Introduction

Both in Poland and in other European countries, there has been a continuous decline in bumblebees and their habitats, starting from the 1980s. The main reasons for this are: the loss and fragmentation of habitats, pesticide usage, disease, parasites, non-native species, climate change, and the interaction between these factors. This situation could cause serious disruptions in the functioning of nature, and economic losses as bees are important pollinators. It is estimated that 35% of the world’s food supply entirely depends on these insects (Klein et al. 2007; Potts et al. 2010).

Bumblebees are particularly important pollinators in greenhouses, e.g. tomatoes. They are also specialized, unlike some other bee species, in causing the flower to vibrate – the phenomenon known as buzz pollination. Moreover, a large number of plant species are mainly or exclusively pollinated by bumblebees. This generates a strong relationship between the survival of plant species and bumblebees’ presence (Corbet et al. 1991). At a moment when a particular insect “vanishes” from the habitat occupied by a plant community the “domino effect” starts, which impacts herbivores and other dependent animal species (Corbet et al. 1991; Goulson 2010; Goulson et al. 2005; Pawlikowski et al. 2016).

In the face of the decreasing amount of agricultural areas suitable for bumblebees, the urban environment is gaining growing interest in the context of their active protection (Goddard et al. 2010). The attractiveness of urban green areas for sustaining diverse resources for bumblebees was recorded in parks in San Francisco (McFrederick and LeBuhn 2005) and suburban areas in Great Britain (Goulson et al. 2002; Osborne et al. 2008).

Cities, from the ecological point of view, are a unique mosaic of habitats with residential, commercial, industrial and infrastructure character, among which there are green areas (Breuste et al. 2008). The green areas of a city are very diverse, and apart of man-made gardens or parks, there are also enclaves of natural and semi-natural plant communities (Chudzicka et al. 1998) with characteristic sets of plant and animal species (Niemela 1999). The essential condition for the bumblebees’ occurrence, regardless of the habitat type, is the availability of food sources (Alford 1975). Therefore, a special role in urban areas is attributed to such green areas, which accumulate vast numbers of flowering plants. The primary, but not the only factor which affects the abundance of bumblebees, is flora richness (Widmer and Schmid-Hempel 1999; Steffan-Dewenter et al. 2002; Diaz-Ferero et al. 2012; Sikora and Kelm 2012; Sikora et al. 2016). It also turns out that not only do habitat characteristics have an impact on bumblebees biodiversity, but also factors in the spatial landscape scale, such as the areas which surround particular habitats (Hatfield and LeBuhn 2007). In the mosaic environment of a big city the attractiveness of green spaces should also be determined by their areas. So far, this relationship has not been strictly defined (Hatfield and LeBuhn 2007; Bates et al. 2011).

The aim of the study was to estimate the influence of the green spaces of big cities on bumblebee species richness, composition, and the relative number of bumblebee communities in Wrocław.

Materials and methods

The study was conducted in 2011 and 2012 on 12 sites of Wrocław (Poland) green areas. These areas, ranging from 8 to 102 ha, were parks, cemeteries, railways and wasteland (Fig. 1). Habitats in each surveyed area were mainly formed by trees and shrubs.

Fig. 1
figure 1

Dispersion of sites in the Wrocław area: C1 – Grabiszyński Cemetery, C2 – Osobowicki Cemetery, P1 – Szczytnicki Park, P2 – Tysiąclecia Park, P3 – Zachodni Park, P4 – Grabiszyński Park, P5 – Południowy Park, P6 – Wschodni Park, R1 – Strzegomska Railway, R2 – Żernicka Railway, W1 – Mikołajskie Wasteland, W2 – Gajowe Wasteland

Full size image

Species richness and abundance of bumblebees were determined by direct observation during 30 min (= sample). Observations were conducted once per month between March and September, from 10:00 am to 5:00 pm, in favourable weather conditions for bee flights (Pawlikowski 2010). Bumblebees were counted on visited plant species. Specimens of Bombus terrestris and Bombus lucorum were counted together, as it is hard to distinguish between them during field surveys. Individual specimens of workers and males castes, impossible to identify in the field, were caught and euthanized with ethyl acetate (method approved by the Regional and General Director of Environmental Protection in Poland for 2011–2012). The specimens are stored in the Department of Plant Protection, the University of Life Sciences, Wrocław.

