Food or nesting place? Identifying factors limiting Wryneck populations

Conceived to combat widescale biodiversity erosion in farmland, agri-environment schemes have largely failed to deliver their promises despite massive financial support. While several common species have shown to react positively to existing measures, rare species have continued to decline in most European countries. Of particular concern is the status of insectivorous farmland birds that forage on the ground. We modelled the foraging habitat preferences of four declining insectivorous bird species (hoopoe, wryneck, woodlark, common redstart) inhabiting fruit tree plantations, orchards and vineyards. All species preferred foraging in habitat mosaics consisting of patches of grass and bare ground, with an optimal, species-specific bare ground coverage of 30–70% at the foraging patch scale. In the study areas, birds thrived in intensively cultivated farmland where such ground vegetation mosaics existed. Not promoted by conventional agri-environment schemes until now, patches of bare ground should be implemented throughout grassland in order to prevent further decline of insectivorous farmland birds.

Zonal tundra vegetation occupies slightly sloped watershed surfaces, weakly convex tops and gentle slopes of moraine hills and ridges with moderate snow cover and loamy soils (plakors). Environmental conditions of such sites are most relevant to macroclimate (Aleksandrova, 1971; Matveyeva, 1998). For the East European sector of the Arctic this vegetation was described in 30–70 years of last century by the Soviet geobotanists V. D. Aleksandrova (1956), ­ V. N. Andreyev (1932), I. D. Bogdanovskaya-Giye­nef (1938), A. A. Dedov (2006), A. E. Katenin (1972), Z. N. Smirnova (1938), who, following the dominant approach, attributed the described associations mainly to the moss vegetation type. In the Asian sector of the Arctic (Yamal and Taymyr peninsulas, Arctic Yakutia, Wrangel Isl.) and in Alaska some associations of zonal communities with Carex bigelowii s. str., C. bigelowii subsp. arctisibi­rica and C. lugens have been described according to Braun-Blanquet approach: Carici arctisibiricae–Hylocomietum alaskani Matveyeva 1994, Dryado integrifoliae–Caricetum bigelowii Walker et al. 1994, Salici polaris–Hylocomietum alaskani Matveyeva 1998, Carici lugentis–Hylocomietum alaskani Sekretareva 1998 ex Kholod 2007, Salici polaris–Sanionietum uncinatae Kholod 2007, Tephrosero atropurpureae–Vaccinietum vitis-idaeae Telyatnikov et Pristyazhnyuk 2012, Festuco brachyphyllae–Hylocomietum alaskani Lashchinskiy ex Telyatnikov et al. 2014. Our study area in the East European tundras (730 km of south–north and 550 km of west–east directions) covers 3 tundra subzones (arctic, typical and southern) and two floristic subprovinces (Kanin-Pechora and Ural-Novaya Zemlya) (Fig. 1). 7 associations (one with 5 subassociations) based upon 101 authors’ relevés as well 95 ones by geobotanists-predecessors were described or validated on plakors and habitats close to these. Zonal communities are comprised by thick multi-species moss layer formed by mesophylous bryophytes (Aulacomnium turgidum, Hylocomium splendens, Ptilidium ciliare, Racomitrium lanuginosum and Tomentypnum nitens), the presence of Carex bigelowii subsp. arctisibirica, Deschampsia borealis or D. glauca in the herb layer, the high abundance of dwarf-shrubs, the presence, but not always, of Dryas octopetala and shrubby willows. Their plant cover is closed or discontinuous with frost-boils (3-component module of patch of bare ground – rim – trough or 2-component one of flat surface – patches of bare ground — see Fig. 2, 3). Zonal syntaxa are the richest in species number, compare to all others because the placor habitats are moderate in such important environmental characters as moisture and nutrition of soil and snow depth. That’s why they contain, with the same constancy and sometimes abundance, some character species of alliances and classes of intrazonal vegetation: Kobresio-Dryadion Nordh. 1943 (dryad fell-fields on well drained snowless or poor snowy habitats with slightly carbonated loamy-gravelly soils at outcrops of bedrock) and Carici rupestris–Kobresietea bellardii Ohba 1974, Loiseleurio-Arctostaphylion Kalliola ex Nordhagen 1943 (dwarf-shrub and dwarf-shrub-lichen (often with Betula nana) communities on sandy soils) and Loiseleurio procumbentis–Vaccinietea Eggler ex Schubert 1960, Rubo chamaemori–Dicranion elongati Lavrinenko et Lavrinenko 2015 (dwarf-shrub-cloudberry-moss (Dicranum elongatum, Polytrichum strictum)-lichen communities of oligotrophic palsa and polygon peatlands) and Oxycocco-Sphagnetea Br.-Bl. et Tx. ex Westhoff et al. 1946. The basic syntaxon, whose communities occupy the placor habitats in the arctic tundra subzone (southern­ variant) is Salici polaris–Polytrichetum juniperini Aleksandrova 1956, described on the Southern Island of Novaya Zemlya (Table 1). Similar syntaxa in the typical tundra subzone are Carici arctisibiricae–Hylocomietum splendentis Andreyev 1932 nom. mut. propos. hoc loco (Table 5, Fig. 14–17) and Dryado octopetalae–Hylocomietum splendentis Andreyev 1932 nom. mut. propos. hoc loco salicetosum nummulariae (Bogdanov­skaya-Giyenef 1938) subass. nov. (stat. nov.), nom. corr. hoc loco, described by us and earlier by I. D. Bogdanov­skaya-Giyenef (1938) and Z. N. Smirnova (1938) on the Kolguyev Isl. (Table 2, Fig. 3, 5, 6); D. o.–H. s. caricetosum redowskianae subass. nov. hoc loco and D. o.