Comparative Analysis of Urban and Rural Ecosystems in the Context of Suitability for Honeybees

1. Introduction

In Poland, the number of bee colonies has reached 2.35 million [1], marking a 100% increase over the past two decades. This growth is accompanied by a 100% increase in the average density of colonies per unit area, which currently stands at 7.7 colonies/km². In comparison, the average density in the 27 EU countries is 4.8 colonies per square kilometer, which is higher than in China (approximately 2 colonies/km²) and India (3.9 colonies/km²), despite these countries collectively accounting for 22% of the global population. The density calculation for China considered the fact that 52% of China's area is covered by mountains [2], where bees are not used. In certain EU countries, the density of colonies per km² is higher than in Poland: Hungary (10.6), Greece (16.8), and Malta (19.2). Turkey, the third largest honey producer in the world after China and the EU, has also a high density of 11.1 colonies per km². All the above calculations are based on [3,4].

It is a commonly held belief that the primary purpose of keeping honeybees is to produce honey. However, it is estimated that the benefits to humanity from bee pollination are significantly greater [5,6,7,8]. In certain regions of the world, notably those engaging in intensive agricultural practices, bees face with progressively challenging living conditions. In these regions, a decline in honeybee colonies has been observed for approximately two decades, necessitating continuous reproduction to maintain the population levels of the colonies at a sustainable level [5,9,10,11]. Intensively farmed appear to be increasingly less hostile to honeybees due to large-scale monocultures of crops that do not provide benefit to bees [12,13,14], as well as a significant increase in the amount of insecticides used in agriculture for crop protection [12]. The occurrence of acute pesticide poisoning clearly demonstrates the negative impact of agriculture on bees. Research has demonstrated that bees exhibit resistance to low to moderate levels of pesticides present in their food [15,16]. However, it remains challenging to discern sublethal effects in bees, particularly given that the Varroa destructor mites are the primary determinant of bee health and longevity [16].

The high densities of bee colonies in some European countries, as well as the threats from the chemicalization of agriculture, mean that an increasing number of bee colonies are being kept in urban areas, where conditions are conducive to the survival of pollinators [17,18]. Urban areas are characterised by a high diversity of nectar- and pollen-bearing plants, which bloom for extended periods, thereby providing food sources for insects. They have a warmer climate zone which has been demonstrated to be conducive to the production of large quantities of nectar. An example of this phenomenon in Poland is Robinia pseudoacacia L., whose flowers frequently undergo frost damage or fail to produce nectar. In the villages, aside from the cultivation of nectariferous crops, there is an absence of initiative to expand the forage base. Conversely, there has been a decline in bee forage, particularly those that start later in the season. Old roadside lime trees are being removed, and in forestry settings, Robinia pseudoacacia L. is being felled. Foresters regard this species as an invasive tree that is not conducive to effective forest management. The population of heather shrubs is declining, and these plants are reaching old age and producing low yields of nectar. This is largely due to restrictions on shrub regeneration management, such as through burning and cutting.

The process of chemisation in agriculture has resulted in the potential risks associated with beekeeping in certain areas. In urban areas, there is either a complete prohibition on the use of pesticides or a significantly reduced scale of their application when compared with rural areas. In addition to their conventional commercial applications, urban apiaries also possess significant educational value. Demonstration apiaries are organised to welcome children and young people for classes and workshops, allowing them to get to know the bees, learn about their lives, and taste the products. In Poland, urban apiaries are subject to the same regulations as other apiaries. These apiaries are typically either private enterprises or funded by local governments or companies as part of a broader marketing strategy.

The issue of whether the development of civilisation, urbanisation, and industrialisation is having a detrimental effect on pollinating insects is a matter of considerable debate. Research to date has focused predominantly on solitary bees. A comparison has been made of bee species richness according to the degree of urbanisation. In urban environments, there is a high abundance of bee species, but social species are more numerous, including polylectic species [19]. The honeybee is an eminently polylectic species, raising concerns that it may dominate these areas, displacing other insects (Apoidea) [20,21]. The adaptability of social insects has been identified as a key factor in their increasing dominance in urban environments. It is notable that the honeybee is well-suited to urban environments, exhibiting adaptability through its ability to fly over long distances and forage for food with greater efficiency compared to other wild bees [22]. Several studies have reported that the honeybee poses a threat to wild bees [23,24,25,26,27,28]. Conversely, other studies showed either no impact or inconclusive results [27].

