Kaman O
    18. Kozubowski TJ
    19. Kursar TA
    20. Lewis OT
    21. Lill J
    22. Marquis RJ
    23. Miller SE
    24. Morais HC
    25. Murakami M
    26. Nickel H
    27. Pardikes NA
    28. Ricklefs RE
    29. Singer MS
    30. Smilanich AM
    31. Stireman JO
    32. Villamarín-Cortez S
    33. Vodka S
    34. Volf M
    35. Wagner DL
    36. Walla T
    37. Weiblen GD
    38. Dyer LA

    (2015) The global distribution of diet breadth in insect herbivores

    PNAS 112:442–447.

    https://doi.org/10.1073/pnas.1423042112

    • PubMed
    • Google Scholar
38. 1. Friedman J
    2. Hastie T
    3. Tibshirani R

    (2010)

    Regularization paths for generalized linear models via coordinate descent

    Journal of Statistical Software 33:1–22.

    • PubMed
    • Google Scholar
39. 1. Futuyma DJ
    2. Moreno G

    (1988) The evolution of ecological specialization

    Annual Review of Ecology and Systematics 19:207–233.

    https://doi.org/10.1146/annurev.es.19.110188.001231

    • Google Scholar
40. 1. Gonzales GB
    2. De Saeger S

    (2018) Author correction: elastic net regularized regression for time-series analysis of plasma metabolome stability under sub-optimal freezing condition

    Scientific Reports 8:9421.

    https://doi.org/10.1038/s41598-018-27563-2

    • PubMed
    • Google Scholar
41. 1. Gu Z
    2. Eils R
    3. Schlesner M

    (2016) Complex heatmaps reveal patterns and correlations in multidimensional genomic data

    Bioinformatics 32:2847–2849.

    https://doi.org/10.1093/bioinformatics/btw313

    • PubMed
    • Google Scholar
42. Software

    1. Hahsler M
    2. Buchta C
    3. Hornik K

    (2020) Seriation: infrastructure for ordering objects using seriation, version version 1.2-9

    R Package.

    https://CRAN.R-project.org/package=seriation
43. Book

    1. Hancock JF
    2. Sjulin TM
    3. Lobos GA

    (2008) Strawberries

    In: Hancock JF, editors. Temperate Fruit Crop Breeding: Germplasm to Genomics. Springer Netherlands. pp. 393–437.

    https://doi.org/10.1007/978-1-4020-6907-9_13

    • Google Scholar
44. 1. Haug K
    2. Cochrane K
    3. Nainala VC
    4. Williams M
    5. Chang J
    6. Jayaseelan KV
    7. O’Donovan C

    (2020) MetaboLights: a resource evolving in response to the needs of its scientific community

    Nucleic Acids Research 48:D440–D444.

    https://doi.org/10.1093/nar/gkz1019

    • PubMed
    • Google Scholar
45. Book

    1. Helsel DR

    (2011) Statistics for Censored Environmental Data Using Minitab and R

    John Wiley & Sons, Inc.

    https://doi.org/10.1002/9781118162729.fmatter

    • Google Scholar
46. 1. Hotti H
    2. Rischer H

    (2017) The killer of socrates: coniine and related alkaloids in the plant Kingdom

    Molecules 22:1962.

    https://doi.org/10.3390/molecules22111962

    • PubMed
    • Google Scholar
47. 1. Jaenike J

    (1990) Host specialization in phytophagous insects

    Annual Review of Ecology and Systematics 21:243–273.

    https://doi.org/10.1146/annurev.es.21.110190.001331

    • Google Scholar
48. 1. Janz N
    2. Nyblom K
    3. Nylin S

    (2001) Evolutionary dynamics of host-plant specialization: a case study of the tribe Nymphalini

    Evolution; International Journal of Organic Evolution 55:783–796.

    https://doi.org/10.1554/0014-3820(2001)055[0783:edohps]2.0.co;2

    • PubMed
    • Google Scholar
49. 1. Johnson CH
    2. Ivanisevic J
    3. Siuzdak G

    (2016) Metabolomics: beyond biomarkers and towards mechanisms

    Nature Reviews. Molecular Cell Biology 17:451–459.

    https://doi.org/10.1038/nrm.2016.25

    • PubMed
    • Google Scholar
50. 1. Kanzawa T

    (1939)

    Studies on Drosophila suzukii (yamanashi kenritsu nouji shikenjo gyomu nenpo)

    Review of Applied Entomology 29:622.

