Research Article

Biotic interactions drive ecosystem responses to exotic plant invaders

See allHide authors and affiliations

Science  29 May 2020:
Vol. 368, Issue 6494, pp. 967-972
DOI: 10.1126/science.aba2225

Exotic plants reduce carbon sequestration

Invasive exotic plants have become a major problem worldwide, with transformational effects on the composition and function of ecosystems. In a multifactorial experiment in New Zealand, Waller et al. show that exotic plants accelerate carbon loss from soils through their interactions with invertebrate herbivores and soil biota (see the Perspective by Urcelay and Austin). They built 160 mini-ecosystems in the field, manipulating interactions among plants, invertebrate herbivores, and soil biota. Key biological and abiotic responses were measured to quantify the relative contribution and interactions of the components of each community, revealing the potential of invasive plants to influence and suppress carbon sequestration through biotic interactions.

Science, this issue p. 967; see also p. 934


Ecosystem process rates typically increase after plant invasion, but the extent to which this is driven by (i) changes in productivity, (ii) exotic species’ traits, or (iii) novel (non-coevolved) biotic interactions has never been quantified. We created communities varying in exotic plant dominance, plant traits, soil biota, and invertebrate herbivores and measured indicators of carbon cycling. Interactions with soil biota and herbivores were the strongest drivers of exotic plant effects, particularly on measures of soil carbon turnover. Moreover, plant traits related to growth and nutrient acquisition explained differences in the ways that exotic plants interacted with novel biota compared with natives. We conclude that novel biological interactions with exotic species are a more important driver of ecosystem transformation than was previously recognized.

Introductions of exotic plants are transforming Earth’s ecosystems and altering the way in which they cycle carbon (C) (17). Successful exotic species can modify C-cycling through several potential interrelated mechanisms, most notably by increases to community productivity driving higher amounts of biomass returned to soil (2) and different traits of exotics compared with native species driving a higher proportion of high-quality biomass to soil (1, 5, 8). Much less attention has focused on how altered biotic interactions in an exotic species’ new range may affect C-cycling. It is well known that organisms above and belowground such as plants, soil biota, and invertebrate herbivores mediate ecosystem processes related to C-cycling and that exotic plants often differ in their interactions with new biota where there is no shared co-evolutionary history (i.e., novel) (57, 9). Thus, a third potential mechanism is that differences in the ways that exotics interact with novel organisms in recipient communities explain changes to C-cycling. However, the relative importance of, and interactions among, these three mechanisms are unknown; most studies have focused on systems invaded by species with high productivity, making it impossible to separate the impacts of community biomass increases from other mechanisms. Both predictive understanding and applied management of C stocks require these three nonmutually exclusive hypothesized mechanisms to be disentangled. Here, we resolve this gap in knowledge with a multifactor, multiresponse experiment aimed at partitioning the drivers of exotic impacts on ecosystem functions related to C-cycling.

We established 160 experimental ecosystems (mesocosms); manipulated interactions among plants, soil biota, and invertebrate herbivores in a fully factorial design; and measured how the traits and biomass of exotic species and biotic interactions affected ecosystem properties and processes (10) (Fig. 1A). Each mesocosm was grown in a 125-L pot (575-mm diameter; Fig. 1B) and comprised one of 20 individual, eight-species plant communities (table S1) varying orthogonally in the proportion of exotic and woody shrub and/or tree species (0 to 100% and 0 to 63%, respectively). These plants were taken from a pool of 20 exotic and 19 native and/or endemic New Zealand plant species (table S2). Soil biota were manipulated using a modified plant–soil feedback approach (11). Each plant species was grown in monoculture in 10-L pots containing field-collected soil for 9 to 10 months (for soil collection details, see table S3), allowing the conditioning of typical associated soil biota for each of the plant species (Fig. 1C). We created “home” soils by taking the conditioned soil from each of the eight representative species in a mesocosm and mixing it together to create a single inoculum. Each home soil mixture was also used as an “away” soil in a different mesocosm that did not contain any of the representative plants in that inoculum (table S4). These soils were intended to increase the relative biomass in inocula of specialized and preferred interaction partners of the resident (or nonresident) plant species, including but not limited to bacteria, fungi, oomycetes, and nematodes. Invertebrate insect herbivore populations were added into half of the mesocosms with home soils and half with away soils. Thirteen invertebrate herbivore species (hereafter simply referred to as “herbivores”) introduced into the mesocosms successfully established, along with seven self-colonizing species, totaling 20 species in all (table S5). All mesocosms were sealed with mesh cages (15% shade factor) designed to retain added herbivores and exclude others from entering (Fig. 1, A and D).

