DNA metabarcoding reveals introduced species predominate in the diet of a threatened endemic omnivore, Telfair’s skink (Leiolopisma telfairii)

Abstract Introduced species can exert disproportionately negative effects on island ecosystems, but their potential role as food for native consumers is poorly studied. Telfair's skinks are endemic omnivores living on Round Island, Mauritius, a globally significant site of biodiversity conservation. We aimed to determine the dietary diversity and key trophic interactions of Telfair's skinks, whether introduced species are frequently consumed, and if diet composition changes seasonally between male and female skinks. We used DNA metabarcoding of skink fecal samples to identify animals (COI) and plants (ITS2) consumed by skinks. There were 389 dietary presence counts belonging to 77 dietary taxa found across the 73 Telfair's skink fecal samples. Introduced taxa were cumulatively consumed more frequently than other categories, accounting for 49.4% of all detections, compared to cryptogenic (20.6%), native (20.6%), and endemic taxa (9.5%). The most frequently consumed introduced species was the ant, Pheidole megacephala, present in 40% of samples. Blue latan palm, Latania loddigesii, was the most frequently consumed endemic species, present in 33% of samples but was only detected in the dry season, when fruits are produced. We found a strong seasonal difference in diet composition explained by the presence of certain plant species solely or primarily in one season and a marked increase in the consumption of animal prey in the dry season. Male and female skinks consumed several taxa at different frequencies. These results present a valuable perspective on the role of introduced species in the trophic network of their invaded ecosystem. Both native and introduced species provide nutritional resources for skinks, and this may have management implications in the context of species conservation and island restoration.