The most famous urban apiary is the one founded in 1980 in Paris, located on the roof of the Opera Garnier. This apiary gave rise to the urban beekeeping industry, and today there are more than 3,000 bee colonies. Paris bee colonies produce an average of 50 kg of honey per year, with peaks of up to 80 kg in seasons where forage is particularly abundant. In contrast, in the rural environment colonies typically yield 30 kg [29]. In the United States, urban beekeeping was not permitted until the 1950s, when regulations began to liberalise. Bans on urban beekeeping are usually motivated by concerns for public safety [30]. A notable increase in the number of urban beekeepers was observed in the US around 2008, likely driven by media attention attracted by reports of 'colony collapse disorder (CCD)' that started in 2006 [31]. As a result, New York City and Los Angeles officially removed restrictions on urban beekeeping in 2010 establishing a significant precedent for cities and municipalities nationwide. This progressive shift was subsequently adopted by Milwaukee, Wisconsin, and Washington DC in 2012. The movement toward more supportive regulations continued with Boston in 2013 and Los Angeles again in 2015.

Bee densities in large cities are higher than in agricultural landscapes. For instance, the average colony density in Paris per km² is 6.5 colonies exceeds that of the rest of the country - 2.5, in London -10 and Brussels -15 [32]. Urban beekeeping has also become an increasingly prevalent practice in other European countries, including Germany and Poland. Most of these colonies are managed, whereas feral colonies are comparatively rare. The average wild colony density in Europe is 0.26/km², which is 25 times lower than in Latin America and Africa [33].

The aim of this study is to compare the rate of development and performance of bees kept in urban and agricultural environments across various regions in Poland.

2. Materials and Methods

2.1. Experimental Setup

The study was conducted in five regions of Poland in 2021 and 2022 (see Table 1 and Figure 1). In each region, we located an experimental apiary in the largest city or most urbanised area (Urban), and control apiaries in agricultural areas at varying distances from experimental apiaries, depending on the region, between 4.5 and 39.5 km (Rural). Tychy was only included in 2021 and Krakow only in 2022. The regions are characterised by a relatively low elevation, with altitudes ranging from 101 to 277 meters above sea level. The mean annual precipitation in Poland is approximately 608.8 mm. Poland is situated within a temperate climate zone, with the average summer temperature of approximately 18 °C and average winter temperature in winter -1.5°C.

We added open-mated sister queens belonging to the Apis mellifera carnica subspecies, born in 2020, to the bee colonies. If any queen died during the experiment, a new one was introduced, unless it was another young queen because of supersedure. The bees were kept in wooden Dadant hives with a frame size of 300 mm in height and 435 mm in width. The urban apiary in Lublin was located on the rooftop of a five-floor building, while the others were located on the ground level. In 2021, 84 colonies were tested (40 in the city and 44 in the countryside), and in 2022, 92 colonies were tested (49 in the city and 43 in the countryside).

