Short?distance nocturnal migration in an island endemic bustard

Migratory birds typically fly long distances in response to large-scale seasonal climate variation. However, most migratory species are partial migrants, and some only short distances, for example take advantage of suitable feeding conditions during the non-breeding season. In spite their distance, such migrations may also be fixed populations as a result evolutionarily adaptive responses local habitat conditions, but causes details these partial, short-distance have been scarcely investigated. Here we analyse patterns insular endemic subspecies Houbara Bustard Chlamydotis undulata fuertaventurae small island Canary archipelago. We tagged 41 individuals with GSM/GPRS loggers tracked them over two annual cycles (2017–2019). found that Canarian Bustards short-distance, migrants. Once mating season was over, more than one-third all males females moved from breeding territories areas where they spent hottest months year. No bird changed its status between years, showed interannual fidelity areas. There were differences productivity green vegetation areas, estimated remote-sensing-based indicators (Normalized Difference Vegetation Index Soil-Adjusted Index). During period, values both indices higher at birds, providing plausible explanation why rather remaining territories. Moreover, lower abandoned by occupied year-round, although latter did not differ those suggesting migration arises variability food availability summer across different One four selected half migrant individuals. This area had mosaic cultivated non-cultivated patches, fallow fields irrigated farmland other Finally, flights performed night. The is globally threatened, given vulnerability collision electric telephone lines migration, recommend burying high-traffic underground. urge protection known current farming condition practices natural shrubland continue provide Bustards. Las aves migratorias suelen volar grandes distancias en respuesta la variación estacional del clima. Sin embargo, mayoría de las especies son migradoras parciales, y algunas solo vuelan cortas, como por ejemplo, para beneficiarse una mayor disponibilidad alimento durante su período no reproductivo. A pesar tratarse movimientos cortos, este tipo migración puede estar bien establecido poblaciones adaptativa condiciones locales hábitat, si causas los detalles dichas migraciones han sido escasamente investigados. Aquí analizamos patrones migratorios subespecie Avutarda Hubara Canaria Lanzarote, tres islas orientales archipiélago canario que es endémica. Basándonos muestra individuos marcados con emisores estudiados dos ciclos anuales entre 2017 2019, encontramos migradora parcial corta distancia. Una vez finalizado el reproductivo, un tercio machos hembras abandonaron área reproducción dirigirse zonas permanecieron meses más calurosos año. Ningún individuo cambió patrón migrador sedentario o viceversa años, mostraron fidelidad interanual sus cría reproductivas. Se encontrado diferencias productividad vegetal reproductivas, estimadas partir índices teledetección (NDVI SAVI). Durante estación reproductiva, valores ambos fueron mayores reproductivas migradores, proporcionando explicación qué dichos migran permanecen zona cría. Es más, hecho SAVI abandonadas migradores fuesen menores hubaras todo año, estas últimas inferiores apoyan nuestra interpretación sugieren mecanismo adaptación abundancia verano distintas zonas. cuatro fue seleccionada mitad migradores. Esta presenta mosaico parcelas cultivadas cultivadas, extensión barbechos cultivos regadío otras Finalmente, se observó vuelos nocturnos. La está globalmente amenazada dada vulnerabilidad colisión tendidos eléctricos telefónicos migración, recomienda soterramiento áreas distribución especie Canarias. Por último, instamos protejan conocidas, garantice prácticas agrícolas actuales vegetación sigan hábitat adecuado Hubara. Partial represents an evolutionary transition residency complete (Berthold 1999, 2001). It has even suggested precursor full 1996). Recent reviews revealed this system much widespread previously reported, exists animal taxa (Cristol et al. Berthold 2001, Newton 2008). being so widespread, promote maintain timescales still debate (Chapman 2011, Reid 2018, Hegemann Akeson & Helm 2020). Further experimental, genomic detailed field studies using modern tracking technology analyses based on up-to-date remote sensing tools necessary determine relative contributions genetic environmental factors (Pettorelli 2005, Ogonowski Conway 2009, Robinson 2010, Liedvogel reported several lekking or polygynous terrestrial (Schroeder 1985, Cade Hoffman 