Fig. 1 Details of the experiment.

(A) Aerial view and (B) orthogonal design of the 20 communities. (C) Plants grown in the initial soil-conditioning stage. (D) Four of the 20 herbivores established in the mesocosms.

In each mesocosm, we measured distinct ecosystem properties and processes that are relevant to C-cycling dynamics (12): above- and belowground biomass; total in situ soil respiration (RT); basal respiration (RB, a measure of soil microbial activity); microbial biomass (RSI, measured as substrate-induced respiration); decomposition (rate of standardized substrate mass loss); soil organic matter; nitrogen (N) availability; herbivore biomass; arbuscular mycorrhizal fungi (AMF) biomass and other fungal biomass [using neutral lipid fatty acid (NLFA) and phospholipid fatty acid (PLFA) biomarkers, respectively]; and bacterial biomass (using PLFA). We measured plant traits representative of fundamental plant life history strategies (13): specific leaf area (SLA), specific root length (SRL), fine-root production, the rate of above- and belowground growth [net primary productivity (NPP)], and the rate of increase in plant height between two sampling periods. We also collected leaf N content (Nmass) data from an external database where available (table S2). SLA is an easily measured trait that correlates positively with relative growth rate (13) and nutrient use efficiency (14) and can directly influence soil C-cycling (15). Root traits such as SRL can directly affect interactions belowground with soil biota that drive changes to ecosystem processes (16). SLA and SRL were measured for each species separately and then scaled to the community level by calculating community-weighted mean trait values using plant biomass from each mesocosm (10). We incorporated more N-fixing exotic species into our communities compared with native species because N-fixers are more common among the New Zealand exotic flora compared with the native flora, and we would expect N-fixing and non-N-fixing exotics to differ in their effects on ecosystem properties (2). Likewise, we might expect differences between herbaceous versus woody species in their impact on functioning (2), so we included a woody gradient orthogonal to proportion exotic.

We used linear mixed-effects models to test our hypothesis that measured ecosystem properties were affected by our treatments, with the proportion of exotic plants in a community, soil and herbivore treatments, interactions among these factors, and plant effect traits (i.e., SLA, weighted by community biomass) as covariates. We used piecewise structural equation modeling (SEM) to explore how exotic species and their traits directly and indirectly (through modifying herbivore biomass, soil microbes, and soil nutrients) influenced C-cycling. To draw the causal links between exotic plant dominance (defined here as a higher richness of exotic compared with native plants), plant traits, herbivores, and soil microbiota to our ecosystem functions in the SEM, we modeled our dependent variables as a function of the experimental responses measured in the mesocosms. We measured SLA and SRL from plants outside of the mesocosms to capture effect traits (17) that were broadly representative of the life history strategies of our plant species but not the plastic responses to our treatments. Thus, even though we weighted trait values by biomass of each plant in the mesocosms, the community-weighted SLA and SRL values used in the hypothesized models and the SEM may over- or underestimate the response of these traits. Finally, to partition the effects of N-fixing exotic species versus non-N-fixing exotics from different plant functional groups, we included the realized proportion of woody and herbaceous N-fixing and non-N-fixing exotic plants into the SEM.

Altered biotic interactions are the strongest drivers of ecosystem functions

Biotic interactions (herbivory and home and/or away soil treatments) modified the impact of exotic species on measures of soil C-cycling (i.e., respiration and decomposition). The presence of herbivores and away soil doubled in situ soil respiration rates in exotic-dominated communities compared with corresponding native communities (RT~proportion exotic*soil origin*herbivore, χ2 = 10.41, p < 0.001; Fig. 2A). Reduced abundance of species-specific soil biota (away soil) in exotic-dominated communities increased soil microbial respiration by 1.5 times compared with the same communities grown in soils cultured to contain more specialist biota (RB~proportion exotic*soil origin, χ2 = 4.77, p < 0.03; Fig. 2B). Soil C:N ratios were also lower in away soils under exotic plants (C:N~proportion exotic*soil, χ2 = 3.87, p < 0.05). Higher proportions of N-fixing plant biomass also explained increases in basal respiration in exotic-dominated communities (Figs. 3 and 4). Herbivore feeding on exotic plants correlated positively with decomposition rates (decomposition~proportion exotic*herbivore, χ2 = 4.29, p < 0.04; Fig. 2C). These results underpin the need to explicitly include biological interactions in C-cycling models to improve predictions of ecosystem and global C budgets.

