Introduction

Populations of many common farmland birds faced a rapid decline throughout the last decades in Europe (Gregory et al. 2019). Several changes in the agricultural landscape are assumed to account for this development. The intensification of agriculture — arable cropland as well as grassland — is characterized by homogenization and might have negatively affected farmland bird populations (Chamberlain et al. 2000; Donald et al. 2001; Geiger et al. 2010b). Rearrangement of parcels leads to increased field sizes and is thereby causing reduced availability of complex habitat structures accompanying fields, e.g. hedgerows and field margins (Benton et al. 2003). In addition, increased use of mineral fertilizers and pesticides may have negatively affected invertebrates (Geiger et al. 2010a), serving as food for many bird species during the reproductive period (Taylor et al. 2006). Ongoing changes in agricultural practices, thus, will also pose a threat to biodiversity of natural flora and fauna in future (Green et al. 2005).

Agricultural landscapes under organic farming seem to provide adequate habitats for many farmland birds in spring and summer during breeding seasons (Chamberlain et al. 1999; Fischer et al. 2011), in autumn during migration (Danhardt et al. 2010) and in winter (Chamberlain et al. 2010; Goded et al. 2018). Organic farming is characterized by high and wide hedgerows accompanying fields, small field sizes and rotational practices (Chamberlain et al. 1999). Thereby, organic farming provides more heterogeneous landscape types, finally leading to increased habitat diversity (Belfrage et al. 2005; Norton et al. 2009). The lack of chemical pesticides and inorganic fertilizers in organic farming is also positively affecting farmland birds (McKenzie and Whittingham 2009; McKenzie et al. 2011). And indeed, the agricultural landscape is still home to farmland birds.

The Grey Partridge (Perdix perdix) is one of the most important indicator species for biodiversity in agricultural areas due to its complex and large habitat requirements (Potts 2012). Another typical inhabitant of agricultural landscape is the Common Pheasant (Phasianus colchicus) (Burger 1988). These two small game species have benefited from traditional agriculture characterized by small farms and yield sizes (Guerrero et al. 2011). The European Red List Assessments showed that Grey Partridge (BirdLife International 2015a) and Common Pheasant (BirdLife International 2015b) are still widespread in Europe. Both species are evaluated as ‘Least Concern’ in Europe; however, while the population trend of Common Pheasants in Europe is increasing (BirdLife International 2015b), the Grey Partridge population trend appears to be decreasing (BirdLife International 2015a). When taking a closer look at the population development of these two bird species within Austria, it is striking that over the past 23 years (1998–2020), there has been a sharp decline in Grey Partridge populations (− 81%), while the Common Pheasant population has decreased less (− 27%) (Teufelbauer and Seaman 2021), raising the question, whether decreasing population trends might be related to the availability of suitable habitats in agricultural landscapes (see Harmange et al. 2019). Moreover, it is unclear, whether the remaining habitats are used by both species (habitat overlap) or exclusively used by either one or the other species. Species-specific knowledge is necessary to gain insight into habitat selection (Harmange et al. 2019); however, interspecific competition can shape habitat choice and thereby the species’ realized niches (Wang et al. 2020). Until now, only a few studies have focussed on the effects of interspecific competition on habitat selection of bird species having a strong association to agricultural areas (Robillard et al. 2012; Rinaud et al. 2020).

Grey Partridges and Common Pheasants prefer agricultural landscapes with a high proportion of cereals (Robertson et al. 1993; Harmange et al. 2019) and diverse crop cultivation (Henderson et al. 2009; Ronnenberg et al. 2016). Both species overwinter in groups, show breeding behaviour in spring, breed on the ground within dense vegetation (Anderson 2002; Potts 2012), and their chicks are insectivorous (Green 1984; Hill 1985; Browne et al. 2006). During breeding season, Grey Partridges seem to prefer wild vegetation (Panek 2002) and Common Pheasants are often found in areas with meadows and tree plantations (Chiatante and Meriggi 2022). However, while Grey Partridges are socially and genetically monogamous bird species (Vaněčková 2011), Common Pheasants are mostly polygynous (Ridley and Hill 1987). In addition, male Common Pheasants display very noticeable by following a ritual consisting of several phases (Czyżowski et al. 2020), compared to male Grey Partridges (Glutz von Blotzheim et al. 1973) in spring. Pheasant cocks look for open areas when calling (Lachlan and Bray 1976), as the display behaviour is often accompanied by visual signals like posture or fluttering of the wings (Czyżowski et al. 2020). In contrast, male Grey Partridges are calling and standing upright in front of a female (Glutz von Blotzheim et al. 1973).

