Monitoring population extinction risk with community science data

Monitoring population extinction risk with community science data | bioRxiv /* */ /* */ <!-- <!-- /*! * yepnope1.5.4 * (c) WTFPL, GPLv2 */ (function(a,b,c){function d(a){return"[object Function]"==o.call(a)}function e(a){return"string"==typeof a}function f(){}function g(a){return!a||"loaded"==a||"complete"==a||"uninitialized"==a}function h(){var a=p.shift();q=1,a?a.t?m(function(){("c"==a.t?B.injectCss:B.injectJs)(a.s,0,a.a,a.x,a.e,1)},0):(a(),h()):q=0}function i(a,c,d,e,f,i,j){function k(b){if(!o&&g(l.readyState)&&(u.r=o=1,!q&&h(),l.onload=l.onreadystatechange=null,b)){"img"!=a&&m(function(){t.removeChild(l)},50);for(var d in y[c])y[c].hasOwnProperty(d)&&y[c][d].onload()}}var j=j||B.errorTimeout,l=b.createElement(a),o=0,r=0,u={t:d,s:c,e:f,a:i,x:j};1===y[c]&&(r=1,y[c]=[]),"object"==a?l.data=c:(l.src=c,l.type=a),l.width=l.height="0",l.onerror=l.onload=l.onreadystatechange=function(){k.call(this,r)},p.splice(e,0,u),"img"!=a&&(r||2===y[c]?(t.insertBefore(l,s?null:n),m(k,j)):y[c].push(l))}function j(a,b,c,d,f){return q=0,b=b||"j",e(a)?i("c"==b?v:u,a,b,this.i++,c,d,f):(p.splice(this.i++,0,a),1==p.length&&h()),this}function k(){var a=B;return a.loader={load:j,i:0},a}var l=b.documentElement,m=a.setTimeout,n=b.getElementsByTagName("script")[0],o={}.toString,p=[],q=0,r="MozAppearance"in l.style,s=r&&!!b.createRange().compareNode,t=s?l:n.parentNode,l=a.opera&&"[object Opera]"==o.call(a.opera),l=!!b.attachEvent&&!l,u=r?"object":l?"script":"img",v=l?"script":u,w=Array.isArray||function(a){return"[object Array]"==o.call(a)},x=[],y={},z={timeout:function(a,b){return b.length&&(a.timeout=b[0]),a}},A,B;B=function(a){function b(a){var a=a.split("!"),b=x.length,c=a.pop(),d=a.length,c={url:c,origUrl:c,prefixes:a},e,f,g;for(f=0;f<d;f++)g=a[f].split("="),(e=z[g.shift()])&&(c=e(c,g));for(f=0;f<b;f++)c=x[f](c);return c}function g(a,e,f,g,h){var i=b(a),j=i.autoCallback;i.url.split(".").pop().split("?").shift(),i.bypass||(e&&(e=d(e)?e:e[a]||e[g]||e[a.split("/").pop().split("?")[0]]),i.instead?i.instead(a,e,f,g,h):(y[i.url]?i.noexec=!0:y[i.url]=1,f.load(i.url,i.forceCSS||!i.forceJS&&"css"==i.url.split(".").pop().split("?").shift()?"c":c,i.noexec,i.attrs,i.timeout),(d(e)||d(j))&&f.load(function(){k(),e&&e(i.origUrl,h,g),j&&j(i.origUrl,h,g),y[i.url]=2})))}function h(a,b){function c(a,c){if(a){if(e(a))c||(j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}),g(a,j,b,0,h);else if(Object(a)===a)for(n in m=function(){var b=0,c;for(c in a)a.hasOwnProperty(c)&&b++;return b}(),a)a.hasOwnProperty(n)&&(!c&&!--m&&(d(j)?j=function(){var a=[].slice.call(arguments);k.apply(this,a),l()}:j[n]=function(a){return function(){var b=[].slice.call(arguments);a&&a.apply(this,b),l()}}(k[n])),g(a[n],j,b,n,h))}else!c&&l()}var h=!!a.test,i=a.load||a.both,j=a.callback||f,k=j,l=a.complete||f,m,n;c(h?a.yep:a.nope,!!i),i&&c(i)}var i,j,l=this.yepnope.loader;if(e(a))g(a,0,l,0);else if(w(a))for(i=0;i (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0];var j=d.createElement(s);var dl=l!='dataLayer'?'&l='+l:'';j.src='//www.googletagmanager.com/gtm.js?id='+i+dl;j.type='text/javascript';j.async=true;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-M677548'); Skip to main content Home About Submit ALERTS / RSS Search for this keyword Advanced Search New Results Monitoring population extinction risk with community science data View ORCID Profile Orlando Acevedo-Charry , View ORCID Profile José M. Ponciano , View ORCID Profile Caroline L. Poli , View ORCID Profile Brian M. Jeffery , View ORCID Profile Robert J. Fletcher Jr. , View ORCID Profile María Ángela Echeverry-Galvis , View ORCID Profile Bette A. Loiselle , View ORCID Profile Scott K. Robinson , View ORCID Profile Miguel A. Acevedo doi: https://doi.org/10.1101/2025.06.19.660569 Orlando Acevedo-Charry 1 School of Natural Resources and Environment , IFAS, University of Florida , Gainesville, FL 32611, USA 2 Department of Wildlife Ecology and Conservation , IFAS, University of Florida , Gainesville, FL 32611, USA 3 Florida Museum of Natural History, University of Florida , Gainesville, FL 32611, USA 4 Center for Latin American Studies and Tropical Conservation Development Program, University of Florida , Gainesville, FL 32611, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Orlando Acevedo-Charry For correspondence: acevedocharry{at}gmail.com José M. Ponciano 5 Department of Biology, University of Florida , Gainesville, FL 32611, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for José M. Ponciano Caroline L. Poli 6 Fish and Wildlife Research Institute , Florida Fish and Wildlife Conservation Commission, 1105 SW Williston Rd, Gainesville, FL 32601, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Caroline L. Poli Brian M. Jeffery 2 Department of Wildlife Ecology and Conservation , IFAS, University of Florida , Gainesville, FL 32611, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Brian M. Jeffery Robert J. Fletcher Jr. 2 Department of Wildlife Ecology and Conservation , IFAS, University of Florida , Gainesville, FL 32611, USA 7 Department of Zoology, Conservation Research Institute, University of Cambridge , Cambridge, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Robert J. Fletcher Jr. María Ángela Echeverry-Galvis 8 Departamento de Ecología y Territorio, Pontificia Universidad Javeriana , Bogotá D.C., Colombia Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for María Ángela Echeverry-Galvis Bette A. Loiselle 2 Department of Wildlife Ecology and Conservation , IFAS, University of Florida , Gainesville, FL 32611, USA 4 Center for Latin American Studies and Tropical Conservation Development Program, University of Florida , Gainesville, FL 32611, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Bette A. Loiselle Scott K. Robinson 3 Florida Museum of Natural History, University of Florida , Gainesville, FL 32611, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Scott K. Robinson Miguel A. Acevedo 2 Department of Wildlife Ecology and Conservation , IFAS, University of Florida , Gainesville, FL 32611, USA Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Miguel A. Acevedo Abstract Full Text Info/History Metrics Supplementary material Preview PDF Abstract The robust estimation of local extinction risk is central to inform management and conservation efforts. Still, estimating this key demographic parameter requires standardized monitoring data that are lacking for most species and systems. The analysis of community science data is emerging as a promising alternative. These expansive datasets leverage observations from multiple volunteers that provide higher temporal and spatial resolution. Nevertheless, the proper analysis of community science data is challenging because it requires accounting for additional complexities in the intrinsic ecological and observational processes. To address this issue, we describe and test a quantitative approach that fits continuous state-space models iteratively to eBird data with the ultimate goal of estimating local persistence probability through time. We evaluated model accuracy by comparing estimates and trends from eBird with those from the endangered Everglades’ snail kite long-term, standardized monitoring project. We also performed two separate sensitivity analyses (temporal and sampling thinning) to assess how robust the persistence estimates are to a reduction in the number of eBird observations available. Our results showed that the temporal trend trajectory of local population persistence estimated from eBird closely matched that from standardized monitoring. Moreover, the trend remained similar even when reducing the amount of eBird data available to 5% of the original data set – a reduction from 258 to 13 weeks or from 7,714 to 385 lists of observations across 5 years of monitoring. Synthesis and applications Our modeling framework provides a robust, computationally efficient, and easy-to-apply tool for monitoring local persistence probability that can support global conservation efforts. This will complement the monitoring of species population viability in places where standardized monitoring is still lacking, but community science observations are common. 1 | INTRODUCTION Monitoring population trends and assessing extinction risk is a priority in applied ecology and conservation ( Bakker & Doak, 2009 ; Callaghan et al., 2024 ; Dennis et al., 1991 ; Fink, Johnston, et al., 2023 ; Neate-Clegg et al., 2020 ; Staples et al., 2005 ). However, accurate trend predictions are often constrained by the limited availability of standardized biodiversity monitoring data that are costly and logistically challenging to gather, particularly for species in understudied regions ( Lees et al., 2022 ; Robinson et al., 2021 ; Simmons et al., 2019 ). Community science data, which is gathered by volunteers and often referred to as citizen or participatory science, has emerged as a promising alternative to address these data gaps ( Amano et al., 2016 ; Crawford et al., 2020 ; Dennis et al., 2017 ; Johnston et al., 2025 ; Kelling et al., 2019 ). Yet, these opportunistic datasets present multiple complexities, requiring advanced modeling approaches to ensure they accurately reflect ecological processes ( Boyd et al., 2023 ; Dennis et al., 2006 ; Johnston et al., 2023 , 2025 ). Community science data has the potential to inform population viability – the minimum conditions required for the long-term persistence of a population ( Bakker & Doak, 2009 ; Crawford et al., 2020 ; Kéry et al., 2010 ; Staples et al., 2005 ; Trouillier et al., 2023 ). However, challenges remain to use community science data for biodiversity monitoring appropriately ( Johnston et al., 2023 ). The opportunistic nature of most community science data initiatives has inherent spatial, observer, and reporting preference biases ( Johnston et al., 2021 ), leading to heterogeneous data that are challenging to analyze ( Johnston et al., 2023 ). Therefore, since more data does not necessarily mean correct representation of true ecological processes ( Boyd et al., 2023 ; Meng, 2018 ), modeling approaches accounting for multiple data complexities are needed to analyze community science data ( Dennis et al., 2006 ; Fink, Johnston, et al., 2023 ; Johnston et al., 2023 , 2025 ). Although recent analytical advances have addressed some of these challenges (Fink, Johnston, et al., 2023 ; Johnston et al., 2021 , 2023 ; Stillman et al., 2023 ; Zhao et al., 2024 ), crucial limitations persist. For example, most of these approaches remain resource-intensive and infeasible to apply in many regions with reduced community science data, scarcity of standardized monitoring data, or limited computer power; thus, limited accessibility to powerful but complex tools continues to impede their broader use ( Johnston et al., 2023 ). In addition, the application of extinction theory to assess population risk from community science data is still conspicuously uncommon ( Bakker & Doak, 2009 ; Ferguson & Ponciano, 2014 ; Staples et al., 2005 ), despite its potential to provide key insights about population trends ( Crawford et al., 2020 ; Dennis et al., 2017 ; Fink, Johnston, et al., 2023 ; Johnston et al., 2025 ; Kéry et al., 2010 ). Therefore, to appropriately leverage community science data, there is still a need for modeling approaches that provide a balance between robustness during the modeling, disentangle observation and ecological processes, ease of application, and computational efficiency. Here we propose and test a risk-based Viable Population Monitoring framework (VPM; Figure 1 ; Staples et al., 2005 ) as a robust and computationally efficient framework to estimate local population persistence from community science data. VPM is a promising tool to estimate the risk of quasi-extinction ( Staples et al., 2005 ), which is defined as the species probability to decrease below pre-established abundance thresholds and become effectively extirpated ( Ginzburg et al., 1982 ; Trouillier et al., 2023 ). In addition, VPM can integrate state-space models that help address the complexities from community science data, including disentangling observation and environmental noise. State-space population dynamic models link observed time series data to latent population processes ( Dennis & Ponciano, 2014 ; Humbert et al., 2009 ; Newman et al., 2014 , 2023 ). These state-space models provide robust estimates of population growth rates, density dependence (when present), and independent variability in ecological and observational processes ( Dennis et al., 2006 ; Dennis & Otten, 2000 ; Evans et al., 2023 ; Hostetler & Chandler, 2015 ; Ponciano et al., 2009 ). The VPM framework uses those estimated population parameters from the state-space models to simulate near-future trajectories and estimate population quasi-extinction probabilities and its mathematical complement, the local persistence probability (φ). Furthermore, the VPM iteratively updates these estimates with new data collection, monitoring the population extinction risk ( Figure 1 ). Download figure Open in new tab Figure 1. The risk-based viable population monitoring (VPM) framework for estimating local persistence probabilities over time. Observed abundance (points) informs a population model fitted from 𝑡 1 to 𝑡 𝑖 (black lines). Using estimated parameters (e.g., growth rate, environmental or observation noise), stochastic trajectories are simulated beyond the vertical dotted line (𝑡 𝑖 ) within moving windows of 𝑆 years (gray ribbons, lines, and shapes). These simulations estimate the probability of counts falling below a quasi- extinction threshold (dashed line, red crosses); the complement is the local persistence probability (φ). The VPM iteratively updates these estimates with new data. Panels ( a-f ) show monitoring progress: abundance from standardized monitoring (SM; left), from community science data (CS; center), and persistence probability estimation φ (right). Probabilities are estimated using illustrative simulation window lengths of 𝑆 3 (3 years, dark gray ribbon, circles), 𝑆 5 (5 years, gray ribbon, squares), and 𝑆 10 (10 years, light gray ribbon, diamonds) across each time with new monitoring data, 𝑖 = 1 in 𝑡 5 ( a ) to 𝑖 = 6 in 𝑡 10 ( f ). Simulated trajectories change between iterations, reflecting changes in persistence estimates before abundance drops below the threshold. Lower abundance estimation in CS matches the dynamics of the population in SM, resulting in similar persistence estimates (filled vs hollowed points). Extending simulations enhances the estimated robustness. To assess the accuracy of the estimated trend in local persistence probability from applying the VPM framework to community science data, we compared our estimates using eBird data to those obtained from standardized monitoring of the endangered Everglade snail kite ( Rostrhamus sociabilis plumbeus , snail kite hereafter; ( Dreitz et al., 2002 ; Reichert et al., 2016 )). This approach addresses major limitations in estimating population extinction risk from community science data because it can be rapidly implemented with limited computation capacity and identifies trends at multiple spatial and temporal scales, while separating observational and ecological processes. The Supplementary Information (hereafter SI) presents a detailed tutorial to allow its broad applicability. 2 | MATERIALS AND METHODS Our workflow consists of five stages that we summarize here and expand in the following sections. First, we downloaded available data from eBird for our species of interest ( Section 2.1 ), filtering the complete checklists (lists hereafter) by their sampling effort information to improve their ecological use during analysis ( Johnston et al., 2021 ). Second, we assigned each list to a spatial grid of sampling units, extracted the maximum observed count reported in any list within the sampling unit during a specific time (e.g., weeks), and constructed the observed time series dataset from eBird ( Section 2.1 ). Third, we applied the VPM framework ( Staples et al., 2005 ) to estimate population viability over time, monitoring population risk with community science data ( Section 2.3 ). This stage included the population dynamics estimation through state-space model methods ( Section 2.2 ). Fourth, leveraging the overlapping standardized monitoring ( Section 2.1 ), we compared estimates of persistence probability between eBird and standardized monitoring of snail kites in a subpopulation that recently expanded the species’ breeding range ( Machado-Stredel et al., 2024 ; Poli et al., 2020 ). Finally, we systematically reduced the eBird dataset, reapplied the VPM framework, and compared these estimates with those from the standardized monitoring dataset. Descriptions of the modeling framework are provided below, and a detailed tutorial is available in Zenodo ( Anonymous 2025 ) and as tutorial file (SI). 2.1 | Study system and time series construction The snail kite is an endangered, wetland-dependent raptor that has been systematically monitored in Florida ( Dreitz et al., 2002 ; Fletcher et al., 2024 ; Martin et al., 2007 ; Reichert et al., 2016 ). This species moves among wetlands, tracking spatially varying environmental resources that include fluctuating densities of apple snails ( Pomacea sp.) as its main prey ( Cattau et al., 2016 ; Machado-Stredel et al., 2024 ; Poli et al., 2022 ; Robertson et al., 2017 ). Systematic monitoring of snail kites includes airboat transects across all known breeding sites, nest surveys, and capture-mark-resight methods; see more details in Fletcher et al. (2024) and references therein. Such standardized monitoring data show a trend of decreases in growth rate in the southern populations and increases in the north ( Reichert et al., 2016 , 2021 ), including recent colonization of wetlands over 100 km north of its breeding range ( Cattau et al., 2016 , 2017 ; Fletcher et al., 2024 ; Machado-Stredel et al., 2024 ; Poli et al., 2020 , 2022 ). These data are ideal to validate the accuracy and performance of our modeling approach because the standardized monitoring ran in parallel with eBird time series records, thus facilitating the comparison of estimates with the most robust monitoring data available (SI- Section 4 ). We downloaded the snail kite eBird data from August 1972 (the first date available for the species in Florida) through November 2024 from https://ebird.org/data/download (SI- Section 3 ). To improve the ecological information included in these data for analysis, we conducted refinements through filtering that constrained the variation of sampling effort in eBird ( Backstrom et al., 2024 ; Johnston et al., 2021 ; Kelling et al., 2019 ; Strimas-Mackey et al., 2023 ). We only included complete lists from "traveling" or "stationary" protocols between January 2018 and November 2024 (≥1000 lists per year, SI-Section 3.4, Figure SI-11), within a sampling effort of ≤5 hours, ≤5 km, ≤10 observers, assigning each list to a hexagonal grid of 100 km 2 area – our spatial sampling units ( Figure 2 and SI-Section 3.3). Download figure Open in new tab Figure 2. Map of spatial eBird sampling of snail kites in Florida emphasizing the hexagonal cell with the highest number of lists. (a) Each hexagonal cell corresponds to an area of ∼100 km 2 , and the bright color represents cells with a