Ecology

Effects of environmental fluctuations on co-culture composition and intermixingWe first tested the effects of fluctuations between anoxic (inducing a mutualistic interaction) and oxic (inducing a competitive interaction) conditions on co-culture composition (quantified as the ratio of consumer-to-producer at the expansion edge) and interspecific mixing (quantified as the number of interspecific boundaries divided by the colony circumference). We expected that, over a series of anoxic/oxic transitions, the ratio of consumer-to-producer at the expansion edge and the degree of intermixing would both decrease (Fig. 1d). To test this, we performed range expansions where we transitioned the environment between anoxic and oxic conditions. While we performed the experiments with defined anoxic and oxic incubation times, our main prediction (i.e., that repeated transitions between anoxic and oxic conditions can induce irreversible pattern transitions that alter co-culture composition and functioning) is independent of the time spent under either of those conditions as far as cells can adjust their metabolism to the new environment (Fig. 1d).As expected, the ratio of consumer-to-producer and the intermixing index both decreased over the series of anoxic/oxic transitions (Fig. 2a, b). The changes in these quantities appear to have two distinct dynamic phases; a first phase with a relatively steep decay and a second phase with a shallower decay. We therefore modeled their dynamics using a two-phase linear regression model [53,54,55]. During the first phase, the ratio of consumer-to-producer decreased significantly more rapidly at pH 7.5 (r2 = 0.90, p = 2 × 10−9, coeff = −0.0374, 95% CI = [−0.038, −0.0368]) than at 6.5 (r2 = 0.94, p = 1 × 10−7, coeff = −0.0103, 95% CI = [−0.0108, −0.0097]) (Fig. 2a). We observed consistent results for the intermixing index, where it also decreased significantly more rapidly at pH 7.5 (r2 = 0.90, p = 2 × 10−9, coeff = −0.0289, 95% CI = [−0.0295, −0.0284]) than at 6.5 (r2 = 0.93, p = 9 × 10−8, coeff = −0.01, 95% CI = [−0.0109, −0.0098]) (Fig. 2b). During the second phase, the change in the ratio of consumer-to-producer did not significantly differ between pH 7.5 (r2 = 0.90, p = 2 × 10−9, coeff = 0.0008, 95% CI = [0.0002, 0.0014]) and 6.5 (r2 = 0.94, p = 1 × 10−7, coeff = 0.0003, 95% CI = [−0.0002, 0.0008]) (Fig. 2a). However, we observed that the decrease in the intermixing index was significantly different between pH 7.5 (r2 = 0.94, p = 2 × 10−9, coeff = 0.0018, 95% CI = [0.0013, 0.0024]) and 6.5 (r2 = 0.94, p = 8 × 10−8, coeff = −0.0019, 95% CI = [−0.0025, −0.0013]). Overall, the final ratio of consumer-to-producer is lower at pH 7.5 (mean = 0.0163, SD = 0.01) than at 6.5 (mean = 0.052, SD = 0.02) (two-sample two-sided t-test; p = 0.03, n = 4) (Fig. 2). Consistently, the final intermixing index is also lower at pH 7.5 (mean = 0.0039, SD = 0.0032) than at 6.5 (mean = 0.0107, SD = 0.0049) (two-sample two-sided t-test; p = 0.05, n = 4) (Fig. 2b).Fig. 2: Dynamics of co-culture composition and intermixing during repeated anoxic/oxic transitions.a Co-culture composition measured as the ratio of consumer-to-producer. b Intermixing between the consumer and producer measured as the intermixing index, where N is the number of interspecific boundaries between the two strains. Experiments were performed at pH 6.5 (strong mutualistic interaction) (magenta data points) or pH 7.5 (weak mutualistic interaction) (cyan data points). Each data point is for an independent replicate (n = 4). The solid black lines are the two-phase linear regression models for pH 6.5, while the dashed black lines are the two-phase linear regression models for pH 7.5. Images of the final expansions after 350 h of incubation at c pH 6.5 and d pH 7.5. The scale bars are 1000 μm.Full size imageThe results described above yielded two important outcomes. First, the modeled two-phase linear regression of the ratio of consumer-to-producer and the intermixing index both depended on the strength of the mutualistic interaction, where the initial rate of decay was faster at pH 7.5 than at 6.5 (Fig. 2a, b). Thus, as the strength of the interdependency increases, the decay in the ratio and the intermixing index slows. Second, at pH 6.5 we never observed the complete loss of the consumer from the expansion edge (i.e., neither the ratio of consumer-to-producer nor the intermixing index reached zero) (Fig. 2a, b), which is counter to our initial expectation (Fig. 1d).We further performed controls under continuous oxic and continuous anoxic conditions (Supplementary Fig. S5). The ratio of consumer-to-producer and the intermixing indices both significantly differed between continuous oxic and continuous anoxic conditions regardless of the pH (two-sample two-sided t-tests; p  More

Fire conditions and particulate and bioaerosol emissionsFire radiative power values estimated from satellite imagery ranged from 6 to 259 MW over three days of burning [19]. Smoke sampled above combusting vegetation contained high concentrations of PM10 (mean ± s.e. 928.4 ± 140.6 µg m−3; Fig. 1). Microbial cells are a component of total bioaerosols, and their abundance can correlate with PM in ambient conditions [24] as well as in wildland fire smoke [6]. However, we observed that only the concentration of viable cells (and not total cells) correlated with PM2.5 and PM10 values (r2 = 0.80, and 0.81, respectively; p  More

