Long-term population monitoring reveals an alarming decline in the endangered endemic Réunion Harrier Circus maillardi

Long-term population monitoring reveals an alarming decline in the endangered endemic Réunion Harrier Circus maillardi | 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 Long-term population monitoring reveals an alarming decline in the endangered endemic Réunion Harrier Circus maillardi View ORCID Profile Alexandre Villers , View ORCID Profile Thomas Cornulier , View ORCID Profile Cyril Eraud , Thibaut Couturier , Pierrick Ferret , Damien Chiron , Nicolas Laurent , View ORCID Profile Aurélien Besnard , View ORCID Profile Vincent Bretagnolle , Marc Salamolard , Valérie Grondin , François-Xavier Couzi , View ORCID Profile Rémi Fay , Steve Augiron doi: https://doi.org/10.1101/2025.11.07.687144 Alexandre Villers 1 Office Français de la Biodiversité, Direction de la Recherche et de l’Appui Scientifique, Service Conservation et Gestion des Espèces à Enjeux , Station de Chizé, 79360 Villiers-en-Bois Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Alexandre Villers Thomas Cornulier 2 Biomathematics and Statistics Scotland, The James Hutton Institute , Aberdeen, AB15 8QH, Scotland, UK Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Thomas Cornulier Cyril Eraud 1 Office Français de la Biodiversité, Direction de la Recherche et de l’Appui Scientifique, Service Conservation et Gestion des Espèces à Enjeux , Station de Chizé, 79360 Villiers-en-Bois Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Cyril Eraud Thibaut Couturier 3 CEFE, Univ Montpellier, CNRS, EPHE-PSL University , IRD, Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Pierrick Ferret 4 SEOR, 13 Ruelle des Orchidées , 97440 Saint-André, La Réunion, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Damien Chiron 4 SEOR, 13 Ruelle des Orchidées , 97440 Saint-André, La Réunion, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Nicolas Laurent 4 SEOR, 13 Ruelle des Orchidées , 97440 Saint-André, La Réunion, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site Aurélien Besnard 3 CEFE, Univ Montpellier, CNRS, EPHE-PSL University , IRD, Montpellier, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Aurélien Besnard Vincent Bretagnolle 5 CEBC-CNRS, UMR 7372, CNRS & La Rochelle Université , 405 route de Prissé-la-Charrière, 79360 Villiers-en-Bois Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Vincent Bretagnolle Marc Salamolard 6 Parc National de La Réunion, Service Préservation des Patrimoines Naturels , 258 rue de La République, 97431 La Plaine des Palmistes, La Réunion Find this author on Google Scholar Find this author on PubMed Search for this author on this site Valérie Grondin 7 2, sentier du Bras Canot , 97 427 Etang Salé, La Réunion Find this author on Google Scholar Find this author on PubMed Search for this author on this site François-Xavier Couzi 8 LPO France, Fonderies Royales, 8 rue du Dr Pujos , CS90263, 17305 Rochefort Find this author on Google Scholar Find this author on PubMed Search for this author on this site Rémi Fay 9 Laboratoire de Biométrie et de Biologie Evolutive, Equipe Santé Publique Vétérinaire , 43 bd du 11 novembre 1918, 69622 Villeurbanne cedex Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Rémi Fay Steve Augiron 4 SEOR, 13 Ruelle des Orchidées , 97440 Saint-André, La Réunion, France Find this author on Google Scholar Find this author on PubMed Search for this author on this site For correspondence: augironstv{at}seor.fr Abstract Full Text Info/History Metrics Data/Code Preview PDF SUMMARY Insular species are disproportionately vulnerable to extinction due to small population sizes, geographic isolation, and high susceptibility to stochastic and anthropogenic pressures. The endemic Réunion Harrier ( Circus maillardi ), the last breeding raptor on La Réunion (Indian Ocean), is currently listed as “endangered” (EN) on the IUCN Red List. Concerns have recently emerged regarding the demographic impacts of secondary poisoning from anticoagulant rodenticides used for rat control in agricultural landscapes, in addition to other threats such as habitat loss and fragmentation, and reduced genetic diversity. Here, we analysed data from three standardized island-wide breeding censuses (1998–2000, 2009–2010, 2017–2019), encompassing 355 sampling points and totalising > 1,500 h of observations, to quantify overall population trends and spatial variation within those. Using generalized additive mixed models accounting for spatial variation in the different census designs, we estimated a significant decrease of -46% (95% CI: -61% to -28%) in the relative abundance of breeding pairs at the island scale over 21 years (equivalent to three generations), which supports its current endangered (EN) status. Decline was more pronounced in the eastern part of the island (-56%, 95% CI: -69% to -40%) compared to the western part (-29%, 95% CI: -52% to +0.3%). Combined with previous work on inbreeding and mutational load, exposure to anticoagulant rodenticides and ongoing threats (habitat changes), these results suggest this harrier population is undergoing an extinction vortex that requires immediate and targeted, spatially explicit, conservation actions. We also recommend that population size continues to be monitored and that knowledge of demographic parameters be improved to guide adaptive management for the species. INTRODUCTION The current biodiversity crisis has led to the extinction of hundreds of species during the past 500 years, and hundreds more are on the verge of extinction ( Ceballos et al ., 2017 ; Ceballos & Ehrlich, 2023 ). While this crisis is affecting all taxa across the globe (see e.g. Alroy, 2015 ; Régnier et al ., 2015 ; McClure et al ., 2018 ), some geographical areas such as islands are more sensitive to extinction risk ( Wood et al ., 2017 ). As a whole, islands account for 61% of known extinctions and 37% of critically endangered species worldwide ( Tershy et al ., 2015 ). Indeed, insular communities are often characterised by low population sizes and geographical isolation, both increasing not only the sensitivity of those populations to environmental stochasticity ( Lande, 1998 ) but also exposing them to genetic drift ( Funk et al ., 2016 ). The latter can result in inbreeding, increasing the emergence and persistence of deleterious mutations ( Keller & Waller, 2002 ). All these mechanisms can have detrimental effects on the viability of insular animal populations, which can then be reinforced by anthropogenic pressures, such as the introduction of invasive species Blackburn et al . (2014) ; Doherty et al . (2021) and habitats modification ( Gillespie et al ., 2008 ). La Réunion, a 2,500 km 2 island located in the Indian Ocean, belongs to the Madagascar hotspot, one of the 25 areas of biodiversity importance identified by Myers et al . (2000) , characterized by a high level of endemism. However, since human occupation of the island initiated in the 18th century, there has been unusually rapid habitat loss, fragmentation, and modification ( Strasberg et al ., 2005 ), as well as the introduction of numerous invasive plant and animal species ( Baret et al ., 2006 ), causing harmful effects to native fauna and flora, notably through predation by feral cats Felis catus and black rats Rattus rattus , or competition and habitat modification by invasive plants such as Miconia calvescens . Thus, during the past 300 years, 73% of the native vegetation was transformed ( Lagabrielle et al ., 2009 ), and 51% of known native bird species were extirpated from the island ( Mourer-Chauviré et al ., 2006 ). Today, nine endemic bird species are still present on the island, the Réunion Harrier Circus maillardi being the only breeding bird of prey remaining on La Réunion ( Safford & Hawkins, 2020 ). This raptor faces similar threats as other birds of prey around the globe ( O’Bryan et al ., 2022 ), such as poaching ( Thomason et al ., 2023 ), collisions with power lines, wind turbines (Bellebaum et al. 2013, Slater et al. 2020) or vehicles ( Bullock et al ., 2024 ). Additionally, secondary poisoning from Anticoagulant Rodenticides (ARs, Coeurdassier et al ., 2019 ), used to control rats’ populations in farmlands, was considered as one of the most severe threats (BirdLife International 2021), given rodents constitute a large proportion of harriers’ diet. Recent genetic