| INTRODUC TI ON
A novel species introduced into a new ecosystem can interact with taxa already present in a variety of ecological roles, for example, as a mutualist (Kaiser-Bunbury et al., 2011), competitor (Cole & Harris, 2011), predator (O'Dowd et al., 2003, prey (Li et al., 2011), or parasite (Arbetman et al., 2013). Introduced species are often associated with network and community restructuring (Memmott et al., 2000;Russo et al., 2014), as well as native biodiversity declines (Clavero & Garcia-Berthou, 2005;Luque et al., 2014) and even ecological collapse (O'Dowd et al., 2003). However, introduced species may have more nuanced effects on ecosystems, including interactions beneficial to native species (Schlaepfer et al., 2011). For example, non-native trees may provide nesting sites to threatened birds (Schlaepfer et al., 2011) and non-native plants may provide floral resources to a range of threatened native pollinators (Baldock et al., 2015).
Introduced species have been studied extensively as invasive predators and herbivores, but their role in the diet of native species has been given less attention. A few studies examine this subject explicitly.
For example, Ando et al. (2013) showed that the critically endangered red-headed wood pigeon, Columba janthina nitens, consumed introduced plants more frequently than native species on the Ogasawara Islands, Japan. Similarly, introduced species were consumed frequently by the Ogasawara buzzard, Buteo buteo oyoshim, with 90% of its diet consisting of introduced animals (Kato & Suzuki, 2005). These small oceanic islands harbor high levels of endemism. Introduced species are typically associated with disproportionately negative effects on island biodiversity (Sax & Gaines, 2009), but are shown to provide nutritional resources to these endemic species. This may be more common than currently acknowledged, with introduced species representing a significant dietary element for native consumers.
Round Island, situated 22.5 km North-East of Mauritius (Figure 1), is a globally significant site of biodiversity conservation and now represents the last remnant of native lowland palm habitat ( Figure 2) in the Mascarenes (Cheke & Hume, 2008). The palm habitat has been recovering since the eradication of goats, Capra aegagrus hircus, in 1979, and rabbits, Oryctolagus cuniculus, in 1986(Cheke & Hume, 2008Merton, 1987). At just 2.19 km 2 , it is home to several reptile species extirpated from mainland Mauritius by introduced species and habitat destruction. Telfair's skinks, Leiolopisma telfairii ( Figure 3), are vulnerable omnivorous reptiles, typically growing to approximately 30 cm in total length, and are endemic to Mauritius.
They became restricted to Round Island by the mid-1800s because of the introduction of non-native predators, such as rats . The species has now been re-introduced to the island Nature Reserves, Ile aux Aigrettes (0.26 km 2 , located 600 m from South-East Mauritius), and Gunner's Quoin (0.7 km 2 , 5 km to the North of Mauritius) . Round Island has been a designated nature reserve since 1957 and has never suffered from introduced terrestrial vertebrate predators, which have caused the extirpation and extinction of multiple Mauritian species F I G U R E 1 Location of the study. The left map shows the location of Round Island in the Indian Ocean. The right map shows the topography of Round Island (5 m contour lines) and the sampling locations for each skink. Symbol shape denotes the season samples were collected: wet = square, dry = circle, unknown = triangle. Symbol color denotes sex of the skinks samples were collected from: green = females, blue = males, black = unknown elsewhere (Cheke & Hume, 2008). Habitat restoration efforts on Round Island since the 1980s have led to the recovery of its reptile populations, which includes seven species, four of which became restricted to the island by the mid-19th century North et al., 1994).
Previous dietary analyses of Telfair's skinks include morphological identification of food items and molecular analyses (Brown et al., 2014;Moorhouse-Gann et al., 2021;Pernetta et al., 2005;Zuël, 2009). Morphological examination of feces shows that Telfair's skinks consume a variety of introduced and native species of fruit, seeds, arthropods, and vertebrates (Pernetta et al., 2005;Zuël, 2009). However, morphological methods of diet analysis can be unreliable and taxonomically imprecise, even when researchers are skilled. These methods also fail to adequately detect small or softbodied prey (Pompanon et al., 2012;Symondson, 2002). Molecular approaches, especially those using high-throughput sequencing (HTS), can provide much greater precision, frequently identifying taxa in fecal samples to species (Pompanon et al., 2012;Symondson, 2002;Taberlet et al., 2012). Previous HTS-based fecal analysis of Telfair's skinks targeting plant (Moorhouse-Gann et al., 2021) and animal (Brown et al., 2014) food resources on Ile aux Aigrettes and Round Island confirmed skinks eat a diverse range of taxa.
Identifying the diet of omnivores is challenging, but a few studies have facilitated the most comprehensive complex dietary assessments to date using DNA metabarcoding (Bonin et al., 2020;De Barba et al., 2014;Robeson et al., 2018;Silva et al., 2019). Trophic generalists may be central to ecological networks and can elicit topdown effects across their entire breadth and depth. Deciphering the structure and dynamics of these interactions is therefore valuable, especially within a conservation context. Telfair's skinks are large, locally abundant trophic generalists endemic to Mauritius Jones, 1993;Vinson & Vinson, 1969), and are therefore likely to exert strong top-down pressures on the ecological network of Round Island.
Here, we aimed to study the complete diet of Telfair's skinks on Round Island by using broad-coverage plant and animal DNA metabarcoding primers. In doing so, we aimed to show: (a) the dietary diversity and key trophic interactions of Telfair's skinks; (b) whether introduced species feature prominently in the diet; (c) whether diet composition changes between seasons; and (d) whether diet composition is different between male and female skinks, which may have implications for conservation management and reintroduction initiatives.