higher number of submitted lists to eBird (note log10 scale). (b) A single cell in northcentral Florida (∼Payne’s Prairie Preserve State Park, ∼85 km 2 ) concentrated over 7,000 lists. eBird records of snail kites in Payne’s Prairie display some popular sites visited within this sampling unit. This sampling unit overlaps with ongoing standardized monitoring surveys. Although several sampling units overlap with ongoing standardized monitoring surveys, we used data from a single sampling unit corresponding to the recently established population in north-central Florida: the Payne’s Prairie subpopulation in Alachua County ( Figure 2 ). After Hurricane Irma flooded this ∼85 km 2 wetland system in 2017, the snail kites established a nesting population in 2018 ( Machado-Stredel et al., 2024 ; Poli et al., 2020 ) and have experienced an estimated steady increase in abundance ( Poli et al., 2020 ), as inferred from eBird data from 2012-2022 (Fink, Auer, et al., 2023). This expansion motivated the inclusion of Payne’s Prairie in the standardized monitoring project of the species ( Poli et al., 2020 ). Payne’s Prairie has 7,414 eBird lists reporting snail kites since 2018 ( Figure 2 ). We took advantage of this overlap between high eBird observations and standardized monitoring datasets in Payne’s Prairie to compare local persistence estimates using these two data sources (see Section 2.3 ). We constructed the time series for the two datasets by extracting the maximum observed count reported in any list of eBird or any standardized monitoring survey per week in the sampling unit of 100 km 2 that overlapped Payne’s Prairie. We assumed this weekly maximum count is a potential minimum number of individuals known to be alive during sampling ( Poli et al., 2020 ). This assumption implies homogeneous detection probability through time. Still, the state-space modeling approach addresses this issue by estimating the observation process from the two datasets separately. Moreover, to evaluate different structures of the weekly maximum counts time series, we randomly thinned the lists by sampling (list-level) and time (week-level) in our sensitivity analyses (SI-Section 7). 2.2 | State-Space population dynamics modelling At the core of our modelling approach is the specification of a population dynamics model that represents the temporal fluctuation of the population of interest appropriately (represented by the straight black lines in Figure 1 ). Such a model should contain deterministic and stochastic components shaping the population dynamics to effectively capture and estimate population persistence probabilities ( Ferguson & Ponciano, 2014 ). This model should also include estimation of the variability of the data due to observation error ( Dennis et al., 2006 ; Dennis & Otten, 2000 ; Evans et al., 2023 ; Hostetler & Chandler, 2015 ; Ponciano et al., 2009 ). We chose a special case of the stochastic Gompertz state-space model (see SI for additional properties of this model ( Dennis et al., 2006 ; Dennis & Ponciano, 2014 )), which is the Exponential Growth State-Space (EGSS) model for density-independent dynamics ( Humbert et al., 2009 ). Let 𝑁 𝑡 be the unobserved population abundance of the species of interest at time 𝑡. The deterministic skeleton of the ecological process is a discrete-time model using the difference equation 𝑁 𝑡 = 𝑁 O exp(𝑎), where 𝑁 O is the initial population and 𝑎 is a positive constant representing maximum population growth rate ( Humbert et al., 2009 ). The stochastic version including observation and environmental noise is a state-space model with unobserved and observed population abundance. The unobserved abundance is assumed to follow a stochastic process, 𝑁 𝑡 = 𝑁 O exp(𝑎 + 𝐸 𝑡 ), where 𝐸 𝑡 ∼Normal(0, 𝜎 2 ) and 𝜎 2 is the variation of environmental noise. Let 𝑋(𝑡) be the unobserved log-abundance of the population at time 𝑡. We end with a continuous-time ordinary differential equation of a Brownian diffusion process ( Humbert et al., 2009 ), 𝑑𝑋(𝑡 𝑖 ) = (ln𝜆)𝑑𝑡 𝑖 + 𝑑𝐵(𝑡 𝑖 ), where ln𝜆 = 𝑎 is the expected change of log-abundance of the unobserved abundance (𝑋(𝑡 𝑖 )) and 𝑑𝐵(𝑡 𝑖 ) is a random perturbation representing environmental variability, 𝑑𝐵(𝑡 𝑖 )∼Normal(0, 𝜎 2 𝑑𝑡 𝑖 ), both occurring at one small-time unit of time 𝑡 𝑖 . The fit of the EGSS model includes the aforementioned process, and the observation model given by 𝑌(𝑡 𝑖 ) = 𝑋𝑡 𝑖 ) + 𝐹 𝑖 . The observed log-abundance time series is 𝑌(𝑡 𝑖 ), while 𝐹 𝑖 is the observation error, 𝐹 𝑖 ∼Normal(0, 𝜏 2 ), with 𝜏 2 being the variation of observation noise. This model allows sampling times to be discontinuous and not equally spaced ( Dennis & Ponciano, 2014 ; Humbert et al., 2009 ). Thus, this density-independent model has four unknown parameters (SI- Section 2.2 ): ln𝜆 (the trend parameter or expected change of ln𝑁 𝑡 in one small unit time 𝑡 𝑖 ), 𝜎 2 (variability of the process noise), 𝜏 2 (variability of the observation noise), and 𝑥 O (the initial log-abundance population). We compared the estimates of these parameters for the two datasets by visually inspecting the time series of the estimates and by computing the Root Mean Square Error (RMSE), where lower values indicate similarity between estimates from both datasets (SI-Section 6.3). Although the state-space model formulates changes in abundance in discrete times of demographic dynamics (e.g., years), implementing the stochastic continuous models of diffusion process allowed us to use weekly observations as our time series data. This is possible because the discrete-time model has an exact continuous-time counterpart in the diffusion process. Specifically, the solution of the continuous-time model matches the solution of the difference equation model at discrete-time points ( Dennis et al., 1991 ; Dennis & Ponciano, 2014 ; Humbert et al., 2009 ). Structural errors of applying a demographic year-based model to weeks have a lower effect on pattern inferences than parameter estimation errors ( Taper et al., 2008 ). In fact, weekly estimation of local persistence from population dynamics models provides valuable insights for the management of the snail kite populations, given that wetland management (flood or drain regimes) and nest initiation rates during breeding season are typically summarized at such a temporal scale ( Fletcher et al., 2021 , 2024 ). 2.3 | Estimation of the local persistence probability (𝛗) We applied a risk-based viable population monitoring framework (VPM, Staples et al. 2005 ; Figure 1 ). To estimate the unknown parameters in the state-space models, we computed restricted maximum likelihood because it improves identifiability for the multivariate normal distribution in the state-space population models ( Dennis & Ponciano, 2014 ; Humbert et al., 2009 ). The risk-based VPM description recommended the simulation of population trajectories in moving windows of one, two, and five generations into the near future using annual time steps ( Staples et al., 2005 ). Generation time estimates for snail kite range from 5-8 years ( Bird et al., 2020 ; Cattau et al., 2017 ). Here, specifically, we compare trends using moving windows of 5, 10, and 25 years. Note that throughout the manuscript we also refer to the length of these moving simulation window sizes in weeks (𝑆 25O , 𝑆 5OO , or 𝑆 125O , respectively), which was our temporal scale for the observation time series input and the local persistence probability estimation. Furthermore, although the specific population response to hydrology regime in the recently colonized wetland is still unknown, periodic droughts (2-10 years apart) influence snail kite population dynamics in historical locations ( Beissinger, 1995 ); thus, these three moving window sizes can capture population responses to drought regimes at a temporal resolution that is meaningful for managers, as well as the potential rapid rates of population decline during assessment of species vulnerability to extinction ( Bird et al., 2020 ). Defining extinction risk during estimation of population viability is imperative for the VPM framework ( Staples et al., 2005 ). Population extinction risk can be the probability of the population to decrease below a pre-established abundance threshold, known as quasi- extinction ( Beissinger, 1995 ; Dennis, 2002 ; Dennis et al., 1991 ; Ginzburg et al., 1982 ; Staples et al., 2005 ; Trouillier et al., 2023 ). Expert opinion is often used to establish an appropriate quasi- extinction threshold. For example, a previous population viability analysis (PVA) of snail kites, at a time when population size fluctuated between 114 and 668 individuals during a twenty-year survey (uncorrected by sampling effort), used <25 individuals as a quasi-extinction threshold for the entire population in Florida ( Beissinger, 1995 ). In contrast, the quasi-extinction threshold was set at one-half of the initial population estimates in examples within the original description of VPM ( Staples et al., 2005 ). An appropriate quasi-extinction threshold should be low enough to estimate population risk accurately, yet high enough to avoid small population phenomena (e.g., Allee effect) or other extinction vortices that could increase the chances of extinction ( Dennis et al., 1991 ; Dennis, 2002 ). Given that our counts came from two different datasets, we assumed that one-half of the observed mean counts for each dataset between January 2018 and November 2024 provides an appropriate balance for quasi-extinction; we did not run our estimates for our studied subpopulation with different values (but see SI-Section 8). Thus, we assumed that a near future population trajectory might be at extinction risk if, at some point, the