Similarities in the choice of habitat used for breeding and foraging might lead to competition for resources available within the agricultural landscape (see Potts 1970; McCrow 1982). It is assumed that interspecific competition increases once habitat overlap of two species extends above a tolerable degree (Pianka 1984), thereby probably affecting individual fitness (Martin and Martin 2001). To overcome this issue, species usually adapt their habitat selection, but it is unclear whether the used habitats of Grey Partridges and Common Pheasants overlap in agricultural landscapes under conventional and organic farming.

To gain insight into interspecific habitat selection of farmland bird species, we collected data on environmental habitat use of Grey Partridge and Common Pheasant. We analysed habitat selection of both species in spring by calculating Manly’s selectivity ratios, focussing on agricultural management systems (organic vs. conventional) and small-scale habitat parameters (agricultural fields). In addition, habitat overlap was studied using Broennimann’s framework of implementing a principal component analysis to originally study niche selection (Broennimann et al. 2012), describing the environmental parameters in the agricultural landscape. We assume that study plots with high habitat heterogeneity and high surface area covered by fields under organic farming are preferably used by Grey Partridges and Common Pheasants. In addition, we assume that Common Pheasants prefer open areas (e.g. agricultural areas without mature crops) in spring, since these sites provide an ideal habitat for cocks’ display behaviour.

Material and methods

Study area

Habitat use of Grey Partridges and Common Pheasants was examined in our study area (1064 ha) located in Lower Austria (48°25′51.35″N, 16°30′51.58″E), northeast of the Austrian capital Vienna in the district of Mistelbach. Within the district, 10–20% of the livestock farms were managed according to organic farming principles in 2017 (Rech et al. 2020). The study area itself is characterized by a flat to slightly hilly landscape and is strongly dominated by arable farmland. The northern region of our study area is dominated by conventional farming, while organic agriculture predominates in the southern part (Fig. 1). The climate within the study area is characterized by a conspicuous lack of precipitation (mean annual precipitation ranges between 400 and 600 mm), hot summers (mean monthly temperature June–August: 20.9 °C) and moderately cold winters (mean monthly temperature December–February: 1.6 °C, Höfler et al. 2020). Two adjacent hunting grounds existed within the study area (see also Cybulska et al. 2020). Predator control (e.g. Red Fox Vulpes vulpes, Carrion Crow Corvus corone) took place in both hunting grounds based on the legal foundation. The hunting season for Grey Partridges and Common Pheasants in the study area starts on 21st September and 1st October and ended on 30th November and 31st December, respectively. Within the whole study area, annual harvest rates of Common Pheasants range between 5 and 6 individuals per 100/ha, while Grey Partridges are not shot in the study site. Neither Grey Partridges, nor Common Pheasants are released in the hunting estates within the study region.

Fig. 1
figure 1

Map of the study area located in Lower Austria. Occurrence of Grey Partridges and Common Pheasants was assessed within 29 study squares in spring 2018

(Source: INVEKOS Schläge Österreich 2018: © Agrarmarkt Austria, CC BY 4.0; Administrative boundaries: © EuroGeographics, Eurostat/GISCO: https://ec.europa.eu/eurostat/web/gisco/geodata/reference-data/administrative-units-statistical-units/countries#countries20, access date: 17-Aug-2022)

Full size image

A grid was created, placed over the study area and occurrence of Grey Partridges and Common Pheasants was determined (see Fig. 1). A grid size of 300 × 300 m (9 ha) was small enough to have an overview of the terrain within each individual square. The open and semi-open landscape allowed for an easy detection of both species, especially because there was no to little vegetation on the agricultural fields at this time of year. In addition, the grid size approximately reflects the home range of Grey Partridges (winter: 14.4 ha, spring: 6.8 ha, Buner et al. 2005) and Common Pheasants (11.1 ha, Draycott et al. 2009) during breeding season. Grid squares were regularly spaced and a total of 29 squares were systematically selected. Unsuitable habitats for both species (e.g. settlements and forests) were avoided. Each square touched other squares only at corner points to minimize spatial autocorrelation. The distance between squares dominated by conventional farming in the northern part of the study site and squares dominated by organic farming in the southern part of the study site varied between a minimum of 424 and a maximum of 4460 m (mean ± SD, 2266 ± 847 m).