Introduced herbaceous plants, such as Achyranthes aspera and
Tridax procumbens, form swathes of invaded habitat in large open clearings between thickets of native trees. Before the 1980s, much of the island suffered deforestation caused by the introduced herbivores, resulting in the loss of all but two hardwood tree species represented by a single individual bois buis, Fernelia buxifolia, tree and a few individuals of acacia indigéne, Gagnebina pterocarpa (Strahm, 1993). Loss of habitat led to extensive soil erosion and created large expanses of barren rock slab over much of the island (Figure 2). Since 2002, there have been extensive efforts to restore the lost hardwood forests and to enhance the natural regeneration of the palm habitat (Jones, 2008). The invertebrate community is poorly studied, with few native species formally identified and described (Moldowan et al., 2016). Several introduced invertebrate species are now established over much of the island, such as the ants Pheidole megacephala and Brachymyrmex cordemoyi, and the webspinner Oligotoma saundersii. Endemic species of arthropod are common in their favored habitats, such as the Round Island stick insect, Apterograeffea marshallae, a herbivore of L. loddigesii (Moldowan et al., 2016), and the Serpent Island centipede, Scolopendra abnormis, a large invertebrate predator (Lewis et al., 2010). Many common invertebrates are yet to be described but are presumed to be native, such as a hyperabundant but undescribed cockroach species.
The vertebrate community consists of regionally important seabird colonies, a remnant endemic reptile assemblage and two introduced land bird species (Cheke & Hume, 2008;. Seven endemic reptile species survive on Round Island because of the absence of introduced predators. Five of these are listed as Threatened on the IUCN Red List: Bojer's skink, Gongylomorphus bojerii, Durrell's Night gecko, Nactus durrellorum, keel-scaled boa, Casarea dussumieri, Round Island day gecko, Phelsuma guentheri, and Telfair's skink, Leiolopisma telfairii, (IUCN, 2020). Additionally, two tortoise species, Aldabra giant tortoise, Aldabrachelys gigantea, and radiated tortoise, Astrochelys radiata, have been introduced to Round Island as "ecological replacements" for extinct Mauritian tortoises, Cylindraspis spp. (Griffiths et al., 2010). As the largest and one of the most abundant of the island's lizards, Telfair's skinks constitute the largest component of animal biomass of any omnivore on the island  and likely have a significant role within the island's food web dynamics.
Broad dry and wet seasons exist in Mauritius (Senapathi et al., 2010). The dry season typically begins in May and is characterized primarily by low rainfall, mean air temperature of ~20.5°C, and stronger winds, with the driest months being September and October.
The wet season typically begins in December and is characterized by much more frequent rainfall, mean air temperature of ~24.5°C, and minimal winds, with the wettest months being January and February (Senapathi et al., 2010).

| Skink sampling on Round Island
Fecal samples were collected in March, June, July, and December 2015 ( Figure 1). Skinks were caught opportunistically by noose or hand after which defecation was induced using a gentle abdominal massage. The fecal samples were placed in polythene bags and dried over silica gel. Telfair's skinks are present over the entire island, but, unfortunately, some areas of the island are too dangerous to capture these fast-moving reptiles. Skinks were released unharmed within ten minutes of capture at the locations where they were caught. Moorhouse-Gann et al., 2021). Due to funding constraints, we were only able to advance 82 samples to sequencing, which were randomly selected for the current study from both dry (40) and wet (42) seasons.

| Primer selection
Animal primers were tested in silico with a broad range of vertebrate and invertebrate taxa using PrimerMiner (Elbrecht & Leese, 2017) and in vitro with DNA extracted from animals sampled on Round Island. BerenF-LuthienR (Cuff et al., 2020) Table S1).
We used the following procedure to identify animal prey in the diet of Telfair's skinks. Polymerase chain reactions (PCRs) used 25 μl reaction volumes containing 5 μl DNA template, 12.5 μl of multiplex PCR mix (Qiagen, Manchester, UK), 2.5 μl of both forward and reverse primers (0.2 μM each), and 2.5 μl of nuclease-free water (Qiagen). PCR conditions are as follows: 95°C for 15 min, 35 cycles of 95°C for 30 s, 54°C for 90 s, and 72°C for 90 s, and 72°C for 10 min, as instructed by the manufacturer (Qiagen). Each sample incorporated a unique combination of molecular identification (MID) tags (Binladen et al., 2007) that allowed for each skink to be identified after pooling and sequencing as per Brown et al. (2014).
These 10-bp fragments were added to both the forward and reverse primers for each sample, and thus, dietary taxon sequences could be assigned to individuals. PCR products were then run through a 2% agarose gel stained with SYBR ® Safe (Thermo Fisher Scientific, Paisley, UK). Twelve negatives were included in each PCR run, 10 PCR negatives, and two extraction negatives. Additionally, two positive controls consisting of a standardized DNA concentration (4 ng/μl) of known invertebrate species likely absent from the study site (Supplementary Information S1) were used to control for tagjumping between samples in the filtering steps detailed below. PCR