Various estimates for great crested newt pond occupancy rates have been published with most relating to site or regional scale assessments. A naïve occupancy rate of 0.13 has been identified for a data set from the northwest of England29, while estimates based on conventional occupancy16 modelling of 0.31 for southeast England and between 0.32 and 0.33 for mid Wales were presented by Sewell et al.13. The only other national data which the authors are aware of are within the Freshwater Habitat National PondNet Study, which estimates a naïve pond occupancy of between 13 and 18%30, and the Amphibian and Reptile Conservation Trust National Amphibian and Reptile Recording Scheme, which suggests a 12% occupancy rate for the UK31. Using data from nearly 5000 ponds sampled across England, here we provide a more extensive national-level analysis while accounting for imperfect detection in the eDNA sampling protocol. Assuming a threshold of just one positive qPCR replicate in a sample, the naïve occupancy estimate of 0.30 is similar to the localised regional estimates made by Sewell et al.13 using direct observation methods. The posterior mean estimates of 0.198 for occupancy are comparable to most other estimates for great crested newt pond occupancy in the UK, but lower than the naïve estimate. The lower modelled estimates of occupancy than the naïve estimate suggest that false positives should not be ignored and need to be accounted for statistically using methodologies such as the eDNAShinyApp package used here23,24,27.The goodness of fit analysis was based on the MCMC output for each site and observed covariate levels in the data set. Some lack of fit was observed, with a predicted peak in amplification at 10 qPCR replicates but an observed peak at 12 qPCR replicates. There are several potential causes for this. For example, variation between laboratories could not be accounted for as these metadata were not made available. The assumption that error rates are the same across all laboratories may therefore not apply and could contribute to poorer model fit. Secondly, we did not consider error rates as functions of covariates, and this may also have contributed to a poorer fit.Stage 1 error was found to be smaller than Stage 2 error for both false positive and false negative error. However, Stage 2 error operates on individual qPCR replicates and not at the site level. If there was no error at Stage 2, we would observe either zero qPCR replicates amplifying or all qPCR replicates amplifying (i.e. 12 in the case of the data presented here). The majority of samples showed zero qPCR amplification (3429 samples), and this was strongly linked to absence of newts. For sites with amplification, we observed a greater number of samples amplifying between 1 and 11 qPCR replicates (1074 samples) than we did amplifying with all 12 qPCR replicates (422 samples). The qPCR replicates that do not amplify in samples containing target DNA are erroneous, even if other replicates within that sample do amplify and contribute to this high Stage 2 false negative error in the model output. Data simulated from the fitted model show that the frequency of samples that contain DNA at Stage 2 amplifying in less than five of the 12 qPCR replicates is very low (Fig. 2b). Given that all replicates need to be erroneous to alter the naïve assignment of a sample containing DNA to negative, Stage 2 false negatives at this sampling level are unlikely. However, this does not rule out Stage 1 false negative error which we estimate to be 5.2% (with wide credible intervals between 0.1% and 25.1%).Higher levels of Stage 2 replication remove lab-based false negative error. If eDNA is present within a sample and a high number of replicates are used, it is highly unlikely that all qPCR replicates will be erroneously negative, even when the false negative rate at the replicate level is high. Conversely, high levels of Stage 2 replication increase the likelihood of false positive error occurring32. Stage 2 false positive results are of greater consequence than the 2% the model output would suggest. Unlike false negative error where all Stage 2 replicates need to be erroneous to change the naïve assignment of occupancy of a sample, when a threshold of one amplifying replicate is applied, only a single replicate needs to be an erroneous to generate a false positive. With 12 qPCR replicates at Stage 2 and a 2% false positive error per replicate, a sample with no DNA present has a 24% chance of producing at least one amplification. Assuming this error is randomly distributed through samples with no DNA present and qPCR replicates, it is more likely that samples with small numbers of replicates amplifying would be erroneous than where large numbers of replicates amplify. This was confirmed in the goodness of fit analysis with the distribution of Stage 2 false positive replicates making up all samples amplifying with one or two positive qPCR replicates, while negligible false positive amplification was seen with four amplifying replicates or above (Fig. 2a). With only a single sample at Stage 1, false positive error is limited to the 1.5% per sample, as per the ({theta }_{10}) value in the occupancy model output.We would recommend that, where possible, results from individual sites are interpreted as a probability of site occupancy, based on modelled outputs such as those produced by the eDNAShinyApp R package23,27. The precision of these models is dependent on sample size. Where sample size is large, a reduced bias and narrower credible interval range is observed24. However, using occupancy modelling, Buxton et al.24 demonstrated that studies that contain only a small number of sites are unlikely to produce accurate and precise estimates. As a result, such assessments will need to continue to rely on a threshold value of amplifying qPCR replicates to define site occupancy. A naïve amplification threshold for assigning occupancy of one positive qPCR replicate is unwise and should be increased to reduce Stage 2 false positive error. Indeed, a threshold of three positive qPCR replicates would reduce false positive error, without increasing false negative error. Alternatively, redistributing the replication between Stage 1 and Stage 224, would also reduce the credible interval width and generate a more precise posterior mean estimate at Stage 1, in turn reducing the uncertainty around the occupancy estimate. A redistribution of replication leading to two samples collected from each site, both analysed using up to six qPCR replicates, as opposed to one sample analysed using twelve qPCR replicates, has been suggested24.Equal weighting of the ten covariates used in the traditional great crested newt HSI assessment25 may be ecologically unrealistic29. This is supported by the observations here, with only some of the HSI covariates identified as important for occupancy. The model applied by the eDNAShinyApp package23,27 successfully identified several covariates known to influence great crested newt occupancy, that are included within the HSI assessment25. These included occurrence of fish, water quality, shade, pond density, macrophyte cover, frequency of drying and geographic area; although our analysis was based on Easting and Northing, rather than the broad-scale suitability map used in deriving the original HSI25. However, several traditionally used HSI variables emerged as unimportant, i.e., waterfowl, terrestrial habitat quality, and area of pond; while ground frost, rainfall, surface wind and land cover type, are not included within the HSI assessment but were important.The importance and influence of the HSI suitability indices of fish, shade, pond density, water quality, macrophyte cover, and frequency of drying on pond occupancy were all as expected with wide literature support25,33,34,35,36,37,38,39,40,41,42,43,44,45,46,47,48. The negative pond occupancy response to climatic covariates of ground frost and high precipitation are supported in relation to annual survival47. ‘It is worth noting that the PIP value for wind speed was only just over the threshold for inclusion as important. Although ponds are shallow with limited stratification possible, wind speed has been shown to influence the distribution of eDNA in deeper waterbodies49,50. Estimating the presence of fish in a pond by direct observation for the traditional HSI may be problematical, and metabarcoding approaches to eDNA surveys which offer information on presence of other species would improve the accuracy of covariates, such as fish presence40. Indeed, assigning “Possible” fish presence within the HSI when scoring a pond accounted for the same percentage (33.1%) of both positive and negative eDNA samples. This suggests that when surveyors are not confident of fish presence, they are using this category in equal proportions for both occupied and unoccupied ponds. Landcover and bedrock were also important for pond level occupancy. This is expected given the importance of terrestrial habitat, and water retention to the species (Figs. S1, S3)35,51. However, with very unbalanced sample sizes between the categories (Figs. S2, S4), and influence of nearby land cover types uncaptured by the data, this variable is difficult to interpret, and we suggest further examination. Nevertheless, the positive associations with woodland and grassland reflect established knowledge of habitat preferences36. Equally, as freshwater predominantly relates to rivers and lakes rather than ponds in the landcover dataset used, negative relationships reflect the lower suitability of these habitats36.Several covariates, however, did not exhibit the expected response for pond occupancy. Terrestrial habitat was not found to be important despite the species being only semi-aquatic, and previous studies emphasising the importance of this variable36. This may be a result of the original Oldham et al.25 terrestrial habitat assessment being simplified into four subjective categories in the ARG UK26 protocol: this may not be nuanced enough to differentiate terrestrial habitat usage using statistical modelling. Waterfowl were not identified by the model as important predictors of great crested newt pond occupancy, where they have been elsewhere29,41, with one study suggesting a positive relationship between waterfowl species richness and great crested newt occupancy40. The lack of importance demonstrated in this data set may indicate that other covariates outweigh waterfowl in terms of occupancy importance, or they may only become important predictors of occupancy at very high waterfowl densities rarely observed in this data set. Similarly, pond area was not found to be an important predictor of pond occupancy. There was no difference in the mean area for occupied or unoccupied ponds; however, no occupied ponds were found above 10,000 m2. We would anticipate that both very small and very large ponds to be unsuitable for great crested newts25,52.Northing but not Easting was found to be an important predictor of pond occupancy. A distribution gradient with latitude is a common feature of biodiversity generally, and in the UK great crested newts are much more patchily distributed in Scotland than in England53,54. Pond occupancy estimates varied by year, with a greater occupancy in 2018 than the other years considered. This is likely linked to climatic conditions and may relate to the timings of ponds drying in relation to eDNA sample collection. This may therefore be an artefact of unoccupied ponds being more likely to dry early in the season and therefore being excluded from occupancy estimates for dry years, or local migration to less suitable habitat if core ponds start to dry, however long term analysis of individuals within a metapopulation shows little support for this47. As a result, in very dry years, we would expect an increase in pond occupancy to be observed in the data. Although average early spring rainfall for England in 2018 was higher than in either 2017 or 2019, rainfall during the main eDNA survey window of May and June was considerably less in 2018 than in the other two years (Fig. S5). Similar variation in year on year occupancy rate has been observed elsewhere30.As with all sampling methods, imperfect detection is a general feature of eDNA surveys. When high levels of qPCR replicates are used, false negative error is predominantly due to failure to collect DNA in a sample rather than failure to detect DNA within the lab. False positive error can occur at both stages and is exaggerated at Stage 2 by high levels of replication; Stage 2 false positive error is most likely in samples with a low proportion of replicates amplifying. We recommend using statistical models to estimate the occupancy of individual sites, taking into consideration sampling error. Failing that, a naïve occupancy threshold of two or three amplifying qPCR replicates, adjusting for total levels of replication, should be applied before assigning a site as occupied or not.With specific reference to great crested newts, we estimate approximately 20% of ponds through their natural range within England are occupied. We estimate that eDNA sampling failed to collect DNA from approximately 5% of sites where it was present. However, if eDNA is collected it is highly unlikely to be missed during the laboratory phase using the present protocol. We estimate that eDNA is erroneously collected in approximately 1.5% of water samples causing Stage 1 false positive results. However, false positives at the laboratory phase were found to be 2% per qPCR replicate; it is likely that this error would account for the majority of samples amplifying with one or two qPCR replicates, as a result these need to be treated with caution. To maximise accuracy, we recommend redistributing replication between the two stages, as is recommended elsewhere, and that thresholds to define a replicate as positive are further examined24,55. It is important to recognise that visual surveys also experience imperfect detection13, with observation errors likely to be similar to or greater than the error experienced using eDNA methods, particularly if the recommendations presented here are put in place to minimise laboratory stage false positive error. The benefits associated with eDNA over traditional methods allowing rapid collection of large scale distribution data are invaluable and should not be devalued in relation to traditional methods15. Although not identified within the models as important predictors, waterfowl, terrestrial habitat, and pond area may remain important habitat features for great crested newts. These covariates may be less important than the other HSI covariates, may not be measured in a sufficiently nuanced way to enable their importance to be identified, or may have influence on a local but not national scale29,40. However, equal weighing of the ten HSI variables is an oversimplification with the effect of some variables, for example pond area, overinflated within the HSI analysis, whereas others are undervalued, for example fish intensity. It is important to measure HSI covariates accurately and consistently to allow them to be utilised in statistical analysis such as this, and a review of the covariates and weighting is warranted now large occupancy data sets are becoming available. More