–H. s. caricetosum arctisibiricae (Koroleva et Kulyugina in Chytrý et al. 2015) subass. nov. (stat. nov.) hoc loco (Table 4, Fig. 4, 9–13) — in the most eastern part of the studied area (Vaygach Isl., Yugorskiy Peninsula and Pay-Khoy Range); D. o.–H. s. typicum subass. nov. hoc loco (Tab­le 3), described by us with use the V. N. Andreyev (1932) relevés on Vangureymusyur Upland (Bolshezemelskaya tundra). In the southern tundra subzone the basic zonal association is Calamagrostio lapponicae–Hylocomietum splenden­tis ass. nov. hoc loco (Table 6, Fig. 20–22). Even small deviations from placor habitat conditions are reflected in the community species composition. In such habitats the following syntaxa are described: Deschampsio borealis–Limprichtietum revolventis Aleksandrova 1956 nom. mut. propos. hoc loco and Flavocet­rario nivalis–Dryadetum octopetalae Aleksandrova 1956 nom. mut. propos. hoc loco on gentle slopes and loamy soils, not in moderate soil moisture, but in wet or, on the contrary, well-drained ecotopes on the Novaya Zemlya (Table 1); Dryado octopetalae–Hylocomietum splendentis caricetosum capillaris subass. nov. hoc loco — on the deluvial tails, in the mid and lower parts of the gentle slopes in Bolshezemelskaya and Malozemelskaya tundras (Table 4, Fig. 2, 7, 8); Oxytropido sordidae–Hylocomietum splendentis ass. nov. hoc loco — in the Pakhancheskaya Bay area (the northern part of the Bolshezemelskaya tundra) on strongly sloping well drained slopes (Table 6, Fig. 18, 19). We attributed these syntaxa to zonal vegetation due to the presence of such taxa of its differential combination as shrub Salix glauca, dwarf-shrub Salix polaris, herbs Bistorta major, Carex bigelowii subsp. arctisibirica, Deschampsia borealis, D. glauca, Eriophorum brachyantherum, Juncus biglumis, Luzula arcuata, Pedicularis lapponica, Petasites frigidus, Poa arctica, Saxifraga hieracifolia, S. hirculus, Stellaria peduncularis, Valeriana capitata, mosses Aulacomnium turgidum, Hylocomium splendens, Ptilidium ciliare, Racomitrium lanuginosum, Tomentypnum nitens and lichens Lobaria linita, Nephroma expallidum, Protopannaria pezizoides, Psoroma hypnorum. This combination of taxa differentiates (by the presence, constancy, abundance) the zonal communities in studied area from vegetation of other classes (Carici rupestris–Kobresietea bellardii, Loiseleurio procumbentis–Vaccinietea, Oxycocco-Sphagnetea) (Table 7). The borders of many species area distribution are held in the East European tundras, so the variation of the community species composition along the latitudinal and longitude gradients is quite natural. Thus, in zonal communities Ledum palustre subsp. decumbens and Salix phylicifolia occur and Betula nana as well as hypoarctic dwarf-shrubs Arctous alpina, Empetrum hermaphroditum, Vaccinium uliginosum subsp. microphyllum­ and V. vitis-idaea subsp. minus are most active only in the southern tundra subzone; Salix polaris (its activity increases to the north) and, in some syntaxa, Dryas octopetala are common in the subzones of typical and arctic tundras. In zonal conditions shrubs Salix glauca, Betula nana (prostrate) and all hypoarctic dwarf-shrubs occur in the typical tundra subzone on the mainland and on Kolguyev Isl., while in the northern part of this subzone on Vaygach Isl. they are already absent, ­except the Vaccinium spp. (with low constancy). In the arctic tundra subzone there are no shrubs and hypoarctic dwarf-shrubs on plakors, while Salix polaris is abundant. We believe that these floristic differences of zonal communities can be considered as markers of their subzonal affiliation. A similar shift in species distribution on the latitudinal gradient is established (Matveyeva, 1998) for the zonal biotopes on Taymyr Peninsula. Some species (Arctagrostis latifolia, Cerastium regelii subsp. caespitosum, Saxifraga oppositifolia, Silene acaulis) have high constancy in zonal communities within the Ural-Novaya Zemlya subprovince, as opposed to the Kanin-Pechora one. Herbs Oxyria digyna, Papaver polare, Parrya nudicaulis, Pedicularis sudetica subsp. arctoeuropaea, Saxifraga cernua and S. cespitosa occur with high constancy only in zonal communities on Novaya Zemlya that brings them closer to syntaxa described in the arctic and typical tundra subzones on Taymyr Peninsula (Matveyeva, 1994, 1998). Already in 1994, N. V. Matveyeva stated the need to describe a new class for zonal vegetation. The name Carici arctisibiricae–Hylocomietea alaskani cl. prov. has been reserved for this class in Yalta’s conference on the classification of Russian vegetation (Lavrinenko et al., 2016), Prague’s “Circumpolar Arctic Vegetation Archive and Classification Workshop” (presentation by N. V. Matveyeva) and “Arctic Science Summit Week 2017” (Lavrinenko et al., 2017). We do not attribute the newly described syntaxa to alliance Dryado octopetalae–Caricion arctisibiricae Koroleva et Kulyugina in Chytrý et al. 2015, which was described at the base of 15 relevés by geobotanists-predecessors (V. N. Andreyev, A. A. Dedov) and as well the 11 ones by E. E. Kulyugina for zonal habitats in the East European tundras (Koroleva, Kulyugina, 2015). At least, it is necessary to revise this alliance, since the name of ass. Pediculari oederi–Dryadetum octopetalae (Andreev 1932) Koroleva et Kulyugina 2015 are not legitimate (nomen superfluum), ass. Salici reticulatae–Dryadetum octopetalae Koroleva et Kulyugina 2015 need to be revised and the rank of the third one (Dryado octopetalae–Caricetum arctisibiricae Koroleva et Kulyugina in Chytrý et al. 2015 was lowered by us (in this paper) to subass. Dry