| Bioinformatics
The Illumina Nano cartridge run generated 750,645 reads. Highthroughput sequencing data for the animal component of Telfair's skink diet followed the bioinformatic process of Drake et al. (2021): FastP (Chen et al., 2018) was used to check the quality of reads, discard poor quality reads (<Q30, <125 bp long or too many unqualified bases, denoted by "N"), trim reads to a minimum length of 300 bp and merge read pairs from Miseq files (R1 and R2). Read pairs were assigned to samples and demultiplexed using Mothur v1.39.5 (Schloss et al., 2009), after which MID-tag and primer ends were removed. Unoise3 (Edgar, 2010) was used to remove replicates, denoise the sequences, and group identical sequences into zeroradius operational taxonomic units (ZOTUs, which are clustered without % identity to avoid multiple species being nested within an OTU). Processed sequences were given taxonomic information from GenBank using BLASTn v2.7.1 (Camacho et al., 2009) with a 93% identity threshold. This threshold was chosen to capture the wide variety of invertebrates on Round Island to genus-or familylevel, most of which have not been barcoded or formally described.
When more than one taxon was assigned to a sequence, we manually checked the feasibility for the presence of each taxon on Round Island by searching published articles, unpublished reports, and personal observations of species accounts. If these manual checks were inconclusive, we assigned the sequence to a higher taxonomic level (genus, family, order, etc.). MEGAN Community Edition v6.18.9 (Huson et al., 2016) was used to analyze the BLAST output and assign taxonomic identities to each ZOTU. Using the lowest evalue (a value estimating the number of hits "expected" by chance when searching a database of a given size-in this instance anything <0.00001), the top hit was assigned to each sequence. Where top hits were taxonomic levels higher than species, these were manually checked and assigned to a feasible taxon or deleted from the analysis if erroneous. ZOTUs that were assigned to the same taxon were aggregated.
Data were cleaned for statistical analysis following the methods set out by Drake et al. (2021): The combined removal of the maximum read count in blanks and negative controls, and reads not meeting a predefined per sample threshold, removes both erroneous reads (laboratory contaminants and sequencing errors) that are likely to occur in low abundances mitigate tag-jumping and bleeding of over-represented taxa into other samples, while utilizing a per sample threshold and those arising through tag-jumping and bleeding of over-represented taxa into other samples removes erroneous reads (laboratory contaminants and sequencing errors) that are likely to occur in low abundances. The maximum read count of known contaminants and other obviously erroneous ZOTUs across the dataset was calculated as a percentage of their respective total sample read count, and any read counts less than this were removed. For this, a threshold of 0.3% was applied, removing low-frequency laboratory contaminants and sequencing errors. Following this, the highest read count within a blank or negative per ZOTU was calculated and any ZOTU reads below this value were removed. In addition, we established an extra per-ZOTU filtering step, which removed remaining erroneous taxa. The per-ZOTU threshold was set to 0.74%.
After these filters were applied, read counts were converted to presence-absence data for each sample. Nine samples were removed due to the absence of any dietary detections, leaving 73 samples to be taken forward for statistical analyses. Bioinformatic analysis for plant sequencing data followed Moorhouse-Gann et al. After animal ZOTUs were given taxonomic information, status of each taxon relative to Round Island was determined for each by manually searching for relevant data in published articles, unpublished reports, and personal species accounts, and then classified as "cryptogenic," "endemic," "introduced," or "native." Cryptogenic species were defined as species that had no clear status, either because of poor taxonomic resolution, or because they may be known natives of the Indian Ocean islands, but their history on Round Island is unknown. Plant status was taken from Moorhouse-Gann (2018; Moorhouse-Gann et al., 2021).