Acute cold shock causes drastic phenotypic alterationsThe duration of cold exposure for young adult hermaphrodite C. elegans at 2 °C is negatively correlated to post-shock survival rates15. Wild-type hermaphrodite worms exposed to a 4-h cold shock (CS) do not initially display high mortality rates (Fig. 1a); this allows observation of a range of phenotypic transitions as they recover from the limited-duration cold stress at their preferred temperature of 20℃. One of the most striking phenotypes exhibited in post-cold shock (post-CS) animals during the recovery period is a dramatic decrease in pigmentation in the normally highly pigmented intestine, so that the body becomes almost entirely clear (Fig. 1b, c)15. This is often accompanied by motor and reproductive disruptions such as mobility loss, withering of the gonad arms, decreased number of internal embryos, and the eventual death of about 30% of the population (Fig. 1a–d)15. It should be noted that these phenotypic responses do not appear to be due to any relative heat shock following the transition from 2 to 20 °C as the expression of GFP-tagged HSP-4 (heat shock protein) is not induced following cold shock (Fig. 1e). Neither is the reduced pigmentation following cold shock due to a period of starvation presumably experienced by the worms while they are at 2 °C. At this extreme cold temperature, the worms enter a “chill coma” in which pharyngeal pumping and virtually all other movement ceases15,16; however, a total absence of food for a similar time period does not induce a comparable clearing phenotype (Supplemental Fig. S1). Interestingly, some CS wild-type animals regain pigmentation after clearing; these worms do not die and display a general reversal of the other negative impacts of cold shock (Fig. 1b)15. We sought to better understand the factors regulating the post-CS recovery program in wild-type worms, focusing particularly on the functional role of pigmentation loss and the genetic components involved in producing it.Figure 1Cold-shocked worms show decrease in survival and characteristic phenotypic alterations. N2 young adult hermaphrodites were shifted from 20 to 2 °C for a 4 h cold shock (CS) and thereafter recovered at 20 °C for 96 h with assessment of (a) survival and (b) phenotypic alterations (n = 177). Death and immobility were assayed by nose tap; worms were considered to be immobile if the tap elicited slight movement in the head region but no other body movement, and dead worms were completely unresponsive (Chi-squared Test for Homogeneity: P  More

Study site and experimental designThe study site, experimental design, and annual sampling protocol have been described in previous publications15,22,47 but a summary will be provided here. The experiment was conducted in a remote study site approximately 20 km northeast of Kangerlussuaq, Greenland, at 67.11° N latitude and 50.34° W longitude, approximately 160 km inland from Baffin Bay. Annual growing season (May through July) mean temperature and total precipitation at the study site during the duration of this experiment (2002–2017) were 8.62 ± 0.20 °C and 43 ± 6.78 mm, respectively47. The surrounding area has functioned as an important caribou (Rangifer tarandus) migration corridor, calving ground, and Indigenous Peoples hunting site for at least approximately 4000 years48, and was designated as a UNESCO World Heritage Site, Aasivissuit—Nipisat, by the United Nations in 2018. Caribou are present in greatest numbers seasonally, with most of the animals that use the site migrating into it during late winter and early spring and migrating out of it in mid to late summer; some male caribou remain at the site through winter. Muskoxen (Ovibos moschatus) are present at the site year-round. Arctic hares (Lepus arcticus) and rock ptarmigan (Lagopus muta) occupy the site in low numbers. In contrast to other locations in the Arctic where they are important herbivores, this site does not harbor voles or lemmings.In June 2002 we erected six exclosures constructed of woven wire fencing material supported by steel t-posts; each exclosure was circular and measured 800 m2. Adjacent to each exclosure, and separated from it by approximately 20–50 m, we located a comparable control site. Exclosure sites and adjacent control sites covered a range of elevations from approximately 275–300 m above sea level. In early May 2003, prior to onset of the plant growing season, we installed passive, open-topped warming chambers constructed of UV neutral glazing material on three plots inside and three plots outside of one exclosure site and three plots inside and four plots outside of a second exclosure site. In early May 2004, we added three warming chambers inside and three warming chambers outside one of the sites equipped in 2003, and we installed an additional three warming chambers on plots inside and three warming chambers on plots outside of a third exclosure site, thus resulting in a total of 12 warmed plots distributed among three exclosure sites and 13 warmed plots distributed among three control (grazed) sites. An ambient (control) plot was located near, but not closer than 2 m to, each warmed plot, thus resulting in 25 warmed plots and 25 ambient plots distributed among three exclosures and adjacent grazed sites. No plot was located closer than 2 m to the edge of any exclosure. Warming chambers were constructed according to the International Tundra Experiment (ITEX) protocol49, were 1.5 m in basal diameter, and encompassed 1.77 m2. Warming chambers were installed in early May each year, anchored to plots using metal garden stakes, and removed annually at the time of vegetation sampling, which was intended to coincide with peak aboveground abundance at mid to late July in most years (except in 2006, when sampling was conducted in mid-June, and in 2003 and 2011 when sampling was conducted in mid-August)47. Warming chambers significantly elevated near surface temperature by approximately 1.5–3.0 °C, and resulted in a non-significant reduction of soil moisture22,50.Vegetation samplingVegetation sampling was conducted non-destructively using a square Plexiglas tabletop point frame on adjustable aluminum legs. The point frame measured 0.25 m2 and was centered within each plot for sampling. The corners of each plot were equipped with hollow aluminum tubes sunk into the soil surface at the cardinal directions, and the legs of the point frame were inserted into these tubes to ensure consistent orientation and location of the frame during sampling. Once the frame was positioned, a steel welding pin was lowered through each of 20 randomly located holes in the point frame tabletop, and each encounter by the tip of the pin with vegetation was recorded until the pin struck soil, litter, or rock. In 2003 and 2004, vegetation was recorded at the species level for deciduous shrubs (Betula nana and Salix glauca) and at the functional group level for graminoids (including grasses, rushes, and sedges of the genera Calamagrostis sp., Poa sp., Festuca sp., Hierochloë sp., Trisetum spicatum, Luzula sp., Carex sp., and Kobresia sp.), forbs, mosses, lichens, and fungi. Beginning in 2005, vegetation was recorded at the species level for forbs, in addition to deciduous shrubs, and at the genus level for lichens (Peltigera sp.), fungi [Calvatia sp.; most likely C. cretacea51], and mosses (Aulacomnium sp.). Graminoids were not resolved to the genus or species levels due to concerns about consistent identification. All taxa were identified in the field by the authors on the basis of visual inspection of live individuals in consultation with reference guides52,53,54,55. In adherence with the Guidelines for Professional Ethics established by the Botanical Society of America, sampling and identification were done non-destructively, and no voucher specimens were collected.Commonness estimationEcologically meaningful estimation of commonness is inherently relative; a taxon is only common or rare in relation to other taxa5. While there exist a considerable array of quantitative indices of commonness56, we opted for one that integrates abundance and occurrence by assigning equal weight to each. Using annual abundance sums obtained during point frame sampling, we calculated commonness for each taxon as the product of its proportional abundance across all plots within each treatment and its proportional occurrence across all plots within each treatment. Hence, the commonness (C) of an individual taxon, i, in a given year, t, can be expressed as the product of its proportional abundance (A) and proportional occurrence (O) in that year:$$C_{it} = A_{it} *O_{it}$$
(1)
in which proportional abundance of taxon i in year t is the sum of point frame pin intercepts, h, for that taxon in that year across all plots sampled that year divided by the total number of point frame pin intercepts, H, of live vegetation biomass recorded across all plots sampled that year:$$A_{it} = h_{it} /H_{t}$$
(2)
and in which proportional occurrence of taxon i in year t is the sum of the number of plots, p, on which point frame pin intercepts of taxon i were recorded in year t divided by the total number of plots, P, sampled in year t:$$O_{it} = p_{it} /P_{t}$$
(3)
This index was used to estimate taxon-specific commonness within each experimental treatment combination (i.e., exclosed ambient, exclosed warmed, grazed ambient, and grazed warmed treatments), as well as across the entire site (sitewide commonness) for derivation of baseline commonness. To derive baseline commonness for subsequent analysis of its contribution to taxon-specific trends in commonness over the course of the experiment, we used sitewide commonness of each taxon in the year 2006. As described above, greater taxonomic resolution beyond functional group was not widely applied in our sampling until the third year of the experiment, 2005. However, we decided against using 2005 as a baseline for commonness at the site because it also happened to be the final year of a two-year outbreak of caterpillar larvae of a noctuid moth, Eurois occulta, that reduced aboveground abundance of nearly all taxa on our plots22,57. Except for the fungus C. cretacea, all taxa, whether recorded by pin intercepts during point-frame sampling or not, were observed on at least one plot under each of the four experimental treatment combinations. The rarest forb in this study, Pyrola grandiflora, was observed on a single plot under each of the exclosed ambient, exclosed warmed, and grazed warmed treatments, and on two plots under the grazed warmed treatment, but was not recorded during point frame sampling of exclosed ambient or grazed ambient plots. Hence, any conclusions about the effects of warming on this species must be limited. Similarly, the lichen Peltigera sp., which was also very rare in this study, was recorded during point frame sampling on plots under each treatment combination, but was not detected by sampling on exclosed warmed plots after 2005 even though it was observed on one exclosed warmed plot after that. This might be considered corroboration of the negative effect on this genus of warming under herbivore exclusion reported in the Results, but caution may also be warranted. The fungus C. cretacea first appeared under the grazed ambient treatment in 2008 and then under the exclosed ambient treatment in 2012, but was not recorded under the grazed warmed or exclosed warmed treatments. This might in and of itself suggest a negative effect of warming on the establishment or occurrence of this species, or fungi in general, and might be consistent with limiting effects of reduced moisture availability under warming. However, we urge caution with this interpretation because fungi may not form fruiting bodies every growing season, and such fruiting bodies may emerge aboveground in different locations from one growing season to the next, thereby potentially confounding repeated detection by sampling methods such as ours.Analysis of experimental treatment effects on plant functional group abundanceWe used a Gaussian generalized linear model (GLM) with an identity link function to analyze variation in functional group abundance among experimental treatment combinations. This GLM included total annual abundance, for the period 2003–2017, of deciduous shrubs (comprising summed abundances of Betula nana and Salix glauca leaf and stem point frame pin intercepts), graminoids (comprising all grass, rush, and sedge tissue point frame pin intercepts), forbs, mosses, lichens, or fungi, in separate models with the two experimental treatments (warming and herbivore exclusion) and their interaction as factors, year as a factor, and day of year of sampling as a continuous covariate. Significance of individual treatment effects of warming and herbivore exclusion, as well as their interaction, was determined based on Wald Chi-square statistics and associated two-tailed P-values (with significance indicated at P ≤ 0.05).Analysis of experimental treatment effects on commonnessAnalyses of commonness data were performed at higher taxonomic resolution than were analyses of abundance data, and so were limited to analysis of data from the last 12 years of the experiment, 2006–2017. Using Eq. (1), commonness was estimated for 14 taxa, including two species of deciduous shrubs, Betula nana and Salix glauca; graminoids, comprising at least eight non-distinguished genera of grasses, rushes, and sedges listed above in the sub-section Vegetation sampling; eight species of forbs, including Equisetum arvense, Stellaria longipes, Cerastium alpinum, Bistorta vivipara, Draba nivalis, Campanula gieseckiana, Viola canina, and Pyrola grandiflora; one genus of moss, Aulacomnium sp.; one genus of fungus, Calvatia sp.; and one genus of lichen, Peltigera sp.We first investigated general characteristics of and treatment effects on commonness across the study site. We examined the skewness of commonness to determine whether the distribution of the 14 focal taxa was significantly right-skewed, indicating greater numbers of rare than of common taxa2. We obtained an estimate of skewness and its standard error across pooled data for the period 2003–2017, derived a 95% confidence interval, and compared it to zero. Next, we examined experimental treatment effects on sitewide commonness. To do this, we used a Gaussian GLM with identity link function to analyze pooled commonness of all taxa for the period 2006–2017, with commonness as the dependent variable and the two experimental treatments and their interaction as factors, year as a factor, taxon as a factor, and day of year of sampling as a covariate. We determined significance of individual treatment effects and their interaction by examining Wald Chi-square statistics, with significance indicated if the two-tailed P ≤ 0.05. We then tested for experimental treatment effects on individual taxa using the same analytical approach, but with taxon-specific commonness as the dependent variable, and treatment and year as factors, with day of year of sampling as a covariate.Analysis of trends in commonness and skewness of commonness over the last 12 years of the experimentWe next investigated whether common and rare taxa displayed different trends in commonness over the course of the last 12 years of the experiment. This was motivated by a presupposition that warming and/or herbivore exclusion might have differentially altered commonness of common vs. rare species. We first examined linear trends in sitewide commonness of all 14 taxa pooled across experimental treatments by testing for significance of linear regressions of taxon-specific commonness vs. year for the period 2006–2017. We then conducted the same analysis for each taxon individually under each experimental treatment combination to determine whether our experimental manipulations contributed to trends differentially in common vs. rare taxa. We then investigated whether the distribution of commonness across the 14 focal taxa displayed directional change over the course of the final 12 years of the experiment, and whether it might have done so differently in relation to experimental treatment combinations. To do this, we tested for significance of linear regressions of treatment-specific skewness of commonness vs. year for the period 2006–2017. Finally, we examined whether trends in commonness were related to baseline commonness for the 13 taxa resolved to the genus or species level, excluding graminoids because this group comprised multiple unresolved genera. This analysis was motivated by interest in determining whether taxa that were common at the beginning of the experiment tended to become more common and taxa that were rare at the beginning of the experiment tended to become rarer, thus indicating that degree of commonness itself might be an important driver of changes in commonness over the course of a multi-annual experiment such as ours. To do this, we fit a non-linear regression model using a von Bertalanffy equation to quantify the relationship between taxon-specific commonness trend (standardized coefficient from the regression of commonness vs. year, ranging between − 1 and 1) and baseline commonness by treatment. This equation took the form:$$Y = 1 – left( {1 – a} right)e^{ – bX}$$
(4)
In which Y = taxon- and treatment-specific commonness trend, estimated in this case using the standardized coefficient from a linear regression of commonness of taxon i under a given experimental treatment combination vs. year; a = the Y-intercept; b = the slope; and X = baseline commonness of taxon i under the same treatment combination in 2006. Significance of regressions for each treatment was determined by calculating an F-statistic using corrected model sums of squares, error sums of squares, model degrees of freedom, and error degrees of freedom. Non-linear regression models were considered significant if the F-associated P ≤ 0.05. More

