Habitat quality and B chromosomes in the yellow‐necked mouse Apodemus flavicollis

Abstract

The correlation between frequency of B chromosomes and quality of habitat was analysed in 680 specimens of yellow‐necked mice, Apodemus flavicollis, collected at six localities in Serbia. Animals with B chromosomes were present at all localities in frequencies ranging from 0.23 to 0.40. Quality of habitat was judged from an index of overall body size (BSI). At the studied localities, the frequency of animals with Bs was negatively correlated with the index of overall body size (r = −0.93), indicating that B chromosomes in this species are mentioned due to the effects that they exert at the level of populations. This finding supports the heterotic model of maintenance of B chromosomes.

Keywords:

• B chromosomes
• Apodemus flavicollis
• habitat quality
• heterotic model
• parasitic model

Introduction

Polymorphism produced by the presence of supernumerary chromosomes characterises almost 15% of species through all major taxonomic categories except for birds. Dispensability, absence of pairing and recombination with members of standard complements, together with non‐Mendelian inheritance are the main shared characteristics of these chromosomes, named B chromosomes (Bs) (Jones & Houben 2003). Although the percentage of mammalian species in which Bs were recorded is small (1.2%), the genus Apodemus is an exception, since as many as six species possess Bs (Vujošević & Blagojević 2004; Wojcik et al. 2004). For a long time, maintenance of Bs in natural populations has been explained by two models. The parasitic model, proposed by Östergren (1945) and defined as selfish by Jones (1985), claims that Bs are maintained by meiotic drive despite detrimental effects on carrier fitness (Jones 1991). On the other hand, the heterotic model of White (1973) suggests that, in the absence of drive, a small number of Bs could confer an adaptive advantage to carriers, while a large number could be detrimental. With rare exceptions, attempts to find the adaptive significance of Bs failed, but it seems that the complexity of such studies has discouraged research in that direction. This contributed to the currently prevailing view that Bs are genomic parasites.

If B chromosomes are deleterious to the carriers, the logical outcome is that their frequency is the highest in an environment which is optimal for the species and where they could be tolerated more easily. This was demonstrated in a number of plant species studied in detail, grasshoppers (for a summary, see Jones & Rees 1982), and characid fish (Neo et al. 2000). However, in two plant species, Dampiera linearis and Allium shoenoprasum (Bousefield & James 1976; Plowman & Bougourd 1994), and one Japanese species of arachnida (Gorlov & Tsurusaki 2000), the frequency of Bs is not habitat‐dependent in this way.

In the populations of yellow‐necked mice, Apodemus flavicollis, B chromosomes are widespread (Zima & Macholán 1995; Kartavtseva 2002). Accumulated data on the correlation between frequency of Bs and environmental variables in A. flavicollis, as well as those for phenotypic variation, indicate that Bs exert their effects at the level of populations (Vujošević & Blagojević 2000) and it is suggested that, at least in certain environments, the presence of Bs could be beneficial for their host (Zima et al. 2003).

While it is not so difficult to define the optimal habitat for certain plant species, for mammals it is. In small mammals, the competitive ability of individuals is closely related to body size (Corbert 1961; Montgomery & Gurnell 1985; Wauters & Dhondt 1989). Individuals with higher competitive ability will tend to occupy better habitats, or, in other words, the quality of habitat for some species could be judged by comparison of parameters related with size, between animals from different populations. The aim of the present work was to establish whether or not there is any correlation between frequency of Bs and quality of habitat (including biotic and abiotic factors).

Materials and methods

A total of 680 (342 females and 338 males) adult specimens of Apodemus flavicollis, collected using Longworth traps at six localities (Table I) in Serbia, were analysed. Part of the specimens from this study had been included in previous studies (Blagojević & Vujošević 1995, 2004; Vujošević & Blagojević 2000). Some differences in number and frequencies of Bs from previously published data are due to elimination of juvenile animals and addition of new specimens collected at the same localities. In this study, only adults with complete eruption of the third molar were included. Chromosomes were prepared directly from bone marrow using the standard procedure. The presence of Bs was confirmed by examining 30 metaphase figures. All animals with more than 48 chromosomes (standard complement) were considered to have Bs.

