ISSN 10623590, Biology Bulletin, 2013, Vol. 40, No. 3, pp. 318–328. © Pleiades Publishing, Inc., 2013. Original Russian Text © I.N. Bolotov, Yu.S. Kolosova, M.V. Podbolotskaya, G.S. Potapov, I.V. Grishchenko, 2013, published in Izvestiya Akademii Nauk, Seriya Biolog icheskaya, 2013, No. 3, pp. 357–367.

ECOLOGY

Mechanism of Density Compensation in Island Bumblebee Assemblages (Hymenoptera, Apidae, Bombus) and the Notion of Reserve Compensatory Species I. N. Bolotova, Yu. S. Kolosovaa, M. V. Podbolotskayab, G. S. Potapova, and I. V. Grishchenkoc a

Institute of Ecological Problems of the North, Ural Branch, Russian Academy of Sciences, nab. Severnoi Dviny 23, Arkhangelsk, 163000 Russia b Lomonosov Northern (Arctic) Federal University, nab. Severnoi Dviny 17, Arkhangelsk, 163000 Russia c Northern Department for Hydrometeorology and Environmental Pollution Monitoring, ul. Mayakovskogo 2, Arkhangelsk, 163020 Russia email: [email protected] Received March 14, 2012

Abstract—The notion of a dynamic compensatory system is discussed, characterized by the alternation of species occupying the leading position in bumblebee assemblages, while the total density of these pollinators in island ecosystems remains at similar levels. The functioning of the compensatory system is regulated by both abiotic factors (the weather and climate) and biotic factors (competition for trophic resources). The sta bility of the system is determined by the presence of reserve compensatory species capable of rapid population growth against the background of depressed abundance of other species under changing environmental con ditions. DOI: 10.1134/S1062359013030035

Under unfavorable conditions, which decrease the species diversity, various compensatory changes take place in the structure of communities (MacArthur and Wilson, 1967; MacArthur et al., 1972; Wright, 1980; Chernov, 2005, 2008; Gonzales and Loreau, 2009). Such compensatory mechanisms are quite pro nounced in island communities (MacArthur et al., 1972; Bennett and Gorman, 1979; Rodda and Dean Bradeley, 2002; Chernov, 2005, 2008). The study of island biocenoses touches extremely complex prob lems such as the structure and dynamics of isolated populations and peculiar features of niches and inter actions between species in qualitatively depleted com munities (Chernov, 1982). It is believed that one of the manifestations of the action of compensatory mechanisms is the phenome non of density compensation, understood as the increase in the abundance and biomass of some spe cies as a consequence of the disappearance or initial absence of other species, their potential competitors (Chernov, 2005). In spite of the availability of many studies analyzing such compensatory phenomena in various animal groups, the space–time compensatory dynamics remains insufficiently investigated (Gonza lez and Loreau, 2009). The Solovetskie Islands in the White Sea were selected as the model area for this study. This archipel ago is rather strongly isolated from the continent (lying at a distance of 20–30 km from the coast). The composition of a number of animal taxa on the archi

pelago is strongly depleted, compared to that of the continental taiga (Bolotov and Podbolotskaya, 2003; Bolotov, 2006; Bolotov and Shutova, 2006; Shwarts man and Bolotov, 2008; Bespalaya et al., 2009, 2011; Bolotov et al., 2011). The cause of this depletion lies in the peculiar features of the genesis of the postglacial biota in this area, which are related to the paleogeog raphy of northwestern Europe (Bolotov and Shutova, 2006). Since the biota is young, macroevolutionary processes are difficult to trace: the formation of inher ited adaptations to the unfavorable climate has been revealed in only a few animal species (Severtsov, 1995). The purpose of this study was to obtain facts that would allow assessing the variation in time of the com pensatory phenomena that take place in island com munities depleted of species. Initially, we assumed that compensatory processes in such communities were rather dynamic. Depending on the weather and cli mate conditions of the seasons, species best adapted to the particular combination of environmental factors should gain competitive advantage. We selected bum blebees as the model system, because it was shown ear lier that the structure of their assemblages can change considerably over longterm intervals (Berezin et al., 1996; Colla and Packer, 2008; Colla, 2010; Dupont et al., 2011).

318

MECHANISM OF DENSITY COMPENSATION IN ISLAND BUMBLEBEE

319

MATERIAL AND METHODS The environmental conditions, fauna, and flora of Solovetskie Islands were characterized earlier, based on the available publications (Prirodnaya sreda …, 2007; Shvartsman and Bolotov, 2008). It has been shown that the vegetation is dominated by coniferous and mixed forests and peat bogs; small areas are occu pied by dry meadows; on maritime terraces of larger islands, sparse forests with Betula tortuosa, black crowberry tundras, and smallgrass meadows form; smaller islands (such as Bol’shoi Zayatskii Island) are entirely of this type of “forest–tundra” aspect. The minimum distance of the islands from the continent is around 25 km. The migrant biota of the archipelago formed mainly at the boundary of the late glacial period and the Early Holocene (Bolotov and Shutova, 2006; Shvartsman and Bolotov, 2008). The material for this study consisted of bumblebees collected on Bol’shoi Solovetskii Island in 2001–2010. The insects were usually collected two times per sea son: in late June to early July and in late July to early August. The principal collecting method was nonse lective insect net capturing of all individuals encoun tered in habitats occupying standard key plots (Pesenko, 1982). The allocation of the collecting areas and habitats have been described in detail earlier by the authors (Prirodnaya sreda …, 2007; Shvartsman and Bolotov, 2008; Podbolotskaya, 2009). The size of the tenyear sample is 15 028 specimens. The insects collected are stored in the Institute of the Ecological Problems of the North, Ural Branch, Russian Acad emy of Sciences (Arkhangelsk). In addition, in early July of 2011, bumblebee nests were sought and exam ined on Bol’shoi Solovetskii Island. The density of bumblebees was estimated by visual censuses (in which all individuals encountered were recorded) performed in late June to early July and in August in standard 2 m2 plots in meadow ecosystems on Bol’shoi Solovetskii Island and on continental dry meadows in the Northern Dvina River basin. For technical reasons, the censuses were performed only in certain years (in 2007, 2009, and 2010 on the island and in 2007 and 2010 on the continent). The relative abundance of species was calculated as the proportion of individuals of each species of the total number of individuals in the sample (Ps, %). The original coefficient K was also used, defined as the ratio between the abundance of Bombus pascuorum and B. jonellus, the two most abundant bumblebee species: Kp/j = Np/Nj,

