More winners than losers over 12 years of monitoring tiger moths (Erebidae: Arctiinae) on Barro Colorado Island, Panama

Data used in the manuscript:

Column Name: Description

Taxon.Code: this code represents the unique identification number used during monitoring for each species of there past 12 years

Species.name: This is the initial species name based on morphological and DNA barcoding data

Proposed.new.name: Recent changes in DNA barcoding data and results from taxonomist suggest these new proposed species names

Tribe: The tribe that each Arctiinae is within

Subtribe: The subtribe that each species is within

numYearObs: The number of years that each species was observed, a maximum of 12 and a minimum of 6

sumAbund: The total number of individuals that were collected over the 12 year period

NumZeroObs: The number of sample that the species was not observed, with a maximum of 47 and minimum of 0

n: Total number of sample days over the 12 year period. Four sample days per year, but we omitted one sample period due to covid

pctZero: the proportion of sample periods that the species was not observed

pctObs: the proportion of sample periods that the species was observed

NumZeroObsYear: the number of years that each species was not observed. As long as a species was observed once out of the 4 sample days, it was counted as observed for that year

pctZeroYear: the proportion of years that species was not observed

pctObsYear: the proportion of years that species was observed

numyear.vec: the mean of the posterior probability distribution for Year in the Negative Binomial Bayesian regression. This served as our measurement of population trend

postCI2.5.numyear.vec: The lower 2.5% of the 95% credible interval from the posterior probability distributon for year

postCI97.5.numyear.vec: The upper 2.5% of the 95% credible interval from the posterior probability distributon for year

p.value.KSTest: p-value from the DHARMa package checking for uniformity of siulated residuals from the bayesian regression. Uses the Komogorov-Smirnov Test (KS Test). P-values greater than 0.05 suggest no deviation from uniformity

p.value.ZInfTest: p value from DHARMa's zero inflation test which compares observed number of zeros with zeros expected from simulations. Values greater than 0.05 suggest no evidence of zero inflation

p.value.DispTest: p-value form DHARMa's dispersion test, which compares the variance of the observed raw residuals against the variance of the simulated residuals. Values greater than 0.5 suggest no evidence of over dispersion

p.value.TempAutoCor: p-value from DHARM's temporal autocorrelation test, which used the Durbin-Watson test on the uniformly scaled residuals. If values are greater than 0.05, it suggest no temporal autocorrelation

looic: leave-one-out cross validation information criteria from "loo" package

r2Bayes: R2 for each Bayesian model was estimated as the variance of the residuals for a given mode divided by the total variance, which is the sum of the variance of fitted vales and the residual variance. The value was subtracted from 1

essyear: Effective sample size, which measure the amount of independent information there is in an autocorrelated chain (Kurschke 2015)

probgreat1: Probability that the trend is greater than one, or the species is increasing based on the Bayesian "degree-of-belief"

Forewing_length: Average forewing length for each species. Averaged across sexes

Thorax_width: Average thorax width for each species. Averaged across sexes

Wing_load: ratio of thorax width to wing length

Wing_tip_angle: Angle of the tip of the forewing

Lightness: Lightness of the dorsal wing

Redness: Redness of the dorsal wing

Greenness: Greenness of the dorsal wing

Blueness: Blueness of the dorsal wing

mean.red: mean red value of dorsal wings

mean.green: mean green value of dorsal wings

mean.blue: mean blue value of dorsal wings

maxtemp.vec: the mean of the posterior probability distribution for average maximum temperature during the sampling month in a Negative Binomial Bayesian regression. This served as our measurement of sensitivity to average maximum temperature

postCI2.5.maxtemp.vec: The lower 2.5% of the 95% credible interval from the posterior probability distribution for the maximum temperature predictor

postCI97.5.maxtemp.vec: The upper 2.5% of the 95% credible interval from the posterior probability distribution for the maximum temperature predictor

mintemp.vec: the mean of the posterior probability distribution for average miniumun temperature during the sampling month in a Negative Binomial Bayesian regression. This served as our measurement of sensitivity to average maximum temperature

postCI2.5.mintemp.vec: The lower 2.5% of the 95% credible interval from the posterior probability distribution for the minimum temperature predictor

postCI97.5.mintemp.vec: The upper 2.5% of the 95% credible interval from the posterior probability distribution for the minimum temperature predictor

avgprecip.vec: the mean of the posterior probability distribution for average precipitation duing the sampling month in a Negative Binomial Bayesian regression. This served as our measurement of sensitivity to average maximum temperature

postCI2.5.avgprecip.vec: The lower 2.5% of the 95% credible interval from the posterior probability distribution for the average monthly precipitation predictor

postCI97.5.avgprecip.vec: The upper 2.5% of the 95% credible interval from the posterior probability distribution for the average monthly precipitation predictor

CVAbund: Coefficient of variation in abundance across each sample period (n=47)

Geographic_range: a count of the number of countries with records for each species logged on internet databases including funet, BOL, EOL, GBIF, and Google scholar

Iridescence: presence/absence of iridescence

Aposematism: presence/absence of aposematism

Clearwing_coverage: % clearwing coverage

Seasonality: a seasonal decomposition of time series as the proportion of standard deviation in seasonal component to standard deviation in detrended time series

Spatial_aggregation: Lloyd index of patchiness (Lloyd, 1967), based on the dispersion of individuals captured in the ten traps across BCI

Final.group: functional grouping based on multivariate and phylogenetic analyses