    • Google Scholar
51. 1. Kenis M
    2. Tonina L
    3. Eschen R
    4. van der Sluis B
    5. Sancassani M
    6. Mori N
    7. Haye T
    8. Helsen H

    (2016) Non-crop plants used as hosts by Drosophila suzukii in Europe

    Journal of Pest Science 89:735–748.

    https://doi.org/10.1007/s10340-016-0755-6

    • PubMed
    • Google Scholar
52. 1. Kessler A
    2. Kalske A

    (2018) Plant secondary metabolite diversity and species interactions

    Annual Review of Ecology, Evolution, and Systematics 49:115–138.

    https://doi.org/10.1146/annurev-ecolsys-110617-062406

    • Google Scholar
53. 1. Kirpich A
    2. Ainsworth EA
    3. Wedow JM
    4. Newman JRB
    5. Michailidis G
    6. McIntyre LM

    (2018) Variable selection in omics data: a practical evaluation of small sample sizes

    PLOS ONE 13:e0197910.

    https://doi.org/10.1371/journal.pone.0197910

    • PubMed
    • Google Scholar
54. 1. Knudsen JT
    2. Eriksson R
    3. Gershenzon J
    4. Ståhl B

    (2006) Diversity and distribution of floral scent

    The Botanical Review 72:1–120.

    https://doi.org/10.1663/0006-8101(2006)72[1:DADOFS]2.0.CO;2

    • Google Scholar
55. Software

    1. Kuhn M

    (2021) Caret: classification and regression training, version version 6.0-90

    R Package.

    https://CRAN.R-project.org/package=caret
56. 1. Lazarevic J
    2. Peric-Mataruga V
    3. Ivanovic J
    4. Andjelkovic M

    (1998) Host plant effects on the genetic variation and correlations in the individual performance of the gypsy moth

    Functional Ecology 12:141–148.

    https://doi.org/10.1046/j.1365-2435.1998.00166.x

    • Google Scholar
57. 1. Lê S
    2. Josse J
    3. Husson F

    (2008) Factominer: an R package for multivariate analysis

    Journal of Statistical Software 25:i01.

    https://doi.org/10.18637/jss.v025.i01

    • Google Scholar
58. 1. Lee JC
    2. Bruck DJ
    3. Curry H
    4. Edwards D
    5. Haviland DR
    6. Van Steenwyk RA
    7. Yorgey BM

    (2011) The susceptibility of small fruits and cherries to the spotted-wing Drosophila, Drosophila suzukii

    Pest Management Science 67:1358–1367.

    https://doi.org/10.1002/ps.2225

    • PubMed
    • Google Scholar
59. 1. Lee JC
    2. Dreves AJ
    3. Cave AM
    4. Kawai S
    5. Isaacs R
    6. Miller JC
    7. Van Timmeren S
    8. Bruck DJ

    (2015) Infestation of wild and ornamental noncrop fruits by Drosophila suzukii (Diptera: Drosophilidae)

    Annals of the Entomological Society of America 108:117–129.

    https://doi.org/10.1093/aesa/sau014

    • Google Scholar
60. 1. Müllner D

    (2013) Fastcluster: fast hierarchical, agglomerative clustering routines for R and python

    Journal of Statistical Software 53:i09.

    https://doi.org/10.18637/jss.v053.i09

    • Google Scholar
61. 1. Nielsen JM
    2. Clare EL
    3. Hayden B
    4. Brett MT
    5. Kratina P
    6. Gilbert MTP

    (2018) Diet tracing in ecology: method comparison and selection

    Methods in Ecology and Evolution 9:278–291.