Fig. 2 Relationship between exotic plants and ecosystem properties.

(A) Total soil respiration. (B) Basal respiration. (C) Decomposition rate. (D) Community-weighted SLA. (E) Realized plant richness. (F) Herbivore biomass. (G) Bacterial biomass and (H) AMF biomass as a function of proportion of exotics planted and our imposed biotic treatments. Only significant predictors are shown; nonsignificant predictors were removed from the full model based on Akaike information criterion (AIC) selection. Regression lines represent the output of the best-fitting mixed model.

Fig. 3 Piecewise structural equation model showing the inferred direct and indirect effects of exotic plant richness, traits and biotic interactions on indicators of ecosystem function.

Plant traits (yellow boxes) are the realized proportional richness of exotic N-fixing and non-N-fixing woody and herbaceous plants, CWSLA, and community-weighted SRL (CWSRL). Productivity (green boxes) is measured as NPP, the rate of plant biomass produced over the course of the experiment; Biotic interactions (blue boxes) include AMF biomass, non-AMF biomass, herbivore biomass, and Gram-positive (G+) and Gram-negative (G) bacterial biomass. Ecosystem processes (gray boxes) are the total plant richness, available N, decomposition, soil organic matter, basal respiration, and total in situ respiration. A priori hypothesized models can be found in table S6. Black arrows indicate significant factors associated with positive effects on response variables and red arrows indicate negative effects. Only significant predictors are shown; nonsignificant predictors were removed from the full model on the basis of AIC selection. Arrow width corresponds to absolute values of relative effect sizes, which can be found, along with standardized coefficients and p values for each path, in table S7.

Fig. 4 Variance partitioned among treatments from hypothesized models.

Bars show the percentage of variation explained by each treatment and their interactions (indicated with a colon in the key) on our measured ecosystem properties.

Ecosystem effects by exotic plants are mediated by leaf traits, not increases to community biomass

Exotic species had 1.8 times higher mean SLA (SLA~plant provenance, χ2 = 44.64, p < 0.001; fig. S3), grew taller 2.5 times faster (height~plant provenance, χ2 = 49.42, p < 0.0001), had 2 times higher leaf N (Nmass~plant provenance, χ2 = 5.05, p < 0.02), and assimilated twice as much C into tissues compared with native plants (NPPshoots~plant provenance, χ2 = 5.24, p < 0.02; NPProots~plant provenance, χ2 = 3.61, p = 0.057), congruent with traits of exotic plants worldwide (8). Root traits (SRL, average diameter of fine roots and the proportion of fine roots), however, did not differ between native and exotic species (SRL~plant provenance, χ2 = 0.25, p = 0.62; average diameter, χ2 = 0.32, p = 0.57; fine roots, χ2 = 0.05, p = 0.82). Exotic plants also increased mean leaf trait values at the community level, because the community-weighted mean SLA (CWSLA) increased with the proportion of exotics in the community (CWSLA~proportion exotic, χ2 = 12.64, p < 0.001; Fig. 2D). High SLA is indicative of a suite of traits allowing for faster growth (13), potentially explaining the larger biomass and faster growth of exotic compared with native plants.

Exotic invaders can increase community biomass as a function of higher resource acquisition (18) and resource use efficiency (19), stronger competitive ability [(20); but not always, see (21) for exceptions], and ability to reach high abundance (22). In our experiment, exotic plants were indeed larger and more competitive than native plants, as evidenced by double the per capita plant biomass (per capita plant biomass~plant provenance, χ2 = 7.66, p < 0.006; fig. S1) and a drop in total plant richness as the proportion of planted exotic species increased (plant richness~proportion exotic, χ2 = 47.36, p < 0.001; Fig. 2E), likely through competitive exclusion. Although species richness declined along the exotic gradient, exotic plants compensated for this loss with increased per capita biomass, resulting in no net change in total community biomass production along the exotic gradient (mesocosm biomass~proportion exotic, χ2 = 0.01, p = 0.92; fig. S2). This result, combined with the lack of direct links from NPP to ecosystem functions in the SEM (Fig. 3), suggests that the effects of exotic plant biomass on ecosystem functions were likely caused by differences in traits, leading to changes in the mean quality of plant inputs. In other words, the total quantity of biomass was the same whether communities were 100% native or 100% exotic. What differed was the relative proportion of high-quality material, which was higher in exotic-dominated communities.