The gathered materials were used to identify the structure of bumblebee communities on the study sites. The structure of distanced communities was described using the number of species (S), average density (At), an index of total species diversity (H′), and an evenness index (J’). Average density (At) was the number of recorded specimens on the site per 30 min of observation. The Shannon and Weaver (1963) was used as a formula to calculate the index of general species diversity. The evenness index of species frequency distribution was calculated using Pielou (1966). The qualitative similarity of the distanced communities (QS) was calculated using the Sørensen (1948), while quantitative similarity (Re) was determined by using the Renkonen (1938). Cody (1970) was used for quantitative and qualitative comparisons of bumblebee communities. The QS and Re indices were selected in order to assess the similarity of the communities according to their adequate representation of species number and density in the communities along an appointed gradient of the surveyed areas’ size. The qualitative and quantitative similarity of habitats was calculated using the MS EXEL program, and the graphic file was generated in PHOTOSHOP CS6 software. Specific structure parameters (S, At, H′, J’) were additionally surveyed in terms of the relationship with the size of the habitats. Structure parameters were calculated according to correlation analysis, with a significance level of p ≤ 0.05. Calculations were made with STATISTICA v 6.0 software. In order to define the surveyed site areas the Wrocław Map of Geographic Information System was used, together with measuring applications available at: www.geoportal.wroclaw.pl.

Results

In the 12 surveyed habitats 14 species of bumblebees (Bombus Latr.) were recorded, in which 3 belong to cuckoo bumblebees (Psithyrus subgenus) (Table 1). In Szczytnicki Park (P1) the highest number of bumblebee species (12) was observed, whereas the lowest species number (5) was recorded on Gajowe Wasteland (W2) and Żernicka Railway (R2).

Table 1 Parameters of bumblebee communities in Wrocław green areas in the phenological seasons 2011–2012
Full size table

The highest average density of bumblebees (At) was recorded in the Gajowe Wasteland area (W2) – 68.5 specimens/30′, while the lowest occurred in the Zachodni Park area (P3) – 13.42 specimens/30′.

On the qualitative diagram (Fig. 2), based on the Sørensen index of species similarity (QS), 7 communities with predominant mutual similarity above 77% were distanced. Those communities were present on the surveyed sites above 30 ha. The other 5 communities, within smaller areas P2, R2, R1, W1 and W2, showed lower mutual similarity. Amongst them, the smallest similarity to the others (less than 50%) was characteristic for the community in area R2.

Fig. 2
figure 2

Diagrams of qualitative similarity (QS) and quantitative similarity (Re) of bumblebee communities according to the gradient of green areas in Wrocław

Full size image

On the quantitative diagram (Fig. 2), based on the Renkonen index (Re), 7 communities with area above 30 ha have been extracted. Among all of the other sites smaller than 30 ha, the most similar community was for the R1, W1 and W2 areas.

On dendrite resemblance based on the qualitative-quantitative Cody’s index (Fig. 3), bumblebee communities clearly concentrate in regard to P6 and C2 communities. Within the range of average T value there are 7 communities from the sites of 30 ha area each.

Fig. 3
figure 3

Dendrite resemblance presenting species diversity of bumblebee communities in green areas of Wrocław. The communities refer to the circle of Cody’s index average values (T)

Full size image

Analyses of communities structure dependence parameters indicated a strong positive correlation between the number of bumblebee species (r = 0.7497) and the size of urban green areas. In the case of species diversity (r = 0.3566) and evenness (r = 0.0344) there was a weak positive correlation, and for average density (r = −0.2677) a weak negative correlation (Fig. 4).

Fig. 4
figure 4

Structure parameters of bumblebee communities according to the gradient of green areas in Wrocław

Full size image

Discussions/Conclusions

The urban agglomerations might be potentially attractive habitats for bumblebees thanks to the mosaic of green areas. In the spatial plan of Wrocław the green spaces are formed in annular-cuneiform, enabling their permanence, which positively influences the entirety of natural factors (Bińkowska and Szopińska 2013). The green areas of Wrocław constitute a fifth of the total city area, and within the city borders there are zones of the European Natura 2000 network.

According to Alford (1975), bumblebee species richness in different areas depends on the food attractiveness of these habitats, and thus the areas with the highest number of bumblebee species must be characterized by specific features. Such a place in Wrocław is Szczytnicki Park, the oldest and largest urban park. Moreover, it was the only area where a bumblebee rare in the country and on the European scale, Bombus humilis, was found.

The factor affecting the number of bumblebee species on Wrocław sites was the area of the green areas. This confirms the assumptions of the theory of insular biogeography (MacArthur and Wilson 1967). The theory states that the bigger the area of habitat, the more bumblebee species can be expected there. Analogical conclusions were drawn in relation to birds (Evans et al. 2009), amphibians (Parris 2006), mammals (Magle et al. 2009), and ground beetles (Sadler et al. 2006) living in urban habitats, and bumblebees living in meadow habitats of the rural landscape (Diaz-Ferero et al. 2012).

During research into bumblebees in Estonia, where the size of rural areas was correlated with the bumblebee numbers, it was pointed out that the bigger the size of a rural farm, the lesser the bumblebee number (Muljar et al. 2010). The above-mentioned theory of insular biogeography explains the matter of the numbers in a way that in a bigger area the specimen dispersion is greater (MacArthur and Wilson 1967). However, the research carried out in urban habitats, like urban gardens and traditional flowerbeds in many kinds of green spaces, does not demonstrate the impact of the area site on bumblebee numbers (Gunnarsson and Federsel 2014). Also, during the surveys undertaken on the Wrocław sites the impact of the site area on bumblebee numbers was not observed.