Organisms do not only use personal but also social information for habitat selection. However, the use and relative importance of these two sources of information for settlement decisions remain unclear for many species. Conservation tools that attract target species to available or restored habitat rely on social information, while additional or interacting effects of personal information may limit their efficiency. Here, we investigated the effects of food availability (individually acquired personal information) and conspecific vocalization (social information) on habitat selection of a regionally endangered, migrating woodpecker, the Wryneck Jynx torquilla. First, we analysed the species' occurrence in relation to food availability and habitat characteristics in general. Then, we conducted a playback experiment, during both the prospecting period and the breeding period, in previously unoccupied plots with different conditions of food availability to disentangle the relative importance of food availability and conspecific vocalization for habitat selection during both the prospecting period and the breeding period. Wryneck occurrence increased with food availability (i.e. ant nest density) and slightly related to food accessibility (i.e. the proportion of bare ground). Food availability varied considerably among habitat types, with the highest values in vineyards and the lowest values in intensively managed grasslands. Within the experimental framework, high food availability increased the probability of both Wryneck prospecting and breeding. In turn, playback had a positive effect on Wryneck prospecting under poor but not favourable food conditions, but not on Wryneck breeding. Finally, playback did not exert a significant effect on breeding success. Practical implication: Both personal and social information seem to affect Wryneck habitat selection behaviour. However, food availability as individually acquired personal information of a habitat cue mainly influenced Wryneck settlement decisions that led to breeding. Consequently, playback as social information through a simulated conspecific social cue is a potentially useful tool for Wryneck conservation, and the risk of attracting Wrynecks into an ecological trap is minimal. Nevertheless, habitat quality, and in particular food availability, are more important and a crucial requirements for successful conservation management.