Exotic plants and plants with higher SLA also supported a greater herbivore biomass (herbivore biomass~plant provenance, χ2 = 26.07, p < 0.001; Fig. 3 and fig. 5), a result similar to that of a previous study conducted with naturally occurring herbivores (23). Faster-growing plants with high SLA typically have tissues with proportionately lower structural carbohydrates and, as such, are more readily mineralizable by herbivores and soil biota (13, 24). Indeed, exotic-dominated communities accumulated more herbivores (herbivore biomass~proportion exotic, χ2 = 16.49, p < 0.001; Fig. 2F), and this was partly driven by the higher mean SLA of these communities (Fig. 3).

Fig. 5 Mean proportion of plant biomass realized relative to proportion planted by provenance and functional group.

Bars represent the mean realized proportion of biomass for exotic and native plants of each plant functional group and provenance along the planted exotic gradient.

Herbivore accumulation on exotic plants explains heightened decomposition rates

Previous reports have shown increases in decomposition rates as high as 117% after exotic invasion, whereas others have shown decreases [for review, see (2)]. Here, decomposition rates increased only in exotic communities in which herbivores were also present (Fig. 2C). This relationship was likely driven by the higher nutritional quality of exotic compared with native plants, with consumed biomass returned to soil by herbivores more rapidly and in a more readily mineralizable form (24). Exotic plants were more palatable, as indicated by their higher SLA, leaf N, and growth rates (24). Further, N-fixing woody and non-N-fixing herbaceous exotics had higher rates of growth despite high herbivore feeding (Fig. 3), indicating that exotics may compensate for damaged tissues more readily. Although it is not unusual for plant–herbivore interactions to increase decomposition when herbivores feed on more palatable plants (24, 25), the accumulation of herbivores on exotic plants (particularly on non-N-fixing herbaceous exotics; Fig. 3), coupled with the high regrowth capacity of those plants, likely explains increased decomposition rates in mesocosms with a high proportion of exotic plants and herbivores added. Accumulation of successful biocontrol arthropods aimed at reducing exotic plant biomass may have similar effects in communities where they have been introduced. Our findings clarify previous idiosyncrasies from studies focused on exotic plant effects on decomposition (2), showing that exotic plants increase decomposition rates, but only when herbivores are present and able to accumulate.

Changes to C-cycling by exotic plants develops through multiple pathways

Basal and in situ respiration rates were elevated when exotic plants interacted with novel soil biota and insect herbivores. This result emerges from multiple potential pathways, the first by alteration of two bacterial taxonomic groups indicative of different C-cycling strategies. Exotic-dominated communities were associated with elevated bacterial biomass (bacterial biomass~proportion exotic, χ2 = 7.54, p < 0.006; Fig. 2G), particularly where N-fixing woody species were present (Fig. 3) and in communities with high mean SLA (bacterial biomass~CWSLA, χ2 = 4.62, p < 0.03). Herbivores depressed bacterial biomass overall (bacterial biomass~herbivore, χ2 = 13.06, p < 0.001; fig. S6), but also reduced the relative proportion of Gram-positive bacteria (fig. S7A), many of which are oligotrophic taxa using slow enzymatic pathways to break down recalcitrant C sources, compared with Gram-negative bacteria (often copiotrophic), which are associated with more rapid utilization of plant-based labile C sources (2628) (Fig. 3). Aboveground herbivory can also change exudation profiles and increase plant C allocation belowground into roots (29). Thus, by increasing the amount of labile C available to soil microbes through exudates (28), herbivores may have shifted the functional attributes of the bacterial community toward those associated with the more rapid breakdown of short-term pools.

Decomposition is faster where non-N-fixing exotic plants reduce AMF

A second pathway by which exotic plants can increase soil C loss is through their reduced interactions with AMF. AMF biomass in soil decreased with exotic dominance (AMF~proportion exotic: χ2 = 46.27, p < 0.001; Fig. 2H) and a greater proportion of non-N-fixing plants (Fig. 3). Communities with lower AMF biomass were associated with faster rates of decomposition (Fig. 3). AMF can stabilize organic matter in soil (30) by rapidly assimilating plant C into extensive intra- and extraradical hyphal structures (31), but much of this hyphal biomass may reside in soil as recalcitrant material and increase aggregate stabilization in soil (32). Non-AMF plants exude more labile C into soil (33, 34), which can increase mineralization of soil organic matter by cooccurring microbes (30, 33, 34). Indeed, suppression of AMF has accelerated soil respiration in previous studies (35). By contrast, ectomycorrhizal plants stimulate C mineralization in soils (33). However, we did not observe increases in C turnover in communities with ectomycorrhizal plants, likely because communities containing these species typically contained only a single ectomycorrhizal individual and made up a relatively small proportion of biomass (~6% where they occurred).