Citizen science program “Hanamaru-maruhana national census”We asked citizens to take bee photographs and send them by e-mails in citizen science program “Hanamaru-Maruhana national census (Bumble bee national census in English)” (http://hanamaruproject.s1009.xrea.com/hanamaru_project/index_E.html)8. We gave citizens previous notice that their photographs were going to be used for scientific studies, and for other non-profit activities on our homepage and flyers. From 2013 to 2016, we collected roughly 5000 photographs taken by citizens. Citizens sent photographs of various bee species, but most of them were bumble bees and honey bees. They have interspecific similarity and intraspecific variation, making it difficult for non-experts to identify species. Since species identification was not a requirement for participants, most citizens sent bee photographs without species identification. These bees were identified by one of the authors, J. Yokoyama. These bees are relatively easy for experts to identify because only two honey bee species and 16 bumble bee species inhabit the Japanese archipelago excluding the Kurile Islands. The consistency of species identification by J. Yokoyama was 95% for 15 bumble bee species, and 97.7% for major six bumble bee species in our test using 100 bumble bee photographs8.Bee photographs used for deep learningFrom bee species observed in citizen science program “Hanamaru-maruhana national census (Bumble bee national census in English)”, we selected two honey bee species and 10 bumble bee species having interspecific similarity and intraspecific variation. Two honey bee species consisted of Apis cerana Fabricius, and A. mellifera Linnaeus. 10 bumble bee species consisted of Bombus consobrinus Dahlbom, B. diversus Smith, B. ussurensis Radoszkowski, B. pseudobaicalensis Vogt, B. honshuensis Tkalcu, B. ardens Smith, B. beaticola Tkalcu, B. hypocrita Perez, B. ignitus Smith, and B. terrestris Linnaeus. To increase training data of B. pseudobaicalensis, we added photographs of B. deuteronymus Schulz to photographs of B. pseudobaicalensis because they can rarely be distinguished using only photographic images (see http://hanamaruproject.s1009.xrea.com/hanamaru_project/identification_E.html for the details of their color patterns). We primarily used photographs taken by citizens from 2013 to 2015 in the citizen science program, but also used photographs taken by citizens in 2016 if the number of photographs for a certain class was small.We cropped a bee part as a rectangle image from a photograph to reduce background effects. We increased the number of photographs by data augmentation (Fig. S1 in Appendix S1 in Supplementary information). Please see Appendix S1 in Supplementary information for the details of “Data augmentation.” We assigned 70, 10, and 20% of the total data of the training dataset, validation dataset, and test dataset, respectively. Please see Appendix S1 in Supplementary information for the details of “Data split and training parameters”.Deep convolutional neural network (DCNN)In this study, we chose a deep convolutional neural network Xception, as it provides a good balance between the accuracy of the model on one hand and a smaller network size on the other. We adopted transfer learning21,22 and data augmentation23 to solve the issue of a shortage of photographs. The Xception network has a depth of 126 layers (including activation layers, normalization layers etc.) out of which 36 are convolution layers. In this study, we employed the pretrained Xception V1 model provided on the Keras homepage. Please see Appendix S1 in Supplementary information for the details of “Xception”, and “Transfer learning.” For the training, we chose a learning rate of 0.0001 and a momentum of 0.9.Species identification by biologistsWe asked 50 biologists to identify the species present in nine photographs selected randomly from the photograph dataset using a questionnaire form. Their professions were forth undergraduate student (16%), Master’s student (14%), Ph.D. student (12%), Postdoctoral fellow (26%), Assistant professor (6%), Associate professor (12%), Professors (6%), and others (8%). Their research organisms were honey bees (6%), bumble bees (14%), bees (6%), insects (12%), plants and insects (12%), plants (22%), and others such as fishes, reptiles, and mammals (28%). 14% of the biologists were studying bumble bees, but they did not need to identify all bumble bee species in their researches because only several species inhabit their study areas. We allowed the biologists to see field guide books, illustrated books, and websites. We did not limit the method or time to identify the species of photographs to simulate the species identification of actual citizen science programs as much as possible, except for asking experts. The experiment was approved by the Ethics Committee in Tohoku University, and carried out in accordance with its regulations. Informed consent was obtained from the biologists.Species identification in species class experiment by XceptionWe conducted species class experiment by categorizing photographs into different classes according to species. A total of 3779 original photographs were used in species class experiment (Table S1 in Appendix S1 in Supplementary information). These photographs were classified into 12 classes according to species. We inputted test dataset to Xception, and recorded their predicted classes.Species identification in color class experiment by XceptionWe conducted color class experiment by categorizing photographs into different classes according to intraspecific color differences. Photographs of B. ardens were classified into the following four classes: female B. ardens ardens, B. ardens sakagamii, B. ardens tsushimanus, and male B. ardens (Table S1 in Appendix S1 in Supplementary information). Photographs of B. honshuensis, B. beaticola, B. hypocrita, and B. ignitus were classified into female and male classes. In trial experiments, we had found that the Xception cannot learn images in minor classes if the number of original photographs in the classes was less than 40. No photographs in the class were predicted correctly, and no photographs in the other classes were predicted as the class. Therefore, in color class experiment, we did not use the photographs of minor classes (B. ardens subspecies: B. ardens sakagamii and B. ardens tsushimanus, male B. honshuensis, and male B. beaticola). Therefore, a total of 3681 original photographs were used in color class experiment (Table S1 in Appendix S1 in Supplementary information). They were classified into 15 classes according to intraspecific color differences in addition to species classes. We inputted test dataset to Xception, and recorded their predicted classes. To compare the total accuracy of color class experiment by Xception with those of other experiments, it was normalized using the number of test data including those of the minor classes, assuming that all test data of the minor classes were misidentified.The accuracy of species identificationWe calculated total accuracy, precision, recall, and F-score in each class. Total accuracy is the number of total correct predictions divided by the number of all test datasets. Note that the total accuracy of color class experiment by Xception was normalized using the number of test data including those of the minor classes. It reduces the total accuracy of color class experiment by Xception, and enables to compare with those by biologists and species class experiment by Xception directly. Precision is the number of correct predictions as a certain class divided by the number of all predictions as the class returned by biologists or Xception. Recall, which is equivalent to sensitivity, is the number of correct predictions as a certain class divided by the number of test datasets as the class. F-score is the harmonic average of the precision and recall, (2 × precision × recall)/(precision + recall).To show the effect of interspecific similarity on the accuracy of species identification, we used confusion matrix. The confusion matrix represents the relationship between true and predicted classes. Each row indicates the proportion of predicted classes in a true class. All correct predictions are located in the diagonal of the matrix, wrong predictions are located out of the diagonal. In species identification by biologists, “Others” class represents cases that they wrote no species name or a species name other than two honey bee species and 10 bumble bee species in the answer column. More