Occurrence of Grey Partridges and Common Pheasants

Grey Partridge and Common Pheasant occurrences were assessed within two field trips, which were performed by a single field observer. The two field trips were conducted in spring between 5th March and 6th April 2018 (see Online Resource, Table S1). The order of monitoring squares was random, but the minimum time interval between the two repeated observations of a study square was at least 7 days (max: 30 days, mean ± SD, 15.1 ± 5.7 days). Sampling effort within and between the two field trips was equal. In case a square was visited in the early morning in the first field trip, the second field trip was made in the evening or vice versa. Thus, both species were counted once in the early morning (starting 30 min before sunrise) and once in the evening (starting 30 min before sunset), respectively. Survey points were located in the centre of every grid square. Thereby, distance between neighbouring survey points was at least 424 m, reducing probability of double counting. When the field observer approached the survey point, birds had 3 min to settle, before the count itself started. Within each square, the landscape was visually observed and listened for spontaneously calling males. In addition, the observer used playbacks of Grey Partridges for 1 min and counted the answering males during the next 3 min after playback (Panek 1998), regardless of whether Grey Partridges were seen or heard before. We are aware that playback calls affect detection probability of Grey Partridges (Kasprzykowski and Goławski 2009; Warren et al. 2018), but since playback was used in all squares, once in the morning and once in the evening, we assume that sampling effort and detection probability of Grey Partridges was uniform. There are no indications that Grey Partridge calls affect Common Pheasant behaviour. Surveys were only conducted during good weather conditions (no rainfall, fog or strong winds). Location and movement of each male Grey Partridge or male Common Pheasant was mapped in a field map to avoid false double counts. The count period lasted for 8 min at each count point.

Land use

We digitally analysed land cover within our 29 study squares by using the software ArcGIS version 10.7 (ESRI) and open government data (INVEKOS Schläge Österreich 2018, © Agrarmarkt Austria, CC BY 4.0) to quantify the surface area (hectare, ha) covered by different habitat types within each square (Table 1). Within each field, the habitat type and its area were determined during fieldwork. Moreover, two local hunters provided information on local farming practices (organic vs. conventional). In Austria, organic farming follows the EU Regulations No 834/2007 and 889/2008, including amendments (Rech et al. 2020). Organic farmers are certificated through a control body and controlled by independent inspection bodies (Rech et al. 2020).

Table 1 Proportions of habitat types covering the 29 study plots and available for both species and number of Grey Partridges (GP) and Common Pheasants (CP) detected within these habitat types
Full size table

To study differences between fields under organic or conventional farming, ground vegetation characteristics at three random points within each agricultural field were determined. Within three 1 × 1-m plots delimited in a habitat type-specific field, ground vegetation density (%) in five height classes (classified as 0, 1–5, 6–25, 26–50, and > 50 cm) covering the soil was measured. We also collected data on additional “border” habitat structures, not linked to organic or conventional farming, that Grey Partridges and Common Pheasants might use, such as hedgerows or field paths (including linear green strips alongside paths). Paved roads, settlements and forests were assumed to be unsuitable habitats for both species. Serving as a proxy of study square land use type heterogeneity, we calculated Shannon–Wiener-indices using the R package “vegan” (Oksanen et al. 2022).