    https://doi.org/10.1111/2041-210X.12869

    • Google Scholar
62. 1. Nosil P

    (2002) Transition rates between specialization and generalization in phytophagous insects

    Evolution; International Journal of Organic Evolution 56:1701–1706.

    https://doi.org/10.1111/j.0014-3820.2002.tb01482.x

    • PubMed
    • Google Scholar
63. 1. Nylin S
    2. Slove J
    3. Janz N

    (2014) Host plant utilization, host range oscillations and diversification in nymphalid butterflies: a phylogenetic investigation

    Evolution; International Journal of Organic Evolution 68:105–124.

    https://doi.org/10.1111/evo.12227

    • PubMed
    • Google Scholar
64. 1. Olazcuaga L
    2. Rode NO
    3. Foucaud J
    4. Facon B
    5. Ravigné V
    6. Ausset A
    7. Leménager N
    8. Loiseau A
    9. Gautier M
    10. Estoup A
    11. Hufbauer RA

    (2019) Oviposition preference and larval performance of Drosophila suzukii

    Environmental Entomology 1:nvz062.

    https://doi.org/10.1093/ee/nvz062

    • Google Scholar
65. 1. Olazcuaga L
    2. Foucaud J
    3. Gautier M
    4. Deschamps C
    5. Loiseau A
    6. Leménager N
    7. Facon B
    8. Ravigné V
    9. Hufbauer RA
    10. Estoup A
    11. Rode NO

    (2021) Adaptation and correlated fitness responses over two time scales in Drosophila suzukii populations evolving in different environments

    Journal of Evolutionary Biology 34:1225–1240.

    https://doi.org/10.1111/jeb.13878

    • PubMed
    • Google Scholar
66. 1. Petschenka G
    2. Agrawal AA

    (2016) How herbivores coopt plant defenses: natural selection, specialization, and sequestration

    Current Opinion in Insect Science 14:17–24.

    https://doi.org/10.1016/j.cois.2015.12.004

    • PubMed
    • Google Scholar
67. 1. Pichersky E
    2. Lewinsohn E

    (2011) Convergent evolution in plant specialized metabolism

    Annual Review of Plant Biology 62:549–566.

    https://doi.org/10.1146/annurev-arplant-042110-103814

    • PubMed
    • Google Scholar
68. 1. Poyet M
    2. Le Roux V
    3. Gibert P
    4. Meirland A
    5. Prévost G
    6. Eslin P
    7. Chabrerie O

    (2015) The wide potential trophic niche of the asiatic fruit fly Drosophila suzukii: the key of its invasion success in temperate Europe?

    PLOS ONE 10:e0142785.

    https://doi.org/10.1371/journal.pone.0142785

    • PubMed
    • Google Scholar
69. 1. Ragland GJ
    2. Almskaar K
    3. Vertacnik KL
    4. Gough HM
    5. Feder JL
    6. Hahn DA
    7. Schwarz D

    (2015) Differences in performance and transcriptome-wide gene expression associated with rhagoletis (Diptera: Tephritidae) larvae feeding in alternate host fruit environments

    Molecular Ecology 24:2759–2776.

    https://doi.org/10.1111/mec.13191

    • PubMed
    • Google Scholar
70. 1. Rane RV
    2. Clarke DF
    3. Pearce SL
    4. Zhang G
    5. Hoffmann AA
    6. Oakeshott JG

    (2019) Detoxification genes differ between cactus-, fruit-, and flower-feeding Drosophila

    The Journal of Heredity 110:80–91.

    https://doi.org/10.1093/jhered/esy058

    • PubMed
    • Google Scholar
71. Software

    1. R Development Core Team

    (2019) R: a language and environment for statistical computing

    R Foundation for Statistical Computing, Vienna, Austria.

    https://www.r-project.org/index.html
72. 1. Rodriguez-Saona C
    2. Vorsa N
    3. Singh AP
    4. Johnson-Cicalese J
    5. Szendrei Z
    6. Mescher MC
    7. Frost CJ