1.United Nations. Transforming our World: The 2030 Agenda for Sustainable Development (United Nations General Assembly, 2015).
Google Scholar 
2.Godfray, H. C. J. et al. Food security: The challenge of feeding 9 billion people. Science (80-) 327, 812–818 (2010).ADS 
CAS 

Google Scholar 
3.FAO, IFAD, UNICEF, WFP & WHO. The State of Food Security and Nutrition in the World 2021. Transforming Food Systems for Food Security, Improved Nutrition and Affordable Healthy Diets for All (FAO, 2021). https://doi.org/10.4060/cb4474en.Book 

Google Scholar 
4.FAO. The Second Report on the State of the World’s Animal Genetic Resources for Food and Agriculture (FAO, 2010). https://doi.org/10.4060/i4787e.Book 

Google Scholar 
5.Gepts, P. Plant genetic resources conservation and utilization: The accomplishments and future of a societal insurance policy. Crop Sci. 46, 2278–2292 (2006).
Google Scholar 
6.McCouch, S. et al. Feeding the future. Nature 499, 23–24 (2013).ADS 
CAS 
PubMed 

Google Scholar 
7.Castañeda-Álvarez, N. P. et al. Global conservation priorities for crop wild relatives. Nat. Plants 2, 16022 (2016).PubMed 

Google Scholar 
8.Esquinas-Alcázar, J. Protecting crop genetic diversity for food security: Political, ethical and technical challenges. Nat. Rev. Genet. 6, 946–953 (2005).PubMed 

Google Scholar 
9.Fernández-Llamazares, Á. et al. Scientists’ warning to humanity on threats to indigenous and local knowledge systems. J. Ethnobiol. 41, 144–169 (2021).
Google Scholar 
10.FAOSTAT. Food and Agriculture Data. (2019). http://www.fao.org/faostat/en/#data/QC. (Accessed: 15th July 2021)11.Lebot, V. Tropical Root and Tuber Crops: Cassava, Sweet Potato, Yams and Aroids. Tropical Root and Tuber Crops: Cassava, Sweet Potato, Yams and Aroids (CABI, 2009). https://doi.org/10.5822/978-1-61091-225-9_2.Book 

Google Scholar 
12.Gade, D. W. Names for Manihot esculenta: Geographical variations and lexical clarification. J. Lat. Am. Geogr. 1, 55–74 (2002).
Google Scholar 
13.McKey, D. & Delêtre, M. The emergence of cassava as a global crop. in Achievng Sustainable Cultivation of Cassava, Vol. 1 (ed. Hershey, C. H.) 3–32 (Burleigh Dodds Science Publishing, 2017). https://doi.org/10.19103/as.2016.0014.04.14.Howeler, R., Lutaladio, N. & Thomas, G. Save and Grow: Cassava. A Guide to Sustainable Production Intensification (Food and Agriculture Organization of the United Nations, 2013).
Google Scholar 
15.Allem, A. C. The origin of Manihot esculenta Crantz (Euphorbiaceae). Genet. Resour. Crop Evol. 41, 133–150 (1994).
Google Scholar 
16.Olsen, K. M. & Schaal, B. A. Evidence on the origin of cassava: Phylogeography of Manihot esculenta. Proc. Natl. Acad. Sci. USA 96, 5586–5591 (1999).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
17.Olsen, K. M. & Schaal, B. A. Microsatellite variation in cassava (Manihot esculenta, Euphorbiaceae) and its wild relatives: Further evidence for a southern Amazonian origin of domestication. Am. J. Bot. 88, 131–142 (2001).CAS 
PubMed 

Google Scholar 
18.Olsen, K. M. SNPs, SSRs and inferences on cassava’s origin. Plant Mol. Biol. 56, 517–526 (2004).CAS 
PubMed 

Google Scholar 
19.Léotard, G. et al. Phylogeography and the origin of cassava: New insights from the northern rim of the Amazonian basin. Mol. Phylogenet. Evol. 53, 329–334 (2009).PubMed 

Google Scholar 
20.Mühlen, G. S. et al. Genetic diversity and population structure show different patterns of diffusion for bitter and sweet manioc in Brazil. Genet. Resour. Crop Evol. 66, 1773–1790 (2019).
Google Scholar 
21.Ménard, L., McKey, D., Mühlen, G. S., Clair, B. & Rowe, N. P. The evolutionary fate of phenotypic plasticity and functional traits under domestication in manioc: changes in stem biomechanics and the appearance of stem brittleness. PLoS ONE 8, e74727 (2013).ADS 
PubMed 
PubMed Central 

Google Scholar 
22.Brown, C. H., Clement, C. R., Epps, P., Luedeling, E. & Wichmann, S. The Paleobiolinguistics of domesticated manioc (Manihot esculenta). Ethnobiol. Lett. 4, 61–70 (2013).
Google Scholar 
23.Isendahl, C. The domestication and early spread of manioc (Manihot esculenta Crantz): A brief synthesis. Lat. Am. Antiq. 22, 452–468 (2011).
Google Scholar 
24.McKey, D., Elias, M., Pujol, B. & Duputié, A. Ecological approaches to crop domestication. in Biodiversity in Agriculture: Domestication, Evolution, and Sustainability (eds. Gepts, P. et al.) 377–406 (Cambridge University Press, 2012). https://doi.org/10.1017/CBO9781139019514.023.25.McKey, D. & Beckerman, S. Chemical ecology, plant evolution and traditional manioc cultivation systems. In Tropical forests, people and food. Biocultural interactions and applications to development (eds Hladik, C. M. et al.) 83–112 (Parthenon Carnforth and UNESCO, 1993).
Google Scholar 
26.Elias, M. & McKey, D. The unmanaged reproductive ecology of domesticated plants in traditional agroecosystems: An example involving cassava and a call for data. Acta Oecol. 21, 223–230 (2000).ADS 