measurements of the air temperature and precipita tion at the Solovki Weather Station in 2000–2010. These data were provided by the Northern Depart ment for Hydrometeorology and Environmental Pol lution Monitoring (Arkhangelsk). The analysis was based on the values of average monthly air tempera ture and monthly total precipitation. The influence of the overwintering conditions was estimated using the following data (from October to April): the sum of the monthly temperature over the period with average daily air temperatures >0°С; the number of thaw days (>0°С) per month; and the average monthly snow cover thickness. The following integral parameters were also involved in the analysis: the sum of tempera tures and total precipitation over the periods with sta ble average daily air temperature >5 or >10°C (the growth season and summer season, respectively); the duration of these periods; and Selyaninov’s hydrother mal coefficient. The integral winter and summer weather and climate conditions were calculated directly by expert meteorologists in certified electronic systems of the Roshydromet under the supervision of I.V. Grishchenko. The significance of the differences between plots in bumblebee abundance was estimated by the nonpara metric Mann–Whitney test (Utest). The significance of differences between the proportions of the number of individuals of particular bumblebee species in the samples was estimated by Fisher’s test (Ftest), taking into account the size of the samples. The correlations between the abundance of particular bumblebee spe cies and the climate were tested by Spearman’s rank correlation test. The abovelisted calculations were performed in the program STATISTICA, version 6.1 (Puzachenko, 2004). Multifactor estimation of the relations of bumble bee species to meteorological parameters was per formed by canonical correspondence analysis. The direct gradient procedure of the analysis was performed in the program CANOCO, version 4.56 (ter Braak, 1996; ter Braak and Šmilauer, 2002). The initial data on the number of bumblebee individuals in samples for each year were transformed by chord transforma tion (Legendre and Gallagher, 2001). As the ecologi cal factors, the meteorological variables used were those that proved especially informative in the course of preliminary correlation analysis. The meteorologi cal data were not transformed. The analysis was focused on interspecies distances with Hill’s scaling. The downweighting of rare species was not corrected. The significance of the canonical axes was determined by the MonteCarlo permutation test.

where N is the number of individuals of the respective species in the sample. The population dynamics of insects is usually espe cially strongly influenced by the weather and climate conditions. To analyze their influence, we selected standard meteorological variables based on regular

RESULTS A total of 13 species of five bumblebee subgenera have been found on the Solovetskie Islands (Kolosova and Podbolotskaya, 2010). The relative abundance of the species has pronounced cyclic interannual dynam

BIOLOGY BULLETIN

Vol. 40

No. 3

2013

BOLOTOV et al. Proportion of the total number of individuals, %

320 100

80

60

40

20

0 2001

2002

2003 1

2004 2

2005 3

2006

Year 4

5

2007 6

2008 7

2009

2010

8

Fig. 1. Longterm dynamics of relative abundance of bumblebees on Bol’shoi Soloveskii Island: 1, B. jonellus; 2, B. pascuorum; 3, B. pratorum; 4, B. lucorum; 5, B. sporadicus; 6, B. hypnorum; 7, five species of the subgenus Psithyrus; 8, the other two species.

ics (Fig. 1). Especially significant are the changes in abundance of the five dominant species: B. jonellus, B. pascuorum, B. pratorum, B. lucorum, and B. sporadicus. The first two can be considered abun dant, or superdominant, species (Chernov, 2005): in 2001–2010, the average proportions of individuals of these species in the samples were 32.9 ± 7.2 and 31.8 ± 5.1%, respectively (hereinafter, the ranges of variation indicate the standard error of the average). The average proportion of the individuals of the spe cies in the samples is 18.2 ± 3.2% for B. pratorum, 5.7 ± 2.1% for B. lucorum, and 4.9 ± 2.1% for B. spo radicus. The average proportion of the individuals of each of the other eight bumblebee species in the sam ples is 6.5 ± 1.8%, including the cleptoparasitic bum blebees of the genus Psithyrus, the average proportion of which in the samples was 4 ± 1.4%. Changes in the abundance of most bumblebee spe cies on Bol’shoi Solovetskii Island reflect changes in the weather and climate conditions (table). Many meteorological parameters of the cold and warm peri ods proved informative. In the course of overwinter ing, the abundance of the majority of bumblebee spe cies (B. pascuorum, B. hypnorum, B. sporadicus, B. bohemicus, B. norvegicus, and B. sylvestris) is nega tively influenced by thaws, especially during the sec ond half of this period (from January to March). This is confirmed by the high values of rank correlation coefficients for correlation with such parameters as the sum of average daily temperatures >0°C, the number of thaw days (>0°C) per month, and the average monthly air temperatures (especially that of March). The abundance of B. lucorum is negatively influenced