    (2011) Tracing the history of plant traits under domestication in cranberries: potential consequences on anti-herbivore defences

    Journal of Experimental Botany 62:2633–2644.

    https://doi.org/10.1093/jxb/erq466

    • PubMed
    • Google Scholar
73. 1. Roughgarden J

    (1972) Evolution of niche width

    The American Naturalist 106:683–718.

    https://doi.org/10.1086/282807

    • Google Scholar
74. 1. Roy A
    2. Walker WB
    3. Vogel H
    4. Chattington S
    5. Larsson MC
    6. Anderson P
    7. Heckel DG
    8. Schlyter F

    (2016) Diet dependent metabolic responses in three generalist insect herbivores Spodoptera spp

    Insect Biochemistry and Molecular Biology 71:91–105.

    https://doi.org/10.1016/j.ibmb.2016.02.006

    • PubMed
    • Google Scholar
75. 1. Scalerandi E
    2. Flores GA
    3. Palacio M
    4. Defagó MT
    5. Carpinella MC
    6. Valladares G
    7. Bertoni A
    8. Palacios SM

    (2018) Understanding synergistic toxicity of terpenes as insecticides: contribution of metabolic detoxification in Musca domestica

    Frontiers in Plant Science 9:1579.

    https://doi.org/10.3389/fpls.2018.01579

    • PubMed
    • Google Scholar
76. 1. Schymanski EL
    2. Jeon J
    3. Gulde R
    4. Fenner K
    5. Ruff M
    6. Singer HP
    7. Hollender J

    (2014) Identifying small molecules via high resolution mass spectrometry: communicating confidence

    Environmental Science & Technology 48:2097–2098.

    https://doi.org/10.1021/es5002105

    • PubMed
    • Google Scholar
77. 1. Seeram NP
    2. Adams LS
    3. Hardy ML
    4. Heber D

    (2004) Total cranberry extract versus its phytochemical constituents: antiproliferative and synergistic effects against human tumor cell lines

    Journal of Agricultural and Food Chemistry 52:2512–2517.

    https://doi.org/10.1021/jf0352778

    • PubMed
    • Google Scholar
78. 1. Sexton JP
    2. Montiel J
    3. Shay JE
    4. Stephens MR
    5. Slatyer RA

    (2017) Evolution of ecological niche breadth

    Annual Review of Ecology, Evolution, and Systematics 48:183–206.

    https://doi.org/10.1146/annurev-ecolsys-110316-023003

    • Google Scholar
79. 1. Silva-Soares NF
    2. Nogueira-Alves A
    3. Beldade P
    4. Mirth CK

    (2017) Adaptation to new nutritional environments: larval performance, foraging decisions, and adult oviposition choices in Drosophila suzukii

    BMC Ecology 17:21.

    https://doi.org/10.1186/s12898-017-0131-2

    • PubMed
    • Google Scholar
80. 1. Simmonds MSJ

    (2001) Importance of flavonoids in insect -- plant interactions: feeding and oviposition

    Phytochemistry 56:245–252.

    https://doi.org/10.1016/s0031-9422(00)00453-2

    • PubMed
    • Google Scholar
81. 1. Simmonds MSJ

    (2003) Flavonoid-insect interactions: recent advances in our knowledge

    Phytochemistry 64:21–30.

    https://doi.org/10.1016/s0031-9422(03)00293-0

    • PubMed
    • Google Scholar
82. 1. Simon J-C
    2. d’Alençon E
    3. Guy E
    4. Jacquin-Joly E
    5. Jaquiéry J
    6. Nouhaud P
    7. Peccoud J
    8. Sugio A
    9. Streiff R

    (2015) Genomics of adaptation to host-plants in herbivorous insects

    Briefings in Functional Genomics 14:413–423.

    https://doi.org/10.1093/bfgp/elv015

    • PubMed
    • Google Scholar
83. Book

    1. Singer MS

    (2008) Evolutionary ecology of polyphagy

    In: Tilmon KJ, editors. Specialization, Speciation, and Radiationthe Evolutionary Biology of Herbivorous Insects. Univ California Press. pp. 29–42.