Google Scholar 
27.Duputié, A., Massol, F., David, P., Haxaire, C. & McKey, D. Traditional Amerindian cultivators combine directional and ideotypic selection for sustainable management of cassava genetic diversity. J. Evol. Biol. 22, 1317–1325 (2009).PubMed 

Google Scholar 
28.Peroni, N., Kageyama, P. Y. & Begossi, A. Molecular differentiation, diversity, and folk classification of ‘sweet’ and ‘bitter’ cassava (Manihot esculenta) in Caiçara and Caboclo management systems (Brazil). Genet. Resour. Crop Evol. 54, 1333–1349 (2007).
Google Scholar 
29.Elias, M. et al. Unmanaged sexual reproduction and the dynamics of genetic diversity of a vegetatively propagated crop plant, cassava (Manihot esculenta Crantz), in a traditional farming system. Mol. Ecol. 10, 1895–1907 (2001).CAS 
PubMed 

Google Scholar 
30.Martins, P. S. Dinâmica evolutiva em roças de caboclos amazônicos. in Scientific Papers of Paulo Sodero Martins 1941–1997: A tribute (eds. Veasey, E. A., Oliveira, G. C. X. & Pinheiro, J. B.) 217–228 (SBG, 2007).https://doi.org/10.1590/s0103-40142005000100013.31.Coomes, O. T. Of stakes, stems, and cuttings: The importance of local seed systems in traditional Amazonian societies. Prof. Geogr. 62, 323–334 (2010).
Google Scholar 
32.Dyer, G. A., González, C. & Lopera, D. C. Informal ‘seed’ systems and the management of gene flow in traditional agroecosystems: The case of cassava in Cauca, Colombia. PLoS ONE 6, e29067 (2011).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
33.Salick, J., Cellinese, N. & Knapp, S. Indigenous diversity of cassava: Generation, maintenance, use and loss among the Amuesha, peruvian upper amazon. Econ. Bot. 51, 6–19 (1997).
Google Scholar 
34.Sambatti, J. B. M., Martins, P. S. & Ando, A. Folk taxonomy and evolutionary dynamics of cassava: A case study in Ubatuba, Brazil. Econ. Bot. 55, 93–105 (2001).
Google Scholar 
35.Heckler, S. & Zent, S. Piaroa manioc varietals: Hyperdiversity or social currency?. Hum. Ecol. 36, 679–697 (2008).
Google Scholar 
36.Delêtre, M., McKey, D. & Hodkinson, T. R. Marriage exchanges, seed exchanges, and the dynamics of manioc diversity. Proc. Natl. Acad. Sci. USA 108, 18249–18254 (2011).ADS 
PubMed 
PubMed Central 

Google Scholar 
37.Sardos, J. et al. Evolution of cassava (Manihot esculenta Crantz) after recent introduction into a South Pacific Island system: The contribution of sex to the diversification of a clonally propagated crop. Genome 51, 912–921 (2008).CAS 
PubMed 

Google Scholar 
38.Ellen, R. & Soselisa, H. L. A comparative study of the socio-ecological concomitants of cassava (Manihot esculenta Crantz) diversity, local knowledge and management in Eastern Indonesia. Ethnobot. Res. Appl. 10, 15–35 (2012).
Google Scholar 
39.Burns, A. E., Gleadow, R., Cliff, J., Zacarias, A. & Cavagnaro, T. Cassava: The drought, war and famine crop in a changing world. Sustainability 2, 3572–3607 (2010).
Google Scholar 
40.Pujol, B., David, P. & McKey, D. Microevolution in agricultural environments: How a traditional Amerindian farming practice favours heterozygosity in cassava (Manihot esculenta Crantz, Euphorbiaceae). Ecol. Lett. 8, 138–147 (2005).
Google Scholar 
41.Mba, R. E. C. et al. Simple sequence repeat (SSR) markers survey of the cassava (Manihot esculenta Crantz) genome: Towards an SSR-based molecular genetic map of cassava. Theor. Appl. Genet. 102, 21–31 (2001).CAS 

Google Scholar 
42.de Oliveira, E. J. et al. Genome-wide selection in cassava. Euphytica 187, 263–276 (2012).CAS 

Google Scholar 
43.Ferguson, M. E., Shah, T., Kulakow, P. & Ceballos, H. A global overview of cassava genetic diversity. PLoS ONE 14, e0224763 (2019).CAS 
PubMed 
PubMed Central 

Google Scholar 
44.Wolfe, M. D. et al. Historical introgressions from a wild relative of modern cassava improved important traits and may be under balancing selection. Genetics 213, 1237–1253 (2019).PubMed 
PubMed Central 

Google Scholar 
45.Bredeson, J. V. et al. Sequencing wild and cultivated cassava and related species reveals extensive interspecific hybridization and genetic diversity. Nat. Biotechnol. 34, 562–570 (2016).CAS 
PubMed 

Google Scholar 
46.Kuon, J. E. et al. Haplotype-resolved genomes of geminivirus-resistant and geminivirus-susceptible African cassava cultivars. BMC Biol. 17, 1–15 (2019).CAS 

Google Scholar 
47.Prochnik, S. et al. The cassava genome: Current progress, future directions. Trop. Plant Biol. 5, 88–94 (2012).CAS 
PubMed 
PubMed Central 

Google Scholar 
48.Rabbi, I. Y. et al. Tracking crop varieties using genotyping-by-sequencing markers: A case study using cassava (Manihot esculenta Crantz). BMC Genet. 16, 115 (2015).PubMed 
PubMed Central 

Google Scholar 
49.Albuquerque, H. Y. G., do Carmo, C. D., Brito, A. C. & de Oliveira, E. J. Genetic diversity of Manihot esculenta Crantz germplasm based on single-nucleotide polymorphism markers. Ann. Appl. Biol. 173, 271–284 (2018).
Google Scholar 
50.Ogbonna, A. C. et al. Large-scale genome-wide association study, using historical data, identifies conserved genetic architecture of cyanogenic glucoside content in cassava (Manihot esculenta Crantz) root. Plant J. 105, 754–770 (2021).CAS 
PubMed 

Google Scholar 
51.Allendorf, F. W. Genetics and the conservation of natural populations: Allozymes to genomes. Mol. Ecol. 26, 420–430 (2017).CAS 
PubMed 

Google Scholar 
52.Morrell, P. L., Buckler, E. S. & Ross-Ibarra, J. Crop genomics: Advances and applications. Nat. Rev. Genet. 13, 85–96 (2012).CAS 

Google Scholar 
53.Ahrens, C. W. et al. The search for loci under selection: Trends, biases and progress. Mol. Ecol. 27, 1342–1356 (2018).PubMed 

Google Scholar 
54.Lotterhos, K. E. & Whitlock, M. C. The relative power of genome scans to detect local adaptation depends on sampling design and statistical method. Mol. Ecol. 24, 1031–1046 (2015).PubMed 

Google Scholar 
55.Lotterhos, K. E. & Whitlock, M. C. Evaluation of demographic history and neutral parameterization on the performance of FST outlier tests. Mol. Ecol. 23, 2178–2192 (2014).PubMed 
PubMed Central 

Google Scholar 
56.Hoban, S. et al. Finding the genomic basis of local adaptation: Pitfalls, practical solutions, and future directions. Am. Nat. 188, 379–397 (2016).PubMed 
PubMed Central 

Google Scholar 
57.Pankin, A., Altmüller, J., Becker, C. & von Korff, M. Targeted resequencing reveals genomic signatures of barley domestication. New Phytol. 218, 1247–1259 (2018).CAS 
PubMed 
PubMed Central 

Google Scholar 
58.Maccaferri, M. et al. Durum wheat genome highlights past domestication signatures and future improvement targets. Nat. Genet. 51, 885–895 (2019).CAS 
PubMed 

Google Scholar 
59.Allaby, R. G., Ware, R. L. & Kistler, L. A re-evaluation of the domestication bottleneck from archaeogenomic evidence. Evol. Appl. 12, 29–37 (2019).PubMed 

Google Scholar 
60.Brown, T. A. Is the domestication bottleneck a myth?. Nat. Plants 5, 337–338 (2019).PubMed 

Google Scholar 
61.Gaillard, M. D. P., Glauser, G., Robert, C. A. M. & Turlings, T. C. J. Fine-tuning the ‘plant domestication-reduced defense’ hypothesis: Specialist vs generalist herbivores. New Phytol. 217, 355–366 (2018).CAS 
PubMed 

Google Scholar 
62.Hillocks, R. J. & Wydra, K. Bacterial, fungal and nematode diseases. In Cassava: Biology, Production and Utilization (eds Hillocks, R. J. et al.) 261–280 (CABI, 2002).
Google Scholar 
63.Jarvis, A., Ramirez-Villegas, J., Campo, B. V. H. & Navarro-Racines, C. Is cassava the answer to African climate change adaptation?. Trop. Plant Biol. 5, 9–29 (2012).
Google Scholar 
64.Hanks, S. K. Genomic analysis of the eukaryotic protein kinase superfamily: A perspective. Genome Biol. 4, 111 (2003).PubMed 
PubMed Central 