analyses also highlighted a high degree of inbreeding which may result in compromised genetic health and reduced adaptative potential ( Bourgeois et al ., 2024 ). In this context, Fay et al . (2023) developed a demographic model of the species’ dynamics, and their results, based on parameters’ values estimated empirically, indicated that the population was declining at an alarming rate, which could eventually lead to a quasi-extinction in the near future (c. 40 years). This work relies however on fragmentary demographic data, principally originating from the eastern part of the island, and therefore the predicted population trajectory requires validation, to precisely quantify ongoing population changes. The absence of reliable data on recent population trends has hindered the objective reassessment of the species’ Red List status. Currently, the IUCN endangered (EN) status of the Réunion Harrier relies on criteria related to its geographic range and habitat (criteria B1ab(iii)), as well as on a population size estimated to number fewer than 250 mature individuals (criteria D), but the decrease in population is only suspected. In 2022, a national conservation action plan (2022-2031) was initiated, which aims to improve the conservation status of the species through the reduction of identified threats. The action plan anticipates the latter to be based upon sound knowledge acquired on its ecology and population dynamics. However, the different surveys that were conducted in the past 25 years differed slightly regarding sampling schemes and protocols, which may lead to hazardous conclusions regarding trends that would be based on raw data. In this context, our study aims at assessing changes in abundance of the Réunion Harrier breeding population over the past 20 years and relies on data collected during three extensive censuses conducted respectively in 1998-2000, 2009-2010, and 2017-2019. Considering results of a genetic studies ( Bourgeois et al ., 2024 ) as well as the high levels of exposure to anticoagulant rodenticides mainly used in human-modified environments ( Coeurdassier et al ., 2019 ), we strongly suspect the presence of geographical contrasts in demographic functioning. Accordingly, we refined our modelling approach by explicitly considering spatial structure in population trend, which could help understanding the underlying demographic processes at play, as well as to spatializing conservation actions. MATERIAL & METHODS Study species The Réunion Harrier is a medium-sized raptor (700 – 1,000g), which was also formerly present on Mauritius but that disappeared soon after the arrival of the first European settlers ( Cheke, 2010 ). It is closely related to the Malagasy Harrier Circus macrosceles , that breeds on Madagascar, from which it diverged c.a. 0.1-0.3 mya ago (0.05 - 0.5 CI, see Oatley et al ., 2015 ). On la Réunion, the species can be encountered throughout the island, with most breeding pairs located between the coast and 1,200m ( Augiron, 2022 ). It originally exploited native forests ( Clouet, 1978 ) that covered most of the island but adapted to the degradation and modification of native habitats by humans. This was achieved, for example, by exploiting the low vegetation of secondary forests as breeding sites, or by preying upon new food resources, such as introduced mammals, which have proliferated significantly in most heavily modified habitats. The black rat in particular thrives in agricultural landscapes, especially in sugar cane fields and near cattle farms, and thus constitutes a significant portion of the harrier’s diet ( Augiron, 2022 ). The most recent census suggests a population size of approximately 150 breeding pairs, and approximately 560 mature individuals in 2009-2010 ( Grondin & Philippe, 2011 ), while Ghestemme et al . (1998) reported a population of 120-180 breeding pairs. Data collection Three censuses were conducted over the past two decades: in 1998-2000 ( Bretagnolle et al ., 2000 ; Ghestemme et al ., 1998 ), 2009-2010 ( Grondin & Philippe, 2011 ) and 2017-2019 ( Chiron & Augiron, 2019 ), hereafter referred to as 1998, 2009 and 2017 surveys. All censuses shared a common field methodology, based on counts of individuals on a set of sampling points (hereafter “SP”) distributed over the island in different habitats (see Fig. 1 ). Download figure Open in new tab Figure 1: Locations of observations points for the three different censuses (1998-2000, 2009-2010, 2017-2019) on La Réunion. The colour of the points indicates whether a breeding pair was observed (red) or not (blue). Point size is proportional to the maximum number of pairs detected at each sampling point during each census. Polygons delineate the eastern and western areas used for trend comparisons. Contour lines represent altitude isopleths every 100 meters, computed from the BD ALTI 2.0 (©IGN) SPs were located near paths, roads, forest tracks or mountain trails to facilitate the accessibility for the observers, and on high observation posts in order to guarantee a large field of view ( Couturier et al ., 2022 ). Over the three censuses, 355 unique SPs were monitored at least once, with 37.5% (n=133) and 15.5% (n=55) of those points that were common to two and three censuses, respectively. On each SP, one or repeated observation sessions (hereafter session) were performed. The total number of sessions was 341 in 1998, 185 in 2009, and 522 in 2017. Each SP was, on average, sampled three times (range: 1-8) over the three census periods. Survey conducted in 2017 had a more extensive coverage of the island, and notably embraced a wider altitudinal gradient than previous censuses. Average altitude was 798 m (range: 2m – 2482 m) in 1998, 756 m (range: 2m – 2327m) in 2009 and 856 m (range: 2m – 2482m) during the last survey. In 2017, if the visual range on a SP, previously surveyed in 1998 and/or 2009, had been obstructed by vegetation or a newly constructed building, the observer was allowed to relocate the SP within a 200 m radius of the original location, and the identity of the SP was preserved over the 3 censuses. Otherwise, a new identity was attributed to the SP. Sessions on each SP started on average 217 ± 103 minutes (SD) after sunrise ( hours). All observations were conducted only during favourable weather conditions (no or low wind and no rainfall or very light showers). If the weather conditions were detrimental to the activity of the harriers or to their detection (fog, heavy showers, strong gusts of wind), the session was stopped, and the interruption duration was subtracted to only consider the duration dedicated to observation. In case the conditions did not improve, the session was cancelled, and a new one was scheduled in the following couple of weeks. The length of observation differed between the first two censuses (120 minutes in 1998 and 2009), and the last one (80 minutes). Overall, the duration of observations ranged from 25 to 300 minutes, with an average duration of 112 ± 26 min in 1998, 114 ± 26 min in 2009 and 80 ± 0 min in 2017. During the first two censuses, the number of observers was limited, and all participants were professionals from conservation organizations. For the third census, a citizen science approach was implemented to significantly increase the number of sampling points ( Couturier et al., 2022 ). Except for salaried employees or experienced ornithologists, all participants received both theoretical and practical training before their participation. This mandatory training helped to standardize ornithological skills (such as identifying sex, age, and the various reproductive behaviours) and ensured a clear understanding of the objectives and the protocol of data collection. Information gathered on the breeding biology of the species before 2010 suggested that the breeding activity peaked between the months of April and July ( Grondin & Philippe 2011 ), but at the time of the first census, such information was unavailable. Consequently, observations conducted in 1998 spanned over the whole year, and only 33% of the samples were collected between April and July. Conversely, 2009 and 2017 censuses focused specifically on this period, with 95% and 84% of samples collected during the peak of breeding activity for the two last censuses respectively. During each session, all birds that could be detected with the naked eye from the SP were