The Ortolan Bunting Emberiza hortulana is a long‐distance migrant that has suffered major population declines across much of its European breeding range. While northern populations are bound largely to farmland, Mediterranean populations are largely confined to habitats subject to recurrent wildfires. Habitat selection of the Ortolan Bunting was assessed in a recently burnt area in Catalonia at landscape and habitat scales. A Zero‐inflated Poisson procedure was used to model the abundance of birds in relation to landscape and habitat variables. The most parsimonious landscape model predicted the highest abundance on south‐facing slopes, with a gradient above 10°. The most parsimonious habitat model showed a positive quadratic effect of bare ground and regenerating oak Quercus spp., with predicted optima for abundance around 20–30% and 20% cover, respectively. There was a clear relationship between predicted abundance of the Ortolan Bunting and post‐fire regenerating oak shrubs. South‐facing, moderately sloping areas were favoured and bare ground was a key feature of the species’ habitat. A matrix combining patches of sparse oak shrubs and patches of bare ground appears to be the optimal breeding habitat in the Mediterranean. The maintenance or provision of similar habitat features, especially patches of bare ground, may prove crucial for the conservation of rapidly declining Ortolan Bunting populations on farmland across temperate Europe.

Distribution and abundance of hollow-bearing trees in urban forest fragments

Landscape heterogeneity and management practices drive habitat preferences of wintering and breeding birds in intensively-managed fruit-tree plantations

Mt. Kisokoma is located in the northern part of the central Japanese Alps. On the southwest-facing slope of Mt. Kisokoma, patches of bare ground are well developed. This paper presents the extending direction of the patches of bare ground in relation to wind direction, and seasonal changes in the amount of retreat of turf scarps measured with erosion bars. Extending direction of patches of bare ground correspond to the direction of the prevailing westerly wind, which is dominant throughout the year. The results show that the prevailing wind controls the extension of patches of bare ground. The turf scarps retreated most in the Baiu-season. Strong wind accompanied by heavy rain is observed several times during the season, and rain splash wash seems to be the dominant cause of retreat. Negligible retreat of the turf scarps in winter suggests that the turf scarps are protected by snow beds in winter. In spring and autumn, the combination of needle ice and wind erosion might have caused some retreat of the turf scarps.

Vegetation of spotted dwarf shrub-sedge-moss and dwarf shrub-lichen communities in the southern tundra subzone of the Taymyr Peninsula is assigned to 3 associations, including 2 new ones. For comparison, we used previously published relevés of ass. Hierochloo alpinae–Hylocomietum splendentis Telyatnikov, Troeva, Ermokhina et Pristyazhnyuk 2019 with 2 subassociations and 3 variants, into which the authors combined spotted dwarf shrub-lichen-moss communities on sandy loam soils and dwarf shrub-lichen communities on sandy soils on the Gydanskii and Tazovskii Peninsulas. As a result of the revision of these syntaxa and their comparison with new materials from the Taymyr Peninsula, we have revised some syntaxonomic decisions. The zonal vegetation is assigned to 2 associations: ass. Carici arctisibiricae–Hylocomietum alaskani Matveyeva 1994 (with regular-cyclic horizontal structure: patches of bare ground – rims – small troughs) and a new ass. Arctagrostio latifoliae–Caricetum arctisibiricae Telyatnikov et al. ex Lavrinenko in Lapina et Lavrinenko ass. nov. hoc loco (with continuous sedge-moss cover) in the class Carici arctisibiricae–Hylocomietea alaskani Matveyeva et Lavrinenko 2023. Dwarf shrub-lichen communities on sandy soils are assigned to a new ass. Asahino chrysanthae–Salicetum nummulariae (Khitun in Telyatnikov et al. 2021) Lapina et Lavrinenko ass. nov. (stat. nov.) hoc loco in the class Loiseleurio procumbentis– Vaccinietea Eggler ex Schubert 1960.

