Fearfulness of geese and swans on cropland in winter: A multi-species Flight Initiation Distance approach
Data files
Jan 21, 2025 version files 55.79 KB
Abstract
Geese and swans are focal species in conservation and in management aimed at reducing crop damage. In the former disturbance should be minimized, and in the latter it is important to know how different species react to scaring activities. Previous research about trade-off between predation risk and foraging in birds often use ‘Flight Initiation Distance’ (FID) as a proxy to compare fearfulness under different circumstances and among species. We studied variation in FID in geese and swans by species, flock size and composition, time of day, and body size (408 scaring trials on agricultural land in winters 2018—2021). In single-species flocks mean FID decreased in the order: bean goose (171 m) > greylag goose (104 m) > whooper swan (102 m) > Canada goose (92 m) > barnacle goose (77 m). In line with predictions based on body mass, the lightest species (barnacle goose) was the least fearful, but contrary to prediction neither of the two heaviest species (whooper swan, Canada goose) was the most fearful. FID was negatively correlated with flock size in bean goose. Flock size and FID did not correlate in greylag, Canada, and barnacle geese. FID did not differ between morning and afternoon in the 4 species with >20 single-species trials (i.e., supposedly hungry versus satiated geese). When in multi-species flocks, FID differed less among species, converging in the 108—138 m range. For example, bean goose FID decreased significantly whereas it increased significantly in barnacle and greylag geese. Barnacle goose (protected from open hunting in EU) was less fearful than species with an open hunting season in the EU, implying that exposure to hunting affect species-specific FID. We show that the level of fearfulness varied among swan and goose species, making it necessary to adopt diverse strategies in conservation as well as crop protection.
README: Fearfulness of geese and swans on cropland in winter: A multi-species Flight Initiation Distance approach
https://doi.org/10.5061/dryad.z8w9ghxq4
Description of the data and file structure
Files and variables
File: Elmberg-et-al-Wildlife-Biology-Fearfulness-of-geese.xlsx
Description:
Variables
- DATE = day of trial
- MONTH = month of trial
- BEFORE/AFTER NOON = trial before or after 12 AM
- TIME = European Central Time (hr,min)
- WHOOPER SWAN = number of whooper swans
- MUTE SWAN = number of mute swans
- GREYLAG GOOSE = number of greylag geese
- BARNACLE GOOSE = number of barnacle geese
- BEAN GOOSE SENSU LATU = number of bean geese (taxon fabalis, rossicus/serrirostris, or unidentified to subspecies)
- CANADA GOOSE = number of Canada geese
- GREATER WHITE-FRONTED GOOSE = number of greater white-fronted geese
- PINK-FOOTED GOOSE = number of pink-footed geese
- TOTAL NUMBER OF GEESE = sum of geese in trail
- TOTAL NUMBER OF GEESE AND SWANS = total number of geese and swans in trial
- SINGLE SPECIES FLOCKS = names of the sole species in trials with just one species (trials with more than one species present are denoted by "n/a")
- FLIGHT ORDER = the sequence in which species took flight in trials with two species or more (trials where this information was not collected are denoted by "n/a")
- FLIGHT INITIATION DISTANCE (FID) = the distance (m) from the approaching observer at which the first geese in a flock took flight.
Methods
Study area, seasons, and crop availability
This study was carried out during three winters (2018/2019 through 2020/2021) in southernmost Sweden (Scania province (Skåne); 55--56oN, 12--15oE) in open landscapes characterized by intensive agriculture. Cereals (wheat, barley, oats, rye), potatoes, sugar beet, oil rape seed (canola), and ley/hay are the main crops. However, also corn (maize), lettuce, broccoli, cale, carrots, beetroot, parsnip, cabbage, quinoa, green peas, yellow peas, yellow onion, red onion, leek, beans, spice herbs, and sileage crops such as alfalfa and grass-legume mixtures are cultivated in the study area.
Scaring trials were conducted by walking towards a flock of geese and/or swans (see ‘scaring trials’ below for details) and were only carried out on agricultural land, including pastures and hayfields. Since sampling was done from November through March, geese were encountered on growing crops (fall-sown cereals and oil rapeseed (canola), or sugar beets or carrots yet to be harvested) as well as in harvested fields. In early and mid-winter, some fields had fall-sown (growing) intercrops such as grasses, legumes, or oil radish. The seasons of study in the winters of 2018--2021 were typical for recent decades in the area when it comes to weather; snow cover was rare, thin, and lasted a few days or a week at the most. Temperatures were above the monthly means for the meteorological reference period 1991—2020 in 14 out of 15 study months (Supplementary information). Consequently, fall-sown (winter-green) crops and spill from harvested cereals and root crops (mainly potatoes, sugar beets, and carrots) were available to grazing birds throughout the winter. The study area is the northmost region in Sweden where all goose species winter regularly.