reported on a field map ( Couturier et al., 2022 ). Additionally, binoculars were used to determine their sex/age, and the potential presence of distinctive signs (plumage coloration, moult, or wing tags), which would make it easier to differentiate individuals spotted in subsequent observations. For all surveys, if one or more individuals exhibited reproductive behaviour, the observer reported it along with the time and location of the event, the sex and age of the birds concerned (according to the possible combinations proposed for each behaviour see Tab. 1 ), and noted the presence of distinctive signs if applicable. The observer then assigned an identifier to the individual or pairs. When the same individuals or pairs were re-observed during one observation session, the same identifier was used to keep track of individual information. In order to remain conservative, an individual or pair was considered a new one only if i) wingtags or the plumage characteristics ensure the identification of a different individual e.g. due to coloration variation, the observation of missing feather(s) or a moult in progress, or the presence of reproductive behaviour in an area without prior observation and if ii) individuals were observed simultaneously, or within a very short period of time at locations far enough apart to exclude double counts. Observers recorded all individuals or pairs with one or more behaviours indicating reproduction following the criteria used to classify the degree of breeding evidence (see Tab. 1 ). At the end of the observation session, an estimate of the minimum number of certain, probable and possible pairs was determined by the observer. Only the “probable” and “certain” categories were used in our analysis, since the category “possible” was not available for the first census. View this table: View inline View popup Download powerpoint Table 1: Standardised criteria applied across all censuses to classify the degree of breeding evidence. For the analyses, only those observations categorised as “certain” or “probable” breeding were retained Statistical analysis In order to accurately estimate the breeding population trends of La Réunion Harrier, we needed to take into account the different sampling schemes that were deployed during the three censuses and which varied both in space, as the location of sampled points changed between censuses, and in time, as some points were sampled only during one of the three censuses, while others were sampled multiple times over the study period. This was achieved thanks to Generalized Additive Mixed Models ( Wood, 2017 , package mgcv 1.9-3 ), which allow modelling non-linear relationships and are especially convenient for estimating spatial dependency. This approach was particularly suited to the data set generated by the three censuses, with varying sampling effort. The response variable was the number of breeding pairs observed ( ptot , i.e. pairs recorded as probably or certainly breeding) at each point and each observation session, interpreted here as a relative abundance index. The number of breeding pairs was modelled as a function of Date , log observation Effort , coordinates of sampling points ( X, Y ), year as factor ( Y earF ) and site identity ( points) as factor ( SiteID ): where f j is a smooth function of covariates ( Wood, 2017 ). More specifically, the date of the observation session was expressed as a Julian date variable “Date”, with 1 st of January corresponding to day 1. This term was specified with a cyclic cubic regression spline, in order to account for the fact that the 31 st of December and the 1 st of January are equivalent from a biological point of view, e.g. in term of the phase of the breeding cycle or expected climatic conditions at this time of the year. The value for the dimension of the basis function was set to k = 5 to match a-priori expectations and limit risk of overfitting. The length of the observation session (log-transformed), “Effort”, was modelled with a thin plate regression spline with dimension also set to a maximum of k = 5. Spatial smoothers i.e. an interaction between latitude and longitude, were also included in f 3 and f 4 to model potential spatial autocorrelation. This term accounted for both the local specificities (e.g. habitat, abiotic conditions) and heterogeneity in the sampling design, in particular the fact that the last survey covered a wider array of environmental contexts. Details on the parametrisation of this term can be found in the supplementary material section (S1) but the retained model included the main interaction between latitude and longitude, as well as smooth terms modelling deviations from the main spatial smooth for each survey SurveyF . Two random intercepts were also included, namely Y earF , a 9-level random effect accounting for yearly variations of abundance, and SiteID , a 346-level random effect accounting for the identity of sampling points, with identity shared between and within surveys to account for inter- and intra-annual repeated counts. We assume a Tweedie distribution (with a log link) that is a probability distribution encompassing the continuous (normal, gamma and inverse Gaussian), discrete (Poisson) and compound (Poisson-gamma) distributions. This distribution has three parameters, the mean ( µ ), a dispersion parameter ( ϕ ) and a power parameter ( p ). In the mgcv package, the value of p is restricted to ]1; 2[, and is estimated simultaneously with other parameters, offering flexibility in the shape of the probability distribution assumed for the response, particularly with respect to overdispersion and excess zeros, when needed. To compute relative population trends, we used model predictions per sampling point, conditional on survey and location, and other covariates fixed at constant values. This can be interpreted as a population index proportional to density, with an arbitrary scale. For the associated confidence limits, we adopted a posterior simulation approach following ( Wood & Goulson, 2017 , chapter 6.10, p. 293). The general idea is to combine predictions and simulations to produce distribution estimates (i.e. including uncertainty of parameters’ estimate) of a quantity of interest, in our case the relative abundance index of breeding pairs of Réunion Harrier. First, using the lpmatrix option in the predict.gam () function, we generated a model matrix X p for prediction. Values of explanatory variables for computing X p were set as follows: coordinates were those of the centroid of each 1km x 1km grid cell, Date and Effort were set at their respective means, random effects Y earF and SiteID were set to a fixed arbitrary level of each factor. This setup was replicated for all surveys (indicated by column SurveyF ), to generate predictions of relative abundance for each survey. Second, we generated a B r matrix where each line was a set of model parameters, corresponding to columns of X p , and each column was a replicate parameter vector. Parameter replicates (n=1000) were simulated from their posterior distribution using the mgcv::rmvn() function, with the variance-covariance matrix estimated with the mgcv :: vcov () function. Finally, the matrix P , containing for each grid cell and survey, 1000 sets of simulated abundance values, was computed such that: Trends were expressed as: To estimate population trends between surveys at the island level, we averaged, for each replicate, the value of the abundance index for all grid cells covering the island. We then compared the trends in population between pairs of surveys following Eq.(3). To represent spatial changes in trends between surveys, we estimated for each pair of surveys and for each grid cell the 95% confidence intervals around the mean trend value computed with Eq.