1993, Schroeder Braun 1993). family Otididae it present Great Otis tarda Little Tetrax tetrax (Alonso 2000, 2020, Palacín 2012, García Morena 2015) well Macqueen's macqueenii (Seddon van Heezik 1996, Sheldon Launay 1998, Combreau 2011a,b, Tourenq 2004, Judas 2006, Mansouri Riou Burnside 2017, 2020), nominate complex (Hingrat 2008, Hingrat Saint Jalme 2005). movements studied North African undulata, which shows distinct wintering Sightings Europe 19th century (de Juana 2016) probably dispersive erratic Moroccan movements. same can said very recent sighting Fuerteventura captive-bred bird, ringed released Morocco (González-Rodríguez 2021). Indeed, following palaeontological evidence, archipelago presumably colonized 130 000–170 000 years ago (Ancochea 1990, Rando 1995), according mitochondrial DNA analysis, diverged genetically around 20 000–25 (Idaghdour 2004). Occasional islands past (Martín 1996), corroborated recently marked 2022). possibility (Hernández Lorenzo 2012), never confirmed individually study carried out, date, restricted three easternmost archipelago: Graciosa. population Lanzarote 440–452 currently subspecies' main stronghold, c. 80% total (Schuster Alonso Ucero classified Globally Vulnerable (BirdLife International 2016), knowledge will help us understand how human pressures threaten survival environment shared millions tourists visiting every show male-biased sexual size dimorphism (del Hoyo 2014; own data subspecies). sexes rarely aggregate flock, requirements (Yang 2002, 2003, 2007, Females visit displaying reproduction fulfil criteria exploded lek Between late autumn early spring, concentrate display specific location remain faithful whole period described differential 2008), showing mostly unimodal, home-ranges, tendency sedentary behaviour, whereas predominantly multimodal, significantly larger migrate similar above-mentioned aspects system, site-fidelity subspecies. used sample (1) whether assumed, perform could qualify true displacements, (2) describe quantify any detected females, examine vary (3) daily timing (4) discuss possible explaining behaviour observed (5) identify stopover sites propose appropriate management strategies. out Islands (29°02 N, 13°37 W; 846 km2). located Atlantic Ocean, 140 km west coast. subtropical–desert (Köppen classification), tempered cold Current nearly permanent northeasterly ‘Trade’ winds. Mean temperatures 18 °C winter 25 summer. Rainfall concentrated December–February, average 110 mm per Summers dry, less 1 precipitation June–August. volcanic origin, mountain massifs. parts island, occur, smoothed topography, sandy semi-desert (Carrascal dominant consists xerophytic shrubs Launaea arborescens, Salsola vermiculata, Lycium intricatum, Suaeda spp. Euphorbia spp., partly modified goat grazing activities, nowadays except few island. (21 females) captured nylon snares, effective harmless capture method 1999; Y. Hingrat, personal communication). number approximately 7% All fitted backpack-mounted, solar-powered (48-g model males, 25-g females; e-obs GmbH, Gruenwald, Germany), soft, elastic band harness material. weight transmitter plus 2.83% body (range 2.54–3.15%) 2.15% 1.81–2.53%). Males randomly case avoid jeopardizing nesting considering threatened species, nest. harmless, viable efficiently foraging meant use fraction sex, because capturing aggregations our high probabilities migratory. Therefore, calculate realistic migration. team, consisting people, remained 300–500 m site ready quickly access soon became entangled snares. immobilized, heads covered minimize stress. processing time individual release 14 min 5–25 min). observe behavioural aberrations release. Two one female plumage biometric characters made suspect immatures excluded sample. Sex established distinctive features (Glutz 1973, Gubin 1992), analysis extracted contour feathers plucked each bird. programmed record GPS locations 5-min intervals 5:00 am 10:59 pm coordinated universal (UTC), i.e. h (summer solstice) 3 (winter before sunrise after sunset. They included three-axis-accelerator (ACC) recorded acceleration unit three-dimensional space, three-graph representation bird's To prevent excessive battery drainage, ACCs 16 s 5 9:00 UTC. behaviours ACC graphs, direct observations 10 December 2018 March 2019 (720 total) 20× 60× telescope distance 1000 m. compared timed graphs. enabled graph corresponded activities Then, sequences train python-based application ACCELERATER (Resheff 2014), classify ACC-based birds. identified period: run, boom call sequence, pre-copulation movements, (feeding bouts alternating walking), preening, resting flying. addition, incubating when she almost immobile