Habitat selection

Habitat selection was studied by calculating Manly’s selectivity ratios (wi) (Manly et al. 2002). Values of wi < 1 indicate avoidance of habitat types, values of wi > 1 indicate a preference of habitat types and values of wi = 1 indicate that the habitat type is used in proportion to the availability in the study site. For easier interpretation, selection ratios are standardized (Bi) and add to 1. With a value ranging from 0 to 1, Bi gives the estimated probability that a habitat type would be the next one selected, if all habitat types were equally available. We assume that habitat types within the 29 squares were available for all animals within the study site because average distance between squares dominated by conventional and organic farming (mean ± SD, 2266 ± 847 m) was similar to dispersal distances found for Grey Partridges (see Šálek and Marhoul 2008; Rymešová et al. 2013) and Common Pheasants (Felley 1996). Therefore, analyses were made at the population level (Manly design I). The relative amount of all different available habitat types within the 29 squares was compared to the proportion of habitat types which were used by Grey Partridges or Common Pheasants. We therefore used the location of each male Grey Partridge and male Common Pheasant detected within the first and the second field trips, which were mapped throughout the field trips in a field map, to assess the used habitat type. Calculations were done by using the R package “adehabitatHS” (Calenge 2006) and estimated significance of selection of a particular habitat type by conducting log-likelihood chi-squared tests, using a corrected Bonferroni level. Frequencies of detected Grey Partridges or Common Pheasants for each habitat type should be five or more for the chi-squared test to be valid; otherwise, results have to be interpreted with caution (Manly et al. 2002). Therefore, we calculated Manly selectivity indices (i) on a coarse level for the two farming practices (organic vs. conventional) and all other remaining habitat types (others) combined (habitat categories, Bonferroni level 0.017, see Table 2) and (ii) on a fine level focussing on the habitat types without culture, winter cereals, lucerne and fallow land separately, and combined for the remaining rare farming cultures (vineyard, green manure, fennel, oilseed rape, poppy, orchard), the remaining rare grasslands (meadow, grazing area) and all other remaining habitat types (hedgerow, field path, roads, forests, water bodies and settlements) (habitat types, Bonferroni level 0.007, Table 2).

Table 2 Manly’s selection ratios and indices for habitat categories and types used by Grey Partridges and Common Pheasants that were detected within the two field trips. Analyses were performed on a coarse level focussing on the habitat categories (organic, conventional and other) and on a fine level focussing on habitat types (without culture, winter cereals, lucerne, fallow land, other farming cultures, other grassland and other)
Full size table

Habitat overlap, equivalency and similarity

More specific habitat analyses on the effect of farming practices in agricultural fields (organic vs. conventional) were then performed using the methodology proposed by Broennimann et al. (2012) and the R package “ecospat” (Di Cola et al. 2017). To avoid duplicate occurrences, pseudo-replication or effects of imperfect detection when analysing the data, we combined the datasets of the two field trips. Therefore, we determined occurrence (presence vs. absence) of Grey Partridges or Common Pheasants for each study square by defining “presence” of the species, when at least one individual of the corresponding species had been recorded once, either throughout the first or throughout the second field trip (see Online Resource, Table S1).

We then analysed habitat overlap of both species across the grid squares. Highly correlated environmental variables were transformed into two uncorrelated linear combinations (axes) by using a principal component analysis (PCA-env, see Broennimann et al. 2012, for detailed information on the procedure). The habitat types included in the PCA were without culture, winter cereals, lucerne and fallow land (Table 1). In addition, the Shannon–Wiener Index of each square was included in the PCA to account for habitat heterogeneity within squares. The PCA scores of the first two axes for both species were projected onto a grid of cells (100 × 100 cells), where each axis summarizes all the selected environmental variables and each cell corresponds to a unique set of environmental conditions (Broennimann et al. 2012). Then, smoothed density of occurrence for each species was estimated using a Gaussian kernel density function and overlap between species was visualized (Di Cola et al. 2017).

Habitat overlap between species and within species over time was quantified by calculating the metric Schoener´s D (Schoener 1970; Broennimann et al. 2012), which ranges from 0 (no overlap) to 1 (total overlap). Afterwards, habitat equivalency and habitat similarity were tested, following a statistical framework proposed by Warren et al. (2008), originally focussing on niche selection and adapted for our purpose to study habitat selection; while the first test predicts that used habitats between species or within species over time are identical, the latter predicts that used habitats are more similar than expected by chance (niche conservatism, Wiens and Graham 2005; Warren et al. 2008). Thereby, habitat equivalency tests only consider the exact space occupied by the species (conservative), whereas habitat similarity tests also account for the surrounding environmental space where the species occur (less conservative) and are performed in both directions (species 1 vs. species 2, species 2 vs. species 1) (for details see Broennimann et al. 2012). For both analyses, observed D values of the species were compared to the distribution of simulated D values, which were based on 1000 iterations. A p-value < 0.05 was used to evaluate significance of habitat equivalency and similarity. Thus, if the observed D values fall within 95% (p > 0.05) of the simulated D values, the hypotheses of habitat equivalency or habitat similarity cannot be rejected (Broennimann et al. 2012). Dynamics of the environmental variables contributing most to the two PC axis and affecting species’ occurrence density were visualized. All analyses were performed using the software R 3.6.2 (R Development Core Team 2019). Mean ± standard deviation (SD) are shown where appropriate.