(3), with the stats :: quantile () function, at 2.5% and 97.5% percentiles. Bird et al . (2020) estimated the generation time of the Réunion Harrier at 6.84 years. Thus, the period of 21 years over which the trend was computed corresponds to the criteria of a change over 3 generations (20.52 years), used by the IUCN Red List. From the trends computed over the three surveys (Eq.3), we estimated the yearly population growth rate λ over the 1998-2019 periods, i.e. 21 years, as well as the mean yearly change (see S3 & S4 ). Models were fitted with R 4.3.2 ( R Core Team, 2025 ) using the mgcv package ( Wood, 2017 ). Following Wood (2011) we used the REML method, instead of the default Generalized Cross Validation (GCV) approach, to estimate smoothing parameters. Model fit was visually assessed using the mgcv :: gam.check () and the gratia :: appraise () functions (see S2) and model residuals were also inspected to detect any problematic spatial pattern (spatial correlation of residuals was estimated with the ncf :: sp.correlog () function ( Bjornstad, 2022 ), see S5 & S6). Overdispersion value of the retained model, based on Pearson residuals, was 0.88 and we did not consider this slight underdispersion as problematic for parameters interpretation ( Zuur et al ., 2009 ). We used the packages dplyr 1.1.4 ( Wickham et al ., 2022 ), sf 1.0-21 ( Pebesma & Bivand, 2023 ), terra 1.8-60 ( Hijmans, 2025 ), tidyterra 0.7.2 ( Hernangómez, 2023 ) and openxlsx 4.2.8 ( Schauberger & Walker, 2014 ) to import and format the data, suncalc 0.5.1 ( Thieurmel & Elmarhraoui, 2022 ) to compute sunrise schedule on la Réunion, the gratia 0.11.1 package ( Simpson, 2024 ) to assess and plot model fit of the GAM, and the following packages ggplot2 4.0.0 ( Wickham, 2016 ), viridis 0.6.5 ( Garnier et al ., 2024 ), ggpubr 0.6.1 ( Kassambara, 2025 ), ggridges 0.5.7 ( Wilke, 2025 ), to produce the different figures. RESULTS Variations of breeding pairs abundances in space and time The fitted model ( Tab. 2 ) indicated a quadratic-like effect of the Date, with the recorded number of breeding pairs per point peaking in mid-June ( Fig. 2a ), and a clear positive effect of the Effort , i.e. length of observation sessions ( Fig. 2b ). View this table: View inline View popup Download powerpoint Table 2: Summary of the model retained to compute abundance index and population trends Download figure Open in new tab Figure 2: Partial effect for the smoother on Julian date (a.) and duration of the observation (b.) from the selected model The main spatial term f ( X, Y ) identified areas of higher abundance, located mostly on the periphery of the island, at intermediate altitudes ( Fig. 3, a & b ). The term modelling the interaction between spatial coordinates and each level of the SurveyF factor f ( X, Y, surveyF ) suggested a complete shift of a longitudinal gradient of abundance over the three surveys ( Fig. 3, c & d ). This change was most visible with the predicted abundance ( Fig. 4a ), which integrates both f ( X, Y ) and f ( X, Y, surveyF ) terms, highlighting the more pronounced decline in the eastern part of the island than in the west. The uncertainty in abundance estimates ( Fig. 4b ) increased in more sparsely sampled areas, such as for the narrow band along the eastern coast where few sample points were conducted, especially during the first census. Download figure Open in new tab Figure 3: a. Partial effect of the main spatial smoother s(X, Y ); b. associated standard errors; c. partial effect of the census-level specific spatial smoother s(X, Y, surveyF ); and d. the associated standard error of this term across the three censuses Download figure Open in new tab Figure 4: a. Spatial distribution of the predicted abundance index for the three censuses; and b. associated standard error of the mean abundance. Red dots indicate the spatial distribution of the sampling points of each census Population trends When comparing changes in the abundance index at the island level (see Fig. 5 a & b ), we found a strong decline between the 1998 and the 2009 census (-32.8% CI95% [-51.5%; -11.1%]), and a more uncertain, and likely weaker decline between 2009 and 2017 with a change of -18.1% CI95% [-43%; +16%]. Between the 1998 and the 2017 survey (-45.8% CI95% [-60.9%;-28.1%]). Over the 12 years period separating the first two surveys, the averaged yearly decline was estimated at -3.2%. Over the 21 years covering the three surveys, the yearly population change reached -2.8%, corresponding to a decline of -44.2% over three generations. Trends computed for the eastern and western parts of the island revealed a decline twice greater for the eastern part over the study period (-56.2% CI95% [-69.2%; -40.2%]), compared to the likely smaller -29% CI95% [-52.4%; +0.3%] decline for the western area. For the eastern part of the island, this translates into a -3.8% average annual decline from 1998 to 2019. Download figure Open in new tab Figure 5: Predicted values of the abundance index, averaged from the 1,000 simulations, for the whole island (left) and for the western and eastern areas. Thin lines connect predicted values across surveys for a given simulation. The dots and associated error bars represent the mean and 95% interval for each survey’s 1000 simulated values. b. Distributions of predicted changes between pairs of censuses for the 1,000 simulations. Mean values and their confidence intervals are shown for each comparison to the right of each distribution. In panel b, “1998” corresponds to the first census (1998-2000), “2009” to the second (2009-2010) and “2017” to the third (2017-2019) DISCUSSION The present work fills an important gap regarding the conservation of the endemic Réunion Harrier, the last bird of prey on this island, by providing the first objective estimation of the breeding population’s trend. Raptors are often considered flagship species, yet many remain poorly studied, and this work also contributes to addressing this global data gap ( Sergio et al ., 2008 ; McClure et al ., 2018 ). This information is indeed crucial to evidence the conservation status of this species, as it is one of the several parameters considered when this status is assessed. At the island level, the decline estimated through the relative change in abundance indices reaches -45% over the period 1998-2017, corresponding to three generations for this species. This value alone reinforces the ‘Endangered’ (EN) status assigned to C. maillardi, as the confidence interval of the population trend over three generations encompasses the -50% threshold of criteria A2b of the IUCN guidelines (IUCN 2012). Moreover, the relative abundance index displays contrasted patterns in terms of population dynamics, with the eastern part of the island presenting a much steeper decline than the western part between the two first censuses (1998-2009), absolute values being almost twice as large (-56% vs -29%). Initially 25% higher in the East, the breeding index is now 15% lower than in the West, illustrating a marked spatial inversion. This suggests that the decline observed at the island level was mostly driven by a collapse of the breeding population of la Réunion Harrier in the eastern part of the island, even if the population breeding in the western part of the island also decreased, though less markedly. While the decrease is significant between 1998 and 2009, it is less pronounced, and more uncertain, for the 2009-2017 period. The most likely value of the trend for this period remains negative, which could be indicative of the persistence of factors acting detrimentally upon the species. These results are based on data collected during three island-wide censuses, spanning 21 years, and a robust statistical modelling approach that accounts for heterogeneity in observation durations and dates, as well as the spatial heterogeneity in the sampling design. The empirical evaluation of population dynamics conducted in the present work is in agreement with the demographic parameters reported by Fay et al . (2023) . Their work indeed demonstrated that the species was characterised by a low fecundity, resulting from small clutch sizes and a very low hatching success compared to other harrier or bird species in general, as well as a low adult survival for a bird of that size. While these values threaten the viability of the Réunion harrier, the causes of these low demographic rates remains challenging to identify, and at best hypothetical. Bourgeois et al . (2024) highlighted that this species displays an overall very low genetic diversity, and one of the highest mutational loads ever recorded in a wild bird population, resulting from a severe bottleneck this population experienced over the last few millennia. During that period, the effective population size dropped from 3,000 to 300 individuals, with a sharper decline in the past century. In several vertebrate species, such erosion of genetic diversity has been shown to decrease individual fitness, with measurable