Bees ( Hymenoptera : Apoidea , Apiformes ) of the Kujawy Lakeland ( central Poland )

Abstract: Bee diversity was studied in 14 habitats in 7 localities in the Kujawy Lakeland (Pojezierze Kujawskie) in central Poland. Additionally, we investigated the species diversity and phenology of bumblebees on red clover (Trifolium pratense). In total, 146 bee species were recorded in the study area, accounting for 30.7% of bee species reported from Poland so far and 46.2% of bee species known from the Wielkopolska-Kujawy Lowland (Nizina Wielkopolsko-Kujawska). These include 14 red-listed species.


Study area
According to the physico-geographical division of Poland (Kondracki 2009), the Kujawy Lakeland (mesoregion 315.57) constitutes the south-eastern part of the Wielkopolska Lakeland (macroregion 315).This is an agricultural mesoregion with fertile soils and a small percentage contribution of forests.It covers an area of 2500 km 2 and includes the "Nadgoplański Park Tysiąclecia" Landscape Park, which is of high conservation value.The Landscape Park includes Natura 2000 sites: Special Protection Area PLB040004 "Ostoja Nadgoplańska", and Special Area of Conservation PLH40007 Lake Gopło.In historical and ethnographic terms, the study area is part of the Kujawy region, situated between the upper Noteć River, Lake Gopło, and the Vistula (Mileska 1994).According to the climatic regionalization of Woś (1999), the Kujawy Lakeland is located in region XV, i.e. the central-Wielkopolska region (the largest one in Poland).It is characterized by a high frequency of days with very warm and cloudy weather, and often ground frost on very cold days with precipitation (Woś 2003).10.Forest, on a potential site of moderately moist, mixed coniferous forest.Not adjacent to Lake Świesz.Tree layer dominated by 40-year-old Pinus sylvestris, and locally Robinia pseudoacacia and Betula pendula.Shrub layer consisting of Sambucus nigra, Frangula alnus, Robinia pseudoacacia, and Rubus sp.Forest edge with abundant Chelidonium majus.Bees were studied there in May-June 2014.
11. Forest glade south of Lake Świesz, it borders with a pool separated by a narrow levee from a fish pond.Numerous shrubs: Sambucus nigra and Prunus padus.Grasses tall and lush, with small contributions of Lamium purpureum, Leontodon hispidus, and Knautia arvensis.Bees were studied there in April-August 2014.
13. Moderately moist hay meadow of the alliance Arrhenatherion elatioris.Characterized by large patches of abundantly flowering Centaurea jacea in summer.Bees were studied there in July-August 2015.
14. Old field bordered with a reed belt, farmland, and alluvial oak-elm-ash forest Ficario-Ulmetum.Major bee forage plants: Vicia cracca and Brassica napus.Bees were studied there in July-August 2015.

Research methods
Faunistic material was collected by catching of noticed bees with an insect net directly on bee forage plants, during flight or at their nesting sites.Bees were caught for 1 h during observations made in favourable conditions for their activity, i.e. between 9 am and 5 pm, at temperatures ranging from 15ºC to 32ºC, with no or very little wind.The field research was conducted in 5 growing seasons, in 2011-2015.The collected material was prepared, identified to species, and deposited in the collection of the Department of Ecology, Institute of Environmental Biology, Kazimierz Wielki University, Bydgoszcz, except for protected species, which were identified alive in the field and next released.During the counting and catching of bees, also the plant species visited by them were recorded.Honey bees were observed at all the analysed sites, but this species was not taken into account because of its anthropogenic origin.
To analyse species richness, a rarefaction curve was constructed (according to Gotelli & Colwell 2001).Additionally, the Shannon index (H') of species diversity was calculated (Magurran 2004): where p i is the proportion of individuals belonging to the ith species in the community of S recorded species.
To estimate the total number of species (the so-called complete species richness), we used the chao2 method (Chao 1987).It is a classic estimator based on single and double observations of occurrence of individual species in the samples.The calculations were made using EstimateS v. 9.1 software (Colwell 2013).
Species names and categories of threat follow those used in the "Fauna of Poland" (Banaszak 2004).
Most of the habitats are described in general, on the basis of own observations in the field (AS-B).Only the description of the plant cover in Mietlica and Ostrówek is based on the literature (Zarządzenie 2015).For each site, we determined its location on the UTM 1 km × 1 km grid by means of Gnomon version 3.3 software (Desmodus, Poland).