2.2. Measurements and Observations

The state of the bee colonies was monitored regularly. Several measurements and observations were made to assess bee colonies’ performance and vital parameters. Overwintering of bee colonies was assessed by the difference in fully occupied combs of bees before (B1) and after wintering (B2). This difference was called the overwintering index (OIndex), a higher index means that the strength of the colony is lower in the spring. Brood count (BC) and colony strength (SC) were assessed based on the Liebefeld method [34]. BC was made four times during the beekeeping season, twice in spring (BC0 and BC1), in full summer (BC2), and before overwintering (BC3). We conducted the first spring brood survey at the start of the goat willow's flowering period (Salix caprea L.) and the second three weeks later. This approach facilitated a comprehensive assessment of the dynamics of brood growth (BGD) within bee colonies. Honey yield from each colony was assessed by weighing the honey super before and after centrifugation (HY). Prior to the main parasite control of Varroa destructor (at the end of July), the infestation of bees (VInf.) was evaluated in colonies using icing sugar to detach the mites from their host, which was then collected in a jar [35]. The mite counts were converted into units per 10 g of bees. Swarming susceptibility (SB) was evaluated using a scoring system ranging from 0 to 3. We did not give any points to families that were swarmed. We gave three points to families that were not swarming, and one or two points to families that were in a mood depending on the level of sophistication. We took a total of fourteen measurements, which reflected the bees' degree of adaptation to the conditions and, at the same time, the suitability of these conditions for the bees' development and productivity.

2.3. Statistical Analysis

The analysis considers three independent variables. The impact of the type of location (urban vs. rural), the location (Warsaw vs. Lublin vs. Pulawy vs. Cracow vs. Tychy), and the year of research (2021 vs. 2022) on fourteen dependent variables were analysed. The data related to traits, with the exception of the amount of brood in the spring and summer, have a non-normal distribution. The data of bee infestation with Varroa destructor were transformed according to the log (x+1) formula. This variable was analysed using the non-parametric Kruskal-Wallis and Mann-Whitney U tests. The data of BC0, BC1, and BC2 for variable type of location and year of survey were analysed using a student t-test and its modification with Welch's t-test. The data for the variable location were analysed using ANOVA and its modification of Hartley’s test for non-homogeneous variances. The identification of homogeneous groups was conducted through the Tukey test (HSD for various N). To present the differences between locations in the individual characteristics of bee colonies graphically, cluster analysis was performed. Spearman's rank correlation coefficients were calculated to identify the relationship between features. All statistical calculations and figures were prepared using Statistica 13 [36].

3. Results

3.1. Effect of Urban vs. Rural Locations on the Functional and Biological Characteristics of Bee Colonies

The traits that differentiated type of location and showed significantly higher values in urban apiaries were brood count in spring BC1 (p ≤ 0.01) and brood grow dynamics BGD (p ≤ 0.01). It is probably true that the consequence of a higher brood count in spring in urban apiaries was a significantly higher infestation of colonies with the Varroa destructor in summer (median respectively 0.1 and 0.0), as evidenced by Mann-Whitney U test (p ≤ 0.01), as demonstrated in Table 2 and Figure S1 (Supplementary). The other characteristics did not differ significantly between apiary types including honey production although it was 6.1 kg higher in urban apiaries.

3.2. The Impact of Geographic Location on the Functional and Biological Characteristics of Bee Colonies

Most traits of bee colonies, as many as 12, varied significantly between locations, with the exception Varroa infestation VInf. and the occurrence of swarming mood SB. Lublin was the most outlier in terms of positive measurement of characteristics, where as many as 6 parameters were significantly higher compared to other locations. These were mainly developmental indices, the number of brood and bees: B1, B2, BC0, BC1, SC1, SC3, OIndex, and production indicator HY (Table 3, Figure 2S, Supplementary). The greatest honey harvest was in the Warsaw region, with a yield of 68.3 kg per colony, which is significantly higher than in the other locations. In addition, the Warsaw region demonstrated the most marked discrepancy in honey yield between urban and rural apiaries, with a value of 20.7 kg (Figure 2) (p ≤ 0.01). In the Lublin region, honey yield was also found to be higher in urban apiaries, although the differences were smaller but statistically significant (p = 0.01). Tychy stands as a unique case where rural apiaries exhibited higher honey production compared to urban apiaries (Figure 2), but the differences were not significant (p = 0.32).

3.2. Influence of the Year of Study on the Functional and Biological Characteristics of Bee Colonies

As demonstrated in Table 4, the environmental conditions present in each year of the study had a significant impact on various developmental and biological parameters of bee colonies. A total of 10 traits differed significantly between years, with 6 traits manifesting at higher levels in 2022: B2, OIndex, BC0, BC1, SC1, and SC3 (Table 4). The only higher values in 2021 were those for SC2 and SB. The remaining characteristics demonstrated negligible fluctuations from year to year.