    https://doi.org/10.1525/california/9780520251328.001.0001

    • Google Scholar
84. 1. Smith CA
    2. Want EJ
    3. O’Maille G
    4. Abagyan R
    5. Siuzdak G

    (2006) XCMS: processing mass spectrometry data for metabolite profiling using nonlinear peak alignment, matching, and identification

    Analytical Chemistry 78:779–787.

    https://doi.org/10.1021/ac051437y

    • PubMed
    • Google Scholar
85. 1. Treutter D

    (2006) Significance of flavonoids in plant resistance: a review

    Environmental Chemistry Letters 4:147–157.

    https://doi.org/10.1007/s10311-006-0068-8

    • Google Scholar
86. 1. Ulaszewska M
    2. Garcia-Aloy M
    3. Vázquez-Manjarrez N
    4. Soria-Florido MT
    5. Llorach R
    6. Mattivi F
    7. Manach C

    (2020) Food intake biomarkers for berries and grapes

    Genes & Nutrition 15:17.

    https://doi.org/10.1186/s12263-020-00675-z

    • PubMed
    • Google Scholar
87. 1. Vinh NX
    2. Epps J
    3. Bailey J

    (2010)

    Information theoretic measures for clusterings comparison: variants, properties, normalization and correction for chance

    Journal of Machine Learning Research 11:2837–2854.

    • Google Scholar
88. 1. Waldmann P
    2. Mészáros G
    3. Gredler B
    4. Fuerst C
    5. Sölkner J

    (2013) Evaluation of the LASSO and the elastic net in genome-wide association studies

    Frontiers in Genetics 4:270.

    https://doi.org/10.3389/fgene.2013.00270

    • PubMed
    • Google Scholar
89. 1. Watanabe K
    2. Kanaoka Y
    3. Mizutani S
    4. Uchiyama H
    5. Yajima S
    6. Watada M
    7. Uemura T
    8. Hattori Y

    (2019) Interspecies comparative analyses reveal distinct carbohydrate-responsive systems among Drosophila species

    Cell Reports 28:2594–2607.

    https://doi.org/10.1016/j.celrep.2019.08.030

    • PubMed
    • Google Scholar
90. 1. Wetzel WC
    2. Whitehead SR
    3. Hillebrand H

    (2020) The many dimensions of phytochemical diversity: linking theory to practice

    Ecology Letters 23:16–32.

    https://doi.org/10.1111/ele.13422

    • PubMed
    • Google Scholar
91. 1. Whitehead SR
    2. Turcotte MM
    3. Poveda K

    (2017) Domestication impacts on plant-herbivore interactions: a meta-analysis

    Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 372:20160034.

    https://doi.org/10.1098/rstb.2016.0034

    • PubMed
    • Google Scholar
92. 1. Whitehead SR
    2. Schneider GF
    3. Dybzinski R
    4. Nelson AS
    5. Gelambi M
    6. Jos E
    7. Beckman NG

    (2022) Fruits, frugivores, and the evolution of phytochemical diversity

    Oikos 2022:8332.

    https://doi.org/10.1111/oik.08332

    • Google Scholar
93. 1. Wilson AJ
    2. Schutze M
    3. Elmouttie D
    4. Clarke AR

    (2012) Are insect frugivores always plant pests? the impact of fruit fly (Diptera: Tephritidae) larvae on host plant fitness

    Arthropod-Plant Interactions 6:635–647.

    https://doi.org/10.1007/s11829-012-9205-4

    • Google Scholar
94. 1. Winde I
    2. Wittstock U

    (2011) Insect herbivore counteradaptations to the plant glucosinolate-myrosinase system

    Phytochemistry 72:1566–1575.

    https://doi.org/10.1016/j.phytochem.2011.01.016

    • PubMed
    • Google Scholar
95. 1. Young Y
    2. Buckiewicz N
    3. Long TAF

    (2018) Nutritional geometry and fitness consequences in Drosophila suzukii, the spotted-wing Drosophila