by thaws from October to January (Fig. 2); the abun dance of this species is also positively correlated with the thickness of the snow cover. B. jonellus is in antiphase to all other bumblebee species; the abun dance of this species is, on the contrary, positively cor related with the frequency of winter thaws (Fig. 3). The parameters that proved informative for the summer period are those of heat supply, especially sum of temperatures over the period with stable average daily air temperatures >10°C and the duration of this period. The abundance of the majority of bumblebees (B. pascuorum, B. hypnorum, B. sporadicus, B. norveg icus, and B. sylvestris) is positively correlated with these parameters. The only exception is B. jonellus, the abundance of which is, on the contrary, negatively cor related with the summer heat supply. The abundance of particular bumblebee species is related to the dynamics of average monthly temperatures of the warm period. For instance, the abundance of B. pas cuorum can be positively influenced by the weather conditions in June and September, and the abundance of B. lucorum can be positively influenced by those of May. The temperatures of July are negatively related to the abundance of B. pratorum, and the total precipita tion of July is negatively related to the abundance of B. lucorum. The calculated index Kp/j reproduces the above listed trends, because the dynamics of the abundance of the two superdominant species are in antiphase to each other (Fig. 3). The trajectories of this index and the sum of temperatures for the summer season in the Lamerey diagram are very similar (Fig. 4). In both BIOLOGY BULLETIN

Vol. 40

No. 3

2013

MECHANISM OF DENSITY COMPENSATION IN ISLAND BUMBLEBEE

321

Proportion of the total number of individuals, % 20 2

R = 87% 15

R2 = 56%

1

10 R2 = 72%

5

2

3 –20

–10

–15

0

–5

5

10

Average monthly air temperature, °C Fig. 2. Dependence of the relative abundance of B. lucorum on Bol’shoi Solovetskii Island on the average air temperature of Octo ber (1), December (2), and January (3).

R2 = 70% 60 R2 = 77%

40

20

1 –12

Proportion of the total number of individuals, %

80

2 –10

–8

–6

–4

–2

0

Average monthly air temperature, °C

Fig. 3. Dependence of the relative abundance of the superdominant bumblebee species B. pascuorum (1) and B. jonellus (2) on Bol’shoi Solovetskii Island on the average air temperature of March. BIOLOGY BULLETIN

Vol. 40

No. 3

2013

322

BOLOTOV et al. 10 1

f(t+1)

2

End of 1st cycle

End of 1st cycle

f(t)

1 0

Beginning of 1st cycle

10 Beginning of 1st cycle

0 Fig. 4. Trajectory of the Kp/j index and summer heat supply on Bol’shoi Solovetskii Island on the Lamerey diagram (logarithmic scale) in 2001–2010: 1, Kp/j index; 2, normalized sum of effective summer temperatures >10°C.

B. lucorum 1.0 F3

F2 B. sporadicus B. sylvestris

F5 B. hypnorum

B. bohemicus –0.6

B. pascurum F4

F1 B. pratorum B. flavidus

B. jonellus

B. norvegicus 1.0

–1.5

Fig. 5. Diagram of the canonical analysis of connections between the longterm dynamics of bumblebee abundance on Bol’shoi Solovetskii Island and meteorological parameters: F1, sum of daily temperatures >0°C in January–March; F2, sum of tempera tures over the period with average daily temperatures >10°C; F3, average snow cover thickness in December; F4, monthly total precipitation in April, F5, average air temperature of March. The data are represented by bumblebee samples for 2001–2010 (15028 specimens). The proper values for canonical axes 1 (horizontal), 2 (vertical), 3, and 4 are 0.188 (F = 4.0, p < 0.05), 0.059, 0.047, and 0.005, respectively. The first and second axes together account for 82% of the total variance. The canonical correlations of bumblebee abundance and environmental parameters for axes 1 and 2 are 0.96 and 0.99, respectively.

cases, a five to sixyear cycle can be traced from 2001 to 2006, followed by an almost completed next cycle. For canonical analysis, ten bumblebee species and five meteorological parameters were selected (Fig. 5). In the ordination diagram, the first pleiad includes B. jonellus and B. flavidus, which are associated with

winter parameters (thaws). Along the abscissa, this pleiad is opposed by another one, associated with the summer heat supply; this second pleiad includes almost all other bumblebee species, with the sole exception of B. lucorum, the abundance dynamics of which is associated with the thickness of the snow cover. BIOLOGY BULLETIN

Vol. 40

No. 3

2013

Meteorological parameter

BIOLOGY BULLETIN

Vol. 40

No. 3

2013 0.72 – 0.71

July September –

–

June

–

May

"

0.72

0.72

Period of stable average daily temperatures >10°C Ditto

–

January Period of stable average daily temperatures >5°C

– –

December

Monthly total precipita July tion, mm

Average monthly tem perature, °C

Selyaninov’s hydrother mal coefficient

Duration, days

Sum of average daily temperatures, °C

–0.73

November

–

–

–

–

–

–

–

–

0.73

0.78*

–

–

–

–

–

–

–0.64

–

–

–

–

–

–

–0.64

–

–

0.89*

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

0.71

–0.7

–0.81*

–

–

–

–

–

–

–

0.76

–

–

–

0.77*

–

–

0.7

–

–

B. jonellus

–0.65

–

–

–

0.73

–

–

–

–

0.7

0.79*

–

–

–

–

–

–0.9*

–0.83*

–0.81*

–

–0.83*

–

–

–0.75

–

–

–

–

–

–

–

0.71

0.79*

0.64

–

–

–

–

–

–

–

–

–

–

–0.71

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–0.71

–

–

–

–

–

–

–

0.66

–0.68

–

–

–

–

–

–

–

–

–

0.72

–

–

–

–

–

–0.66

–

–0.81*

–

–

–

–0.89*

–

0.66

–0.9*

–

0.68

–

–

–

–

–

–

–

0.71

–

–

–

–

–

–0.66

–

–0.71

–

–

–

–0.79*

–

–

–0.79*

–

–

–

–

Kp/j

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–

–0.71

–

–0.68

–

–

–

–

–

–

–

0.78*

0.84*

–

–

–

–

–

–

–

–0.79* –0.92*

–

–

–

–0.8* –0.89*

–

0.64

–0.87* –0.75

–

–

B. B. B. B. B. Psithyrus lucorum sporadicus bohemicus norvegicus sylvestris

Note: B. flavidus, which displays no significant correlation with any of the meteorological parameters analyzed, as well as B. distinguendus, B. muscorum, and B. quadricolor, which are represented by single collected individuals, are excluded; minus (–) indicates insignificant correlation. * Significance level of correlation coefficients p < 0.01 (in the other cases, significance level p < 0.05).