3.2. Correlation Between Individual Traits

The highest positive correlation coefficient (0.77) was observed between SC1 and BC1, indicating a relationship between the brood quantity and the number of workers in spring. Meanwhile, the highest negative correlation was identified between BC0 and BGD, suggesting a link between the number of brood in early spring and the growth dynamics of the bee population. This coefficient indicates that bee colonies with lower brood numbers in the post-winter period revealed a greater increase in multiplication compared to those with higher brood counts in spring (Table 6). Bees’ infestation by Varroa destructor mites was found to be correlated with the spring brood population. Honey yield throughout the season was positively correlated with brood area and worker numbers during the entire season, and negatively correlated with the number of combs in bee colonies during winter and early spring.

Table 5. Sperman correlation coefficients between individual developmental traits and production parameters. Only significant relationships were considered.

Trait*	B1	B2	OIndex	BC0	BC1	BGD	SC1	BC2	SC2	HY
B2	0.67
OIndex	0.22	-0.47
BC0	0.41	0.62	-0.29
BC1	0.38	0.48	-0.20	0.60
BGD	-0.20	-0.32		-0.63
SC1	0.52	0.63	-0.18	0.60	0.77
SC2								0.38
HY	-0.39	-0.19				0.19		0.20	0.26
VInf.			-0.17	0.28	0.25
BC3			0.17
SC3	0.43	0.34		0.22	0.22		0.45			-0.25
SB			0.18	-0.28	-0.28		-0.22		0.22	-0.17

* Abbreviations explained in Table 1.

Table 5. Sperman correlation coefficients between individual developmental traits and production parameters. Only significant relationships were considered.

Trait*	B1	B2	OIndex	BC0	BC1	BGD	SC1	BC2	SC2	HY
B2	0.67
OIndex	0.22	-0.47
BC0	0.41	0.62	-0.29
BC1	0.38	0.48	-0.20	0.60
BGD	-0.20	-0.32		-0.63
SC1	0.52	0.63	-0.18	0.60	0.77
SC2								0.38
HY	-0.39	-0.19				0.19		0.20	0.26
VInf.			-0.17	0.28	0.25
BC3			0.17
SC3	0.43	0.34		0.22	0.22		0.45			-0.25
SB			0.18	-0.28	-0.28		-0.22		0.22	-0.17

* Abbreviations explained in Table 1.

Table 6. The mean values of the traits in the groups of colonies were classified according to BC0 values relative to the median BC0 which is 28.4 dm2.

Grup of colonies	N	Trait*
		BC0 (dm2)	BC1 (dm2)	BGD (%)
BCO < Median	79	17.9	48.6	284
BCO > Median	81	39.9	68.5	176

* Abbreviations explained in Table 1.

Table 6. The mean values of the traits in the groups of colonies were classified according to BC0 values relative to the median BC0 which is 28.4 dm2.

Grup of colonies	N	Trait*
		BC0 (dm2)	BC1 (dm2)	BGD (%)
BCO < Median	79	17.9	48.6	284
BCO > Median	81	39.9	68.5	176

* Abbreviations explained in Table 1.

4. Discussion

A study was conducted to ascertain the distinguishing characteristics of apiaries located in urban and agricultural areas. It was found that the distinguishing parameter is the number of brood in spring BC1, brood growth rate BGD, and bee infestation with Varroa destructor parasites VInf. The results demonstrated that urban apiaries exhibited higher values for these parameters (Table 2). This suggests that in urban environments, at the beginning of the season, bee colonies experience accelerated development. This finding has been corroborated in other studies [15,19,37,38,39,40]. In four out of five regions, higher honey yields were reported in urban areas, although this was statistically proven in only two locations. These outcomes have been corroborated by other research studies (see references [38,41,42]. In Tychy region, honey production was dominated by rural apiaries due to favourable foraging conditions. However, it should be noted that in this study, data were collected over only one season.