| Statistical analyses
Statistical analyses were conducted in R Statistical Software v4.1.0 (R Core Team, 2021) after data were converted to presence/absence within each sample. Basic characteristics of the diet were quantified by measuring frequency of occurrence. We aimed to reveal whether there were significant differences in the mean frequency of occurrence of dietary taxa from different taxonomic kingdoms (animals, plants) or status relative to Round Island (cryptogenic, endemic, introduced, native), hereafter "status." Data were not normally distributed (Shapiro-Wilk test for normality: W = 0.64, p = <.001), and we therefore used two nonparametric Kruskal-Wallis tests, one each for kingdom and status, to determine whether there were significant differences in average consumption between categories of each variable.
We also wanted to quantify dietary diversity and show whether our samples could be used to sufficiently represent the broad dietary patterns of Telfair's skinks. Sample size and effort-based standardization poorly represent the true diversity of communities because they fail to account for the species-abundance distribution of the community being sampled (Cao et al., 2007;Roswell et al., 2021).
We therefore used coverage-based rarefaction and extrapolation rather than asymptotic species-accumulation curves (Chao & Jost, 2012;Roswell et al., 2021) and robustly estimated species diversity using Hill diversity (Hill, 1973;Roswell et al., 2021). We define Hill diversity by the equation, where D is diversity, S is number of species, p i is the proportion of all individuals that belong to species i, r i is the rarity of species i, defined as 1/p i , and ι is the exponent determining the rarity scale on which the mean is taken (Bullen, 2003;Hill, 1973;Roswell et al., 2021). Hill diversity is the generalized mean species rarity, and the exponent ι determines the sensitivity of the equation to rare species. ι of 1 uses the arithmetic mean rarity, or species richness (Hill-richness), and is very sensitive to the rarest species; ι of 0 uses the geometric mean rarity, or the exponential of Shannon's entropy (Hill-Shannon), and responds to both high and low rarity species; and ι of −1 uses the harmonic mean rarity, or the inverse of Simpson's index (Hill-Simpson), and is most sensitive to the relative abundance of common species (Roswell et al., 2021). Coverage is a measure of how completely a community has been sampled and is an estimated proportion of the sampled individuals in the community that belong to species already detected (Chao & Jost, 2012). For example, a coverage of 0.85 denotes that 15% of the individuals in the community being sampled belong to species that have not been found. We computed these metrics in R package "iNEXT" (Hsieh et al., 2016).
Variation in diet composition was visualized with nonmetric multidimensional scaling (NMDS) in the "vegan" package (Oksanen et al., 2019) using the "metaMDS" function on a matrix of Jaccard distances, where we extracted three dimensions. Data were plotted using package "ggplot2" (Valero-Mora, 2010). To illuminate whether sex, season, or their interaction affects Telfair's skink diet, R package "mvabund" was used (Wang et al., 2012). Multivariate generalized linear models (MGLMs) were run using the "manyglm" function with a Monte Carlo resampling method and "binomial" error family. The "step" function facilitated model selection where we selected the lowest AIC value denoting which model was most supported given the data.

| RE SULTS
There were 389 dietary presence counts belonging to 77 dietary taxa found across the 73 Telfair's skinks samples. Of these, 37 of 38 plant taxa were resolved to species due to extensive barcoding of the Round Island flora. The invertebrates of Round Island have not been described as extensively, and of the 39 dietary taxa detected, 20 were resolved to species, nine to genus, nine to family, and one to order. The invasive ant P. megacephala and a cryptogenic braconid wasp, Heterospilus sp., were the most frequently detected taxa, present in almost 40% of all Telfair's skink samples (Table 1;   Table S2). (1) D = S ∑ i=1 p i r i 1 TA B L E 1 Taxonomic information, frequency of occurrence F O (%), and status relative to Round Island (cryptogenic, endemic, introduced, native) for all dietary taxa occurring in two or more Telfair's skink fecal samples
Moreover, our diversity estimates suggest Telfair's skinks consume many species infrequently instead of consuming taxa evenly.
This study achieved a greater taxonomic resolution compared to previous molecular analyses of Telfair's skink diet (Brown et al., 2014;Pernetta et al., 2005) There are an estimated 46,000 Telfair's skinks on Round Island , with an estimated 210 skinks per ha island-wide. This represents a major component of total animal biomass. Given the abundance and size of Telfair's skinks, these results show they are likely to represent a major top-down pressure on the ecological network through their dietary generalism.

