Polytypic chromosomal variation in Triturus boscai (Urodela: Salamandridae)

Résumé Variation polytypique des chromosomes de Trjturus boscai (Urodèle : Salamandridés). Nous avons analysé la variation chromosomique de populations de l’espèce endémique ibérique Triturus boscai moyennant l’étude de la distribution de l’hétérochromatine et de la quantité d’ADN. Nous présentons des arguments en faveur de l’existence de 2 groupes de populations dont les hybrides montrent une fertilité réduite.

However, some other authors have criticized this kinship based on its possible mating with T alpestris which belongs to the Mesotriton group (Spurway, 1953;Bucci-Innocenti et al, 1983).
The studies on this newt, although very scarce, include some devoted to the analysis of the chromosome complement (Herrero, 1982a,b;Bucci-Innocenti et al, 1983). However, there is no information about intraspecific variability in natural populations.
In this paper we have explored the intraspecific chromosomal variation of T boscai according to 2 features (heterochromatin distribution and DNA content) in different populations of the Iberian peninsula. The differences reported in the C-banding pattern of this species are particularly significant for this genus which shows a high degree of chromosome stability.

MATERIAL AND METHODS
The populations studied are summarized in figure 1. We collected ! 10 males and 10 females from each population, excepting those belonging to Cenicientos, Monfrague and those of the La Vera region which were sampled during 3 consecutive yr, totaling 30 individuals for each population per year.
Specimens were injected with a 0.3% colchicine solution. After 7 h, they were sacrificed and mitotic chromosomes were prepared directly from both intestine and testes and meiotic chromosomes were obtained from testes. In every case the material was fixed in ethanol : acetic acid (3:1) for at least 24 h.
The C-banding technique applied was that reported by Sumner (1972) with some modifications : mitotic and meiotic chromosomes were incubated in Ba(OH) 2 at 60 °C for 15 min and 30 min, respectively. The C-banding pattern of each individual was obtained after analyzing 5 mitotic metaphases. The chromosomes were paired in the karyotype according to their size and C-banding pattern. Chiasma distribution was determined by observation of 20 diplotene cells for each individual.
Nuclear DNA content of the different individuals was estimated from blood smears stained by Feulgen's reaction. Measurements were taken with a Vickers M-85 integrating microdensitometer at a wavelength of 550 nm. In every case blood smears of B!fo bufo were used as a standard. Its DNA content was taken as equal to 100 so that the DNA content of the T boscai samples were calculated in relative units. For each individual at least 100 nuclei of erythrocytes were measured to avoid a standard error of the mean higher than 1%. The mean values of DNA content were compared with a Student's t-test.

RESULTS
The chromosome complement of T boscai consists of 24 chromosomes of decreasing size, where 4 pairs are submetacentric and the remainder are metacentric (Herrero, 1982a).

Heterochromatin distribution
The C-banding patterns of each individual within a given population are identical. However, the patterns show some remarkable differences between distinct populations. One of them is found in the individuals from Madrigal de la Vera, Valverde de la Vera, Villanueva de la Vera and Losar de la Vera (see fig 1). This pattern not have the same distribution. Pair 10 presents a series of thin bands restricted to the long arm that appears as 3 bands when the chromosome is highly condensed; pair 11 has none and pair 12 presents a single one close to the centromeric region (fig 2a). The distance between the pericentric bands located on a given arm varies from 1 chromosome pair to another and sometimes between both arms of the same chromosome. Thus they can be observed as a single band when the chromosomes are highly condensed.
Subterminal bands appear on the long arms of pairs 2 and 11 and on the short arm of pair 8. There are terminal bands on both arms of pair 8 and on the short arm of pair 6. Moreover, pair 7 has an interstitial band on its long arm (fig 2a). The remaining populations showed differences in the heterochromatin distribution affecting mainly chromosome pairs 8 and 11. In this case, the telomeric bands of pair 8 are not present and pair 11 shows two pericentric bands on both sides of the centromere but lacks the subterminal band on the long arm (fig 2b).
On the other hand, we must emphasize that the centromeric index and relative lengths of these pairs do not show any differences between populations and coincide with those previously described by Herrero (1982a). Moreover, there is a correspondence between the C-banding patterns reported here and those previously reported (Herrero, 1982b;Bucci-Innocenti et al, 1983). The pattern corresponding to La Vera populations coincides with that described by Herrero (1982b), where individuals La Vera were also analyzed. The second pattern coincides with that reported by Bucci-Innocenti et al (1983) who do not give details on the geographic origin of the specimens studied.