predicted weekly high counts decline below a dataset-specific threshold of one-half of the observed mean; . We did not use auxiliary information (i.e., same 𝑁 c from literature or expert opinion, like <25 from Beissinger, 1995 ) because that would assume that the quasi-extinction threshold is fixed and the randomization process came from the data sampling (design-based), instead of treating the quasi-extinction threshold as a realization of a data-generating process (model-based; see Boyd et al., 2023 ; Williams & Brown, 2019 ). To formalize our procedure of estimating persistence in each week 𝑖, we simulated 𝑀 = 50,000 trajectories in the three moving simulation windows (𝑆 25O , 𝑆 5OO , 𝑆 125O ). In each simulated trajectory 𝑚 = 1, 2, …, 𝑀, we defined the indicator function , where 𝑁 𝑠 is the simulated count for each week 𝑠 in 𝑆 after the week of persistence estimation 𝑖 within each window of simulation (𝑆 25O , 𝑆 5OO , 𝑆 125O ), while 𝑁 𝑐 is the assumed threshold for quasi-extinction in each dataset. Then, for each simulated trajectory 𝑚 𝑖 of length 𝑆, the estimated local persistence probability is given by . In other words, the probability of persistence at each time 𝑖 is calculated by the number of times the simulated counts fall below the critical threshold in the following moving window. The estimated probability of persistence E[φ 𝑖 ] for the week 𝑖 are provided by the empirical mean, while the variability of the estimates was captured by its standard deviation and interquartile range values of the simulations. We iterated the process for all the following weeks 𝑖 of available high counts observations (𝑖 ≥ 1) in both datasets, repeating the model fitting, estimating parameters, simulating 𝑀 𝑖 trajectories for the three moving windows of lengths 𝑆, and estimating φ 𝑖 with its variability (SD, first, and third quartile). This iterated process was conducted for all weeks with observations between December 2019 (week 101, 𝑖 = 1) to November 2024 (week 360) for the eBird (258 weeks) and the standardized monitoring datasets (32 weeks). The potential density-dependent and density-independent dynamics of the population can be contrasted in each time 𝑖 by fitting the continuous density-dependent dynamic model ( Dennis & Ponciano, 2014 ), and evaluating the density-dependent parameter in the Gompertz state-space model (𝑏 ≈ 0, see SI- Section 2 ). We conducted this contrast initially (SI-Section 5.4), finding that density-dependent dynamic, suggested by preference of the Ornstein- Uhlenbeck diffusion process model over the Brownian diffusion process model (EGSS), was negligible for both datasets across the entire interval of persistence estimation (SI-Section 6.1, Figure SI-63). To compare the trend of the probability of local persistence estimated from eBird (φ̂ eBird ) with that estimated from standardized monitoring data (φ̂ SM ), we followed a threefold approach. First, we visually inspected the time series of the estimates. Then, we performed linear regression and computed the associated Pearson correlation coefficient (ρ; see Figures SI- 74,80,85). We interpreted the fit of the linear models close to the 1:1 identity plot and higher values of the Pearson correlation between φ̂ eBird and φ̂ SM as indicative of similar trends of persistence probabilities estimated from the two datasets. Finally, we calculated the Root Mean Square Error (RMSE) of the absolute difference between φ̂ eBird and φ̂ SM (SI-Section 6.2). We interpreted lower RMSE values as indicative of, on average, close persistence estimates between the two time series datasets. We conducted two separate sensitivity analyses at the temporal (weeks) and sampling (lists) levels (SI-Section 7). In both we randomly reduced 5% of the eBird data iteratively, from 95% to 5%, ending with a total of 38 eBird reduced datasets (19 temporal, 19 sampling) of weekly high counts data. Temporal thinning was applied to the eBird time series of weekly high counts described above. On the other hand, sampling thinning was applied to the original data of 7,414 complete lists, constructing different time series of weekly high counts. With these reduced datasets, we estimated φ 𝑖 and its corresponding variability over time for the given simulation moving windows of two generations (𝑆 5OO ). As mentioned before, this window might reflect periodic droughts, every 2-10 years, affecting the population. In addition, this simulated window had the best compromise between high correlation and low RMSE between the two datasets (SI-Section 6.2). Similarly to the full eBird data comparison, we compared the relationship of φ̂ eBird from each eBird reduced data set with 𝛟̂ SM through visual inspection of the resulting time series, performing linear regressions with the corresponding Pearson correlation coefficient, and calculating the RMSE. Again, we interpreted a fit of the linear model close to the 1:1 identity plot, higher values of Pearson coefficient, and lower RMSE values as validation indicators that the trend of φ̂ eBird is similar to φ̂ SM . All our analyses were conducted using the R software v4.3.3 (R Core Team, 2024). Specifically, we filtered the eBird data with the packages auk ( Strimas-Mackey et al., 2018 ) and tidyverse ( Wickham et al., 2019 ), managing dates with lubridate ( De Valpine et al., 2017 ). We generated the hexagonal grid of spatiotemporal subsampling with the package dggridR ( Barnes, 2023 ). We also adjusted the R scripts published by Humbert et al. (2009) and Dennis & Ponciano (2014), which require the package MASS ( Venables & Ripley, 2002 ). In addition to figures using basic R or ggplot2 ( Wickham, 2016 ) within tidyverse , we used the packages maps ( Becker et al., 2022 ) and gridExtra ( Auguie, 2017 ) to load maps and generate composite figures, respectively, and ggpubr ( Kassambara, 2023 ) to include some linear equations within the figures. 3 | RESULTS 3.1 | Predicted abundance and model parameters estimates The predicted weekly high counts of snail kites in Payne’s Prairie, using independent EGSS models, closely matched the observed high counts in both eBird and standardized monitoring datasets ( Figure 3 ). The predicted trajectory from standardized monitoring was highly correlated with the observed weekly high counts (linear model; intercept: −0.63 ± 1.08 SE; slope = 0.929 ± 0.030 SE, p < 0.001; Figures 3 , SI-23). The EGSS model for eBird data predicted lower weekly high counts than standardized methods (linear model; intercept: 2.34 ± 0.79 SE; slope = 0.074 ± 0.022, p = 0.004; Figures 3 ). Nonetheless, the estimated parameters for environmental noise (𝜎e 2 ≅ 0.03) and initial population size (weekly high count at 𝑡 = 1; 𝑒 𝑥̂0 ≅ 1) were identical between the two datasets. The expected change in log-abundance per week showed a lower increase in eBird and a higher change in standardized monitoring weekly counts . Unsurprisingly, observation noise was about an order of magnitude greater in the eBird data compared to the standardized monitoring data 0.027). We expected this strong difference due to the higher variation of the sampling effort in eBird compared to standardized surveys. This pattern stands across the 32 weeks of estimation with overlap between the datasets (SI-Section 6.3), with same estimation between datasets for initial population (𝑒 𝑥̂0 ), close estimation for environmental noise (𝜎̂ 2 ) and trend parameters , while the higher difference for observation noise estimates (𝜏̂ 2 ; Table 1 , SI-Section 6.3). Download figure Open in new tab Figure 3. Weekly high counts observed (triangles and points) and predicted (solid lines) of snail kites in Payne’s Prairie from standardized monitoring (orange) and eBird (blue). Predicted values correspond to exponential growth state-space (EGSS) models fitted for the two datasets during the initial population dynamic in north-central Florida (from week 1, January 2018 to week 101, December 2019, which was the first week of local persistence estimation, dotted vertical line and gray arrows). Although eBird underpredicts the weekly counts, the estimated population parameters were very similar between the two datasets (SI-Section 6.3), with the main difference being higher observation noise in the models fitted with eBird data compared to the models fitted with standardized monitoring data ( Table 1 ). View this table: View inline View popup Download powerpoint Table 1. Restricted maximum likelihood estimates for the parameters in the exponential growth state- space models used to estimate the local probability of persistence from two datasets. 