Although AMF are ubiquitous and associate with most land plants, there is wide variation in the degree to which plants invest in AMF (36), which may be species specific (36) and differ depending on provenance (37) or environmental conditions (38). A meta-analysis has shown that native plants have lower AMF colonization when grown near or after exotic plants, but provenance did not predict differences in AMF abundance among exotic plants (37). Here, exotic non-N-fixing plants caused the biggest declines in plant richness (Fig. 3), which may have reduced the availability of potentially better hosts for AMF. N-fixing plants did not reduce AMF biomass, likely reflecting their high reliance on AMF to meet the high phosphorus demand of N fixation (39). AMF were also strongly reduced by non-N-fixing woody plants (Fig. 3), possibly reflective of suppression by ectomycorrhizal plants and associated fungi (40). Finally, direct suppression of AMF hyphae has been observed in soils with a high abundance of Gram-negative bacteria (41), which were elevated in our exotic-dominated communities. Thus, increases in potentially antagonistic bacteria or other biota may explain the concomitant reductions in AMF biomass in exotic-dominated communities, or reduced biomass may reflect lower investment in AMF by the exotic plants grown here. AMF can buffer C emissions as long as decomposition rates remain low (42, 43), suggesting that decreases in AMF biomass under exotic plants represents another avenue by which C dynamics shift after invasion.

Increased microbial respiration rates in exotic-dominated communities grown in soils with depressed levels of species-specific soil biota (i.e., away soils) (Fig. 2B) were also likely driven by functional changes to bacterial communities. Plants grown in soils including species-specific soil biota (i.e., home soils) had a higher proportion of Gram-positive bacteria (Gram-positive bacteria~soil, χ2 = 3.94, p < 0.047; fig. S7B), a predominantly oligotrophic group containing many important, often specialized plant pathogens and recalcitrant litter specialists (27). Although we cannot determine whether soil biota were beneficial or harmful to plants, community-level reductions in plant biomass in home soil support the notion of more highly specialized pathogen loads or poor-quality mutualists (mesocosm plant biomass~soil, χ2 = 8.80, p < 0.003; Fig. 4 and fig. S4). Regardless of their effects on live plants, oligotrophic taxa often have slower growth and respiratory loss rates (27, 28) and may have slowed respiratory losses in home soils. In fact, increases in Gram-positive bacteria were associated with higher soil organic matter (Fig. 3), likely explaining the higher C:N in home soils. By contrast, the higher basal respiration rates in away soils likely reflect the relatively rapid breakdown of labile plant material by more prevalent copiotrophic taxa. This provides a temporal aspect to invasion effects on C dynamics. If home soils (containing more specialist soil biota) represent more established invasions and away soils newer invasions (containing no specialist soil biota), then we can expect faster turnover of recently assimilated C in new invasions, shifting to slower turnover and greater storage of longer-term C pools over time as specialist bacteria increase in abundance.

Our results could inform projects aimed at using nature-based solutions to fight climate change, such as the Trillion Trees Program ( Fast-growing exotic plants may be ideally suited for plantings when the goal is to sequester C aboveground, but exotic plants may have more variable effects belowground compared with native plants. Currently, N-fixing trees from the genus Acacia are among the most commonly planted plantation trees globally (44). Results from our experiment, which included A. dealbata, suggest that these types of trees may indeed increase soil C stocks, particularly in soils previously occupied by those species (i.e., home soils) through their positive effects on oligotrophic, specialist soil biota. However, these gains may be reduced over time through respiration associated with increases in copiotrophic taxa such as those that we have observed (Fig. 3), particularly in new or frequently harvested plantations. Our results also suggest that soil C stability may vary depending on whether exotic plants accumulate herbivores and/or fungal saprotrophs regardless of woodiness.