effects on survival or breeding parameters ( Kruuk et al ., 2002 ; Huisman et al ., 2016 ; Clarke et al ., 2024 ) which can ultimately affect population trends ( Blomqvist et al ., 2010 ). A second factor relates to the exposure to anticoagulant rodenticides (ARs) that could negatively affect the species over the whole island. Since the late 1970’s, these molecules have been widely used in sugar cane fields, but also in other crops and close to human habitations, to control populations of rats. Beyond their impact on agriculture ( Stenseth et al ., 2003 ), rats are also reservoirs for spirochetes bacteria which responsible for leptospirosis ( Guernier et al ., 2016 ), a zoonosis that can sometimes be fatal to humans ( Allyn et al ., 2024 ). Given rats make up a significant portion of the Réunion Harrier’s diet, particularly in agricultural and urban areas ( Augiron, 2022 ), these raptors can therefore become secondarily exposed to ARs, when feeding on contaminated rats. Coeurdassier et al . (2019) showed that rodenticides were massively encountered in harriers found dead, with 62% of the birds displaying a liver concentration in rodenticides known to be lethal in other raptors. Exposure to ARs, which increased between 1999 and 2015, may therefore contribute to additional mortality in both adults and juveniles. Moreover, though the effects of sub-lethal exposure of ARs remains less documented, they are suspected to reduce not only survival rates but also fecundity ( Salim et al ., 2014 ; Gomez et al ., 2022 ). One striking result of the present work is the differential population trends detected between the eastern and western parts of the island. This pattern mirrors the two genetic clusters identified by Bourgeois et al . (2024) , located respectively in the western and eastern regions. While the species presents an overall low genetic diversity, the eastern sub-population exhibits an even lower level. These genetic clusters may originate from restricted dispersal, either in terms of distance or frequency of dispersal events. In general, this pattern may be the result of a high degree of philopatry naturally occurring in the species, or through environmental constrains, such as landscape composition or natural barrier, which can limit movements, even in flying vertebrates such as bats ( Pinaud et al ., 2018 ) or birds ( Martensen et al ., 2008 ). For example, a small passerine, the Réunion Grey White-eye Zosterops borbonicus , is characterised by an exceptionally high level of intra-island genetic differentiation that appears to be mostly shaped by restricted dispersal, and to a lesser extent landscape composition ( Milá et al ., 2010 ; Bertrand et al ., 2014 ). In the Réunion Harrier, very few dispersal events have been documented in this species so far ( Augiron, 2022 ), whether with wing-tagged individuals or through satellite tracking. It remains however unclear whether this limited dispersal is due to habitat features, e.g. the volcanic range separating the eastern and western parts of the island, or a consequence of the extensive habitat changes that have occurred since human settlement. Approximately 50% of the island is now covered by cultivated areas, urban areas, or secondary vegetation ( Strasberg et al ., 2005 ), which may have further disrupted dispersal processes. The loss of key landscape elements, together with the presence of built-up areas, may limit the ability of birds to move across this altered landscape, thereby reinforcing the spatial isolation of genetic clusters. Conversely, no spatialised information is currently available on the usage of rodenticides which would allow to detect a differential utilisation of these molecules between the eastern and the western parts of the island and thus explain the contrasting population trends in these two areas. It is however to be noted that Coeurdassier et al . (2019) found a heterogeneous spatial pattern for exposed individuals, with most exposed birds found in the east. This could suggest, if the probability of discovering a dead or injured harrier is spatially homogeneous, that birds are indeed more exposed in this part of the island. A plausible, though speculative, explanation to this higher exposure risk could reside in the availability of rodents. In effect, rainfall is known to affect rat abundance in tropical regions ( Madsen & Shine, 1999 ), and both areas display striking different precipitation regimes (2,132 mm/year in the east vs 886 mm/year in the west). The eastern side would therefore sustain higher densities of rats (exposed to ARs) that harriers would consume more often than in the west, exposing them to a higher risk of lethal and sub-lethal effects. Taken together, genetic erosion and rodenticide exposure are likely to interact, creating a self-reinforcing process typical of extinction vortices ( Fagan & Holmes, 2006 ), where a high mortality rate reduces population size and dispersal opportunities, thereby limiting genetic diversity, which in turn further decreases fitness and exacerbates demographic decline. To halt the decline and prevent the species from approaching the critical threshold of extinction, it is urgent to suppress or strongly reduce the impact of those suspected detrimental factors. Given the state of conservation of the Réunion Harrier, it seems appropriate to initiate complementary conservation actions simultaneously, while continuing to update the knowledge on its ecology and demography. The a priori most tractable option would be to act upon genetic diversity. Genetic rescue has proven to be efficient in several conservation programs worldwide ( Bouzat et al ., 2009 ; Heber et al ., 2013 ; Wright et al ., 2014 ; Lewanski et al ., 2025 ), though the ecology of the species must be carefully considered ( Nichols et al ., 2024 ), in particular to ensure that a sufficient number of individuals is released. For the Réunion Harrier, this could be achieved through the translocation of individuals ( Martínez-Cruz, 2011 ) either collected from nature, e.g. by switching eggs or chicks between nests located in the east and in the west, or a carefully designed captive breeding program ( Heinrichs et al ., 2019 ; Witzenberger & Hochkirch, 2011 ). On the other hand, reducing the impact of poisoning by rodenticides, which would probably be necessary to ensure the success of any reinforcement program, can only be envisaged as a long-term strategy. In effect, given socio-economic as well as human-health considerations, reducing AR use in a way that preserves harrier populations while limiting the exposition of human populations to leptospirosis, and in the meantime reducing damages to crops, would probably require years of cross-sectoral effort before being implemented. The spatial abundance index produced by this work highlights hotspots for the species, while data from the last census (2017–2019) allow explicit estimation of detection probability and total population size, including the breeding population ( Couturier et al ., 2022 ). These data can be used to produce spatial predictions of densities across the island, that can be shared with land-use and agro-environmental administrative services. Such cartographic outputs could help prioritising conservation actions in space, including experimental rodenticide reduction trials, translocation planning to enhance genetic mixing, and providing evidence for their inclusion in forthcoming prefectural decrees. Maintaining rodenticide use in high-density breeding zones could otherwise have dramatic medium-term impacts on the population, making their protection a priority. At the same time, improving our understanding of demographic variation remains essential, as ecological and demographic data are still scarce even for the world’s most imperilled raptors ( McClure et al ., 2023 ). In this perspective, establishing a new study site in the western region would provide valuable data from a distinct ecological context, thereby helping to identify the factors driving population dynamics ( Robinson et al ., 2014 ), allowing a more robust assessment of threats, predicting more accurate predictions of population trajectories, and designing of adaptive conservation strategies ( Wiens et al ., 2017 ; Winship et al ., 2013 ). The standardised protocol, relying heavily upon citizen involvement, constitute an invaluable tool to keep monitoring population