Period of bumblebee activity

Average snow cover thickness, cm

April

– – –

March –

– –0.89*

January

–

–0.64

–

December

–

–0.77*

–

–

–0.64

–

–

–

–

–0.84*

October

January–March

–

December

–0.75

January–March –

–

December November

–

B. B. B. pascuorum hypnorum pratorum

November

Period

Monthly total precipita November tion, mm January

Average monthly tem perature, °C

Number of thaw days

Overwinter Sum of average daily ing of bum temperatures > 0°C blebees

Stage

Spearman rank correlation coefficient between abundance of bumblebees and meteorological parameters on Bol’shoi Solovetskii Island

MECHANISM OF DENSITY COMPENSATION IN ISLAND BUMBLEBEE 323

324

BOLOTOV et al.

The average density of bumblebees in ecosystems of Bol’shoi Solovetskii Island in 2007 (60.9 ± 6.0 ind./100 m2, SD = 82.3, n = 188) and in 2010 (70.5 ± 6.7 ind./100 m2, SD = 96.7, n = 207) was the same (Utest: p = 0.13). In 2009, this parameter (105.8 ± 7.2 ind./100 m2, SD = 76.1, n = 113) was 1.5 times as high as in 2007 and 2010 (Utest: p < 0.001). Comparison of the data on the density of bumble bees on Bol’shoi Solovetskii Island and on continental dry meadows in the Northern Dvina River basin has shown the following: in 2007, the abundance of individ uals on continental meadows (46.7 ± 5.5 ind./100 m2, SD = 51.8, n = 90) was no different from that in island ecosystems (Utest: p = 0.79). In 2009, the density of bumblebees on the continent (108.0 ± 12.4 ind./100 m2, SD = 124.1, n = 100) was also no different (Utest: p = 0.95). In 2010, this parameter on continental mead ows (11.1 ± 1.7 ind./100 m2) was lower by a factor of 6 (Utest: p < 0.001). DISCUSSION The number of bumblebee species living on the Solovetskie Islands is smaller by 40–50% than the number of species in particular faunas of the continen tal taiga (Bolotov and Podbolotskaya, 2003; Podbo lotskaya, 2009; Kolosova and Podbolotskaya, 2010). The depleted species composition opens considerable possibilities for colonization of the environment by certain most adapted species. This process leads to a number of compensatory phenomena (Chernov, 2005): increased abundance of particular species, expansion of niches (multidominance), and, on smaller islands, also increased relative species diversity of assemblages (Podbolotskaya, 2009). For some bumblebee species, the isolation of the Solovetskie Islands, removed from the continent by 20–30 km, is no strong obstacle to dispersal. In some years, the materials we collected contained single specimens of B. distinguendus, B. muscorum, and B. quadricolor, which probably are modern invaders. In the literature, migrations of B. lucorum females over the Gulf of Fin land, Baltic Sea, are reported (Mikkola, 1984). How ever, the majority of bumblebee species (especially those of the subgenus Psythirus) probably formed island populations as early as the late glacial period or Early Holocene, when the Solovetskie Islands were connected with the continent or located very close to it (Bolotov and Podbolotskaya, 2003; Shvartsman and Bolotov, 2008). This statement complies with the clas sic notions of Skorikov (1922) and Panfilov (1957) about the rather poor abilities of many bumblebee spe cies in crossing large obstacles, especially vast spaces of water. Individuals of B. jonellus and B. pascuorum in bum blebee samples from Bol’shoi Solovetskii Island over the period from 2001 to 2010 account for about two thirds of the total. Because of the broad ecological tol

erance and a number of preadaptations of these spe cies, they are especially successful at colonizing island habitats (Bolotov and Podbolotskaya, 2003; Podbo lotskaya, 2009). Accordingly, they are species that determine, above all, the aspect and variation of par ticular assemblages. Furthermore, both these species belong on Bol’shoi Solovetskii Island to the category of multidominants (Chernov, 2005): they colonize practically all kinds of biocenoses present on the island. In different years, the aspect of bumblebee assemblages in the same areas often changes cardi nally, according to the specifics of the population dynamics of B. jonellus and B. pascuorum over the whole island. The changes in the proportions of the dominant bumblebee species by abundance are cyclic, with a period of five or six years, which is roughly equivalent to half of the standard 11year solar activity cycle, which has a considerable influence on most nat ural processes. We have at our disposal estimations of bumblebee density on Bol’shoi Solovetskii Island that represent three time sections (Fig. 1): superdominance of B. jonellus (the proportion of the individuals of this species was 50% of the bumblebees collected) accom panied by high abundance of B. pratorum (31%) and medium abundance of B. pascuorum (14%) (2007); decreased abundance of B. jonellus (36%) accompa nied by high abundance of B. pratorum (28%) and B. pascuorum (24%) (2009); and B. pascuorum indi viduals prevailing in the sample (37%) accompanied by medium abundance of the four other dominant species (11–18%) (2010). At the opposite phases of the population dynamics of B. jonellus in 2007 and 2010, the density of bumble bees proved equal; on the descending branch of the cycle in 2009, it was somewhat higher. On the whole, the interannual variation of the average density of bumblebees on Bol’shoi Solovetskii Island was rather low (60–100 ind./100 m2). For three years, the density of bumblebees on the islands was the same as (or higher than) that on the continent, in compliance with the notion of density compensation (Chernov, 2005). Remarkably, in 2010, when on the continent an abnormally dry and hot summer resulted in a decrease in bumblebee abundance by a considerable factor, on the islands this parameter remained at the same level as in 2007 (according to the Northern Department for Hydrometeorology and Environmental Pollution Monitoring, the absolute values of the temperature abnormalities on the islands were smaller). The preliminary results of our analysis of the rela tions between the dynamics of bumblebee abundance on Bol’shoi Solovetskii Island and the parameters of the climate in 2001–2006 have already been reported earlier (Podbolotskaya, 2009; Kolosova and Podbo lotskaya, 2010). Analysis of tenyear series has shown that some local dependences cannot be traced from them (e.g., the dependence of the number of bumble bee species found on the heat supply); however, the BIOLOGY BULLETIN