3.2 | Estimated local persistence probability (𝛗̂) Local persistence probability (φ̂) estimated from eBird data revealed a similar overall decreasing pattern to that estimated from the standardized monitoring data ( Figure 4 ). The correlation between the two estimates was strongest when the population was projected one generation ahead (𝑆 25O , Pearson ρ = 0.959; Figure 4a ), slightly lower for two generations (𝑆 5OO , ρ = 0.934; Figure 4b ), and weakest for five generations (𝑆 125O , ρ = 0.903; Figure 4c ). However, the RMSE was lowest for 𝑆 125O (ρ = 0.005) compared to 𝑆 5OO (ρ = 0.039) or 𝑆 25O (ρ = 0.103), indicating closer agreement between the datasets at longer projection intervals (SI-Section 6.2). The decreasing trend through time in φ̂ was consistent among datasets and simulation windows during the monitoring of the population extinction risk, though overall lower φ̂ for shorter than longer windows (compare Figure 4a and 4c ). This might be related to stochastic simulated trajectories length, with a higher proportion of simulated weeks ending below the quasi- extinction threshold, resulting in overall lower φ̂. The higher temporal resolution of eBird data allowed for the estimation of φ̂ when data from the standardized monitoring was lacking. This lack of data occurred because the prairie was too dry to conduct airboat surveys. Download figure Open in new tab Figure 4. Estimated local persistence probability (φ̂) during five-years of monitoring (December 2019 to November 2024) after simulation trajectories for three moving windows ( a : 𝑆 250 , b : 𝑆 500 , c : 𝑆 1250 ). For each week of estimation, a density-independence model (EGSS) was fitted for two observation time series datasets, estimating parameters, simulating 50,000 trajectories, and estimating local persistence probability. Variation in φ̂ (points) is reflected by the + one standard deviation (thin light line ranges) and the first and third quartiles (coarse line ranges). Higher temporal resolution is evident in eBird ( n = 258) when compared with standardized monitored data ( n = 32). See extended figures SI-69 to SI-85 (SI- Sections 6.4 to 6.6) with the observed high counts time series and the figure SI-63 (SI-Section 6.1) for the inclusion of potential density-dependent dynamics. The declining pattern between φ̂ from standardized monitoring and eBird datasets is highly correlated (ρ > 0.9), although the best compromise between high correlation and low RMSE was identified in a simulation window of two generations, 10 years (𝑆 500 ). 3.3 | Sensitivity analyses to varying time series length and sampling Overall, the correlation in the trend from the two data sources was consistently above ρ ≥ 0.90 with a temporal reduced datasets of 95-30% (about 47 to 18 weeks per year) and 20% (10 weeks per year) of the eBird high counts data ( Figures 5 , SI-88, SI-90). Similarly, the correlation in the trend was also above ρ ≥ 0.90 with a sampling reduced datasets of 95-50% ( Figures 5 , SI-94, SI-96). Moderate to high correlation, 0.65 ≤ ρ < 0.90, was detected with a temporal reduced dataset of 25% (13 weeks per year) and 15% (8 weeks per year), and a sampling reduced dataset of 45-25% (51-48 weeks or 670-370 lists per year). Few overlapping weeks between the datasets limited the correlation estimates in a temporal reduced datasets of 10-5%. A sampling reduction to the 20-5% of eBird (297 to 74 lists per year) still had a positive correlation (ρ ≥ 0.50; Figures 5 , SI-94, SI-96). Remarkably, the declining trend of φ̂ was still visually evident even with 5% of the data retained, either at temporal or sampling level, with all reduced datasets having RMSE >0.18 ( Figure 5 , SI-Section 7). These results suggest that the framework is particularly robust and provides meaningful estimates of the probability of persistence even when available data is significantly reduced. Download figure Open in new tab Figure 5. Data reduction as sensitivity analyses. Estimation of local persistence probability (φ̂) from standardized monitoring data (orange; φ̂ SM ) is replicated in each comparison with estimation of local persistence probability from eBird (blue; φ̂ eBird ) under temporal (left) and sampling (right) data reduction. The sequence of panels represents reduction from the complete eBird dataset ( Figure 4c ) to different percentages of randomly sampled data by sampling ( a-e ; see extended figures in SI-Section 7.1.1) and temporal thinning ( f-j ; see extended figures in SI-Section 7.1.2). The n in parentheses indicates the number of weeks from eBird reduced data and overlapping benchmark weeks. Trends of φ̂ are accompanied with the relationship of φ̂ eBird (ordinate) and φ̂ SM (abscissa). A linear model of this relationship is fitted and deployed with a blue line, including confidence intervals in gray ribbon (ρ below the equations shows Pearson correlation coefficient); the diagonal line is the line of identity (1:1). Although model accuracy varies with decreasing data and some overestimation of φ̂ eBird compared to φ̂ SM suggests less conservative estimation, the trends are still similar even with as little as 5% of the original eBird weekly high counts (about three weeks or 74 lists per year). 4 | DISCUSSION We show that a risk-based viable population monitoring (VPM) approach is a promising method for estimating local persistence probability over time using community science data. We compared local persistence estimates from eBird with those from standardized monitoring using an endangered wetland raptor as a case study. Both datasets revealed remarkably similar trends, even when eBird weekly observed counts were reduced from 258 to 13 weeks (around 5% of the original dataset) or from 7,714 to 385 lists across the 5 years of monitoring ( Figures 4 , 5). While persistence estimates using eBird were higher when compared to those estimated from standardized monitoring at lower values, our modeling approach effectively assessed population extinction risk over time. The similarity in persistence estimates between the two datasets stems from applying continuous stochastic state-space models within the VPM framework, which offers flexibility in representing ecological processes. These models estimate population parameters using time series of uneven time intervals ( Dennis & Ponciano, 2014 ; Humbert et al., 2009 ), as in our datasets of weekly counts (SI-Section 7), and assess the risk of declining below a quasi- extinction threshold for each period with data. This combined flexibility avoids reliance on arbitrary thresholds and directly quantifies weekly persistence probability from a model-based approach ( Boyd et al., 2023 ; Williams & Brown, 2019 ). Defining population-size thresholds is a common issue in PVA applications, which has been solved using a percentage of either the carrying capacity ( Ginzburg et al., 1982 ) or the initial population ( Taper et al., 2008 ; Trouillier et al., 2023 ). We used one-half of the observed mean counts for each dataset, but other percentages of population size can be applied with our method (SI-Section 8; Dennis, 2002 ; Ferguson & Ponciano, 2014 ; Trouillier et al., 2023 ). Despite underpredicting weekly counts compared to standardized monitoring ( Figures 3 , SI-23), eBird data captured population risk trends reliably ( Table 1 , Figure 4 ). Underprediction arises from spatial coverage bias ( Royle et al., 2007 ) and observation noise during sampling ( ; SI-Section 6.3; Poli et al. 2020 ). eBird observers are sampling only a portion of the population surveyed by standardized methods ( Poli et al., 2020 ; SI-Section 4). The observers’ expertise in eBird is also directly related to detection probability ( Fink et al., 2020 ; Johnston et al., 2018 ). Taken together, counts submitted to the platform likely do not represent all individuals available to be sampled in the population. These biases, however, are less critical for assessing population risk than for estimating absolute abundance because persistence probability estimation uses population process parameters to simulate future trajectories rather than analyze direct change in abundance prediction under confounding observation error. The similar trend, high correlation, and lower RMSE in risk predictions between datasets ( Figures 4 , 5) further supports the robustness of our approach ( Williams & Brown, 2019 ). Lower persistence probabilities in our case study do not necessarily equate to local extirpation. Snail kites in Florida exhibit spatially structured dynamics moving among wetlands in response to environmental variability ( Martin et al., 2007 ; Poli et al., 2022 ; Reichert et al., 2016 ). This spatial structuring may explain why none of the local persistence probability estimates from systematic surveys identified density-dependent dynamics (SI-Figure SI-47). Density- dependence may act on broader spatiotemporal scales or be less relevant for an expanding population ( Phillips, 2009 ), as is the case of snail kites in Payne’s Prairie – likely not reaching carrying capacity yet ( Poli et al., 2020 ). Our findings for the eBird dataset align with density- independent dynamics observed in this population (Figure SI-63), matching the estimates from this community science platform with those from standardized monitoring data ( Figures 4 , 5 ). Our approach can be extended to incorporate spatial structure, density dependence, biotic interactions, and environmental variables to capture nuanced dynamics in a population ( Dennis & Otten, 2000 ; Ferguson et al., 2017 ; Ives et al., 2003 ). Current methods using eBird data to estimate population trends focus on encounter rates during breeding seasons or unidirectional change in relative abundance ( Fink et al., 2020 ; Fink, Johnston, et al., 2023 ; Horns et al., 2018 ; Neate-Clegg et al., 2020 ). In contrast, we derived weekly estimates of population persistence following a monitoring viability approach ( Bakker & Doak, 2009 ; Staples et al., 2005 ), applying continuous state-space models to estimate parameters and simulate trajectories over temporal windows relevant to management ( Fletcher et al., 2021 , 2024 ; Staples et al., 2005 ). Our approach likely captures population dynamics that are more sensitive to environmental change. For instance, while the snail kite eBird status and trends data products suggest a population increase in north-central Florida (Fink, Auer, et al., 2023), our models reveal a decline in local persistence from the two overlapping datasets ( Figure 4 ). This discrepancy likely reflects seasonal and post-colonization dynamics ( Phillips, 2009 ; Yackulic et al., 2015 ), which seems overlooked by the discrete models used in the eBird status and trends products (Fink, Johnston, et al., 2023 ). Although, on average, eBird shows a rise in abundance (2012: 0 individuals, 2022: ∼14) in Payne’s Prairie, which is also evident in the standardized monitored counts (2012: 0, 2022: ∼100), our models provide insights into the nuanced dynamics of the population trend estimating a probabilistic measure of viability through iterative count-based PVAs ( Bakker & Doak, 2009 ; Staples et al., 2005 ). The declining trend likely reflects snail kite response to environmental variables such as hydrology ( Ferguson et al., 2017 ; Fletcher et al., 2021 ; Robertson et al., 2017 ), among other common population dynamics during range expansion shifts ( Phillips, 2009 ; Yackulic et al., 2015 ). Focusing on local persistence probabilities derived from continuous state-space models, our approach complements existing methods by offering a risk-based framework for population monitoring using community science data. This can be used by researchers and managers to identify new sites or providing early warnings of population change in unmonitored subpopulations and prioritize further standardized monitoring (SI-Section 8; Anonymous 2025 ). Integrated population models combining community science data with other demographic approaches improve abundance and temporal dynamics estimates ( Fletcher et al., 2019 ; Stillman et al., 2023 ; Zhao et al., 2024 ). The problem arises when standardized monitoring is scarce or non-existent. Moreover, some advanced quantitative approaches, while useful (e.g., Fink, Johnston, et al., 2023 ; Johnston et al., 2025 ; Stillman et al., 2023 ), are computationally intensive and, hence, less accessible to a wider audience of ecologists that may lack the computational resources to apply them ( Johnston et al., 2023 ). Our approach addresses these issues by providing an easy-to-use practical tool to assess persistence and population decline risks over time (i.e., monitoring extinction risk). Thus, it is potentially extendible where standardized monitoring is still absent but community science is available, enabling adaptive management during monitoring ( Bakker & Doak, 2009 ; Staples et al., 2005 ). Our two sensitivity analyses (SI-Section 7) confirmed the robustness of our approach when applied to different scenarios like lower counts (by sampling thinning) or lower weeks with data (temporal thinning), fostering potential transferability to other species. Our approach can track population trends and accurately estimate persistence probabilities, even if eBird counts are underpredicted. The method could be an excellent tool for near-term forecasting of population extinction risk at the local spatial scale, but it is not relevant for understanding the demographic mechanism driving the population viability. This means that although useful, this approach does not replace long-term standardized monitoring for population conservation; instead, it will provide a complementary tool. Standardized biodiversity monitoring programs are still exceptions rather than the rule globally, thus leverage on this readily available community science data is a great opportunity. eBird, particularly, includes over 10,000 bird species worldwide, a monthly release of data, and an increasing trend of observers’ participation. Our modeling framework provides a computationally efficient, fast, and easy-to-use capability (see the tutorial in the SI). This is especially useful in places where standardized monitoring needs reinforcement, and community science observations are already common and increasing. AUTHOR CONTRIBUTIONS Orlando Acevedo-Charry, José Miguel Ponciano, and Miguel A. Acevedo conceived the ideas, Caroline L. Poli, Brian M. Jeffery, Robert J. Fletcher Jr., María Ángela Echeverry-Galvis, Bette A. Loiselle and Scott K. Robinson provided feedback to designing the study. Caroline L. Poli, Brian M. Jeffery, Scott K. Robinson, José Miguel Ponciano, Orlando Acevedo-Charry, Miguel A. Acevedo, and Robert J. Fletcher Jr. collected the data; Orlando Acevedo-Charry and José Miguel Ponciano analyzed the data, and interpretation was conducted by all authors; Orlando Acevedo-Charry, Miguel A. Acevedo, and José Miguel Ponciano led the writing of the manuscript. All authors contributed critically to the draft and gave final approval for publication. Statement of inclusion Our study brings together authors from several different countries (Colombia, Guatemala, USA), including scientists based in various regions. All authors were engaged early on with the research and study design to ensure that the diverse perspectives they represent were considered from the onset. Whenever relevant, literature published by scientists from the underrepresented regions was cited. Unsuccessful efforts were made to consider relevant work published in a language different than English; we are planning to address this caveat in future research by submitting the application used in the present study in journals in Spanish language. We begin this effort including the Abstract in Spanish, Portuguese, and Chinese (simplified) for the online version of the article and plan to propose workshops in academic meetings and across institutions. CONFLICT OF INTEREST STATEMENT The authors have no conflict of interest to declare. DATA AVAILABILITY STATEMENT eBird data is freely available (under request) on the website https://ebird.org/science/use-ebird-data . The R language source code with data for the eBird example and standardized benchmark, as well as a full tutorial describing the procedure to estimate persistence probability in the Supplementary Information is available in Zenodo https://doi.org/10.5281/zenodo.15475635 (Acevedo-Charry et al., 2025). ACKNOWLEDGEMENTS We are in debt to the many volunteers reporting observations under good practices or reviewing eBird data to maintain quality of the platform; they are the underlying force that provides support to “the power of the people” by community, participatory, or citizen science data. Orlando Acevedo-Charry received departmental funding at the University of Florida, such as the College of Agricultural and Life Science (CALS) Dean’s Award, the Doris and Earl Lowe and Verna Lowe CALS Scholarship, and the School of Natural Resources and Environment Robin E. Nadeau Graduate Research Award. Snail kite monitoring was implemented under IACUC#202008334 and received funding from the US Army Corps of Engineers (W912HZ-20-2-0033), South Florida Water Management District (9500008729), Florida Fish and Wildlife Conservation Commission (13416), and the St John’s River Water Management District (34941). Talita Ribeiro and Min Zhao provided corrections to the Portuguese and Chinese abstracts, respectively. Jackson Martin and Jay Kalogiros-Pepper locked in on providing comments to improve the manuscript. We also thank two anonymous reviewers whose comments improved the manuscript. We acknowledge that the work described in this article was carried out on the ancestral territory of the late Potano (Timucua) and Seminole (Alachua Seminole) people. Funder Information Declared United States Army Corps of Engineers, https://ror.org/05w4e8v21 , W912HZ-20-2-0033 South Florida Water Management District, https://ror.org/055rxf725 , 9500008729 Florida Fish and Wildlife Conservation Commission, https://ror.org/03y5msf78 , 13416 St John's River Water Management District , 34941 REFERENCES 1. ↵ Amano , T. , Lamming , J. D. L. , & Sutherland , W. J . ( 2016 ). Spatial gaps in global biodiversity information and the role of citizen science . BioScience , 66 ( 5 ), 393 – 400 . doi: 10.1093/biosci/biw022 OpenUrl CrossRef 2. ↵ Anonymous , F. P. R . ( 2025 ). Monitoring population extinction risk with community science data . Zenodo . doi: 10.5281/zenodo.15475635 OpenUrl CrossRef 3. ↵ Auguie , B. ( 2017 ). gridExtra: Miscellaneous Functions for “Grid” Graphics . R Package version 2.3. https://CRAN.R-project.org/package=gridExtra 4. ↵ Backstrom , L. J. , Callaghan , C. T. , Worthington , H. , Fuller , R. A. , & Johnston , A . ( 2024 ). Estimating sampling biases in citizen science datasets . Ibis, ibi . 13343 . doi: 10.1111/ibi.13343 OpenUrl CrossRef 5. ↵ Bakker , V. J. , & Doak , D. F . ( 2009 ). Population viability management: Ecological standards to guide adaptive management for rare species . Frontiers in Ecology and the Environment , 7 ( 3 ), 158 – 165 . doi: 10.1890/070220 OpenUrl CrossRef 6. ↵ Barnes , S. K. ( 2023 ). dggridR: Discrete Global Grids . https://CRAN.R-project.org/package=dggrid R 7. ↵ Becker , R. A. , Wilks , A. R. , Brownrigg , R. , Minka , T. P. , & Dechmyn , A. ( 2022 ). maps: Draw Geographical Maps . R Package version 3.4.1. https://CRAN.R-project.org/package=maps 8. ↵ Beissinger , S. R . ( 1995 ). Modeling extinction in periodic environments: Everglades water levels and Snail Kite population viability . Ecological Applications , 5 ( 3 ), 618 – 631 . doi: 10.2307/1941971 OpenUrl CrossRef 9. ↵ Bird , J. P. , Martin , R. , Akçakaya , H. R. , Gilroy , J. , Burfield , I. J. , Garnett , S. T. , Symes , A. , Taylor , J. , Şekercioğlu, Ç. H., & Butchart, S. H. M. ( 2020 ). Generation lengths of the world’s birds and their implications for extinction risk . Conservation Biology , 34 ( 5 ), 1252 – 1261 . doi: 10.1111/cobi.13486 OpenUrl CrossRef 10. ↵ Boyd , R. J. , Powney , G. D. , & Pescott , O. L . ( 2023 ). We need to talk about nonprobability samples . Trends in Ecology & Evolution , 38 ( 6 ), 521 – 531 . doi: 10.1016/j.tree.2023.01.001 OpenUrl CrossRef PubMed 11. ↵ Callaghan , C. T. , Santini , L. , Spake , R. , & Bowler , D. E . ( 2024 ). Population abundance estimates in conservation and biodiversity research. Trends in Ecology & Evolution , S 0169534724000375 . doi: 10.1016/j.tree.2024.01.012 OpenUrl CrossRef PubMed 12. ↵ Cattau , C. E. , Fletcher Jr , R. J. , Kimball , R. T. , Miller , C. W. , & Kitchens , W. M . ( 2017 ). Rapid morphological change of a top predator with the invasion of a novel prey . Nature Ecology & Evolution , 2 ( 1 ), 108 – 115 . doi: 10.1038/s41559-017-0378-1 OpenUrl CrossRef PubMed 13. ↵ Cattau , C. E. , Fletcher , R. J. , Reichert , B. E. , & Kitchens , W. M . ( 2016 ). Counteracting effects of a non-native prey on the demography of a native predator culminate in positive population growth . Ecological Applications , 26 ( 7 ), 1952 – 1968 . doi: 10.1890/15-1020.1 OpenUrl CrossRef PubMed 14. ↵ Crawford , B. A. , Olds , M. J. , Maerz , J. C. , & Moore , C. T . ( 2020 ). Estimating population persistence for at-risk species using citizen science data . Biological Conservation , 243 , 108489 . doi: 10.1016/j.biocon.2020.108489 OpenUrl CrossRef 15. ↵ De Valpine , P. , Turek , D. , Paciorek , C. , Anderson-Bergman , C. , Temple Lang , D. , & Bodik , R. ( 2017 ). Programming with models: Writing statistical algorithms for general model structures with NIMBLE . Journal of Computational and Graphical Statistics , 26 ( 2 ), 403 – 413 . doi: 10.1080/10618600.2016.1172487 OpenUrl CrossRef 16. ↵ Dennis , B . ( 2002 ). Allee effects in stochastic populations . Oikos , 96 ( 3 ), 389 – 401 . doi: 10.1034/j.1600-0706.2002.960301.x OpenUrl CrossRef Web of Science 17. ↵ Dennis , B. , Munholland , P. L. , & Scott , J. M . ( 1991 ). Estimation of growth and extinction parameters for endangered species . Ecological Monographs , 61 ( 2 ), 115 – 143 . doi: 10.2307/1943004 OpenUrl CrossRef Web of Science 18. ↵ Dennis , B. , & Otten , M. R. M . ( 2000 ). Joint effects of density dependence and rainfall on abundance of San Joaquin Kit Fox . The Journal of Wildlife Management , 64 ( 2 ), 388 . doi: 10.2307/3803237 OpenUrl CrossRef 19. ↵ Dennis , B. , & Ponciano , J. M . ( 2014 ). Density-dependent state-space model for population- abundance data with unequal time intervals . Ecology , 95 ( 8 ), 2069 – 2076 . doi: 10.1890/13-1486.1 OpenUrl CrossRef PubMed 20. ↵ Dennis , B. , Ponciano , J. M. , Lele , S. R. , Taper , M. L. , & Staples , D. F . ( 2006 ). Estimating density dependence, process noise, and observation error . Ecological Monographs , 76 ( 3 ), 323 – 341 . OpenUrl CrossRef Web of Science 21. ↵ Dennis , E. B. , Morgan , B. J. T. , Brereton , T. M. , Roy , D. B. , & Fox , R . ( 2017 ). Using citizen science butterfly counts to predict species population trends . Conservation Biology , 31 ( 6 ), 1350 – 1361 . doi: 10.1111/cobi.12956 OpenUrl CrossRef PubMed 22. ↵ Dreitz , V. J. , Nichols , J. D. , Hines , J. E. , Bennetts , R. E. , Kitchens , W. M. , & Deangelis , D. L . ( 2002 ). The use of resighting data to estimate the rate of population growth of the snail kite in Florida . Journal of Applied Statistics , 29 ( 1–4 ), 609 – 623 . doi: 10.1080/02664760120108854 OpenUrl CrossRef 23. ↵ Evans , K. O. , Davis , J. B. , & Wang , G . ( 2023 ). Assessing long-term dynamics of non- breeding Brown Pelican ( Pelecanus occidentalis ) populations using Christmas Bird Count data . Ibis , 165 ( 1 ), 55 – 68 . doi: 10.1111/ibi.13105 OpenUrl CrossRef 24. ↵ Ferguson , J. M. , & Ponciano , J. M . ( 2014 ). Predicting the process of extinction in experimental microcosms and accounting for interspecific interactions in single-species time series . Ecology Letters , 17 ( 2 ), 251 – 259 . doi: 10.1111/ele.12227 OpenUrl CrossRef PubMed 25. ↵ Ferguson , J. M. , Reichert , B. E. , Fletcher , R. J. , & Jager , H. I . ( 2017 ). Detecting population– environmental interactions with mismatched time series data . Ecology , 98 ( 11 ), 2813 – 2822 . doi: 10.1002/ecy.1966 OpenUrl CrossRef PubMed 26. Fink , D. , Auer , A. , Johnston , A. , Strimas-Mackey , M. , Ligocki , S. , Robinson , O. , Hochachka , W. , Jaromczyk , L. , Crowley , C. , Dunham , K. , Stillman , A. , Davies , I. , Rodewald , A. D. , Ruiz- Gutierrez , V. , & Wood , C . ( 2023 ). eBird Status and Trends, Data Version: 2022; Released: 2023 . Cornell Lab of Ornithology , Ithaca , New York . doi: 10.2173/ebirdst.2022 OpenUrl CrossRef 27. ↵ Fink , D. , Auer , T. , Johnston , A. , Ruiz-Gutierrez , V. , Hochachka , W. M. , & Kelling , S. ( 2020 ). Modeling avian full annual cycle distribution and population trends with citizen science data . Ecological Applications , 30 ( 3 ). doi: 10.1002/eap.2056 OpenUrl CrossRef 28. Fink , D. , Johnston , A. , Strimas-Mackey , M. , Auer , T. , Hochachka , W. M. , Ligocki , S. , Oldham Jaromczyk , L. , Robinson , O. , Wood , C. , Kelling , S. , & Rodewald , A. D. ( 2023 ). A double machine learning trend model for citizen science data . Methods in Ecology and Evolution , 14 ( 9 ), 2435 – 2448 . doi: 10.1111/2041-210X.14186 OpenUrl CrossRef 29. ↵ Fletcher , R. J. , Beatty , M. A. , Elmquist , L. , Jeffery , B. M. , Poli , C. L. , & Robertson , E. P . ( 2024 ). An invasive prey and changing climate interact to shape the breeding phenology of an endangered predator . Global Change Biology , 30 ( 8 ), e17478 . OpenUrl CrossRef PubMed 30. ↵ Fletcher , R. J. , Hefley , T. J. , Robertson , E. P. , Zuckerberg , B. , McCleery , R. A. , & Dorazio , R. M . ( 2019 ). A practical guide for combining data to model species distributions . Ecology , 100 ( 6 ), e02710 . doi: 10.1002/ecy.2710 OpenUrl CrossRef 31. ↵ Fletcher , R. J. , Robertson , E. P. , Poli , C. , Dudek , S. , Gonzalez , A. , & Jeffery , B . ( 2021 ). Conflicting nest survival thresholds across a wetland network alter management benchmarks for an endangered bird . Biological Conservation , 253 , 108893 . doi: 10.1016/j.biocon.2020.108893 OpenUrl CrossRef 32. ↵ Ginzburg , L. R. , Slobodkin , L. B. , Johnson , K. , & Bindman , A. G . ( 1982 ). Quasiextinction probabilities as a measure of impact on population growth . Risk Analysis , 2 ( 3 ), 171 – 181 . doi: 10.1111/j.1539-6924.1982.tb01379.x OpenUrl CrossRef 33. ↵ Horns , J. J. , Adler , F. R. , & Şekercioğlu , Ç. H. ( 2018 ). Using opportunistic citizen science data to estimate avian population trends . Biological Conservation , 221 , 151 – 159 . doi: 10.1016/j.biocon.2018.02.027 OpenUrl CrossRef 34. ↵ Hostetler , J. A. , & Chandler , R. B . ( 2015 ). Improved state-space models for inference about spatial and temporal variation in abundance from count data . Ecology , 96 ( 6 ), 1713 – 1723 . doi: 10.1890/14-1487.1 OpenUrl CrossRef 35. ↵ Humbert , J. , Scott Mills , L. , Horne , J. S. , & Dennis , B . ( 2009 ). A better way to estimate population trends . Oikos , 118 ( 12 ), 1940 – 1946 . doi: 10.1111/j.1600-0706.2009.17839.x OpenUrl CrossRef 36. ↵ Ives , A. R. , Dennis , B. , Cottingham , K. L. , & Carpenter , S. R . ( 2003 ). Estimating community stability and ecological interactions from time-series data . Ecological Monographs , 73 ( 2 ), 301 – 330 . doi: 10.1890/0012-9615(2003)073[0301:ECSAEI]2.0.CO;2 OpenUrl CrossRef Web of Science 37. ↵ Johnston , A. , Fink , D. , Hochachka , W. M. , & Kelling , S . ( 2018 ). Estimates of observer expertise improve species distributions from citizen science data . Methods in Ecology and Evolution , 9 ( 1 ), 88 – 97 . doi: 10.1111/2041-210X.12838 OpenUrl CrossRef 38. ↵ Johnston , A. , Hochachka , W. M. , Strimas-Mackey , M. E ., Ruiz Gutierrez , V. , Robinson , O. J. , Miller , E. T. , Auer , T. , Kelling , S. T. , & Fink , D. ( 2021 ). Analytical guidelines to increase the value of community science data: An example using eBird data to estimate species distributions . Diversity and Distributions , 27 ( 7 ), 1265 – 1277 . doi: 10.1111/ddi.13271 OpenUrl CrossRef 39. ↵ Johnston , A. , Matechou , E. , & Dennis , E. B . ( 2023 ). Outstanding challenges and future directions for biodiversity monitoring using citizen science data . Methods in Ecology and Evolution , 14 ( 1 ), 103 – 116 . doi: 10.1111/2041-210X.13834 OpenUrl CrossRef 40. ↵ Johnston , A. , Rodewald , A. D. , Strimas-Mackey , M. , Auer , T. , Hochachka , W. M. , Stillman , A. , Davis , C. L. , Ruiz-Gutierrez , V. , Dokter , A. M. , Miller , E. T. , Robinson , O. , Ligocki , S. , Jaromczyk , L. O. , Crowley , C. , Wood , C. L. , & Fink , D . ( 2025 ). North American bird declines are greates where species are most abundant . Science , 388 ( 6746 ), 532 – 537 . OpenUrl CrossRef PubMed 41. ↵ Kassambara , A. ( 2023 ). ggpubr: “ggplot2” based publication ready plots . https://CRAN.R-project.org/package=ggpubr 42. ↵ Kelling , S. T. , Johnston , A. , Bonn , A. , Fink , D. , Ruiz-Gutierrez , V. , Bonney , R. , Fernandez , M. , Hochachka , W. M. , Julliard , R. , Kraemer , R. , & Guralnick , R . ( 2019 ). Using semistructured surveys to improve citizen science data for monitoring biodiversity . BioScience , 69 ( 3 ), 170 – 179 . doi: 10.1093/biosci/biz010 OpenUrl CrossRef PubMed 43. ↵ Kéry , M. , Royle , J. A. , Schmid , H. , Schaub , M. , Volet , B. , Häfliger , G. , & Zbinden , N . ( 2010 ). Site-occupancy distribution modeling to correct population-trend estimates derived from opportunistic observations . Conservation Biology , 24 ( 5 ), 1388 – 1397 . doi: 10.1111/j.1523-1739.2010.01479.x OpenUrl CrossRef PubMed 44. ↵ Lees , A. C. , Haskell , L. , Allinson , T. , Bezeng , S. B. , Burfield , I. J. , Renjifo , L. M. , Rosenberg , K. V. , Viswanathan , A. , & Butchart , S. H. M . ( 2022 ). State of the World’s birds . Annual Review of Environment and Resources , 47 ( 1 ), 231 – 260 . doi: 10.1146/annurev-environ-112420-014642 OpenUrl CrossRef 45. ↵ Machado-Stredel , F. , Atauchi , P. J. , Nuñez-Penichet , C. , Cobos , M. E. , Osorio-Olvera , L. , Khalighifar , A. , Peterson , A. T. , & Fletcher Jr. , R. J . ( 2024 ). The roles of abiotic and biotic factors in driving range shifts: An invasive Pomacea snail facilitates Rostrhamus sociabilis (Snail Kite) northward range expansion . Ornithology , 141 , ukae019. doi: 10.1093/ornithology/ukae022 OpenUrl CrossRef 46. ↵ Martin , J. , Kitchens , W. M. , & Hines , J. E . ( 2007 ). Importance of well-designed monitoring programs for the conservation of endangered species: Case study of the snail kite . Conservation Biology , 21 ( 2 ), 472 – 481 . doi: 10.1111/j.1523-1739.2006.00613.x OpenUrl CrossRef PubMed Web of Science 47. ↵ Meng , X.