changes in the future, in particular to assess the efficiency of conservation actions that will be implemented. These needs are precisely those anticipated in the National Species Action Plan (2022-2031) for the Réunion Harrier, to which the present work directly contributes. Authors’ contribution Alexandre Villers: Protocol design, Supervision, Data curation, Analytical methodology, Formal analysis, Writing - original draft. Thibaut Couturier: Protocol design, Project management and administration (2017), Data curation, Field data sampling, Reviewing and Editing Pierrick Ferret: Protocol design, Data curation, Field data sampling, Reviewing and Editing Damien Chiron: Protocol design, Project management and administration (2018-2019), Data curation, Field data sampling, Reviewing and Editing Nicolas Laurent: Protocol design, Field data sampling, Design of web interface and data management, Reviewing and Editing Thomas Cornulier: Protocol design, Analytical methodology, Formal analysis, Reviewing and Editing Aurélien Besnard: Protocol design, Analytical methodology, Reviewing and Editing Vincent Bretagnolle: Field data sampling, Data curation, Reviewing and Editing Marc Salamolard: Project management and administration (1998-2000), Field data sampling, Reviewing and Editing Valérie Grondin: Project management and administration (2009-2010), Field data sampling, Data curation, Reviewing and Editing François-Xavier Couzi: Field data sampling, Project Administration, Reviewing and Editing Rémi Fay: Analytical methodology, Reviewing and Editing Cyril Eraud: Analytical methodology, Reviewing and Editing Steve Augiron: Protocol design, Funding acquisition and management, Project management and administration (2016-2019), Field data sampling, Supervision, Data curation, Formal analysis, Writing - original draft. Additional Information Competing financial interests The authors declare no competing financial interests. Access to data and code You can find the data used in this manuscript, the code to analyse these data and to generate the figure at this link. Supplementary material S1 Parametrisation of the spatial term and model specifications We set the value of the basis dimension of the spatial term to k = 100, in order to ensure that a sufficient number of degrees of freedom would be available for this smoother. We compared four different specifications of the spatial term and retained the one that gave a priori the best fit, as Generalized Additive Modelling often requires selecting models based on expert knowledge ( Pedersen et al ., 2019 ). Commonly used method based on AIC to select model prove to be poorly effective. The first model included a surveyF fixed effect factor and a spatial smoother common to the whole period of census with the term s ( X, Y, k = 100). Two models included a smooth factor interaction between coordinates and the census period s ( X, Y, k = 100) + s ( X, Y, SurveyF, bs = ” sz ” , k = 100) or surveyF + s ( X, Y, k = 100) + s ( X, Y, by = SurveyF, m = 1 , k = 100), which allowed to model the deviations from the main spatial term applied to each level of the survey factor, with independent degrees of smoothness / wiggliness for each level. The m = 1 option indicates that the marginal thin plate regression spline basis of this term will penalise the squared first derivative of the function, rather than the second derivative which limits co linearity between the global smoother s ( X, Y ) and the group-specific terms s ( X, Y, by = SurveyF ). Code is available through a github repository, see below. S2 Model fit Download figure Open in new tab Figure 6: Assesment of best model fit S3 Population growth rate and yearly rate of change S4 Mean yearly values of λ computed from the best model View this table: View inline View popup Download powerpoint S5 Spatialised residuals Download figure Open in new tab Figure 7: Residuals values for each of the three censuses S6 Spatial correlograms of residuals Download figure Open in new tab Figure 8: Spatial correlograms of residuals of the selected model ACKNOWLEDGMENT We sincerely thank all the observers, volunteers, and professionals, who took part in the various counts over a period of 20 years, making it possible to compile a dataset that enabled the definition of a temporal indicator of the breeding population, essential for the conservation of this endangered species. We are grateful to the members of the Société d’Etudes Ornithologiques de la Réunion for organizing, simultaneous harrier counts across the entire island. We also extend our thanks to the many funding partners who supported this long-term monitoring programme over the past two decades: the Regional Council of the Réunion (grant), the European Union (FEDER POT and FEDER Grant n°RE0005162), the French Ministry of Environment (DIREN-Réunion), the DEAL Réunion, Aerowatts, the Etang-Salé municipality, EDF and Tereos. The authors are grateful to Dr. C. Morrison for carefully reviewing an earlier version of the manuscript and providing valuable feedback. Funder Information Declared European Union, https://ror.org/019w4f821 , RE0005162 European Fund Footnotes Contacts: Alexandre Villers or Steve Augiron https://github.com/aquila06/Circus_maillairdi_PopulationDecline/tree/415f696c1fe3c6bc8d5adaca51b5ac1fa025050a References ↵ Allyn , J. , Miailhe , A.F. , Delmas , B. , Marti , L. , Allou , N. , Jabot , J. & Reignier , J . ( 2024 ) Severe leptospirosis in tropical and non-tropical areas: A comparison of two french, multicentre, retrospective cohorts . PLOS Neglected Tropical Diseases 18 , e0012084 . OpenUrl ↵ Alroy , J . ( 2015 ) Current extinction rates of reptiles and amphibians . Proceedings of the National Academy of Sciences 112 , 13003 – 13008 . OpenUrl Abstract / FREE Full Text ↵ Augiron , S. ( 2022 ) Plan national d’actions 2022-2031 en faveur du busard de maillard Circus maillardi . Tech. rep., SEOR . ↵ Baret , S. , Rouget , M. , Richardson , D.M. , Lavergne , C. , Egoh , B. , Dupont , J. & Strasberg , D . ( 2006 ) Current distribution and potential extent of the most invasive alien plant species on la réunion (indian ocean, mascarene islands). Austral Ecology 31 , 747 – 758 . OpenUrl ↵ Bertrand , J.A.M. , Bourgeois , Y.X.C. , Delahaie , B. , Duval , T. , García-Jiménez , R. , Cornuault , J. , Heeb , P. , Milá , B. , Pujol , B. & Thébaud , C . ( 2014 ) Extremely reduced dispersal and gene flow in an island bird . Heredity 112 , 190 – 196 . OpenUrl CrossRef PubMed Web of Science ↵ Bird , J.P. , Martin , R. , Akcakaya , H.R. , Gilroy , J. , Burfield , I.J. , Garnett , S.T. , Symes , A. , Taylor , J. , Sekercioglu , C.H. & Butchart , S.H.M . ( 2020 ) Generation lengths of the world’s birds and their implications for extinction risk . Conservation Biology 34 , 1252 – 1261 . OpenUrl CrossRef PubMed ↵ Bjornstad , O.N . ( 2022 ) ncf: Spatial Covariance Functions . R package version 1.3-2. ↵ Blackburn , T.M. , Essl , F. , Evans , T. , Hulme , P.E. , Jeschke , J.M. , Kühn , I. , Kumschick , S. , Marková , Z. , Mrugała , A. , Nentwig , W. , Pergl , J. , Pyšek , P. , Rabitsch , W. , Ricciardi , A. , Richardson , D.M. , Sendek , A. , Vilà , M. , Wilson , J.R.U. , Winter , M. , Genovesi , P. & Bacher , S . ( 2014 ) A unified classification of alien species based on the magnitude of their environmental impacts . PLoS Biology 12 , e1001850 . OpenUrl CrossRef PubMed ↵ Blomqvist , D. , Pauliny , A. , Larsson , M. & Flodin , L.A . ( 2010 ) Trapped in the extinction vortex? strong genetic effects in a declining vertebrate population . BMC Evolutionary Biology 10 , 33 . OpenUrl PubMed ↵ Bourgeois , Y. , Warren , B.H. & Augiron , S . ( 2024 ) The burden of anthropogenic changes and mutation load in a critically endangered harrier from the reunion biodiversity hotspot, circus maillardi . Molecular Ecology 33 . ↵ Bouzat , J.L. , Johnson , J.A. , Toepfer , J.E. , Simpson , S.A. , Esker , T.L. & Westemeier , R.L . ( 2009 ) Beyond the beneficial effects of translocations as an effective tool for the genetic restoration of isolated populations . Conservation Genetics 10 , 191 – 201 . OpenUrl ↵ Bretagnolle , V. , Ghestemme , T. , Thiollay , J. & Attie , C . ( 2000 ) Distribution, population size and habitat use of the réunion marsh harrier, circus m. maillardi . Journal of Raptor Research 34 , 8 – 17 . OpenUrl ↵ Bullock , H.E. , Panter , C.T. & Miller , T.A . ( 2024 ) Conservation