Vol. 40

No. 3

2013

MECHANISM OF DENSITY COMPENSATION IN ISLAND BUMBLEBEE

most significant patterns of the variation of assem blages in connection with the weather and climate conditions of the seasons remain unchanged. According to the data obtained and the results of our mathematical analyses, the mechanism of popula tion density compensation in bumblebee assemblages on Bol’shoi Solovetskii Island works as follows. In years with unfavorable winter conditions (thaws that occur from January to March are the most important factor), part of the overwintering female bumblebees die out. Their mortality leads to a decrease in the num ber and density of the nests started in spring, and this decrease, in turn, determines the rather low density of foraging workers. Consequently, the food supply of the entomophilous plants is not fully utilized by bumble bees, and a surplus of available trophic resources is formed. Under these conditions, B. jonellus gets a competi tive advantage, because this species is distinguished by early nest founding, small family size, and extremely rapid family development, which allows it to develop two generations in some years even in the Subarctic (Meidell, 1968; Douglas, 1973; Kolosova and Podbo lotskaya, 2010). Thus, Douglas (1973) observed in northern Norway the emergence of young males and females of B. jonellus in late June to the first ten days of July; after that, their numbers decreased and the last males were recorded on July 18. During the third ten day period of July, that author recorded young females that had little flight experience, which collected nec tar. Anatomical analysis showed the presence of maturing ovaries in their organisms, warranting the conclusion that the females started making new nests and breeding the second generation during the same summer. Two peaks in the abundance of workers have also been recorded; their high abundance in August was caused by the development of families founded by reproductive individuals of the new generation. On the Solovetskie Islands, we have also periodi cally recorded the emergence of reproductive individ uals of the new generation of B. jonellus in late June to early July (Kolosova and Podbolotskaya, 2010). Male bumblebees occur in June on the Solovetskie Islands also in B. pratorum, which also belongs to the subgenus Pyrobombus, but they are rather scant (Kolosova and Podbolotskaya, 2010). In a nest of B. jonellus, found on July 7, 2011, on the southern coast of Bol’shoi Solovetskii Island, we found a founding female, eight workers, and one male. The nest contained 37 cocoons, 24 of them with larvae. The presence of free trophic resources in combination with the depressed abundance of other bumblebee species may contribute to the successful development of the second generation of B. jonellus on Bol’shoi Solovetskii Island. In addition, it was shown earlier that the repro ductive success of bumblebees can be influenced by the food supply (Goulson et al., 2002; Elliott, 2009). Douglas (1973) suggested that two generations give B. jonellus an evolutionary advantage, decreasing the BIOLOGY BULLETIN

Vol. 40

No. 3

2013

325

competition for food, because families of the second generation emerge in late summer and, thus, have more trophic resources at their disposal. But this is not true, because, on the contrary, these families are the ones that encounter the problem of lacking free trophic resources more often than others. The prob lem lies in the fact that it is during the second half of the season that the families of the other bumblebee species reach their maximum abundance, unless some of the potential limiting factors depress their popula tions. In the latter case, a surplus of free resources is formed, which is utilized by B. jonellus families of the second generation. The lability of the life cycle allows B. jonellus to flourish at the periphery of the adaptive zone of forest bumblebee species, including smaller islands, the hypoarctic tundra, and some foresttundra areas (Chernov, 1966; Shvartsman et al., 2003; Kolosova and Potapov, 2011). At the same time, in taiga habi tats, this species is not abundant (Bolotov and Kolos ova, 2006; Kolosova, 2007). Its facultative bivoltine life cycle should be interpreted as a preadaptation, which allowed the species to disperse northwards suc cessfully. But the formation of this feature was proba bly related to the increase in the heat supply of the sea sons in the course of the dispersal of the originally montane species over the plane taiga and southern Subarctic, rather than to the competition between this species and other bumblebee species. The rather small size of this species allows it to forage successfully on the small flowers of dwarf shrubs of the family Eri caceae, which are dominant in the plant cover of most phytocenoses on the Solovetskie Islands (Kiseleva et al., 2005; Prirodnaya sreda …, 2007), as well as on the forest tundra and hypoarctic tundra. However, these preadaptations are insufficient for successful colonization of arctic landscapes. Even in typical tun dras, the abundance of B. jonellus sharply decreases, and this species is represented among the collected materials by a few individuals, less numerous than rep resentatives of the subgenus Alpinobombus (Kolosova and Potapov, 2011). On the whole, the possibility of a direct increase in the number of bumblebee families, and, thus, abun dance of bumblebee individuals in the same season is extremely important for their colonization of the Sub arctic and the northern temperate zone. The abun dance of founding females can be considerably influ enced not only by unfavorable environmental condi tions during overwintering, but also by sharp changes in weather conditions in early summer. For instance, in the environs of Arkhangelsk in 1996, female bum blebees emerged from their overwintering sites in May and started founding their nests. Suddenly, the tem perature sharply dropped, accompanied by an abun dant snowfall and ice formation. As a result, extremely low bumblebee abundance was recorded here in the following summer: we collected fewer than 100 speci mens at different stations during the fieldwork period

