Wading birds have declined globally with particularly in western Europe. Multiple species are now on the IUCN Red list, with northern lapwing (Vanellus vanellus) near-threatened and declining. Historically, habitat degradation, including from wetland drainage and agricultural intensification, has contributed to population declines. More recently, declines have been attributed to poor breeding success due to unsustainably high rates of predation on eggs and chicks, by avian and mammalian predators. In the UK, the red fox (Vulpes vulpes) is a major mammalian predator of waders, but the Eurasian badger (Meles meles) has increased in range and abundance, and can occur at high densities, with potential for acute local predation impacts on vulnerable wader populations. However, factors affecting rates of badger predation on wader nests remain unexplored. We investigate what these factors might be using data from six years of lapwing nest monitoring at a breeding site in northeast Scotland. The overall probability of badger predation was above 0.1 when mean daily temperature was below 4 °C during the preceding 7 days, dropping close to zero when above 10 °C. Badger predation on lapwing clutches also increased with earthworm availability, and inter-annual effects were observed matching variations in temperature, whereby intense badger predation in 2021 coincided with unseasonably cold temperatures and low lapwing breeding productivity. This highlights the potential for weather forecasting to be used to deploy pre-emptive non-lethal management strategies to mitigate badger predation impacts on lapwing nests.

Lapwing nest direct observation

To monitor lapwing nests, potential wader habitat was surveyed on foot, during the wader breeding season (March to July; earliest start date = 30 March; latest end date = 3 July) over six years, between 2018 and 2023. Routine vehicular farm surveys were used initially to locate nests at the start of the breeding season every year by observing adult activity, including nest concealment and protection behaviours commonly indicative of nesting (Królikowska et al., 2016). Survey start points were additionally selected from previous observations of activity and nests, as lapwing are highly nest-site faithful and philopatric (Thompson et al. 1994; Jarrett et al., 2017). Once a lapwing nest was discovered, the date, time, location, number of eggs, field number, nest number, nest stage/outcome (still nesting, hatched, failed), and reason for failure where applicable, were recorded on Epicollect, version 4.2.0 (Epicollect5, 2022). A rough estimate of likely hatch date was noted when nests were found during the egg-laying phase, based on the presumption of laying one egg per day followed by a 27 day incubation period (Witherby et al., 1940). Failed nests were inspected for field signs of predation, and predator identity was ascertained where possible. Badger predation was determined using multiple field sign cues. Firstly, nocturnal predation (timings determined from regular surveying) implicated mammalian rather than avian predators (Bolton et al., 2007). Nocturnal predation events were substantially less likely to be due to foxes than badgers as a result of lethal fox control that was conducted on site for the entirety of the study. Direct evidence of badger predation included signs of exceptional disturbance at the nest site, crushing damage to surrounding nest vegetation, whole clutch removal, evidence of eating in-situ and distinctive claw marks/scratchings around the predated nests in instances where the terrain enabled this to be seen. These signs of a high level of nest disturbance were characteristic of badger, and absent when nest predation was due to other mammalian species e.g. stoats or weasels (Green et al., 1987, Valkama & Currie, 1999; Whittingham et al., 2002; Stillman et al. 2006). which These signs differ, for example, from partial clutch removal, large eggshell remains and large pecked holes in eggs without tooth marks indicative of corvids or other avian predators (Green et al., 1987; Jackson & Green, 2000; Wallander et al. 2006). These signs were consistent with photographic evidence of active predation obtained from nests monitored with trail cameras. If evidence was deemed not clear, predator identity was not attributed to species but recorded as unknown, unknown mammal or unknown avian predation accordingly.

Nests were marked with a short stake (bearing the ID number), which was placed ≥1.5 m away to the north of the nest. Stakes were purposefully unobtrusive to prevent an impact of marking itself on nest outcomes (Kragten & de Snoo, 2008). Subsequent observations were regularly conducted using binoculars or spotting scopes, from a vantage point approximately 100-150 m away from the nest, to avoid disturbance. This allowed observers to establish an “end” date for each nest, i.e., the point at which the nest either hatched or failed. Evidence of hatching included small chippings of eggshell in the nesting material (Berg et al., 1992), the presence of hatched chicks within the nests, and of identifiable young broods based on regular observations (e.g. where broods could be clearly distinguished by age, the presence of ringed adults, or newly hatched broods in fields where only the one hatched nest was present).Where it was not possible to ascertain an outcome, nest fate was assigned as ‘unknown’. Following hatching, observers continued to monitor broods as accurately as possible, including by colour-ringing chicks and some adults (variable across study years) in addition to regular surveys of the whole farm, recording all observed chicks (including age estimatiio based on size and apparent feather development) and adults. Final observations were conducted as soon as possible after fledging, to prevent inflation of fledgling counts due to influx from neighbouring breeding areas (as cautioned by Bolton et al., 2011).

Trail camera surveys

From 2018 to 2021, a sample of nests were monitored with a limited number of trail cameras based on whether they were likely to be obscured when observed from afar with binoculars or spotting scopes. From 2022 a greater availability of trail cameras meant more nests were monitored with cameras applied if available regardless of nest location. Cameras were set to capture still images on low sensitivity, to preserve battery life.. From 2021 onwards, these were Browning TM Recon Force Elite HP4 (One Browning Place, Morgan, Utah, 84050, USA); and between 2018 and 2021, either Ltl Acorn 6310 (Ltl Acorn, 4100 Gannett Avenue, Des Moines, Iowa, 50321, USA) or Reconyx Hyperfire 2 Covert IR (3828 Creekside Lane, Ste 2 Holmen, Wisconsin, 54636, USA). Once nests had completed, cameras were retrieved and re-deployed. Images were subsequently reviewed, the time of hatch or failure and the outcome of each nest were recorded, together with predator identity (where possible) for each predation event. Predation events were identified via images of egg removal, of the predator’s head within the nest cup, or other clear signs of active predation.

Weather data

Weather data were sourced from the onsite GWSDF Auchnerran Automatic Weather Station (AWS), deployed and remotely managed by the James Hutton Institute, Banchory, Aberdeenshire. Temperature (°C) was collected using a 107 Air temperature probe, and rainfall (mm) using a ARG 100 tipping bucket Rain Gauge (0.2 mm/tip) (Campbell Scientific, 80 Hathern Road, Shepshed, LE12 9GX, UK). Owing to various contemporaneous and time-lagged effects of weather on badger activity documented in other studies (e.g., Noonan et al., 2015), weather variables were summarised as mean daily temperature and total rainfall across two time periods: i) 7-day interval before hatch or failure; and ii) 30-day interval before hatch or failure.

Invertebrate sampling and land use data

Potential additional covariates relating to land use and management were sourced from farm records at field level (Table 1) and included in statistical models. These included invertebrate abundance data and field classifications under the Agri-environment Climate Scheme (AECS) (full details for the latter given in SI Table 1). Invertebrate data were derived from soil sampling in wet weather and conducted annually in June 2018, 2019, and 2021. Two samples with dimensions 20 cm x 20 cm x 20 cm per hectare were collected from all fields (651 samples in total); all invertebrates in each sample were identified to at least family level where possible, and number of individuals counted. This procedure yielded relative abundance data for invertebrates overall (INV), and for earthworms (EARTH) separately, providing a measure of the availability of each to badgers during the wader breeding season.