Species diversity and dominance structure of wild bees of the Kujawy Lakeland
In total, 6312 wild bees were caught and observed.All species and numbers of specimens recorded at individual research sites are listed in Table 1.
The list of recorded bee species in the Kujawy Lakeland is incomplete, as the mean species accumulation curve (Fig. 2) for the total of 6312 recorded individuals still shows an increasing trend.The expected number of species calculated on the basis of the chao2 estimator was 190 (95% CI = 166-242).The percentage contribution of caught species to the total estimated number (76.8%) indicates that the collected material is representative, as it is based on observation of at least 70% of species possible to detect (Williams et al. 2001).The high Shannon index (3.5)attests to a high bee species diversity in the study area.
The total numbers of wild bee species in individual habitats were as follows: 70 on xerothermic grassland, 64 on the synanthropic site, 56 on sandy grassland 'a', 54 on the roadside, 40 on hay meadow/pasture, 38 in the forest glade, 35 on the periodically waterlogged meadow, 29 in willow thickets, 26 on sandy grassland b, 19 on the moderately moist meadow, 17 on red clover 'a', 14 on red clover field 'b', 14 in the historical manor park, 11 in the old field, 11 in the rape field, and 6 in the forest (Fig. 4).

Bombus jonellus (Kirby)
Usually classified as an Boreal-Alpine species, in Poland frequently recorded also at cold and moist sites.New record: Świesz, forest, 28 Jun 2014, 1 ♀.