DNA content
The results from DNA cytophotometric measurements are shown in table I. They do not show significant differences between the populations studied. DNA content is almost 4 times higher than that of Bufo bufo. The absolute value has been calculated considering 14 pg/N for B bufo (Bachmann, 1970).

Meiotic behaviour
At diplotene, bivalents form from 1 to 3 chiasmata in terminal, subterminal or interstitial positions (Herrero and L6pez-Ferndndez, 1986). However, in one population (Valverde de la Vera) 4 males showed a strong incidence of desynapsis since at diplotene several univalents or bivalents with a single terminal chiasma were always observed (fig 3). These individuals, in spite of the normal size of their gonads, were completely sterile since no further meiotic phases were scored in the preparations that were devoid of any spermatozoa. Interestingly, these individuals may be considered as hybrids between the 2 forms described according to their C-banded pattern since, at least, they were heterozygous for pairs 8 and 11 (fig 3).

Chromosomal divergence
Triturus is a group where no gross chromosome rearrangements have occurred (Mancino et al, 1977) except those referred to pair 1 of the Neotriton group (Sims et al, 1984), or the pericentric inversion described in some populations of T italicus (Ragghianti et al, 1980). Many authors support the existence of small rearrangements as main events in the chromosome evolution of this genus (Mancino et al, 1977;Macgregor et al, 1983). However, no clear evidence for this has yet been obtained. For this purpose, studies on intraspecific chromosomal variation in closely related species is desirable, particularly since small rearrangements would be clearly revealed because of the similar characteristics of chromosome constitution in the groups analyzed. However, the chromosomal differentiation found in the T alpestris complex (Herrero et al, 1989) only refers to the amount of heterochromatin. This also seems to be the case of T boscai. The 2 C-banded patterns found cannot be easily explained by simple rearrangements : the morphology and size of chromosomes are preserved. The differences only refer to 2 pairs (8 and 11) which seem to undergo subtle changes in the amount and distribution of heterochromatin. Accordingly DNA values are not significantly altered, although small DNA variations could not be detected with this method. King (1980) suggested a euchromatin-heterochromatin transformation process for explaining these phenomena in Litoria. However, no further evidence has supported his argument. The alternative model by Macgregor and Sessions (1986) on the growth and dispersion of satellite DNA sequences sequestered in heterochromatin regions of newts of the genus Triturus seems more consistent. According to their model, satellite DNA sequences or heterochromatic regions would arise at centromere positions wherefrom they would be dispersed through the genome by successive amplifications and insertions based on unequal sister chromatid exchange, chromosomal rearrangements or some other molecular mechanisms. Although we do not have information on satellite DNA sequences located in the heterochromatic regions of T boscai, this model could fit in. However, an important drawback for this hypothesis stems from the short evolutionary time in which the heterochromatin differences in T boscai would have occurred, in comparison to the evolutionary times for which the model was proposed.

Heterozygosity and infertility
Whatever the origin of the chromosome differences described, a polytypic variation affecting the Iberian populations of T boscai is clearly shown. However, in some areas where populations showing distinct C-band patterns could meet we have found sterile hybrids. The meiotic behaviour of these individuals suggests the existence of mechanisms that prevent normal completion of meiosis. In fact no spermatid nuclei are formed. These results clearly suggest that although hybrids may form and become adults they are sterile. As a consequence, the chromosome differences described may uncover other functions that promote reproductive isolation between both forms.
In summary, Iberian populations of T boscai are distributed in at least two groups according to their C-banding patterns : one group is restricted to the La Vera region, located in the Tietar Valley of the Gredos Mountains and the other one extends throughout the remainder of the geographical distribution of this species.
Moreover, in places where both groups meet, hybrids present a high chromosomal instability affecting the meiotic behaviour and resulting in a high degree of sterility. ACKNOWLEDGMENTS I appreciate very much the helpful comments and suggestions of Dr G de la Vega. I would also like to thank Dr Navarrete for allowing me the use of the Vickers M-85 Integrating Microdensitometer, and J Garcia Herranz and J Dorda Dorda for their technical assistance. The specimens have been collected with official permission obtained from regional authorities. This work was supported by DGICYT PB 880010 Project (Spain) and by a Cooperative Project (Spain-Great Britain)