Results

Occurrence of Grey Partridges and Common Pheasants

Within the first field trip, Grey Partridges and Common Pheasants were detected in 41.4% (n = 12) and 65.5% (n = 19) of all study squares, respectively (see Online Resource, Table S1). The number of individual Grey Partridges detected within a square of 300 × 300 m ranged from 0 to 5 (in total: 31 individuals), with a Grey Partridge density of 11.9 individuals/km2 in the 29 study squares. The number of Common Pheasants detected within a square ranged from 0 to 3 (in total: 27 individuals; density, 10.3 individuals/km2). Both species were detected within less study squares during the second field trip. Grey Partridges and Common Pheasants were detected on 31.0% (n = 9) and 58.6% (n = 17) squares (see Online Resource, Table S1). While the number of Grey Partridges counted within the 29 study square decreased (in total: 12 individuals, range: 0 to 3 individuals per square; density: 6.5 individuals/km2), the number of Common Pheasants increased (in total: 36 individuals, range: 0–4 individuals per square; density: 13.8 individuals/km2) from the first to the second field trips. When combining the data of the two field trips, a total of 43 Grey Partridges and 63 Common Pheasants were detected within 13 and 24 study squares, respectively.

Land use and habitat heterogeneity

Almost half of the surface of all 29 study squares (mean ± SD, 4.4 ± 3.1 ha per square) was covered by agricultural land under conventional farming (see Table 1). Slightly less surface (mean ± SD, 3.6 ± 3.4 ha per square) was covered by agricultural land under organic farming. The remaining surface area of the 29 study squares (mean ± SD, 1.0 ± 0.9 ha per square) was covered by other land use categories (e.g. hedgerows: mean ± SD, 0.6 ± 0.9 ha per square, field paths: mean ± SD, 0.3 ± 0.2 ha per square of total surface area). Roads, forests, water bodies and settlements covered each less than 1% of the total 29 study square surface area (together 1.7%). In total, only 5 study squares did not contain any arable field under organic farming at all. In contrast, all 29 study squares did contain arable fields under conventional farming. Habitat heterogeneity within study squares varied (Shannon–Wiener-indices: range: 0.9–2.2, mean ± SD, 1.4 ± 0.4).

Almost one third of the surface of all 29 study squares was covered by arable land, which had been prepared for cultivation, but crops had not yet started to grow when field work was conducted (without culture: mean ± SD, organic: 1.3 ± 1.8 ha per square; conventional: 1.3 ± 1.7 ha per square, Fig. 2). Winter cereals (e.g. winter wheat Triticum aestivum) were cultivated on 28.8% of the total surface area (mean ± SD, organic: 1.2 ± 1.7 ha per square; conventional: 1.4 ± 1.4 ha per square). Winter cereals had just started to germinate and grow in spring. Hence, vegetation density measured at different height classes in fields under organic and conventional management was similar (Fig. 2). About 10.5% of the total surface area was covered by lucerne (Medicago sativa) (mean ± SD, organic: 0.5 ± 1.1 ha per square; conventional: 0.4 ± 1.2 ha per square). Agricultural fields with lucerne under organic farming had a high vegetation density at height class 1–5 cm, whereas a high proportion of agricultural fields with lucerne under conventional farming were barely covered with any vegetation (Fig. 2).