Table I. Localities at which animals were collected, together with their vegetational units

Locality	UTM coordinates	Vegetational units	Number of animals
			Females	Males
Jastrebac Mt.	EP30	Fagetum montanum	132	115
Ada	CQ82	Salici‐Populetum	42	56
Goč Mt.	DP82	Quercetum petrae	9	9
Avala Mt.	DQ64	Quercetum frainetto‐cerris	75	81
Fruška Gora Mt.	DR00	Quercetum petrae	13	16
Cer Mt.	CQ84	Querco‐Carpinetum frainetto‐cerris	71	61
Total			342	338

Cranial variables, skull length (SL) and skull width (SW), were measured using digital callipers accurate to 0.05 mm. Body length (total body length minus tail length) was measured to the nearest millimetre. To estimate habitat quality, an index of body size (BSI) was used. This index of overall body size (Dobson 1992) is represented by the first principal component (PC1) extracted from log‐transformed values of body length, skull length and skull width.

Results and discussion

Animals with Bs were found at all localities studied in the following frequencies: Ada 0.23, Avala Mt. 0.25, Goč Mt. 0.39, Cer Mt. 0.30, Jastrebac Mt. 0.39 and Fruška Gora Mt. 0.40. In the whole sample, animals with one B chromosome were the most frequent (71.7%), while there were 23.6% with two Bs, 4.2% with three Bs, but only 0.5% with five Bs.

The first principal component (PC1) described over 87.1% of the variation in body measurements. Therefore, it was possible to use PC1 as an index of overall body size. Body size index was negatively correlated (r = −0.93) with frequency of animals with Bs (Figure 1). Levene's test showed that differences among localities in BSI (F _(5,647) = 5.43, P<0.001) were statistically significant.

Figure 1 Correlation between frequency of animals with Bs and body size index.

Contrary to the predictions of the parasitic model of maintenance of Bs in populations, according to which the highest B frequencies would be expected in those localities with more favourable habitats, the frequency of animals with Bs was the lowest in the habitat estimated as the optimal one from the index of overall body size. Thus, the index of overall body size was the smallest at the locality, Fruška Gora Mt., where the frequency of animals with Bs was the highest. This result, together with the fact that the mechanism of accumulation has never been elucidated in this species, support the heterotic model of B chromosomes maintenance. In a population of A. flavicollis from the Czech Republic, Zima et al. (2003) detected significant regression between the frequency of B chromosomes and body mass. They found that specimens with higher body mass (overwintered adults) possessed a distinctly higher frequency of Bs than those born in the year of collection (mainly juveniles). This trend was explained by better survival of individuals with B chromosomes during the cold winter. Both the results of Zima et al. (2003) and those of the present study point to greater survival of B carriers in unfavourable conditions.

The choice of habitat that best satisfies an individual's basic ecological needs could be of crucial importance for animal survival. Habitat selection is based on differences in the fitness of different phenotypes in different habitats. Our results show that the fitness of animals with Bs increases in habitats which are not optimal for the species. The presence of Bs could be seen as cryptic genetic variability that gives an opportunity to the species to widen the edges of occupied habitats. Animals are not entirely able to follow a free choice of habitat that optimizes their fitness, due to despotic behaviour, i.e. territorial defence of a high quality habitat (the “ideal despotic” distribution of Fretwell & Lucas 1970; Fretwell 1972). Despotic behaviour is influenced by the size of animals, which changes in general in the presence of Bs.

The frequency of A. flavicollis individuals with Bs is at an imperfect equilibrium. This is characterised by variation in the frequency of specimens with Bs during the year, but stability in frequency from year to year (Vujošević 1992; Blagojević & Vujošević 1995; Vujošević & Blagojević 1995).

In A. flavicollis the frequency of animals with Bs increases with altitude and is negatively correlated to the average temperature (Vujošević & Blagojević 2000). This has rather complicated the pattern of geographical distribution (Wójcik et al. 2004). The trend of increase in frequency of animals with Bs towards more unsuitable conditions of habitat is supported by the present study. The existence of specific DNA profiles together with differential and increased expression of three genes, found in the presence of Bs, also signifies that Bs in this species interact with the rest of the genome, which does not necessarily produce genetic conflict (Tanić et al. 2000, 2005). It seems that the presence of Bs is widening the genetic variability in this species, enabling it to occupy successfully more habitats and to extend its specific response to changes in the landscape.

Acknowledgements

This work was supported by the Ministry of Science and Environmental Protection of Serbia, Contract No. 143011.