Agriculture of varying management intensity dominates fragmented tropical areas and differentially impacts organisms across and within taxa. We examined impacts of local and landscape characteristics on four groups of ants in an agricultural landscape in Chiapas, Mexico comprised of forest fragments and coffee agroecosystems varying in habitat quality. We sampled ground ants found in leaf litter and rotten logs and arboreal ants found in hollow coffee twigs and on tree trunks. Then using vegetation and agrochemical indices and conditional inference trees, we examined the relative importance of local (e.g. vegetation, elevation, agrochemical) and landscape variables (e.g. distance to and amount of nearby forest and rustic coffee) for predicting richness and abundance of ants. Leaf litter ant abundance increased with vegetation complexity; richness and abundance of ants from rotten logs, twig-nests, and tree trunks were not affected by vegetation complexity. Agrochemical use did not affect species richness or abundance of any ant group. Several local factors (including humus mass, degree of decay of logs, number of hollow twigs, tree circumference, and absence of fertilizers) were significant positive predictors of abundance and richness of some ant groups. Two landscape factors (forest within 200 m, and distance from forest) predicted richness and abundance of twig-nesting and leaf litter ants. Thus, different ant groups were influenced by different characteristics of agricultural landscapes, but all responded primarily to local characteristics. Given that ants provide ecosystem services (e.g. pest control) in coffee farms, understanding ant responses to local and landscape characteristics will likely inform farm management decisions.

ABSTRACTThe leaf litter of tropical wet forests is replete with itinerant ant nests. Nest movement may help ants evade the constraints of stress and disturbance and increase access to resources. I studied how nest relocation and environmental factors may explain the density, size, and growth of leaf litter ant nests. I decoupled the relationships among litter depth, food abundance, and nest availability in a 4‐mo manipulation of food and leaf litter in a community of litter‐nesting ants in a lowland wet forest in Costa Rica. Over 4 mo, 290 1 m2treatment and control plots were sampled without replacement. Nest densities doubled in response to food supplementation, but did not decrease in response to litter removal or stress (from litter trampling). The supplementation of food increased the utilization of less favored nesting materials. In response to food supplementation and litter trampling, arboreal ants established nests in the litter, and growth rates of the most common ants (Pheidolespp.) increased. Colony growth was independent of colony size and growth rates of the most abundant ants. In general, I conclude that litter‐nesting ant density is driven primarily by food limitation, that nest relocation behavior significantly affects access to resource and the demographic structure of this community, and that nest fission may be a method to break the growth–reproduction trade‐off.

Fire can have negative effects on the ant community by reducing species abundance through direct mortality, changes in resource availability, or foraging activity. Fire can also have positive effects, especially for opportunistic species preferring open or disturbed habitats. We assessed the direct effects of a large-scale fire on ant communities in open habitats (grassland and Carex) and moist forested peatland (birch and alder) sites in Biebrza National Park, testing three hypotheses: (i) the large-scale fire had more significant effects on ant nest density in forests than in open habitats, (ii) the post-fire ant diversity changes within sites are stronger in forests than open habitats, and (iii) ant species preferring disturbed habitats are favoured by the fire event. The fire had negative effects on ant nest density only in the Carex and grassland sites but not in the birch and alder sites, suggesting that fire had a stronger impact in open habitats than in forests. Temporal post-fire ant diversity changes within sites were stronger in forests than in open habitats. We observed higher beta diversity changes between the first and second year of the study in the burned forest sites due to colonisation, indicating a greater fire impact on species community composition followed by a higher recolonisation rate. Ant species preferring disturbed habitats were favoured by the fire. The seed-eating ant species Tetramorium caespitum, a thermophilous and opportunistic species, dominated the burned grassland site. This contrasts with other species, e.g., Lasius alienus, for which nest density decreased after fire, underlining the importance of food resource availability as a major driver of community changes after fire. Our study also underlines the importance of periodic biodiversity monitoring in conservation areas for assessing the recovery of the original status after disturbances and revealing possible habitat changes endangering the survival of local biotic communities.

Does Cattle Grazing Affect Ant Abundance and Diversity in Temperate Grasslands?