Fig. 2
figure 2

Ground vegetation density at different height classes in agricultural fields under organic and conventional farming

Full size image

The remaining arable area was covered by the farming cultures vineyards (mean ± SD, organic: 0.1 ± 0.3 ha per square; conventional: 0.5 ± 1.0 ha per square, see Online Resource, Fig. S1), green manure (mainly field been Vicia faba and phacelia Phacelia tanacetifolia) (mean ± SD, organic: 178.5 ± 755.3 m² per square; conventional: 1983.8 ± 10,497.4 m² per square, Fig. S1), fennel (Foeniculum vulgare) (mean ± SD, organic: 0.2 ± 0.5 ha per square, Fig. S1), oilseed rape (Brassica napus) (mean ± SD, conventional: 0.1 ± 0.3 ha per square, Fig. S1), poppy (Papaver somniferum) (mean ± SD: conventional, 0.1 ± 0.4 ha per square, Fig. S1) and orchard (mean ± SD, conventional: 132.5 ± 498.9 m² per square, Fig. S1). Fallow land occurred within 17 study squares and covered 5.1% of the whole study surface area (mean ± SD: organic, 0.1 ± 0.3 ha per square; conventional: 0.3 ± 0.7 ha per square). Vegetation density of different height classes of fallow land under organic and conventional farming was similar (Fig. 2). The other rare grassland types occurring within our study site were meadows (mean ± SD: organic, 0.1 ± 0.3 ha per square; conventional: 0.1 ± 0.2 ha per square, Fig. S1) and grazing areas (mean ± SD: organic, 0.1 ± 0.3 ha per square, Fig. S1).

Habitat selection

Grey Partridges preferred habitats under organic farming and avoided habitat under conventional farming (Table 2). There is also an indication that Grey Partridges avoid other habitat, but as only a single Grey Partridge was detected within this habitat category, the result has to be interpreted with caution. Habitats under organic farming (Bi = 0.709) were more than two times as likely to be selected by Grey Partridges as habitats under conventional farming (Bi = 0.207, Table 2). In contrast, Common Pheasants seemed not to select specific habitat categories; indicating a proportional use to their availability (Table 2). Most of the Grey Partridges and Common Pheasants were detected within fields covered by winter cereals, followed by fields without culture, fields covered by lucerne and fallow land (in descending order, Table 2). However, neither Grey Partridges nor Common Pheasants seemed to prefer specific habitat types per se.

Overlap and similarity in habitat selection

The first two axes of the PCA captured 28.5% and 19.7% of the variation within the environmental variables (Fig. 3). The land cover variable “winter cereals.org” and “without culture.con” contributed most to PC1, whereas “Shannon-diversity index”, “fallow land.org” and “without culture.org” contributed most to PC2 (Online Resource, Table S2).

Fig. 3
figure 3

Correlation plot showing the contribution of environmental variables on the two main axes of the principal component analysis (PCA) which describes the environmental space in the study site

Full size image

Habitat types used by Grey Partridges showed a relative high overlap with habitats used by Common Pheasants (D = 64.2%, see Fig. 4). The observed D value of Grey Partridges and Common Pheasants fell within the density of 95% of simulated D values, thus, the hypothesis of retained habitat equivalency could not be rejected (p = 0.484, Online Resource, Fig. S2). Habitats used by Grey Partridges were not more similar to habitats used by Common Pheasants than expected by chance (p = 0.067, Fig. S2), but used habitats of Common Pheasants were more similar to habitats of Grey Partridges than expected by chance (p = 0.002, Fig. S2).

Fig. 4
figure 4

Occupancy of the two-dimensional environmental space used by a Grey Partridges, b Common Pheasants and c habitat overlap between species (Grey Partridge in light grey and Common Pheasant in grey, overlap in dark grey) monitored in spring (5th March until 6th April 2018). The grey gradient in figure (a) and (b) corresponds to the increase in occurrence density of the species. The solid line in the habitat overlap plot (c) corresponds to the limit of the environmental space, which was available in the 29 study squares. The dashed line corresponds to the 75% most frequently available conditions

Full size image

Habitat dynamics of Grey Partridges and Common Pheasants were affected by environmental variables (Fig. 5). Density of occurrence of both species shifted towards squares with high habitat heterogeneity, with a more pronounced effect in highly heterogeneous squares being detected in Grey Partridges (Fig. 5). While Grey Partridges seemed to occur on study squares with a high amount of surface area not yet covered by any vegetation, covered by winter cereals and fallow land under organic farming, Common Pheasants seemed to prefer squares without culture under conventional farming (Fig. 5).