letter: Raptor collisions in built environments . Journal of Raptor Research 58 . ↵ Ceballos , G. & Ehrlich , P.R . ( 2023 ) Mutilation of the tree of life via mass extinction of animal genera . Proceedings of the National Academy of Sciences 120 . ↵ Ceballos , G. , Ehrlich , P.R. & Dirzo , R . ( 2017 ) Biological annihilation via the ongoing sixth mass extinction signaled by vertebrate population losses and declines . Proceedings of the National Academy of Sciences 114 , E6089 – E6096 . OpenUrl Abstract / FREE Full Text ↵ Cheke , A . ( 2010 ) The timing of arrival of humans and their commensal animals on western indian ocean oceanic islands . Phelsuma 18 , 38 – 69 . OpenUrl ↵ Chiron , D. & Augiron , S. ( 2019 ) Programmefeder ecopap : écologie etconservation du papangue ( Circusmaillardi ) - rapport final. Tech. rep., SEOR . ↵ Clarke , J.G. , Smith , A.C. & Cullingham , C.I . ( 2024 ) Genetic rescue often leads to higher fitness as a result of increased heterozygosity across animal taxa . Molecular Ecology 33 . ↵ Clouet , M. ( 1978 ) Le busard de maillard ( Circus aeruginosus maillardi ) de l’île de la réunion . L’Oiseau et R.F.O pp. 96 – 106 . ↵ Coeurdassier , M. , Villers , A. , Augiron , S. , Sage , M. , Couzi , F.X. , Lattard , V. & Fourel , I . ( 2019 ) Pesticides threaten an endemic raptor in an overseas french territory . Biological Conservation 234 , 37 – 44 . OpenUrl ↵ Couturier , T. , Chiron , D. , Ferret , P. , Laurent , N. , Cornulier , T. , Villers , A. , Besnard , A. , Salamolard , M. , Couzi , F.X. & Augiron , S. ( 2022 ) Sciences participatives sur l’île de la réunion: quand le busard de maillard Circus maillardi (j. verreaux, 1862) fédère bénévoles et professionnels . Naturae pp. 261 – 283 . ↵ Doherty , T.S. , Hays , G.C. & Driscoll , D.A . ( 2021 ) Human disturbance causes widespread disruption of animal movement . Nature Ecology & Evolution A trouver . ↵ Fagan , W.F. & Holmes , E.E . ( 2006 ) Quantifying the extinction vortex . Ecology Letters 9 , 51 – 60 . OpenUrl CrossRef PubMed Web of Science ↵ Fay , R. , Ferret , P. , Chiron , D. , Schaub , M. & Augiron , S . ( 2023 ) First demographic insights reveal high extinction risk of an endemic raptor species: the reunion harrier . Journal of Avian Biology 2023 . ↵ Funk , W.C. , Lovich , R.E. , Hohenlohe , P.A. , Hofman , C.A. , Morrison , S.A. , Sillett , T.S. , Gha-lambor , C.K. , Maldonado , J.E. , Rick , T.C. , Day , M.D. , Polato , N.R. , Fitzpatrick , S.W. , Coonan , T.J. , Crooks , K.R. , Dillon , A. , Garcelon , D.K. , King , J.L. , Boser , C.L. , Gould , N. & Andelt , W.F . ( 2016 ) Adaptive divergence despite strong genetic drift: genomic analysis of the evolutionary mechanisms causing genetic differentiation in the island fox (urocyon littoralis) . Molecular Ecology 25 , 2176 – 2194 . OpenUrl CrossRef ↵ Garnier , Simon , Ross , Noam , Rudis , Robert , Camargo , Pedro , A. , Sciaini , Marco , Scherer & Cédric ( 2024 ) viridis(Lite) - Colorblind-Friendly Color Maps for R . Viridis package version 0.6.5 . ↵ Ghestemme , T. , Portier , E. & Corre , M.L . ( 1998 ) Recensement de la populationde papangue de la réunion Circus maillardi maillardi , densité et distribution des couples reproducteurs . Tech. rep., SEOR . ↵ Gillespie , R.G. , Claridge , E.M. & Roderick , G.K . ( 2008 ) Biodiversity dynamics in isolated island communities: interaction between natural and human-mediated processes . Molecular Ecology 17 , 45 – 57 . OpenUrl CrossRef PubMed Web of Science ↵ Gomez , E.A. , Hindmarch , S. & Smith , J.A . ( 2022 ) Conservation letter: Raptors and anticoagulant rodenticides . Journal of Raptor Research 56 , 147 – 153 . OpenUrl ↵ Grondin , V. & Philippe , J.S. ( 2011 ) Plan de conservation du busard de maillard Circus maillardi . Tech. rep., SEOR, Biotope, pour la DEAL, la Région Réunion, Aérowatt et la Ville de l’Etang-Salé . ↵ Guernier , V. , Lagadec , E. , Cordonin , C. , Le Minter , G. , Gomard , Y. , Pagès , F. , Jaffar-Bandjee , M.C. , Michault , A. , Tortosa , P. & Dellagi , K. ( 2016 ) Human leptospirosis on reunion island, indian ocean: Are rodents the (only) ones to blame? PLOS Neglected Tropical Diseases 10 , e0004733 . OpenUrl PubMed ↵ Heber , S. , Varsani , A. , Kuhn , S. , Girg , A. , Kempenaers , B. & Briskie , J . ( 2013 ) The genetic rescue of two bottlenecked south island robin populations using translocations of inbred donors . Proceedings of the Royal Society B: Biological Sciences 280 , 20122228 . OpenUrl PubMed ↵ Heinrichs , J.A. , McKinnon , D.T. , Aldridge , C.L. & Moehrenschlager , A . ( 2019 ) Optimizing the use of endangered species in multi-population collection, captive breeding and release programs . Global Ecology and Conservation 17 , e00558 . OpenUrl ↵ Hernangómez , D . ( 2023 ) Using the tidyverse with terra objects: the tidyterra package . Journal of Open Source Software 8 , 5751 . OpenUrl ↵ Hijmans , R.J . ( 2025 ) terra: Spatial Data Analysis . R package version 1.8-60. ↵ Huisman , J. , Kruuk , L.E.B. , Ellis , P.A. , Clutton-Brock , T. & Pemberton , J.M . ( 2016 ) Inbreeding depression across the lifespan in a wild mammal population . Proceedings of the National Academy of Sciences 113 , 3585 – 3590 . OpenUrl Abstract / FREE Full Text ↵ Kassambara , A . ( 2025 ) ggpubr: ’ggplot2’ Based Publication Ready Plots . R package version 0 . 6 . 1 . OpenUrl ↵ Keller , L. & Waller , D.M . ( 2002 ) Inbreeding effects in wild populations . Trends in Ecology & Evolution 17 , 230 – 241 . OpenUrl ↵ Kruuk , L.E.B. , Sheldon , B.C. & Merilä , J . ( 2002 ) Severe inbreeding depression in collared flycatchers ( ficedula albicollis ) . Proceedings of the Royal Society of London. Series B: Biological Sciences 269 , 1581 – 1589 . OpenUrl CrossRef PubMed Web of Science ↵ Lagabrielle , E. , Rouget , M. , Payet , K. , Wistebaar , N. , Durieux , L. , Baret , S. , Lombard , A. & Strasberg , D . ( 2009 ) Identifying and mapping biodiversity processes for conservation planning in islands: A case study in réunion island (western indian ocean) . Biological Conservation 142 , 1523 – 1535 . OpenUrl ↵ Lande , R . ( 1998 ) Demographic stochasticity and allee effect on a scale with isotropic noise . Oikos 83 , 353 . OpenUrl CrossRef Web of Science ↵ Lewanski , A.L. , Linderoth , T. , Thompson , G. , Tringali , A. , Angell , E. , Bowman , R. & Fitzpatrick , S.W . ( 2025 ) Translocations contribute to population rescue in an imperiled woodpecker . Proceedings of the National Academy of Sciences 122 . ↵ Madsen , T. & Shine , R . ( 1999 ) Rainfall and rats: Climatically-driven dynamics of a tropical rodent population . Australian Journal of Ecology 24 , 80 – 89 . OpenUrl CrossRef Web of Science ↵ Martensen , A.C. , Pimentel , R.G. & Metzger , J.P . ( 2008 ) Relative effects of fragment size and connectivity on bird community in the atlantic rain forest: Implications for conservation . Biological Conservation 141 , 2184 – 2192 . OpenUrl ↵ Martínez-Cruz , B . ( 2011 ) Conservation genetics of iberian raptors . Animal Biodiversity and Conservation 34 , 341 – 353 . OpenUrl ↵ McClure , C.J. , Westrip , J.R. , Johnson , J.A. , Schulwitz , S.E. , Virani , M.Z. , Davies , R. , Symes , A. , Wheatley , H. , Thorstrom , R. , Amar , A. , Buij , R. , Jones , V.R. , Williams , N.P. , Buechley , E.R. & Butchart , S.H . ( 2018 ) State of the world’s raptors: Distributions, threats, and conservation recommendations . Biological Conservation . ↵ McClure , C.J.W. , Buij , R. , Thorstrom , R. , Vargas , F.H. & Virani , M.Z . ( 2023 ) The world’s most imperiled raptors present substantial conservation challenges . Journal of Raptor Research 57 . ↵ Milá , B. , Warren , B.H. , Heeb , P. & Thébaud , C . ( 2010 ) The geographic scale of diversification on islands: genetic and morphological divergence at a very small spatial scale in the mascarene grey white-eye (aves: Zosterops borbonicus ) . BMC Evolutionary Biology 10 , 158 . OpenUrl PubMed ↵ Mourer-Chauviré , C. , Bour , R. & Ribes , S . ( 2006 ) Recent avian extinctions on re union (mascarene islands) from paleontological and historical sources . Bulletin-British Ornithologists Club 126 , 40 . OpenUrl ↵ Myers , N. , Mittermeier , R.A. , Mittermeier , C.G. , da Fonseca , G.A.B. & Kent , J. ( 2000 ) Biodiversity hotspots for conservation priorities . Nature 403 , 853 – 858 . OpenUrl CrossRef PubMed Web of Science ↵ Nichols , S. , Ewen , J.G. , Gottelli , D. , Grueber , C.E. , Santure , A.W. , Trask , A. & Brekke , P . ( 2024 ) Genetic rescue attempt in a small, inbred population of a wild endangered passerine . Biological Conservation 290 , 110430 . OpenUrl ↵ Oatley , G. , Simmons , R.E. & Fuchs , J . ( 2015 ) A molecular phylogeny of the harriers (circus, accipitridae) indicate the role of long distance dispersal and migration in diversification . Molecular Phylogenetics and Evolution 85 , 150 – 160 . OpenUrl CrossRef PubMed ↵ O’Bryan , C.J. , Allan , J.R. , Suarez-Castro , A.F. , Delsen , D.M. , Buij , R. , McClure , C.J.W. , Rehbein , J.A. , Virani , M.Z. , McCabe , J.D. , Tyrrell , P. , Negret , P.J. , Greig , C. , Brehony , P. & Kissling , W.D . ( 2022 ) Human impacts on the world’s raptors . Frontiers in Ecology and Evolution 10 . ↵ Pebesma , E. & Bivand , R . ( 2023 ) Spatial Data Science: With Applications in R . Chapman and Hall/CRC . ↵ Pedersen , E.J. , Miller , D.L. , Simpson , G.L. & Ross , N . ( 2019 ) Hierarchical generalized additive models in ecology: an introduction with mgcv . PeerJ 7 , e6876 . OpenUrl CrossRef PubMed ↵ Pinaud , D. , Claireau , F. , Leuchtmann , M. & Kerbiriou , C . ( 2018 ) Modelling landscape connectivity for greater horseshoe bat using an empirical quantification of resistance . Journal of Applied Ecology 55 , 2600 – 2611 . OpenUrl ↵ R Core Team ( 2025 ) R: A Language and Environment for Statistical Computing . R Foundation for Statistical Computing , Vienna, Austria . ↵ Régnier , C. , Achaz , G. , Lambert , A. , Cowie , R.H. , Bouchet , P. & Fontaine , B . ( 2015 ) Mass extinction in poorly known taxa . Proceedings of the National Academy of Sciences 112 , 7761 – 7766 . OpenUrl Abstract / FREE Full Text ↵ Robinson , R. , Morrison , C. & Baillie , S.R . ( 2014 ) Integrating demographic data: towards a framework for monitoring wildlife populations at large spatial scales . Methods in Ecology and Evolution . ↵ Safford , R. & Hawkins , F . ( 2020 ) The Birds of Africa: Volume VIII: The Malagasy Region: Madagascar, Seychelles, Comoros, Mascarenes . Bloomsbury Publishing . ↵ Salim , H. , Noor , H.M. , Omar , D. , Hamid , N.H. , Abidin , M.R.Z. , Kasim , A. , Salim , H. , Rawi , C.S.M. & Ahmad , A.H . ( 2014 ) Sub-lethal effects of the anticoagulant rodenticides bromadiolone and chlorophacinone on breeding performances of the barn owl ( Tyto alba ) in oil palm plantations . Slovak Raptor Journal 8 , 113 – 122 . OpenUrl ↵ Schauberger , P. & Walker , A. ( 2014 ) openxlsx: Read, write and edit xlsx files . ↵ Sergio , F. , Newton , I. & Marchesi , L . ( 2008 ) Top predators and biodiversity: much debate, few data . Journal of Applied Ecology 45 , 992 – 999 . OpenUrl Web of Science ↵ Simpson , G.L . ( 2024 ) gratia: An r package for exploring generalized additive models . Journal of Open Source Software 9 , 6962 . OpenUrl ↵ Stenseth , N.C. , Leirs , H. , Skonhoft , A. , Davis , S.A. , Pech , R.P. , Andreassen , H.P. , Singleton , G.R. , Lima , M. , Machang’u , R.S. , Makundi , R.H. , Zhang , Z. , Brown , P.R. , Shi , D. & Wan , X . ( 2003 ) Mice, rats, and people: the bio-economics of agricultural rodent pests . Frontiers in Ecology and the Environment 1 , 367 – 375 . OpenUrl CrossRef Web of Science ↵ Strasberg , D. , Rouget , M. , Richardson , D.M. , Baret , S. , Dupont , J. & Cowling , R.M . ( 2005 ) An assessment of habitat diversity and transformation on la réunion island (mascarene islands, indian ocean) as a basis for identifying broad-scale conservation priorities . Biodiversity and Conservation 14 , 3015 – 3032 . OpenUrl ↵ Tershy , B.R. , Shen , K.W. , Newton , K.M. , Holmes , N.D. & Croll , D.A . ( 2015 ) The importance of islands for the protection of biological and linguistic diversity . BioScience 65 , 592 – 597 . OpenUrl CrossRef ↵ Thieurmel , B. & Elmarhraoui , A . ( 2022 ) suncalc: Compute Sun Position, Sunlight Phases , Moon Position and Lunar Phase . R package version 0.5.1. ↵ Thomason , E. , Wallen , K. & Katzner , T . ( 2023 ) Social and biological perspectives to investigate and address illegal shooting of raptors . Global Ecology and Conservation 46 , e02631 . OpenUrl ↵ Wickham , H . ( 2016 ) ggplot2: Elegant Graphics for Data Analysis . Springer-Verlag New York . ↵ Wickham , H. , François , R. , Henry , L. & Müller , K. ( 2022 ) dplyr: A Grammar of Data Manipulation . R package version 1.0.9. ↵ Wiens , J.D. , Schumaker , N.H. , Inman , R.D. , Esque , T.C. , Longshore , K.M. & Nussear , K.E . ( 2017 ) Spatial demographic models to inform conservation planning of golden eagles in renewable energy landscapes . Journal of Raptor Research 51 , 234 – 257 . OpenUrl PubMed ↵ Wilke , C.O . ( 2025 ) ggridges: Ridgeline Plots in ’ggplot2’ . R package version 0.5.7. ↵ Winship , A.J. , O’Farrell , M.R. & Mohr , M.S . ( 2013 ) Management strategy evaluation applied to the conservation of an endangered population subject to incidental take . Biological Conservation 158 , 155 – 166 . OpenUrl ↵ Witzenberger , K.A. & Hochkirch , A . ( 2011 ) Ex situ conservation genetics: a review of molecular studies on the genetic consequences of captive breeding programmes for endangered animal species . Biodiversity and Conservation 20 , 1843 – 1861 . OpenUrl ↵ Wood , J.R. , Alcover , J.A. , Blackburn , T.M. , Bover , P. , Duncan , R.P. , Hume , J.P. , Louys , J. , Meijer , H.J. , Rando , J.C. & Wilmshurst , J.M . ( 2017 ) Island extinctions: processes, patterns, and potential for ecosystem restoration . Environmental Conservation 44 , 348 – 358 . OpenUrl CrossRef ↵ Wood , S. ( 2017 ) Generalized additive models: an introduction with R . Chapman & Hall/CRC , 2nd edn . ↵ Wood , S.N . ( 2011 ) Fast stable restricted maximum likelihood and marginal likelihood estimation of semiparametric generalized linear models . Journal of the Royal Statistical Society: Series B (Statistical Methodology ) 73 , 3 – 36 . OpenUrl CrossRef Web of Science ↵ Wood , T.J. & Goulson , D . ( 2017 ) The environmental risks of neonicotinoid pesticides: a review of the evidence post 2013 . Environmental Science and Pollution Research 24 , 17285 – 17325 . OpenUrl CrossRef PubMed ↵ Wright , D.J. , Spurgin , L.G. , Collar , N.J. , Komdeur , J. , Burke , T. & Richardson , D.S . ( 2014 ) The impact of translocations on neutral and functional genetic diversity within and among populations of the seychelles warbler . Molecular Ecology 23 , 2165 – 2177 . OpenUrl CrossRef ↵ Zuur , A. , Ieno , E. , Walker , N. , Smith , G. & Saveliev , A . ( 2009 ) Mixed Effects Models and Extensions in Ecology With R . Springer . View the discussion thread. Back to top Previous Next Posted November 09, 2025. Download PDF Data/Code 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 Long-term population monitoring reveals an alarming decline in the endangered endemic Réunion Harrier Circus maillardi 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 Long-term population monitoring reveals an alarming decline in the endangered endemic Réunion Harrier Circus maillardi Alexandre Villers , Thomas Cornulier , Cyril Eraud , Thibaut Couturier , Pierrick Ferret , Damien Chiron , Nicolas Laurent , Aurélien Besnard , Vincent Bretagnolle , Marc Salamolard , Valérie Grondin , François-Xavier Couzi , Rémi Fay , Steve Augiron bioRxiv 2025.11.07.687144; doi: https://doi.org/10.1101/2025.11.07.687144 Share This Article: Copy Citation Tools Long-term population monitoring reveals an alarming decline in the endangered endemic Réunion Harrier Circus maillardi Alexandre Villers , Thomas Cornulier , Cyril Eraud , Thibaut Couturier , Pierrick Ferret , Damien Chiron , Nicolas Laurent , Aurélien Besnard , Vincent Bretagnolle , Marc Salamolard , Valérie Grondin , François-Xavier Couzi , Rémi Fay , Steve Augiron bioRxiv 2025.11.07.687144; doi: https://doi.org/10.1101/2025.11.07.687144 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 (7619) Biochemistry (17642) Bioengineering (13865) Bioinformatics (41863) Biophysics (21410) Cancer Biology (18548) Cell Biology (25437) Clinical Trials (138) Developmental Biology (13359) Ecology (19863) Epidemiology (2067) Evolutionary Biology (24288) Genetics (15587) Genomics (22467) Immunology (17704) Microbiology (40301) Molecular Biology (17142) Neuroscience (88447) Paleontology (666) Pathology (2825) Pharmacology and Toxicology (4815) Physiology (7634) Plant Biology (15109) Scientific Communication and Education (2042) Synthetic Biology (4285) Systems Biology (9812) Zoology (2268)