-L . ( 2018 ). Statistical paradises and paradoxes in big data (I): Law of large populations, big data paradox, and the 2016 US presidential election . The Annals of Applied Statistics , 12 ( 2 ). doi: 10.1214/18-AOAS1161SF OpenUrl CrossRef 48. ↵ Neate-Clegg , M. H. C. , Horns , J. J. , Adler , F. R. , Kemahlı Aytekin , M. Ç. , & Şekercioğlu, Ç. H. ( 2020 ). Monitoring the world’s bird populations with community science data . Biological Conservation , 248 , 108653 . doi: 10.1016/j.biocon.2020.108653 OpenUrl CrossRef 49. ↵ Newman , K. , Buckland , S. T. , Morgan , B. J. T. , King , R. , Borchers , D. L. , Cole , D. J. , Besbeas , P. , Gimenez , O. , & Thomas , L . ( 2014 ). Modelling Population Dynamics: Model Formulation , Fitting and Assessment using State-Space Methods. Springer New York . doi: 10.1007/978-1-4939-0977-3 OpenUrl CrossRef 50. ↵ Newman , K. , King , R. , Elvira , V. , De Valpine , P. , McCrea , R. S. , & Morgan , B. J. T. ( 2023 ). State-space models for ecological time-series data: Practical model-fitting . Methods in Ecology and Evolution , 14 ( 1 ), 26 – 42 . doi: 10.1111/2041-210X.13833 OpenUrl CrossRef 51. ↵ Phillips , B. L . ( 2009 ). The evolution of growth rates on an expanding range edge . Biology Letters , 5 ( 6 ), 802 – 804 . doi: 10.1098/rsbl.2009.0367 OpenUrl CrossRef PubMed Web of Science 52. ↵ Poli , C. L. , Dudek , S. J. , Jeffrey , B. M. , Robertson , E. P. , & Fletcher , R. J . ( 2020 ). Recent breeding range expansion of the endangered Snail Kite ( Rostrhamus sociabilis ) in Florida . The Wilson Journal of Ornithology , 132 ( 1 ), 183 . doi: 10.1676/1559-4491-132.1.183 OpenUrl CrossRef 53. ↵ Poli , C. L. , Robertson , E. P. , Martin , J. , Powell , A. N. , & Fletcher , R. J . ( 2022 ). An invasive prey provides long-lasting silver spoon effects for an endangered predator . Proceedings of the Royal Society B: Biological Sciences , 289 ( 1977 ), 20220820 . doi: 10.1098/rspb.2022.0820 OpenUrl CrossRef PubMed 54. ↵ Ponciano , J. M. , Taper , M. L. , Dennis , B. , & Lele , S. R . ( 2009 ). Hierarchical models in ecology: Confidence intervals, hypothesis testing, and model selection using data cloning . Ecology , 90 ( 2 ), 356 – 362 . doi: 10.1890/08-0967.1 OpenUrl CrossRef PubMed Web of Science 55. R Core Team. ( 2024 ). R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing. https://www.R-project.org 56. ↵ Reichert , B. E. , Fletcher , R. J. , & Kitchens , W. M . ( 2021 ). The demographic contributions of connectivity versus local dynamics to population growth of an endangered bird . Journal of Animal Ecology , 90 ( 3 ), 574 – 584 . doi: 10.1111/1365-2656.13387 OpenUrl CrossRef PubMed 57. ↵ Reichert , B. E. , Kendall , W. L. , Fletcher , R. J. , & Kitchens , W. M . ( 2016 ). Spatio-temporal variation in age structure and abundance of the endangered snail kite: Pooling across regions masks a declining and aging population . PLoS ONE , 11 ( 9 ), e0162690 . doi: 10.1371/journal.pone.0162690 OpenUrl CrossRef PubMed 58. ↵ Robertson , E. P. , Fletcher , R. J. , & Austin , J. D . ( 2017 ). The causes of dispersal and the cost of carry-over effects for an endangered bird in a dynamic wetland landscape . Journal of Animal Ecology , 86 ( 4 ), 857 – 865 . doi: 10.1111/1365-2656.12676 OpenUrl CrossRef PubMed 59. ↵ Robinson , W. D. , Errichetti , D. , Pollock , H. S. , Martinez , A. , Stouffer , P. C. , Shen , F.-Y. , & Blake , J. G. ( 2021 ). Big bird plots: Benchmarking Neotropical bird communities to address questions in ecology and conservation in an era of rapid change . Frontiers in Ecology and Evolution , 9 , 697511 . doi: 10.3389/fevo.2021.697511 OpenUrl CrossRef 60. ↵ Royle , J. A. , Kéry , M. , Gautier , R. , & Schmid , H . ( 2007 ). Hierarchical spatial models of abundance and occurrence from imperfect survey data . Ecological Monographs , 77 ( 3 ), 465 – 481 . doi: 10.1890/06-0912.1 OpenUrl CrossRef Web of Science 61. ↵ Simmons , B. I. , Balmford , A. , Bladon , A. J. , Christie , A. P. , De Palma , A. , Dicks , L. V. , Gallego- Zamorano, J., Johnston, A., Martin , P. A. , Purvis , A. , Rocha , R. , Wauchope , H. S. , Wordley , C. F. R. , Worthington , T. A. , & Finch , T . ( 2019 ). Worldwide insect declines: An important message, but interpret with caution . Ecology and Evolution , 9 ( 7 ), 3678 – 3680 . doi: 10.1002/ece3.5153 OpenUrl CrossRef PubMed 62. ↵ Staples , D. F. , Taper , M. L. , & Shepard , B. B . ( 2005 ). Risk-based viable population monitoring . Conservation Biology , 19 ( 6 ), 1908 – 1916 . doi: 10.1111/j.1523-1739.2005.00283.x OpenUrl CrossRef 63. ↵ Stillman , A. N. , Howell , P. E. , Zimmerman , G. S. , Bjerre , E. R. , Millsap , B. A. , Robinson , O. J. , Fink , D. , Stuber , E. F. , & Ruiz-Gutierrez , V. ( 2023 ). Leveraging the strengths of citizen science and structured surveys to achieve scalable inference on population size . Journal of Applied Ecology , 60 ( 11 ), 2389 – 2399 . doi: 10.1111/1365-2664.14512 OpenUrl CrossRef 64. ↵ Strimas-Mackey , M. , Hochachka , W. M. , Ruiz Gutierrez , V. , Robinson , O. J. , Miller , E. T. , Auer , T. , Kelling , S. , Fink , D. , & Johnson , A . ( 2023 ). Best practices for using eBird data . doi: 10.5281/zenodo.3620739 OpenUrl CrossRef 65. ↵ Strimas-Mackey , M. , Miller , E. , & Hochachka , W. ( 2018 ). auk: eBird Data Extraction and Processing in R . https://cornelllabofornithology.github.io/auk/ 66. ↵ Taper , M. L. , Staples , D. F. , & Shepard , B. B . ( 2008 ). Model structure adequacy analysis: Selecting models on the basis of their ability to answer scientific questions . Synthese , 163 ( 3 ), 357 – 370 . doi: 10.1007/s11229-007-9299-x OpenUrl CrossRef 67. ↵ Trouillier , M. , Meyer , K. M. , Santini , L. , & Pe’er , G . ( 2023 ). A comparison of population viability measures . Ecology and Evolution , 13 ( 1 ), e9752 . doi: 10.1002/ece3.9752 OpenUrl CrossRef 68. ↵ Venables , W. N. , & Ripley , B. D . ( 2002 ). Modern Applied Statistics with S (Fourth Edition) . Springer International Publishing . 69. ↵ Wickham , H. ( 2016 ). ggplot2: Elegant Graphics for Data Analysis . Springer-Verlag. https://CRAN.R-project.org/package=terra 70. ↵ Wickham , H. . , Averick , M. , Bryan , J. , Chang , W. , McGowan , L. D. , François , R. , Grolemund , G. , Hayes , A. , Henry , L. , Hester , J. , Kuhn , M. , Pedersen , T. L. , Miller , E. , Bache , S. M. , Müller , K. , Ooms , J. , Robinson , D. , Seidel , D. P. , Spinu , V. , … Yutani , H . ( 2019 ). Welcome to the tidyverse . Journal of Open Source Software , 4 ( 43 ), 1686 . OpenUrl CrossRef 71. ↵ Williams , B. K. , & Brown , E. D . ( 2019 ). Sampling and analysis frameworks for inference in ecology . Methods in Ecology and Evolution , 10 ( 11 ), 1832 – 1842 . doi: 10.1111/2041-210X.13279 OpenUrl CrossRef 72. ↵ Yackulic , C. B. , Nichols , J. D. , Reid , J. , & Der , R . ( 2015 ). To predict the niche, model colonization and extinction . Ecology , 96 ( 1 ), 16 – 23 . doi: 10.1890/14-1361.1 OpenUrl CrossRef PubMed 73. ↵ Zhao , Q. , Latif , Q. S. , Nuse , B. L. , Pavlacky , D. C. , Kilner , C. L. , Ryder , T. B. , & Latimer , C. E . ( 2024 ). Integrating counts from rigorous surveys and participatory science to better understand spatiotemporal variation in population processes . Methods in Ecology and Evolution , 2041-210X.14368. doi: 10.1111/2041-210X.14368 OpenUrl CrossRef View the discussion thread. Back to top Previous Next Posted June 22, 2025. Download PDF Supplementary Material Email Thank you for your interest in spreading the word about bioRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Monitoring population extinction risk with community science data Message Subject (Your Name) has forwarded a page to you from bioRxiv Message Body (Your Name) thought you would like to see this page from the bioRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Monitoring population extinction risk with community science data Orlando Acevedo-Charry , José M. Ponciano , Caroline L. Poli , Brian M. Jeffery , Robert J. Fletcher Jr. , María Ángela Echeverry-Galvis , Bette A. Loiselle , Scott K. Robinson , Miguel A. Acevedo bioRxiv 2025.06.19.660569; doi: https://doi.org/10.1101/2025.06.19.660569 Share This Article: Copy Citation Tools Monitoring population extinction risk with community science data Orlando Acevedo-Charry , José M. Ponciano , Caroline L. Poli , Brian M. Jeffery , Robert J. Fletcher Jr. , María Ángela Echeverry-Galvis , Bette A. Loiselle , Scott K. Robinson , Miguel A. Acevedo bioRxiv 2025.06.19.660569; doi: https://doi.org/10.1101/2025.06.19.660569 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Ecology Subject Areas All Articles Animal Behavior and Cognition (7618) Biochemistry (17635) Bioengineering (13859) Bioinformatics (41846) Biophysics (21401) Cancer Biology (18534) Cell Biology (25422) Clinical Trials (138) Developmental Biology (13352) Ecology (19860) Epidemiology (2067) Evolutionary Biology (24285) Genetics (15582) Genomics (22463) Immunology (17700) Microbiology (40298) Molecular Biology (17141) Neuroscience (88424) Paleontology (666) Pathology (2825) Pharmacology and Toxicology (4813) Physiology (7633) Plant Biology (15107) Scientific Communication and Education (2042) Synthetic Biology (4284) Systems Biology (9808) Zoology (2267)