Google Scholar 
65.Meng, X. & Zhang, S. MAPK cascades in plant disease resistance signaling. Annu. Rev. Phytopathol. 51, 245–266 (2013).CAS 
PubMed 

Google Scholar 
66.Champion, A., Kreis, M., Mockaitis, K., Picaud, A. & Henry, Y. Arabidopsis kinome: After the casting. Funct. Integr. Genomics 4, 163–187 (2004).CAS 
PubMed 

Google Scholar 
67.Lenser, T. & Theißen, G. Molecular mechanisms involved in convergent crop domestication. Trends Plant Sci. 18, 704–714 (2013).CAS 
PubMed 

Google Scholar 
68.Gepts, P. The contribution of genetic and genomic approaches to plant domestication studies. Curr. Opin. Plant Biol. 18, 51–59 (2014).PubMed 

Google Scholar 
69.Ceballos, H. et al. Discovery of an amylose-free starch mutant in cassava (Manihot esculenta Crantz). J. Agric. Food Chem. 55, 7469–7476 (2007).CAS 
PubMed 

Google Scholar 
70.Jennings, D. L. & Iglesias, C. Breeding for crop improvement. in Cassava: Biology, Production and Utilization (eds. Hillocks, R. J., Thresh, J. M. & Bellotti, A.) 149–166 (CABI, 2002). https://doi.org/10.18520/cs/v114/i02/256-257.71.Meyer, R. S. & Purugganan, M. D. Evolution of crop species: Genetics of domestication and diversification. Nat. Rev. Genet. 14, 840–852 (2013).CAS 
PubMed 

Google Scholar 
72.Meyer, R. S., DuVal, A. E. & Jensen, H. R. Patterns and processes in crop domestication: An historical review and quantitative analysis of 203 global food crops. New Phytol. 196, 29–48 (2012).PubMed 

Google Scholar 
73.Elias, M., Lenoir, H. & McKey, D. Propagule quantity and quality in traditional Makushi farming of cassava (Manihot esculenta): A case study for understanding domestication and evolution of vegetatively propagated crops. Genet. Resour. Crop Evol. 54, 99–115 (2007).
Google Scholar 
74.Zohary, D. Unconscious selection and the evolution of domesticated plants. Econ. Bot. 58, 5–10 (2004).
Google Scholar 
75.Lamberti, G., Gügel, I. L., Meurer, J., Soll, J. & Schwenkert, S. The cytosolic kinases STY8, STY17, and STY46 are involved in chloroplast differentiation in Arabidopsis. Plant Physiol. 157, 70–85 (2011).CAS 
PubMed 
PubMed Central 

Google Scholar 
76.Pujol, B. et al. Evolution under domestication: Contrasting functional morphology of seedlings in domesticated cassava and its closest wild relatives. New Phytol. 166, 305–318 (2005).PubMed 

Google Scholar 
77.Halkier, B. A. & Gershenzon, J. Biology and biochemistry of glucosinolates. Annu. Rev. Plant Biol. 57, 303–333 (2006).CAS 
PubMed 

Google Scholar 
78.Doebley, J. F., Gaut, B. S. & Smith, B. D. The molecular genetics of crop domestication. Cell 127, 1309–1321 (2006).CAS 
PubMed 

Google Scholar 
79.Purugganan, M. D. & Fuller, D. Q. The nature of selection during plant domestication. Nature 457, 843–848 (2009).ADS 
CAS 
PubMed 

Google Scholar 
80.An, F. et al. Domestication syndrome is investigated by proteomic analysis between cultivated cassava (Manihot esculenta Crantz) and its wild relatives. PLoS ONE 11, e0152154 (2016).PubMed 
PubMed Central 

Google Scholar 
81.Alves, A. A. C. Cassava botany and physiology. in Cassava: Biology, Production and Utilization (eds. Hillocks, R. J., Thresh, J. M. & Bellotti, A.) 67–89 (CABI, 2002). https://doi.org/10.1079/9780851995243.0067.82.Alves, A. A. C. & Setter, T. L. Response of cassava leaf area expansion to water deficit: Cell proliferation, cell expansion and delayed development. Ann. Bot. 94, 605–613 (2004).PubMed 
PubMed Central 

Google Scholar 
83.Nielsen, R. & Signorovitch, J. Correcting for ascertainment biases when analyzing SNP data: Applications to the estimation of linkage disequilibrium. Theor. Popul. Biol. 63, 245–255 (2003).PubMed 
MATH 

Google Scholar 
84.Arnold, B., Corbett-Detig, R. B., Hartl, D. & Bomblies, K. RADseq underestimates diversity and introduces genealogical biases due to nonrandom haplotype sampling. Mol. Ecol. 22, 3179–3190 (2013).CAS 
PubMed 

Google Scholar 
85.Alves-Pereira, A. et al. A population genomics appraisal suggests independent dispersals for bitter and sweet manioc in Brazilian Amazonia. Evol. Appl. 13, 342–361 (2020).PubMed 

Google Scholar 
86.Bradbury, E. J. et al. Geographic differences in patterns of genetic differentiation among bitter and sweet manioc (Manihot esculenta subsp. esculenta; Euphorbiaceae). Am. J. Bot. 100, 857–866 (2013).PubMed 

Google Scholar 
87.Kates, H. R. et al. Targeted sequencing suggests wild-crop gene flow is central to different genetic consequences of two independent pumpkin domestications. Front. Ecol. Evol. 9, 618380 (2021).
Google Scholar 
88.Talavera, A., Soorni, A., Bombarely, A., Matas, A. J. & Hormaza, J. I. Genome-wide SNP discovery and genomic characterization in avocado (Persea americana Mill.). Sci. Rep. 9, 20137 (2019).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
89.Barrett, R. D. H. & Hoekstra, H. E. Molecular spandrels: Tests of adaptation at the genetic level. Nat. Rev. Genet. 12, 767–780 (2011).CAS 
PubMed 

Google Scholar 
90.Ross-Ibarra, J., Morrell, P. L. & Gaut, B. S. Plant domestication, a unique opportunity to identify the genetic basis of adaptation. Proc. Natl. Acad. Sci. USA 104, 8641–8648 (2007).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
91.Ogbonna, A. C., Braatz de Andrade, L. R., Mueller, L. A., de Oliveira, E. J. & Bauchet, G. J. Comprehensive genotyping of a Brazilian cassava (Manihot esculenta Crantz) germplasm bank: insights into diversification and domestication. Theor. Appl. Genet. https://doi.org/10.1007/s00122-021-03775-5 (2021).Article 
PubMed 
PubMed Central 

Google Scholar 
92.McKey, D., Cavagnaro, T. R., Cliff, J. & Gleadow, R. Chemical ecology in coupled human and natural systems: People, manioc, multitrophic interactions and global change. Chemoecology 20, 109–133 (2010).CAS 

Google Scholar 
93.Clement, C. R., de Cristo-Araújo, M., Coppens d’Eeckenbrugge, G., Alves Pereira, A. & Picanço-Rodrigues, D. Origin and domestication of native Amazonian crops. Diversity 2, 72–106 (2010).
Google Scholar 
94.Peña-Venegas, C. P., Stomph, T. J., Verschoor, G., Lopez-Lavalle, L. A. B. & Struik, P. C. Differences in manioc diversity among five ethnic groups of the Colombian Amazon. Diversity 6, 792–826 (2014).
Google Scholar 
95.Moreira, P. A. et al. Diversity of treegourd (Crescentia cujete) suggests introduction and prehistoric dispersal routes into Amazonia. Front. Ecol. Evol. 5, 150 (2017).
Google Scholar 
96.Clement, C. R. et al. Origin and dispersal of domesticated peach palm. Front. Ecol. Evol. 5, 148 (2017).
Google Scholar 
97.Mutegi, E. et al. Genetic structure and relationships within and between cultivated and wild sorghum (Sorghum bicolor (L.) Moench) in Kenya as revealed by microsatellite markers. Theor. Appl. Genet. 122, 989–1004 (2011).CAS 
PubMed 

Google Scholar 
98.Roullier, C., Rossel, G., Tay, D., McKey, D. & Lebot, V. Combining chloroplast and nuclear microsatellites to investigate origin and dispersal of New World sweet potato landraces. Mol. Ecol. 20, 3963–3977 (2011).CAS 
PubMed 

Google Scholar 
99.Alves-Pereira, A. et al. Patterns of nuclear and chloroplast genetic diversity and structure of manioc along major Brazilian Amazonian rivers. Ann. Bot. 121, 625–639 (2018).PubMed 
PubMed Central 