Study species and species-specific predictions
Greylag goose is a common breeder in wetlands near agricultural land throughout the study area, whereas Canada goose and barnacle goose are scarce breeders. All three breeding species are resident, thus occurring throughout winter. Residents of these species are reinforced in numbers by conspecifics arriving from the northeast to spend the winter in the study area. Bean goose (Anser fabalis (taxa fabalis and serrirostris/rossicus)) is an abundant non-breeding visitor from November through March. Pink-footed goose and greater white-fronted goose are regular in fall, winter, and spring, but are not as abundant as the aforementioned species. Lesser white-fronted goose and red-breasted goose are very rare but annually occurring in the study area in the non-breeding season. Whooper swan (Cygnus cygnus) and mute swan (Cygnus olor) are scarce but widespread local breeders, being more numerous in winter. In other words, the study area has a speciose set of large grazing avian herbivores present on agricultural land throughout winter, making it well suited for studies of the present type.
With respect to the species later observed during the course of the study, two of the general predictions listed in the introduction can be made species-specific: 1) FID should be longer in large-bodied than in smaller-bodied species. Prediction: FID mute swan > whooper swan > Canada goose > greylag goose > bean goose > pink-footed goose > white-fronted goose > barnacle goose) (cf. Blumstein 2006, body mass data from Cramp and Simmons 1977), and 2) FID should be longer in species hunted in the EU. Prediction: FID longer in greylag, bean, pink-footed, white-fronted, and Canada geese than in barnacle goose and whooper swan.
Scaring trials
We drove rural roads slowly by car with the purpose of spotting flocks of geese and/or swans foraging in agricultural fields. Days with very strong winds and reduced visibility due to rain, fog, or snowfall were avoided. Whenever geese and/or swans were sighted, we stopped at a discrete distance before they were alarmed (stretched necks, wing-flapping, walking away) and made sure that there were not any obvious ongoing external disturbances. Next, we ensured that sight was free between the observer and the majority of the individuals in the flock (cf. Mayer et al. 2019). The distance between the car and the flock (‘detection distance’ sensu Weston et al. 2012) varied and the primary goal was to park the car so that its presence did not initiate agitation in the flock.
From immediately outside the car, we counted and identified geese and/or swans in the flock using a spotting scope. Next, the scaring person started walking slowly, calmly, and silently towards the flock, in a straight direction whenever possible. We used plain clothing and left the spotting scope behind. When the first birds in the flock took off, we stopped to record the observer’s GPS position (first GPS fix), while at the same time noting the spot from which the first birds had left. Next, we walked to the latter spot to obtain a second GPS fix. Flight initiation distance (FID) is the distance between the first and second GPS fixes.
When a flock was scared off a field, we used binoculars to note its flight direction and, if possible, where it landed. We did this to avoid scaring the same birds more than once the same day. We regard the risk of repeated scaring (and hence habituation) as very small. For example, the study area typically hosts hundreds of thousands of geese during the winter months (Haas and Nilsson 2019), and we tried to visit new areas from day to day. When possible, we recorded the flight order of the different species in multi-species flocks.
We discarded cases during which geese or swans were disturbed by something else after the start of our approach (e.g., cars, airplanes, dogs, noises). We also discarded three trials for which a totally unrealistic FID was obtained, most likely due to handwriting errors when filling out the field protocol. This left us with a total of 408 successful scaring trials, of which 146 were carried out before noon and 262 in the afternoon (Table 1).
To test the last of the hypotheses (see Introduction), scaring trials were grouped as either ‘morning’ (before 12 AM) or ‘afternoon trials’ (after 12 AM) (Table 1). Although geese may forage at night under some circumstances, most if not all geese in our study area in winter spend nights in roost sites where they cannot forage much (lakes, iced-over wetlands, sheltered sea bays; Olsson, Månsson and Elmberg, unpublished).
We also noted crop stage and crop type in fields from which geese were scared, but due to sample size restrictions it was not possible to analyze these data statistically when broken down by species, flock type, flock size, and time of day. Data from all three winters of study were pooled before subsequent analyses, as they were all benign for geese weather-wise.
Restrictions: bean geese, pink-footed goose, and mute swan
Although the two bean goose taxa fabalis and serrirostris (here regarded as subspecies) can be identified when seen well, the latter is not always the case. During our field work there were individuals that could not be identified to subspecies in 41 of the 102 trials involving bean geese. Out of the 102 bean goose trials there were 53 that positively had fabalis in the flock and 41 that had serrirostris. In many cases flocks contained both subspecies. For these reasons we subsequently did not treat the two subspecies separately; hence all trials involving bean geese are termed as concerning bean goose in the wide sense (Anser fabalis sensu latu) (Table 2). Mute swan and pink-footed goose were recorded in just one trial each and were excluded from further analyses.