days, moving feeding. downloaded through phone network collated MOVEBANK (https://www.movebank.org), without having recapture More million study, covering For defined home-range (delimited first last display, day attempt, successful not). non-nesting area. Non-breeding rest year, once compare date (calculated separately arrived left (a well-defined spot majority displays) nest (there cases re-nesting among females), modal within preferred mean represent places time, influenced outlier errors. home-ranges computed 95% kernel means density estimation methods (Worton 1989), reference bandwidth ‘href’ smoothing parameter estimations (Kie Kie 2013, Schuler ‘reproducible home ranges’ package (rhr) R software. movement separate (non-overlapping) (Newton sedentary. consecutive second year overlapped 40% flight usually clearly large locations, detectable amplitude waves. Stays distinguishable checked visual inspection movement. considered stayed least Flight speed calculated (metres) divided duration (seconds). outward (calendar Julian hereafter JD) flew area, return returned Migration travel site. flights, including periods sites. data, 30 excluding days reduce effect proximity budget. check differed patterns, surveys seasons. proportions migrated centre captured. Using figure quantified alone, censuses (19 January–23 February) (22 May–21 June). These teams people driving vehicles low (15–20 km/h) along tracks roads throughout stopping frequently scan vantage points 20×–60× telescopes. As detectability survey effort intensive survey, numbers counted smaller. Surveys began dawn ended dusk, break midday (10:30 4:30 UTC), rest. From counts, difference called inferred sex ?2 tests test migrants (Supporting Information Table S1). variables expressed ± standard deviation (sd). problems relating distributions, sizes large, non-parametric statistics. Mann–Whitney's U look dates, speed. Wilcoxon's signed-rank parameters vs. periods, budget relationship stopovers examined Kruskal–Wallis test, directionality Rayleigh’s uniformity test. Mann–Whitney Wilcoxon wilcox.test function, kruskal.test function. plotted ARCGIS, software R, version 3.6.3 (https://www.r-project.org) statistical analyses, related directions, analysed ORIANA (Kovach Computing Services processed series NDVI spectral 27 Sentinel-2A Sentinel-2B multi-spectral instruments month January 2019. launched 23 June 2015 followed 7 enabling image acquisition days. free-cloud available images species’ requirements. Images obtained Copernicus Open Access Hub (https://scihub.copernicus.eu), started May US Geological Survey (https://earthexplorer.usgs.gov). Most level-1C top-of-atmosphere products, projected cartographic geometry WGS84 datum, Universal Transverse Mercator projection, Zone28North. ensure highest quality Sentinel-2 quantitative land monitoring, geometrically rectified atmospherically corrected, converting pixel into surface reflectance (Level 2) applying Sen2Cor processor (Main-Knorn Sola 2018) deliver Level-2A bottom-of-atmosphere products. Given flat topography life-cycle interest (i.e. areas) apply topographic correction. bands (B2: 497 nm B12: 2020 nm) then resampled capabilities spatial resolution nearest-neighbour approach, atmospheric (B1, B9 B10) analysis. red (B4: 664 near infrared (B8a: 865 nm), period. By time-series information, averaged (means sd) 10-m pixels contained Land Parcel Identification System Lanzarote. IT aerial satellite photographs recording agricultural parcels Member States. key control mechanism under Common Agricultural Policy accessible Spain Geographical Parcels entire Spanish territory Guarantee Fund (https://www.sigpaccanarias.org/sigpac/visor/ Islands). Crop surfaces map (https://www.gobiernodecanarias.org/agricultura/temas/mapa_cultivos/lanzarote/). Seven (36.8%) 19 (Fig. 1). highly variable 12 (63.2%) sedentary, staying 0.25 (sd = 0.15; range 0.06–0.54 km, n 12) 11.06 3.65; 4.80–17.11 7) males. 0.24 0.06; 0.18–0.29 km) 17 (they 11.35 sd 8.22; 1.53–27.55 17), noted Methods cannot Nevertheless, able assess pattern (46 57 S1) (169 227 (182 257 S1), either (respectively, 0.063, df 1, P 0.802; 0.232, 0.630; results suggest prefer spend alternative attraction exerted illustrated Figure 2 explained below, section Habitat Among seven (57%) Data six females. None study. (92% individuals) (mean overlap 52.1% 42.2% 61.6% 47.2% major phenology (Table S2). Only her remarkably exactly years. September 1), earlier (W 23, 0.044; test). significant 1; W 138, 0.106; result, longer 136 89 62, 0.050; length stay change (V 38, 0.154; 10, sexes; signed rank apparent success days; 4) later incubation (n 16; 18, 0.049; 46, 0.53; Mann–