Fig. 5
figure 5

Habitat dynamics of the environmental variables contributing most to the two PC axis and affecting species´ occurrence density of Grey Partridge and Common Pheasant. Smoothed and rescaled species densities (ranging between 0 and 1) along environmental variables are shown for Grey Partridge (dashed lines) and Common Pheasant (dotted lines) ranges. The solid black contour line delimits the 100th quantile of the density at the land use category

Full size image

Discussion

Grey Partridges and Common Pheasants are typical inhabitants of agricultural landscapes (Potts 2012; Ronnenberg et al. 2016), and we showed in our study that the agricultural landscape was used by Grey Partridges and Common Pheasants and that habitats used by both species overlapped to a high extent. Grey Partridges preferred squares with agricultural farmland under organic farming, while Common Pheasants did not show a clear preference for organic or conventional farming per se. Study squares with high habitat heterogeneity seemed to have a higher importance to Grey Partridges in comparison to Common Pheasants. However, results have to be interpreted with caution, as both species were observed only during two field trips in spring. Datasets of the two field trips were combined to avoid that habitat overlap, and similarity analyses were affected by imperfect detection; an issue that can also be accounted for by analysing data using occupancy models (MacKenzie et al. 2006), when an adequate number of sampling replicates is available (MacKenzie and Royle 2005; Paniccia et al. 2018). In addition, the two main axes of the principal component analyses, which described the environmental variables to study habitat similarities of both species explained only a part (48.2%) of the environmental variation in the study squares.

Grey Partridges usually form family groups in late summer and stay together in coveys during winter (Potts 2012). In spring, they start to separate again; males show courtship display behaviour; and finally breeding pairs are formed (Potts 2012; Rinaud et al. 2020). Common Pheasants concentrate on traditional areas, which provide food and shelter to overwinter in groups (Gates and Hale 1974). Cock dispersal from wintering areas, as well as display of territorial behaviour starts in late March (Gates and Hale 1974). Survival of both species is dependent on habitat structures, providing not only shelter from severe weather conditions and predators (Gottschalk and Beeke 2014), but also access to food (Anderson 2002; Bro et al. 2004). Since predation risk close to or along linear or rare habitat structures (e.g. hedgerows and field boundaries) is high (Morris and Gilroy 2008; Černý et al. 2020), extensive usage of such structures might be especially risky for species roosting on the ground. Using telemetry, Rantanen et al. (2010) showed that probability of survival of Grey Partridges in spring increased when breeding pairs stayed in crop fields, whereas survival decreased when they stayed on field margins. Maladaptation, which was proven for newly released Grey Partridge pairs preferring high risk habitats (Rantanen et al. 2010), seemed not to occur in wild birds in our study site, since density of occurrence of Grey Partridges and Common Pheasants was high in squares with high habitat heterogeneity. In contrast, a study performed in peri-urban farmland in the vicinity of the city of Vienna (Austria) showed that Grey Partridges’ selected habitats with low habitat diversity (Hille et al. 2021).

Differences in vegetation density at various height classes in agricultural fields under organic and conventional farming were barely visible (except in lucerne) in our study; thereby indicating that vegetation height might not be the most important factor affecting habitat choice of Grey Partridges and Common Pheasants in spring. Instead, other factors (e.g. food availability) associated with farming practices might have affected habitat suitability. Food of both species mainly consists of grains and leaves of cereals during winter (Gates and Hale 1974; Huss 1983), but insect diet is of high importance during chick rearing (e.g. sawfly larvae, Browne et al. 2006; Potts 2012; Warren et al. 2017). Thus, habitats providing such food items are necessary to successfully raise chicks (Potts 2012).