326

BOLOTOV et al.

from July to August. The abundance of bumblebees was restored only in the following year. The abundance gradation of B. jonellus is especially successful in cold summer seasons, when the sum of effective temperatures is at most 1200°C, while the total duration of such seasons is usually less than 90 days (Podbolotskaya, 2009; Kolosova and Podbo lotskaya, 2010). By contrast, in warm seasons, when the duration of the season and the sum of effective temperatures is higher than the abovegiven values, other species are usually dominant by abundance, especially B. pascuorum. This is explained by the more favorable conditions of warm seasons for the develop ment of bumblebee families; these conditions can probably partly compensate for the negative conse quences of winter thaws. This suggestion is confirmed by the results of canonical analysis (Fig. 5), in which the parameters of thaw frequency in winter along the horizontal axis are opposed to the parameters of summer heat supply and, consequently, B. jonellus is opposed to almost all other bumblebee species. B. lucorum occupies in the canon ical analysis diagram a special position; the abundance of this species is clearly regulated by other factors. For B. lucorum, the most important factors are the absence of thaws during the first half (October to January) of the overwintering period, greater snow cover thickness during the coldest winter months, and dry conditions in summer. This polyzonal species, like B. jonellus, rather successfully colonizes the Subarctic, where in some bumblebee assemblages of the forest tundra and hypoarctic tundra, it is among the abundant species or even superdominants (Kolosova and Potapov, 2011). In these zones, it is distributed preferably along sea coasts and in river deltas, because of the welldrained landscapes and presence of sandy soils, which make it easier to nest and provide for dry conditions in the nests. This species nests underground, at a depth of around 18 cm (Skorikov, 1922), and the length of the entrance shaft of its nest is greater than 40 cm (Løken, 1973). As a result of their increased density, species strongly dominant by abundance in island assem blages, in effect, compensate for the depleted number of species in the biota (Chernov, 2005). It has been found that the role of such species in assemblages can change considerably with time. It can be concluded that there is a dynamic compensation system, in which leading positions in assemblages are occupied by alter nating species. The work of this system is regulated by both abiotic factors (weather and climaterelated) and by biotic ones (competition for trophic resources). Since bumblebees make up around 99% of the total abundance of the superfamily Apoidea in the Solovetskie Islands, they are the principal pollinators of entomophilous plants on the archipelago; i.e., they represent a single functional group of the community. It is at the scale of such groups that the density com

pensation effect manifests itself especially clearly (McGradySteed and Morin, 1999). In the studied island bumblebee assemblages, there is a kind of “reserve” of potential compensatory spe cies. During the periods of depressed populations in other species, these species can become dominant in the structure of the assemblages and quickly compen sate for the decreased total abundance of the assem blages, rising it to a certain optimal (or maximal?) level. It is proposed to treat them as reserve compensa tory species. The presence of such species increases the hardiness of assemblages to changing environmen tal conditions. On the Solovetskie Islands, this com pensatory mechanism provides for the stable pollina tion of entomophilous plants during fluctuations of the weather and climate. Thus, in the vast meadows created on reclaimed lands as early as the 19th century and possessed by the Solovetskii Monastery, legumes (Trifolium, Vicia, Lathyrus) still retain an important part in the composition of the herbaceous layer, although they have not been reseeded for many decades (Prirodnaya sreda …, 2007). The results of this study comply with the theoreti cal notions about natural cycles, according to which natural cyclicity is defined as a system of intercon nected synchronized changes in the longterm dynamics of the biocenosis and in its physiographic surroundings, and the natural cycle is the temporal and territorial unit of this cyclicity (Shnitnikov, 1969; Maksimov and Erdakov, 1985; Maksimov, 1989). Our data also comply with the concept of stable attractive states of the biota, which are formed as a response of the biome to particular combinations of abiotic and biotic environmental factors (Zakonomernosti polu vekovoi …, 2000). (We retained here the interpretation of the term “biota” used by the authors of that con cept, although it is incorrectly interpreted in their study, meaning, doubtlessly, the cenotic system of the taiga, rather than the composition of taxa; the impor tance of strictly distinguishing between these terms was emphasized by Chernov (1984).) Such states are recognized based also on abiotic (weather and cli materelated) factors. ACKNOWLEDGMENTS The authors are grateful to M.Yu. Gofarov, S.A. Iglovskii, Yu.A. Shvakov, and many others who took part in the fieldwork collecting bumblebees on the Solovetskie Islands in different years. This study was supported by the President of the Russian Federation (project no. MD4164.2011.5), the Russian Foundation for Basic Research (project nos. 070400313, 100400897, and 110498817), the Ural Branch, Russian Academy of Sciences (project nos. 12P51014 and 12M452062), the Federal Special Purpose Program “Personnel,” and the Official State Program “Thematic Plan of Higher Education Institutions” (project no. 546152011). BIOLOGY BULLETIN