Google Scholar 
100.Siqueira, M. V. B. M. et al. Genetic characterization of cassava (Manihot esculenta) landraces in Brazil assessed with simple sequence repeats. Genet. Mol. Biol. 32, 104–110 (2009).CAS 
PubMed 
PubMed Central 

Google Scholar 
101.Allem, A. C. The origins and taxonomy of cassava. in Cassava: Biology, Production and Utilization (eds. Hillocks, R. J., Thresh, J. M. & Bellotti, A.) 1–16 (CABI, 2002). https://doi.org/10.1079/9780851995243.0001.102.Barbieri, R. L., Gomes, J. C. C., Alercia, A. & Padulosi, S. Agricultural biodiversity in southern Brazil: Integrating efforts for conservation and use of neglected and underutilized species. Sustainability 6, 741–757 (2014).
Google Scholar 
103.Khoury, C. K. et al. Increasing homogeneity in global food supplies and the implications for food security. Proc. Natl. Acad. Sci. USA 111, 4001–4006 (2014).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
104.Peroni, N. & Hanazaki, N. Current and lost diversity of cultivated varieties, especially cassava, under swidden cultivation systems in the Brazilian Atlantic Forest. Agric. Ecosyst. Environ. 92, 171–183 (2002).
Google Scholar 
105.Peroni, N. & Martins, P. S. Influência da dinâmica agrícola itinerante na geração de diversidade de etnovariedades cultivadas vegetativamente. Interciencia 25, 22–29 (2000).
Google Scholar 
106.Doyle, J. J. & Doyle, J. L. A rapid DNA isolation procedure for small quantities of fresh leaf tissue. Phytochem. Bull 19, 11–15 (1987).
Google Scholar 
107.Poland, J. A., Brown, P. J., Sorrells, M. E. & Jannink, J. L. Development of high-density genetic maps for barley and wheat using a novel two-enzyme genotyping-by-sequencing approach. PLoS ONE 7, e32253 (2012).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
108.Andrews, A. FastQC: A Quality Control Tool for High Throughput Sequence Data. http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (2010).109.Catchen, J., Hohenlohe, P. A., Bassham, S., Amores, A. & Cresko, W. A. Stacks: An analysis tool set for population genomics. Mol. Ecol. 22, 3124–3140 (2013).PubMed 
PubMed Central 

Google Scholar 
110.Langmead, B. & Salzberg, S. L. Fast gapped-read alignment with Bowtie 2. Nat. Methods 9, 357–359 (2012).CAS 
PubMed 
PubMed Central 

Google Scholar 
111.Li, H. A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics 27, 2987–2993 (2011).CAS 
PubMed 
PubMed Central 

Google Scholar 
112.Li, H. et al. The Sequence Alignment/Map format and SAMtools. Bioinformatics 25, 2078–2079 (2009).PubMed 
PubMed Central 

Google Scholar 
113.Danecek, P. et al. The variant call format and VCFtools. Bioinformatics 27, 2156–2158 (2011).CAS 
PubMed 
PubMed Central 

Google Scholar 
114.Luu, K., Bazin, E. & Blum, M. G. B. pcadapt: An R package to perform genome scans for selection based on principal component analysis. Mol. Ecol. Resour. 17, 67–77 (2017).CAS 
PubMed 

Google Scholar 
115.R Core Team. A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2018). https://www.r-project.org/. (Accessed: 15th January 2018).116.Fariello, M. I., Boitard, S., Naya, H., SanCristobal, M. & Servin, B. Detecting signatures of selection through haplotype differentiation among hierarchically structured populations. Genetics 193, 929–941 (2013).PubMed 
PubMed Central 

Google Scholar 
117.Weir, B. S. & Cockerham, C. C. Estimating F-statistics for the analysis of population structure. Evolution (N. Y. ) 38, 1358–1370 (1984).CAS 

Google Scholar 
118.Keenan, K., McGinnity, P., Cross, T. F., Crozier, W. W. & Prodöhl, P. A. DiveRsity: An R package for the estimation and exploration of population genetics parameters and their associated errors. Methods Ecol. Evol. 4, 782–788 (2013).
Google Scholar 
119.Bonhomme, M. et al. Detecting selection in population trees: The Lewontin and Krakauer test extended. Genetics 186, 241–262 (2010).PubMed 
PubMed Central 

Google Scholar 
120.Reynolds, J., Weir, B. S. & Cockerham, C. C. Estimation of the coancestry coefficient: Basis for a short-term genetic distance. Genetics 105, 767–779 (1983).CAS 
PubMed 
PubMed Central 

Google Scholar 
121.Sabeti, P. C. et al. Genome-wide detection and characterization of positive selection in human populations. Nature 449, 913–918 (2007).ADS 
CAS 
PubMed 
PubMed Central 

Google Scholar 
122.Scheet, P. & Stephens, M. A fast and flexible statistical model for large-scale population genotype data: Applications to inferring missing genotypes and haplotypic phase. Am. J. Hum. Genet. 78, 629–644 (2006).CAS 
PubMed 
PubMed Central 

Google Scholar 
123.Gautier, M., Klassmann, A. & Vitalis, R. rehh 2.0: A reimplementation of the R package rehh to detect positive selection from haplotype structure. Mol. Ecol. Resour. 17, 78–90 (2017).CAS 
PubMed 

Google Scholar 
124.Cingolani, P. et al. A program for annotating and predicting the effects of single nucleotide polyorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin). 6, 1–13 (2012).
Google Scholar 
125.Ten Blake, J. A. quick tips for using the Gene Ontology. PLoS Comput. Biol. 9, e1003343 (2013).ADS 
PubMed 
PubMed Central 

Google Scholar 
126.Quinlan, A. R. & Hall, I. M. BEDTools: A flexible suite of utilities for comparing genomic features. Bioinformatics 26, 841–842 (2010).CAS 
PubMed 
PubMed Central 

Google Scholar 
127.Alexa, A. & Rahnenführer, J. TopGO: Enrichment analysis for Gene Ontology. R package version 2.44.0. (2021).128.Osuna-Cruz, C. M. et al. PRGdb 3.0: A comprehensive platform for prediction and analysis of plant disease resistance genes. Nucleic Acids Res. 46, D1197–D1201 (2018).CAS 
PubMed 

Google Scholar 
129.Camacho, C. et al. BLAST+: Architecture and applications. BMC Bioinformatics 10, 421 (2009).PubMed 
PubMed Central 

Google Scholar 
130.Paquette, S. R. Useful Functions for (Batch) File Conversion and Data Resampling in Microsatellite Datasets. https://cran.r-project.org/package=PopGenKit (2012).131.Excoffier, L. & Lischer, H. E. L. Arlequin suite ver 3.5: A new series of programs to perform population genetics analyses under Linux and Windows. Mol. Ecol. Resour. 10, 564–567 (2010).PubMed 

Google Scholar 
132.Frichot, E., Mathieu, F., Trouillon, T., Bouchard, G. & François, O. Fast and efficient estimation of individual ancestry coefficients. Genetics 196, 973–983 (2014).PubMed 
PubMed Central 

Google Scholar 
133.Jombart, T., Devillard, S. & Balloux, F. Discriminant analysis of principal components: A new method for the analysis of genetically structured populations. BMC Genet. 11, 94 (2010).PubMed 
PubMed Central 

Google Scholar 
134.Pritchard, J. K., Stephens, M. & Donnelly, P. Inference of population structure using multilocus genotype data. Genetics 155, 945–959 (2000).CAS 
PubMed 
PubMed Central 

Google Scholar 
135.Frichot, E. & François, O. LEA: An R package for landscape and ecological association studies. Methods Ecol. Evol. 6, 925–929 (2015).
Google Scholar 
136.Jombart, T. & Ahmed, I. Genetics and population analysis. Adegenet 1.3-1: New tools for the analysis of genome-wide SNP data. Bioinformatics 27, 3070–3071 (2011).CAS 
PubMed 
PubMed Central 

Google Scholar  More

Teasing apart the factors that influence prey choice and foraging tactics in the wild poses formidable logistical challenges because of multiple confounding features. For example, a particular type of prey may be rarely consumed not because of predator aversion, but because that prey type is more difficult to find or to capture than some other kind of prey22. Similarly, predators may key in on specific types of prey based on dietary preferences, prey size, or abundance23,24,25. The method of bait deployment that we adopted circumvents many of those problems, by standardising prey abundance, observability, and ease of capture by the predator. Under these conditions, free-ranging crocodiles from toad-sympatric versus toad-naïve populations showed substantial differences in foraging tactics and bait choice. In toad-naïve populations, crocodiles took equal numbers of treatment (toad) baits and control (chicken) baits, and frequently took baits located on land as well as over water. In contrast, crocodiles in toad-sympatric populations generally avoided toad baits in all locations and foraged primarily in the water rather than on land. Both of these shifts—in prey types and foraging locations—conceivably reduce the vulnerability of crocodiles to fatal ingestion of highly toxic cane toads.The relatively rapid ( More