Other researchers point out that floristic richness is a factor influencing the bumblebees abundance and their species diversity (Widmer and Schmid-Hempel 1999; Steffan-Dewenter et al. 2002; Diaz-Ferero et al. 2012; Parde and Philipott 2014; Salisbury et al. 2015; Wood et al. 2015). However, on the Wrocław study sites the number of visited flowering plants did not correlate with the size of the site r = 0.0792 (Fig. 5). Therefore, the quality factors like the degree of food attractiveness and abundance in the habitat seem to be more important than the number of flowering plants. The bumblebees abundance depends primarily on the quality, availability and stability of food sources in the inhabited area. Comparing the bumblebees’ density in urban parks, cemeteries and allotments, the highest value was presented in species-rich flower gardens (Ahrne 2008). In such places, with a high diversity of flowering plants, bumblebee density can reach 8 specimens per 100 m2 (Matteson and Langellotto 2009), while on cultivated fields in England only 4–5 specimens per 100 sq. m (Carvell et al. 2004) and less than 1 specimen on 100 sq. m on meadow habitats in Sweden (Öckinger and Smith 2007). Observed trophic interactions confirm the principle that a few but adequately selected plant species as food sources might provide suitable conditions for the development of numerous bumblebee families. This has significant importance in rural environments, where the economic use of bumblebees is mainly focused on plant monocultures (Goulson 2010).

Fig. 5
figure 5

Relationship between the number of visited flowering plants (P) and the size of green areas in Wrocław

Full size image

Isolation in extremely fragmented city habitats is a serious threat for bumblebee families. Smaller populations are more vulnerable to extinction due to environmental changes (Frankham et al. 2002). It was confirmed that populations which cannot freely move between habitats are forced to inbreed, and this leads to a decrease in the condition of the whole population (Krebs 2009). Habitat fragmentation may cause a situation where too small sites with food plants will not be attractive for some bumblebee species (Mayer et al. 2012).

Within the city space the sequences of green areas along roads should allow migration. In the Wrocław agglomeration such a function can be fulfilled by railway sites with attractive ruderal plants. Plants from the Fabaceae family could be a perfect pollen source for visiting bumblebees (Goulson and Darvill 2004). The value of these sites is undoubtedly increased by the presence of one of the rarest and endangered bumblebee species in Poland – Bombus veteranus.

International efforts to preserve the environment are focused mainly on large ecosystems, usually very little altered by human activity. Less attention is paid to small green spaces in cities, in neighbourhoods where people live and work. However, urban parks and open green areas are essentially important for the quality of life in a more urbanised society (Chiesura 2004). The protection of the wild bee population makes sense only when it is carried out on a landscape scale, which requires coordination and cooperation at a regional level (Richards 1993). The first step towards effective management in the urban environment is the recognition of landscape factors on a patch scale, which affects biodiversity (Angold et al. 2006). The next step is to tackle a social challenge – combining the practice of management with the achievements of scientific theories (Cadenasso and Pickett 2008). Considering the obtained results, large green urban areas should be created of a size of at least 30 ha. Such sites are similar to each other in terms of habitat attractiveness and provide conditions for the most diversified bumblebee species communities.

Such sites will be pollinator shelters if managed properly. In the most potentially attractive habitats mowing, and herbicide and insecticide usage should be restricted. On lawns and green spaces with blooming clovers (Trifolium spp.), birdsfoot trefoil (Lotus corniculatus), self-heals (Prunella spp.) and knapweeds (Centaurea spp.), the maintenance treatments should consider the appropriate height of cutting, providing constant flower food sources. For urban parks special management plans should be developed in order to determine the rules of habitat use which consider the active protection of pollinators. Margin areas with shrubs, tall grass and touchwood should be in every park as nesting habitats for pollinating insects. In order to increase the potential of lawns and green spaces, mixes of native flowering species should be introduced in the form of blooming stripes and flower rich meadows. In habitats such as urban parks, which are used for recreation and sport, an effective solution would be the introduction of flowerbeds, which are attractive for bumblebees. The flowerbeds, including native and exotic plant species, would play a natural, aesthetic and educational role.

On sites smaller than 30 ha the bumblebee communities were dominated by abundant species with a wider polyphagous spectrum. However, small green areas can play an important role as refuges and allow migration to all pollinators, especially when places are enriched with a mosaic of various food plants. In each habitat it is possible to introduce even small alterations beneficial for pollinators. Undoubtedly, the protection strategies should be mainly based on blooming species diversity and richness (Sutcliffe and Plowright 1988).