Intensive colony development in urban apiaries led to higher levels of the Varroa destructor mite infestation during the summer months. The spread of the parasite is favoured by colony development, as has been described in other studies [43,44,45,46]. However, infestation levels in urban apiaries have not exceeded colony-threatening values of 3/10 g bees, [47,48], although some studies already consider 1/10 g bees as a critical level [49,50].

However, the 14 developmental, biological, and productivity parameters of the bee colonies were found to be most dependent by the different geographical parameters. In most cases, 12 parameters showed significant variation between geographical locations (Table 3). The apiary in Warsaw demonstrated the highest honey yield (68.3 kg/colony), this was a noteworthy achievement that outperformed both other regions and the rural control apiary. The average yield in Warsaw was 20.7 kg higher than in the rural apiary. The apiary in Warsaw exhibited the lowest overwintering strength, while simultaneously achieving the highest honey yield among of all the locations studied. This phenomenon may be attributed to the excessive workload placed on the bees. The existence of geographical variation in living conditions for bees is not a novel finding, as evidenced by other publications [51,52,53,54]. In the Lublin, Pulawy, or Cracow region, the predominance of urban apiaries in honey production was less pronounced compared to Warsaw.

The urban apiary in Lublin demonstrated exceptional performance in the majority of the 14 parameters under consideration. It was characterised by notably dynamic colony development during the spring period and the presence of stronger colonies throughout the season. This phenomenon may be attributed to the specific location of the apiary, which was the sole facility situated on the roof of a five-storey building. This particular location provides effective protection from wind while also ensuring optimal sun exposure. The distinctiveness of this region is reflected in Figure 3.

The climatic conditions at the macro and micro scale and forage conditions in which the experimental colonies were tested had a very significant effect on the parameters of the bee colonies. The distribution of the data did not permit a Multifactor ANOVA to be conducted, however, Figure 2, Figure 3 and Figure 4, and Figures S2 and S3 demonstrate that an interaction of factors of location and year of testing can occur.

In the temperate climate zones, seasonal variation in bee traits plays a crucial role. This assertion is corroborated by studies [55,56,57]. This finding is consistent with the results of the present study, which demonstrated that climatic conditions in each year significantly influenced bee overwintering B1, OIndex, spring colony development BC0, BC1 and honey yield. In 2022, the values of the key traits of colony development were higher than in 2021. These fluctuations can be attributed to various bee responses to weather conditions, such as the number of flightless days when bees are unable to forage, as well as presence of forage plants that do not produce nectar during cold weather, such as Robinia pseudoacacia, which flowers briefly [55].

The environmental impact on bees was so significant in the experiment that the correlation between brood and bee numbers and honey production was low at 0.2 and 0.26, respectively. Much higher correlation coefficients were obtained by other researchers, but they kept their apiaries in one place. The correlation coefficients for brood and honey production were 0.73 [58], 0.55-0.62 [59] and 0.96 [60]. However, the correlation coefficients between the number of bees and honey production were 0.71 [58], 0.54-0.64 [59] and 0.96 [60].

5. Conclusions

A comparative analysis of urban and rural apiaries revealed significant disparities in key parameters. These included brood numbers, colony populations, and, regrettably, Varroa destructor infestation level. These parameters exhibited higher values in urban apiaries, indicating a potentially accelerated development of urban bee colonies. Urban apiaries were distinguished by higher honey yields in four out of five cities, although in only two cases the results were statistically significant. Tychy region was distinguished by its superior honey production in rural apiaries; however, it should be noted that the study was conducted over a single season, which could have influenced the results. Despite the substantial growth of colonies in urban areas, the infestation levels of the Varroa destructor mite did not exceed levels that would threaten colony collapse, suggesting good bee health. Geographical location had the greatest impact on the 14 parameters studied. The highest honey yield was recorded in Warsaw, despite the bees were weaker there in winter. Seasonal changes in the weather, microclimatic conditions and the availability of forage had a significant impact on bee development and population numbers of colonies. Climatic conditions in different years had a substantial impact on the overwintering of bees and their spring development.