Bombus soroeensis soroeensis (Fabricius)
Very rare subspecies in Poland, which is its western limit and the eastern limit of more abundant subspecies B. s. proteus.Associated with forest habitats.New record: Lubsin, forest glade, 30 Aug 2014, 1 ♀; Świesz, forest, 28 Jul 2014, 1 ♀.Species richness and seasonal dynamics of red clover by bumblebees Three-year observations (2011-2013) of bumblebees on two red clover fields provided data on 10 species.In field 'a', two more species were recorded than in field 'b'.Considering the total numbers of individual bee species in the three years, in both fields eudominants included Bombus terrestris, B. lucorum, B. muscorum, B. lapidarius, and B. pascuorum.Before the second moving of red clover, more bumblebees were recorded than before the first harvest.An exception is field 'b', where bumblebee density in 2011 was identical in both periods.The markedly lower density of bumblebees at this site than at the other sites was due to sporadic grazing of cattle in the field in 2011.
It is noteworthy that in 2011 and 2013 in these fields the number of bumblebee species before the second moving was higher than before the first harvest.In 2012, species richness was the same in both periods.
Species composition of bumblebees differed between years.Bombus muscorum was most abundant in field 'a' (before both harvests) and 'b' (before the first harvest) in 2011, and also in field 'a' (before the second moving) in 2013.In nearly all the other instances, before both harvests, the dominant species was B. terrestris, except before the second moving in field 'b', where in 2011 B. pascuorum was the most frequent, and in 2013 B. sylvarum (Table 2).DISCUSSION AND CONCLUSIONS The results of this study in the Kujawy Lakeland supplement our knowledge of the bee fauna of the Wielkopolska-Kujawy Lowland.In the study area, 146 species of wild bees were recorded.They represent 29 genera, and 7 families, accounting for 30.7% of bee species reported from Poland so far (Banaszak 2004) and 46.2% of bee species known from the Wielkopolska-Kujawy Lowland (Banaszak 2010c).Incomplete list of bee species can be explained for example by missing of Andrena fulvida in spite of the recording of its cleptoparasite Nomada opaca.The relatively high environmental value of the investigated habitats is confirmed by the presence of red-listed species of wild bees, which in total account for about 10% of all the recorded species.
As compared to the fauna of Poland (Banaszak 2004) and of the Wielkopolska-Kujawy Lowland (Banaszak 2010c), in this study the percentage contributions of species of 2 bee families were the markedly higher: the Andrenidae and Apidae (Fig. 5).In contrast, lower contributions were recorded for the families Colletidae, Halictidae, and Anthophoridae, while the contributions of the families Megachilidae and Melittidae were very much like generally in Poland and in the Wielkopolska-Kujawy Lowland.
A particularly high species diversity of wild bees was observed in xerothermic grassland at the archaeological site in Mietlica, located within the Landscape Park.A comparable result was recorded in xerothermic grassland on the Budzyń Esker in the Wielkopolska National Park, where 69 bee species were found (Banaszak 1983).It is surprising that in both localities, although markedly different in area (grassland near Lake Gopło 0.2 ha and grassland near Lake Budzyńskie about 1.3 ha), nearly identical numbers of species were recorded.In comparison, in the proposed "Folusz" nature reserve near Szubin, 88 species were found in an area of 3.0 ha (Banaszak et al. 2004).In both "Folusz" and Mietlica, the eudominant was the notably oligolectic Andrena vaga, which collects pollen from flowers of Salix spp.(Bischoff et al. 2003, Vanderplanck et al. 2003), flowering on nearby meadows.
The high faunistic richness of xerothermic grassland probably results from the species richness of bee forage plants at that site.The dominance of Andrena vaga in the total catch and on xerothermic grassland results from its very numerous nesting there.Its colonies covered in total about 73.5 m 2 .To determine the dynamics of its population, as well as the sex ratio of the host and its parasite, bees were counted in a belt 3.5 m wide along 21.0 m of the embankment length (Table 3).The flight season of A. vaga lasted from mid-April till mid-June.It is noteworthy that males of this mining bee species emerged slightly earlier, which was not recorded because of unfavourable weather conditions.In the colony, females and males reached the highest numbers of 338 and 125, respectively.As the host numbers decreased, the numbers of their nest parasite, Nomada lathburiana, increased.The most frequent wild bees on xerothermic grassland included also Bombus terrestris and Anthophora plumipes.The former collected nectar and pollen from 6 plant species, while the latter was observed on Salvia pratensis, which flowered abundantly in May.The study site in Mietlica seems very interesting from the point of view of nature conservation, so it is necessary to undertake measures to remove successively the self-sown oaks at the foot of the slope.Another problem is the uncontrolled penetration of the slope by local inhabitants riding quads, which probably have a negative effect on nesting of e.g. the large local population of A. vaga.
On the synanthropic site, spring species prevailed: Osmia rufa, Andrena haemorrhoa, and Anthophora plumipes.Their high contribution was associated with rich food resources: http://rcin.org.plflowering fruit trees Malus domestica and Prunus spp. as well as Taraxacum officinale in the herb layer.At this site, nests of Anthophora plumipes and Osmia rufa were found in a clay wall of a farm building.We observed also Melecta albifrons, which is a nest parasite of A. plumipes.
The periodically waterlogged meadow proved to be a perfect site for Macropis europaea, which clearly dominated there (33.0%).It is a narrowly oligolectic bee species, adapted to waterlogged conditions.Its females collect pollen and oils for the offspring e.g. from Lysimachia vulgaris, which flourished in the meadow.
Sandy grassland was characterized by a high contribution of thermophilous bee species: Heriades crenulatus and Dasypoda altercator.Among bumblebees, the most abundant was Bombus lapidarius.The 3 species mentioned above visited flowers of Cenataurea stoebe and C. scabiosa.Sandy grassland b showed a different dominance structure of bees.The most frequent was Bombus lapidarius, followed by Halictus sexcinctus, and Heriades crenulatus.This can be explained by the fact that ecological succession is more advanced there than on sandy grassland a.
On the moderately moist meadow, the major species were Bombus lapidarius, B. pratorum, and Andrena pilipes.This site was visited twice in summer (31 Jul 2015 and 15 Aug 2015), and those species were observed on flowers of Centaurea jacea.Forest communities located in the vicinity were probably favourable sites for Bombus pratorum.
The extremely high density of Heriades crenulatus in comparison with ubiquitous H. truncorum requires comment.H. crenulatus is sub-Ponto-Mediterranean species, in Poland infrequent.In other countries of the Central Europe it is local species also and its density is higher only at not disturbed areas (Přidal & Veselý 2011).
The major bumblebee species on red clover included Bombus terrestris, B. lucorum, B. muscorum, and B. lapidarius.Similar results were reported by Błażejewska et al. (1961), Ruszkowski & Biliński (1968), Biliński (1977), Anasiewicz (1976), andBanaszak (1984).All these Bombus species are common in Poland.The small differences in dominance structure of bumblebees between the cited publications and this study depend on the types of habitats surrounding the red clover fields.
When comparing the numbers of bumblebee species before the first and second moving, we did not confirm the trend reported by Biliński (1977), who observed more Bombus spp.before the first harvest than before the second one.The seasonal dynamics of red clover pollination by bumblebees in individual years probably depended on weather, which was not analysed here.

Fig. 1 .
Fig. 1.Map of the investigated area of the Kujawy Lakeland, with location of study sites (description in the main body of the article).

Fig. 2 .
Fig. 2. Expected species richness based on the rarefaction curve for the Kujawy Lakeland.Confidence intervals are marked with a broken line.

Fig. 4 .
Fig. 4. Number of species of Apiformes at the particular study sites in the Kujawy Lakeland.

Fig. 5 .
Fig. 5. Percentage contributions of individual families to the total number of wild bee species in the Kujawy Lakeland, as compared to the bee fauna of Poland and the Wielkopolska-Kujawy Lowland.

Table 1 .
Number of species and individuals of Apiformes at the study sites in the Kujawy Lakeland.

Table 2 .
Abundance of bumblebees (Bombus spp.) in red clover fields in the Kujawy Lakeland.

Table 3 .
Population dynamics of Andrena vaga and its cleptoparasite Nomada lathburiana, observed in an area of 73.5 m 2 of xerothermic grassland in the "Nadgoplański Park Tysiąclecia" Landscape Park in 2015.