| Prevalence of introduced taxa
Overall, introduced taxa formed the primary component of Telfair's skink diet as measured by frequency of occurrence. The majority of dietary detections and richness were of introduced taxa, accounting for almost half in both cases. Therefore, this study illuminates that introduced taxa have become a large part of the diet of a globally threatened endemic species. However, for some taxa it is unclear whether skinks rely on them for nutrition, and this is a broader issue in dietary metabarcoding studies because sequencing data cannot convey nutritional information (Alberdi et al., 2019;Lamb et al., 2019). For example, the introduced ant P. megacephala is present in 39.7% of samples, but may be a distasteful meal for Telfair's skinks.
On Round Island, P. megacephala is hyperabundant and found in every habitat type in this study. Predation may not provide a costeffective nutritional reward to an unspecialized ant-eating vertebrate given that the ant is very small compared to Telfair's skinks  (Schmidt, 2009). If these ants truly are deleterious to skinks, their high frequency of occurrence in the diet could be explained by accidental consumption. Accidental consumption may occur when skinks consume food items that have been colonized by ants, which typically occurs rapidly on Round Island. Another explanation is through secondary predation, which entails detection of food items in the digestive system of primary skink prey. Both accidental consumption and secondary predation may complicate interpretation of dietary analyses using HTS (Robeson et al., 2018;Silva et al., 2019;Tercel et al., 2021). Nevertheless, even accidental ingestion of some species could provide nutritional benefits to skinks.
With roughly half of all dietary detections originating from introduced species, non-native taxa appear to be a dominant part of Telfair's skink diet. It may be that the original components of the diet have been lost after Round Island suffered severe habitat destruction and have been subsequently replaced by non-native species.
Equally, the availability or nutritional value of non-native species may be relatively higher than existing native food.

| Seasonal and sex differences
The presence of plant species in the diet of the skinks solely or pri- Exactly how the tissue types of A. indicum may differentially benefit male and female skinks requires further study.
Understanding the nutritional requirements between sexes could be an important factor governing the success of skink translocations. Although we broadly see that male and female skinks consume the same species, we show that female skinks might rely more on certain species during the breeding season, which is a pivotal period in any reintroduction program.

| Limitations
The general limitations of dietary metabarcoding have been reviewed extensively by other authors (Alberdi et al., 2019;Lamb et al., 2019;Nielsen et al., 2018;Taberlet et al., 2018), but we also identified some study-specific limitations. This study converts sequence data to presence/absence and subsequently frequency of occurrence. We believe this is the most robust interpretation of sequencing data, because sequencing output only very weakly correlates with biomass in a sample (Deagle et al., 2019;Lamb et al., 2019). Nevertheless, frequency of occurrence therefore also omits how much biomass is consumed in each sample and, thus, a dietary taxon may appear frequently between samples but not contribute proportionately to the nutrition of the consumer.
As discussed above, the very high prevalence of introduced ants in Telfair's skink diet is difficult to explain ecologically with any certainty. These have not been observed to be directly eaten by the skinks, but are ubiquitous over Round Island, and colonize food resources rapidly. Moreover, a very frequently found tiny (<2 mm) cryptogenic braconid wasp, Heterospilus sp., seems unlikely to be actively preyed upon by adult Telfair's skinks. Accidental consumption or secondary predation might explain these detections, as has been seen in other dietary metabarcoding studies (Silva et al., 2019), and have been identified as a potential source of error that may disproportionately complicate the interpretation of dietary analyses of omnivores . With an aim to tease apart some of these issues, we conducted a co-occurrence analysis (Supplementary Information 3; Figure S1) but found no clear ecological patterns that explain these detections. Indeed, co-occurrence analyses may be used as an exploratory element in ecological studies but cannot provide strong evidence to support ecological hypotheses in this context (Blanchet et al., 2020), and may not facilitate interpretation .

| Concluding remarks
Our study represents one of only a few complete dietary analyses of omnivores using DNA metabarcoding (but see De Barba et al., 2014;Ducotterd et al., 2021;Robeson et al., 2018;Silva et al., 2019) and the first study examining the omnivorous diet of a threatened endemic reptile. We found that Telfair's skinks consume a few species regularly and many species rarely. We also found that Telfair's skinks shows that many introduced species of animal and plant contribute positively to providing nutritional subsidies to a globally threatened endemic omnivore. Positive effects of introduced species must therefore be weighed up against potential negative consequences of colonization for the ecosystem. This is pertinent for conservation managers to consider when restoring native habitats and controlling introduced species, especially when threatened animal species may be consuming introduced taxa in the absence of lost native food resources.

CO N FLI C T O F I NTE R E S T
The authors declare there are no conflicts of interest. org/0000-0003-0820-3278