The pathways by which N-fixing versus non-N-fixing exotics affect ecosystem properties related to C-cycling appear to differ, and their relative effects differed between woody and herbaceous plants. N-fixing woody exotics disproportionately affected ecosystem functions relative to their proportional biomass in communities (Fig. 5), which has also been observed in other systems (45). Despite their low richness, the five exotic N-fixing woody species with high survival drove the biggest increases in fungal and bacterial biomass (Fig. 3), likely explaining much of the increased microbial respiration rates along the exotic gradient (Fig. 3). Non-N-fixing species contributed to significant decreases in AMF biomass, which may be driven by their exploitative life history strategies [i.e., fast growth (46) and high SLA; Figs. 2 and 3]. Herbaceous N-fixing exotics had no detectable effects in their communities, whereas non-N-fixing woody species had few (but strong) effects, perhaps because these species did not reach maturity, whereas many of the herbaceous plants grew quite large and usually flowered. Thus, exotic effects on ecosystem functions consistently result from differences in the way that they interact with other organisms in their environment compared with their native counterparts. Our experiment reveals multiple pathways through which these effects occur, influenced in different ways by various plant traits, including leaf traits, plant functional group (i.e., woody or herbaceous, N-fixing or not), and their typical microbial associations.

Taken together, these results suggest that shifts in fungal and bacterial function in exotic-dominated communities with herbivores and more generalist soil biota may underpin increased respiration rates through more rapid microbial mineralization of organic matter. Soil organic matter is primarily derived from plants and developed and turned over through interactions among soil microbiota (47), the abiotic environment (48), and living plant root exudates (49), which can differ substantially depending on plant composition (50). Indeed, our analyses indicate that it is not simply the large increases in exotic plant biomass that drive changes to C-cycling, but also differences in the types of inputs compared with natives. Exotic N-fixers drive increases in faster-cycling bacterial taxa, whereas high-quality exotic leaf tissues promote biomass of herbivores that reduce slower-cycling relative to faster-cycling taxa (Fig. 3 and fig. S7).

Understanding the factors regulating ecosystem responses in invaded ecosystems is complicated by the myriad of factors that influence C gains and losses. Our manipulative experiment showed strong interdependence among three leading hypotheses used to explain drivers of exotic plant impacts on ecosystem properties related to C-cycling. Exotic plants, which make up the greatest proportion of biomass in many invaded communities, have traits (e.g., high SLA and Nmass) that influence the way that they interact with herbivores and soil biota to drive ecosystem properties. Our results show that exotic and native plants differ in common bacterial and fungal pathways of C storage and release, and these differences are driven by trait differences such as SLA, woodiness, and the ability to fix N. Exotic plants, particularly woody species that fix N, increase bacterial and non-AMF biomass associated with storage and breakdown of C pools. By contrast, non-N-fixing exotic herbs substantially reduce AMF biomass, which accelerates soil C loss compared with communities dominated by native plants. In addition to their effects on fungi and bacteria, herbaceous exotic plants increase herbivore biomass, which we linked to increases in decomposition, an indicator of C-cycling. Our study provides an unprecedented community-scale understanding of how the transformation of ecosystems by exotic species depends on complex species interactions, emphasizing the need for whole-ecosystem approaches.

Supplementary Materials

Materials and Methods

Figs. S1 to S7

Tables S1 to S7

References (5267)

References and Notes

  1. See the supplementary materials and methods.
Acknowledgments: We thank D. Wardle and E. T. Kiers for valuable comments and J. Allen, N. Allen, A. Barrington, G. Boitt, J. Breitmeyer, J. Bufford, J. Burson, D. Conder, R. Cresswell, D. Dash, L. Dickie, B. Fairbrass, C. Ferguson, B. Ganley, T. Glare, A. Holyoake, D. Jack, C. Johns, J. Johns, E. Jones, B. Kwan, S. Larsen, H. Lea, I. Luxford, F. Martoni, L. Meachen, A. McKinnon, G. McSweeny, M. O’Callaghan, A. Puértolas, B. Richards, S. Richardson, H. Ridgeway, R. Scott, M.-R. Shadbolt, K. Slattery, R. Vardy, R. Wainer, D. Waller, M. Waller, A. Wakelin, and S. Wilson for technical support. Funding: This research was primarily supported by Tertiary Education Commission funding to the BioProtection Research Centre. Author contributions: Conceptualization: W.J.A., L.M.C., I.A.D., J.M.T., L.P.W.; Funding acquisition: B.I.P.B., L.M.C., I.A.D., J.M.T., S.A.W.; Investigation: W.J.A., F.M.F., J.E.H., N.K., K.H.O., G.S.S., L.P.W.; Data analysis: W.J.A., I.A.D., J.M.T., L.P.W.; Supervision: B.I.P.B., L.M.C., I.A.D., J.M.T.; Writing original draft: L.P.W.; Writing, review, and editing: all authors. Competing interests: The authors declare no competing interests. Data and materials availability: The data reported here have been deposited in Dryad (51).

Stay Connected to Science

Navigate This Article