In organic farming, the usage of chemical inputs (e.g. insecticides, herbicides and fungicides) and fertilizers is restricted. The reduced use of synthetic pesticides is associated with an increase of invertebrate densities and weed cover; thereby positively affecting availability of insects for insectivorous birds and chick survival rates (Chiverton 1999). And indeed, a meta-analysis showed that organic farming has a positive effect on birds, predatory insects (e.g. carabids and spiders) and weed abundance (Bengtsson et al. 2005; Fuller et al. 2005; but see also Gabriel et al. 2010). Grey Partridges, known to be affected by pesticides causing dramatic reductions in food availability (Kuijper et al. 2009), might therefore have occupied study squares mainly consisting of agricultural fields under organic farming in our study site. Nesting sites of Grey Partridges can be located within cereal fields in open landscapes (Bro et al. 2000), permanent vegetation (Panek 2002; Gottschalk and Beeke 2014) and unmanaged habitats (Černý et al. 2020). Mechanical techniques to control weeds in cereal fields, mainly used in organic farming, can however be problematic during breeding seasons, as nests can be destroyed and chicks or incubating females can be killed (Newton 2004). Pesticides can, however, also affect farmland birds directly, if pesticide-treated seeds are consumed by the animals during the sowing season (Fernández-Vizcaíno et al. 2022). Several experimental studies showed that seed-treatment can affect sexual hormones (Fernández-Vizcaíno et al. 2020), offspring development (Gaffard et al. 2022) and reproductive performance (Lopez-Antia et al. 2018). In addition, imidacloprid-treated seeds consumed by granivorous birds (e.g. Grey Partridge) can have lethal effects (Millot et al. 2017). A study of Draycott et al. (2009) has shown that Common Pheasants preferred rotational set-asides planted with seed mixtures of grasses and forbs during brood-rearing, thereby indicating the importance of insect-rich foraging habitats. In addition, Common Pheasant harvest rates seem to be positively related to the proportion of agricultural fields under organic farming (Holá et al. 2015). In our study, Common Pheasants seemed to rather use agricultural fields without culture under conventional than under organic farming. The usage of this habitat type by Common Pheasants cannot be explained by a more pronounced selection of study squares with low vegetation, as vegetation height of agricultural fields without culture under conventional and organic farming was rather similar. However, the detection of Common Pheasants within these sites might be explained by two, non-exclusive phenomena: (i) Common Pheasants might either be less specialized in their habitat and food selection (Potts 1970) and thus have not shown any preference for agricultural sites under organic or conventional farming per se or (ii) interspecific competition for habitats might have affected habitat choice of Common Pheasants (Wang et al. 2018, 2020) and Grey Partridges. To test the mechanisms of interspecific competition between both species in more detail, field experiments (e.g. using telemetry) have to be conducted to compare habitat selection on sites, where both species occur individually or together (see also Rinaud et al. 2020) and subsequent effects on fecundity and survival have to be monitored (Martin and Martin 2001). Common Pheasants can compete with Grey Partridges for nesting habitats (McCrow 1982; Bro et al. 2000; Anderson 2002) and invertebrate food resources (Potts 1970; Gates and Hale 1974; Green 1984), but can also transmit parasites (Tompkins et al. 1999, 2000). Assuming that interspecific competition between Grey Partridges and Common Pheasants could exist and affect population trends (Vrezec 2006; McCrow 1982), awareness should be raised with regard to large-scale captive-reared and released pheasants (Bicknell et al. 2010).

The availability of suitable habitat is of high importance to reduce the sharp decline in farmland bird populations (Harmange et al. 2019). Resources provided within these habitats (e.g. nesting sites and food items) can be used by a variety of bird species and interspecific competition for breeding and foraging habitats can affect a species’ habitat selection, breeding success (Martin and Martin 2001) and thereby population trends (Vrezec 2006). Whether purely promoting organic farming is the key to halt the decline in Grey Partridge populations is, however, under debate, since arable land under “environmentally friendly management” (Marja et al. 2014) seems to provide enough suitable habitat for farmland birds (Kragten and de Snoo 2008). Moreover, some studies found no differences in Grey Partridge abundance between organic and conventional farming (Chamberlain et al. 1999, 2010), thereby indicating that a variety of other factors can also strongly affect abundance in agricultural landscapes. Measures that can be carried out by both, conventional and organic farmers (Bengtsson et al. 2005), are for example, the extent of pesticides and fertilizers used (Chiverton 1999), the choice of crop diversity (Ronnenberg et al. 2016), the crop sequence patterns (Joannon et al. 2008) and the provisioning of additional habitat structures (e.g. hedgerows: Batary et al. 2010; uncultivated habitats: Černý et al. 2020; manure heaps: Šálek et al. 2020). Indeed, recent studies showed that positive effects of organic farming are highlighted especially in homogeneous landscapes (Danhardt et al. 2010) or in regions with intensive agriculture (Kirk et al. 2020). Thus, even agricultural landscapes under conventional farming can provide valuable habitats for farmland birds, if overall landscapes heterogeneity is high (Benton et al. 2003; Smith et al. 2010).