Vol. 40

No. 3

2013

MECHANISM OF DENSITY COMPENSATION IN ISLAND BUMBLEBEE

REFERENCES Bennett, A.F. and Gorman, G.C., Population density and energetics of lizards on a tropical island, Oecologia, 1979, vol. 42, pp. 339–358. Berezin, M.V., Beiko, V.B., and Berezina, N.B., Analysis of structural changes in the population of bumblebees (Bom bus, Apidae) in the Moscow oblast for the last 40 years, Zool. Zh., 1996, vol. 75, no. 2, pp. 212–221. Bespalaya, Yu.V., Bolotov, I.N., and Zubrii, N.A., Topical groups of mollusks in the lakes of Bol’shoy Solovetsky Island (Solovetskiy Archipelago, White Sea, northwestern Russia), Inland Water Biol., 2009, vol. 2, no. 2, pp. 177– 186. Bespalaya, Yu.V., Bolotov, I.N., and Usacheva, O.V., Struc ture and species diversity of topical groups of mollusks in lakes of the Solovetsky Islands and Onega Peninsula, north western Russia, Russ. J. Ecol., 2011, vol. 42, no. 2, pp. 143– 150. Bolotov, I.N., Butterflies (Lepidoptera, Diurna) of Solovetsky Islands (northwest Russia, White Sea), Zool. Zh., 2006, vol. 85, no. 8, pp. 943–949. Bolotov, I.N. and Kolosova, Yu.S., Trends in the formation of biotopic complexes of bumblebees (Hymenoptera, Api dae: Bombini) in northern taiga karst landscapes of the west ern Russian Plain, Russ. J. Ecol., 2006, vol. 37, no. 3, pp. 156–166. Bolotov, I.N. and Podbolotskaya, M.V., Local fauna of bumblebees (Hymenoptera: Apidae, Bombini) of the Euro pean north of Russia: Solovetsky Islands, Vestn. Pomorsk. Univ., Ser. Estestv. Tochn. Nauki, 2003, no. 1 (3), pp. 74–87. Bolotov, I.N. and Shutova, E.V., Patterns of formation of island fauna of butterflies (Lepidoptera, Diurna) at the northern forest boundary in the region of Pleistocene con tinental glaciation (by the example of White Sea Islands), Biol. Bull., 2006, vol. 33, no. 3, pp. 260–268. Bolotov, I.N., Zubrii, N.A., Tsyvareva, E.P., and Khristo forova, N.S., The species composition of ground beetles (Coleoptera, Carabidae) of Solovetsky Islands, Vestn. Pomorsk. Univ., Ser. Estestv. Tochn. Nauki, 2011, no. 2, pp. 45–52. Chernov, Yu.I., The complex of anthophilous insects in the tundra zone, in Voprosy geografii (Problems of Geography), vol. 69: Organizmy i prirodnaya sreda (Organisms and the Natural Environment), Moscow: Mysl’, 1966, pp. 76–97. Chernov, Yu.I., Pathways and sources of formation of the fauna of small islands of Oceania, Zh. Obshch. Biol., 1982, vol. 43, no. 1, pp. 35–47. Chernov, Yu.I., Species diversity and compensatory effects in communities and biological systems, Zool. Zh., 2005, vol. 84, no. 10, pp. 1221–1238. Chernov, Yu.I., Ekologiya i biogeografiya. Izbrannye raboty (Ecology and Biogeography: Selected Works), Moscow: KMK, 2008. Colla, S.R., COSEWIC Assessment and Status Report on the RustyPatched Bumble Bee Bombus affinis in Canada, Ottawa: Committee on the Status of Endangered Wildlife in Canada, 2010, pp. 1–34. www.sararegistry.gc.ca/sta tus/status_e.cfm Colla, S.R. and Packer, L., Evidence for the decline of east ern North American bumblebees, with special focus on Bombus affinis Cresson, Biodiv. Conserv., 2008, vol. 17, pp. 1379–1391. BIOLOGY BULLETIN

Vol. 40

No. 3

2013

327

Douglas, J.M., Double generations of Bombus jonellus sub borealis Rich. (Hym. Apidae) in an Arctic summer, Ento mol. Scand., 1973, no. 4, pp. 283–284. Dupont, Y.L., Damgaard, C., and Simonsen, V., Quantita tive historical change in bumblebee (Bombus spp.) assem blages of red clover fields, PLoS ONE, 2011, vol. 6, no. 9, pp. 1–7. Elliott, S.E., Surplus nectar available for subalpine bumble bee colony growth, Environ. Entomol., 2009, vol. 38, no. 6, pp. 1680–1689. Gonzalez, A. and Loreau, M., The causes and conse quences of compensatory dynamics in ecological commu nities, Annu. Rev. Ecol. Evol. Syst., 2009, vol. 40, pp. 393– 414. Goulson, D., Hughes, W.O.H., Derwent, L.C., and Stout, J.C., Colony growth of the bumblebee, Bombus ter restris, in improved and conventional agricultural and sub urban habitats, Oecologia, 2002, vol. 130, pp. 267–273. Kiseleva, K.V., Novikov, V.S., Oktyabreva, N.B., and Cher enkov, A.E., Opredelitel’ sosudistykh rastenii Solovetskogo arkhipelaga (Vascular Plants of the Solovetsky Islands: An Identification Guide), Moscow: KMK, 2005. Kolosova, Yu.S., Fauna and ecology of bumblebees (Hymenoptera, Apidae, Bombus) of forest ecosystems of the northern taiga of the Russian Plain, Extended Abstract of Cand. Sci. (Biol.) Dissertation, Syktyvkar: Inst. Biol. Komi NTs UrO RAN, 2007. Kolosova, Yu.S. and Podbolotskaya, M.V., Population dynamics of bumblebees (Hymenoptera, Apidae, Bombus Latr.) on the Solovetsky Archipelago: the results of ten years of monitoring, Tr. Rus. Entomol. Obshch., 2010, vol. 81, no. 2, pp. 135–141. Kolosova, Yu.S. and Potapov, G.S., Bumblebees (Hymenoptera, Apidae) of the forest–tundra and tundra in the northeastern Europe, Zool. Zh., 2011, vol. 90, no. 8, pp. 959–965. Legendre, P. and Gallagher, E.D., Ecologically meaningful transformations for ordination of species data, Oecologia, 2001, vol. 129, pp. 271–280. Løken, A., Studies of Scandinavian bumble bees (Hymenoptera, Apidae), Norsk Entomol. Tidsskrift, 1973, vol. 20, pp. 1–218. MacArtur, R.H., Diamond, J.M., and Karr, J.R., Density compensation in island faunas, Ecology, 1972, vol. 53, no. 2, pp. 330–342. MacArtur, R.H. and Wilson, E.O., The Theory of Island Biogeography, Princeton: Princ. Univ. Press, 1967. Maksimov, A.A., Prirodnye tsikly. Prichiny povtoryaemosti ekologicheskikh protsessov (Natural Cycles: Causes of Recurrence of Ecological Processes), Leningrad: Nauka, 1989. Maksimov, A.A. and Erdakov, L.N., Tsiklicheskie protsessy v soobshchestvakh zhivotnykh (bioritmy, suktsessii) (Cyclic Processes in Animal Communities (Biorhythms and Suc cessions)), Novosibirsk: Nauka, 1985. McGradySteed, J. and Morin, P.J., Biodiversity, density compensation, and the dynamics of populations and func tional groups, Ecology, 1999, vol. 81, no. 2, pp. 361–373. Meidell, O., Bombus jonellus (Kirby) (Hym., Apidae) has two generations in a season, Norsk Entomol. Tidsskrift, 1968, vol. 14, pp. 31–32.