Cues Cattle Use to Avoid Stepping on Crested Wheatgrass Tussocks

Frequencies of plants with extrafloral nectaries were determined for two elevations in Jamaica. Extrafloral nectaries were found on 0.28 of the plants at sea level (Happy Grove, Portland) and 0.00 of the plants at 1310 m (Whitfield Hall, St. Thomas). Ant abundance, as indicated by discovery of and recruitment to baits, was greater at the lower elevation site. However, despite the apparent absence of plants with extrafloral nectaries, there were abundant ants at 1310 m. MUCH EVIDENCE suggests that extrafloral nectaries attract insects which defend plants against herbivores and/or seed predators (Elias and Gelband 1975; Bentley 1976, 1977a, b; Keeler 1977). Extrafloral nectaries are glands which are located anywhere on a plant except those sites involved in pollination. These glands produce an aqueous solution containing sugars and other compounds (Baker and Baker 1975, Bentley 1977a, Keeler 1977). At present, only Bentley (1976) has studied the distribution of extrafloral nectaries in a natural habitat. She reported a positive correlation between frequency of plants with extrafloral nectaries and ant abundance in tropical dry forest in Guanacaste, Costa Rica. I report extrafloral nectary frequency and ant abundance from two sites in Jamaica. Two sites were compared: Happy Grove, Portland, Jamaica (sea level, approximately 1700 mm annual rainfall, mean annual temperature 260C, mean monthly temperatures 24-28?), and Whitfield Hall, near Hagley Gap, St. Thomas (1310 m up Blue Mountain, 3500 mm annual rainfall, mean annual temperature 220 C, mean monthly temperatures 16-32?). Since there are no weather records for the specific sites, values are extrapolated from U.S. Weather Bureau (1966) and Clarke (1974). Significant human disturbance was seen at both sites. Studies were conducted close to trails used daily by local people. The transects at both sites ran from under the forest canopy (presumably second growth) out into partially open areas (early second growth at Happy Grove, coffee fields at Whitfield Hall). Frequency of plants with extrafloral nectaries was determined in four transects at each site. At approximately every meter along each transect, plants were scored as having or lacking extrafloral nectaries and ants. The presence of extrafloral nectaries was determined by observing ants feeding in a stereotyped manner, and then locating the nectary. Once a species was determined to have extrafloral nectaries, it was scored as such on subsequent encounters. Ants were present at most extrafloral nectaries; of the 70 plants with extrafloral nectaries observed, only 10 (14%) did not have ants on them. The frequency of plants with extrafloral nectaries determined by this method is an underestimate, since some species may produce extrafloral nectar at other times of the year (e.g., fruit nectaries). Ant abundances were estimated by using baits of canned corned beef and local commercial jelly. A pile of each food about 1 cm in diameter was placed on a separate piece of plastic, 25 cm2, in the litter at each station. Time until arrival of the first ant, type of ant, peak number of ants responding and number of ant species attracted were recorded at each bait. Representative ants from baits at both sites and from foliar nectaries of plants at Happy Grove were collected. The experiments were carried out for three hours each, from 09.00-12.00 hrs in December 1977. Results from the extrafloral nectary transects are given in table 1. No species with extrafloral nectaries was observed at Whitfield Hall. At Happy Grove, 28 percent of the observed plants had extrafloral nectaries. These differences are statistically significant (T-test; p p> 0.01 and p-0.005, respectively). Ants collected from baits at Hatpy Grove included Paratrechina longicornis (Latreille), Solenopsis sp. 152 BIOTROPICA 11 (2): 152-154 1979 This content downloaded from 207.46.13.75 on Fri, 08 Jul 2016 05:47:31 UTC All use subject to http://about.jstor.org/terms (not S. geminata) , and Tetramorium guineense (F.); ants collected at Whitfield Hall were P. longicornis, Crematogaster brevispinosa Mayr., and Monomorium floricola (Jerdon). Ants collected from extrafloral nectaries at Happy Grove were P. longicornis, Solenopsis sp. (same as above), and Wassmannia auro-

Pollinating insect populations, essential for maintaining wild plant diversity and agricultural productivity, rely on (semi)natural habitats. An increasing human population is encroaching upon and deteriorating pollinator habitats. Thus the population persistence of pollinating insects and their associated ecosystem services may depend upon on man-made novel habitats; however, their importance for ecosystem services is barely understood. We tested if man-made infrastructure (railway embankments) in an agricultural landscape establishes novel habitats that support large populations of pollinators (bees, butterflies, hoverflies) when compared to typical habitats for these insects, i.e., semi-natural grasslands. We also identified key environmental factors affecting the species richness and abundance of pollinators on embankments. Species richness and abundance of bees and butterflies were higher for railway embankments than for grasslands. The occurrence of bare (non-vegetated) ground on embankments positively affected bee species richness and abundance, but negatively affected butterfly populations. Species richness and abundance of butterflies positively depended on species richness of native plants on embankments, whereas bee species richness was positively affected by species richness of non-native flowering plants. The density of shrubs on embankments negatively affected the number of bee species and their abundance. Bee and hoverfly species richness were positively related to wood cover in a landscape surrounding embankments. This is the first study showing that railway embankments constitute valuable habitat for the conservation of pollinators in farmland. Specific conservation strategies involving embankments should focus on preventing habitat deterioration due to encroachment of dense shrubs and maintaining grassland vegetation with patches of bare ground.