    Ecology and Evolution 8:2842–2851.

    https://doi.org/10.1002/ece3.3849

    • PubMed
    • Google Scholar
96. 1. Yu G
    2. Smith DK
    3. Zhu H
    4. Guan Y
    5. Lam TTY

    (2017) Ggtree: an R package for visualization and annotation of phylogenetic trees with their covariates and other associated data

    Methods in Ecology and Evolution 8:28–36.

    https://doi.org/10.1111/2041-210X.12628

    • Google Scholar
97. 1. Zagrobelny M
    2. Bak S
    3. Rasmussen AV
    4. Jørgensen B
    5. Naumann CM
    6. Lindberg Møller B

    (2004) Cyanogenic glucosides and plant-insect interactions

    Phytochemistry 65:293–306.

    https://doi.org/10.1016/j.phytochem.2003.10.016

    • PubMed
    • Google Scholar
98. 1. Zeng L
    2. Zhang N
    3. Zhang Q
    4. Endress PK
    5. Huang J
    6. Ma H

    (2017) Resolution of deep eudicot phylogeny and their temporal diversification using nuclear genes from transcriptomic and genomic datasets

    The New Phytologist 214:1338–1354.

    https://doi.org/10.1111/nph.14503

    • PubMed
    • Google Scholar
99. 1. Zou H
    2. Hastie T

    (2017) Regularization and variable selection via the elastic net

    Journal of the Royal Statistical Society Series B 67:301–320.

    https://doi.org/10.1111/j.1467-9868.2005.00503.x

    • Google Scholar

Author details

1. Laure Olazcuaga

   1. UMR CBGP (INRAE-IRD-CIRAD, Montpellier SupAgro), Campus International de Baillarguet, Montferrier, France
   2. Department of Agricultural Biology, Colorado State University, Fort Collins, United States

   Contribution

   Conceptualization, Resources, Funding acquisition, Investigation, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Raymonde Baltenweck

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0001-9100-1305

2. Raymonde Baltenweck

   Université de Strasbourg, INRAE, Colmar, France

   Contribution

   Conceptualization, Resources, Data curation, Validation, Investigation, Methodology, Writing – original draft, Writing – review and editing

   Contributed equally with

   Laure Olazcuaga

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-8228-1517

3. Nicolas Leménager

   UMR CBGP (INRAE-IRD-CIRAD, Montpellier SupAgro), Campus International de Baillarguet, Montferrier, France

   Contribution

   Conceptualization, Resources, Investigation, Methodology, Writing – original draft, Writing – review and editing

   Competing interests

   No competing interests declared

4. Alessandra Maia-Grondard

   Université de Strasbourg, INRAE, Colmar, France

   Contribution

   Conceptualization, Resources, Validation, Investigation, Methodology

   Competing interests

   No competing interests declared

5. Patricia Claudel

   Université de Strasbourg, INRAE, Colmar, France

   Contribution

   Conceptualization, Resources, Investigation, Methodology

   Competing interests

   No competing interests declared

6. Philippe Hugueney

   Université de Strasbourg, INRAE, Colmar, France

   Contribution

   Conceptualization, Resources, Supervision, Funding acquisition, Validation, Investigation, Methodology, Writing – original draft, Project administration, Writing – review and editing

   Contributed equally with

   Julien Foucaud

   Competing interests

   No competing interests declared

7. Julien Foucaud

   UMR CBGP (INRAE-IRD-CIRAD, Montpellier SupAgro), Campus International de Baillarguet, Montferrier, France

   Contribution

   Conceptualization, Resources, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing – original draft, Project administration, Writing – review and editing

   Contributed equally with

   Philippe Hugueney

   For correspondence

   julien.foucaud@inrae.fr

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-2272-3149

• 1,040

  views

• 186

  downloads

• 7

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.

Citations by DOI

• 7

  citations for umbrella DOI https://doi.org/10.7554/eLife.84370