328

BOLOTOV et al.

Mikkola, K., Migration of wasp and bumblebees queens across the Gulf of Finland (Hymenoptera: Vespidae and Apidae), Notulae Entomol., 1984, vol. 64, pp. 125–128. Panfilov, D.V., Experience of reconstruction of paleogeog raphy of Northern Eurasia in the Quaternary based on modern bumblebee fauna, in Vopr. paleobiogeografii i bios tratigrafii. Tr. I sessii paleontol. obshch. (Problems of Paleo biogeography and Biostratigraphy: Transactions of the I Session of the Paleontological Society), Moscow: Gos geoltekhizdat, 1957, pp. 97–106. Pesenko, Yu.A., Printsipy i metody kolichestvennogo analiza v faunisticheskikh issledovaniyakh (Principles and Methods of Quantitative Analysis in Faunistic Studies), Moscow: Nauka, 1982. Podbolotskaya, M.V., Fauna and ecology of bumblebees (Hymenoptera, Apidae: Bombus) of Solovetsky Islands, Extended Abstract of Cand. Sci. (Biol.) Dissertation, Syk tyvkar: Inst. Biol. Komi NTs UrO RAN, 2009. Prirodnaya sreda Solovetskogo arkhipelaga v usloviyakh menyayushchegosya klimata (The Natural Environment of the Solovetsky Archipelago in a Changing Climate), Shvartsman, Yu.G. and Bolotov, I.N., Eds., Yekaterinburg: Izd. UrO RAN, 2007. Puzachenko, Yu.G., Matematicheskie metody v ekolog icheskikh i geograficheskikh issledovaniyakh (Mathematical Methods in Ecological and Geographical Studies), Mos cow: Akademiya, 2004. Rodda, G.H. and DeanBradley, K., Excess density com pensation of island herpetofaunal assemblages, J. Biogeogr., 2002, vol. 29, pp. 623–632. Severtsov, A.S., Population as an object of natural selection, in Ekologiya populyatsii: struktura i dinamika. Mater. soveshch. (Population Ecology: Structure and Dynamics.

Proc. Meet.), Moscow: IPEE im. A.N. Severtsova, Ros. Akad. Nauk, 1995, Ch. 1, pp. 42–62. Shnitnikov, A.V., Vnutrivekovaya izmenchivost’ komponentov obshchei uvlazhnennosti (Intracentennial Variation of the Overall Moisture Components), Leningrad: Nauka, 1969. Shvartsman, Yu.G., Bolotov, I.N., and Iglovskii, S.A., The current geoecological state of landscapes of Mezenskaya tundra, Vestn. Pomorsk. Univ., Ser. Estestv. Tochn. Nauki, 2003, vol. 1, no. 3, pp. 42–55. Shvartsman, Yu.G. and Bolotov, I.N., Prostranstvennovre mennaya neodnorodnost’ taezhnogo bioma v oblasti pleistot senovykh materikovykh oledenenii (Spatiotemporal Hetero geneity of the Taiga Biome in the Pleistocene Continental Glaciation Area), Yekaterinburg: Izd. UrO RAN, 2008. Skorikov, A.S., Palearctic bumblebees. Part 1. General biol ogy (with inclusion of zoogeography), Izv. Sev. Oblastn. Stants. Zashchity Rast. Vredit., 1922, vol. 4, no. 1, pp. 5– 160. ter Braak, C.J.F., Unimodal Models to Relate Species to Environment, Wageningen: DLOAgricultural Mathematics Group, 1996. ter Braak, C.J.F., and Šmilauer, P., CANOCO Reference Manual and CanoDraw for Windows User’s Guide: Software for Canonical Community Ordination (Version 4.5), Ithaca; New York: Microcomputer Power, 2002. Wright, J.S., Density compensation in island avifaunas, Oecologia, 1980, vol. 45, pp. 385–389. Zakonomernosti poluvekovoi dinamiki bioty devstvennoi taigi Severnogo Predural’ya (Patterns of HalfCentennial Dynamics of the Biota of the Northern TransUral Virgin Taiga) Vasil’ev, A.G., Ed., Syktyvkar: Goskomstat Respub liki Komi, 2000.

BIOLOGY BULLETIN

Vol. 40

No. 3

2013