Biochemical genetics of chromosome forms of Venezuelan spiny rats of the proechimys guairae and proechimys trinitatis superspecies

Spiny rats from Venezuela show an extensive karyotypic diversification (2n=24 to 2n=62) and little morphological differentiation. This study reports genetic distance, heterozygosity and polymorphism based upon 22 loci in semispecies and allospecies of the Proechimys guairae superspecies from N Central Venezuela, as compared with Proechimys urichi, a member of the Proechimys trinitatis superspecies from eastern Venezuela. Four chromosome forms of the P. guairae complex are included, each characterized by karyotypes of 2n=46 (Fundamental Number=72), 2n=48 (FN=72), 2n=50 (FN=72) and 2n=62 (FN=74). Proechimys urichi has a distinetive karyotype of 2n=62 (FN=88). The overall mean value of Nei's genetic identity index for all pair-wise comparisons is I=0.942±0.011. Mean identity within the P. guairae complex is Ī=0.969±0.033. Mean identity between P. urichi and members of that complex is Ī=0.889±0.011. Within the P. guairae complex, increased genetic divergence is correlated with higher karyotypic divergence. Heterozygosity varies from H=0.059 to H=0.153, with a mean value of H0.059. The mean percent of polymorphic loci is P=18.2±3.9 after the ‘0,95%’ polymorphism criterion, and P=20,5±5.2 after the ‘0.99%’ criterion. These results are compared with similar data from fossorial and non-fossorial rodents. Spiny rats are non-fossorial, forest-dwelling rodents which have undergone a speciation process with little genetic divergence and extensive chromosome rearrangements.


Introduction
Spiny rats of the genus Proeehimys are common rodents of the lowland and premontane tropical forests of Venezuela.With the exception ofP.guyannensis (= P. cherriei, Reig, Trartier & Barros, 1979), and P. canicollis (Aguilera et al. 1979) most of the species of Venezuelan spiny rats are so similar morphologically that many of them have recently been incorrectly assigned to a single species, P. semispinoms (Handley, 1976).
Previous studies (Reig, Kiblisky & L6big, 1970;Reig & Useche, 1976) have shown that there is a wide range of chromosomal diversification in Proechimys spiny rats, chromosome numbers varying from 2n = 24 to 2n = 62.More recent studies (Reig et al. 1978(Reig et al. , 1979a(Reig et al. , 1979b) ) suggests that three species groups can be distinguished among Venezuelan spiny rats of the genus Proechimys occurring N of the Orinoco River.The first group herein called the P.
guairae superspecies, lives in N central and N W Venezuela, inhabiting both sides of the Andes and the central section of the Cordillera de la Costa (Fig. 1).This species complex comprises at least three sibling species, P. poliopus (2n = 42), P. guairae (2n = 44-50), and P.n.sp.(2n = 62, FN = 74).The second group, is represented by P. trinitatis and P. urichi, and these species occur in eastern Venezuela and on Trinidad.Morphologically, this group has been supposed to be so similar to the former one that its members have been regarded as conspecifics.However, there are very distinctive karyotypic differences between this group (2n = 62) and those of the P. guairae superspecies (2n = 42-62), which include different chromosomal morphology and fundamental numbers (Reig et al., in prep.).The third group is represented by P. canicollis (2n = 24), which may belong to a Peruvian-Colombian group of species.
The purpose of this study is to assess the biochemical genetic differentiation that exists among the karyomorphs of the P. guairae superspecies, and to compare these data with a representative of the chromosomally different, but morphologically similar P. trinitatis superspecies, i.e., P. urichi.

Material and methods
Animals were trapped in 1976 and 1977 and transported alive to the laboratory.Blood plasma, red cells and internal organs (kidney, heart, liver, striated muscle) were preserved at -80°C.Most of the rodents were killed several days after their removal from the field.Others were kept alive from two to four weeks before processing.In August, 1977, all the samples were Mr-mailed to Davis, California, where f'mat processing and the electrophoretic runs were performed.
Horizontal starch gel electrophoresis was used to separate the soluble proteins.Biochemical procedures and recipes are essentially those of Ayala et al. (1972) and Selander et al. (1971) and will be reported in full elsewhere.Gel patterns were scored by a 'blind procedure', where karyotypic morph of the sample was unknown to the scorer.
Genetic nomenclature is as follows: loci are numbered sequentially according to the distance migrated in the gel, starting with the least anodal one; alleles 90 are given letters in alphabetical order, starting with the one that migrated the least.Proteins that migrated cathodally are given an asterisk.

Results
The populations sampled cover a rather extensive area in north-central, north-eastern and north-western Venezuela (Fig. 1).The diploid complement and fundamental number of each are listed in Table 1.La Trilla (T), Aragua, and San Esteban (E), Carabobo, are considered two populations of the same species with 2n = 46 of the P. guairae complex.San Antonio (A), Miranda, with 2n = 48, San Carlos (C), Cojedes, with 2n = 50, and Barinitas (B), Barinas, with 2n = 62, are all distinct karyomorphs of the P. guairae superspecies.Gu~icharo (G), Monagas, although having the same chromosome number as Barinitas (2n = 62), is regarded as a different species (P.urichi) not included in P. guairae, attending to its geographical distribution and fundamental number.P. guairae together with P. g. ochraceus (O) (2n = 44) and P. poliopus (P) (2n = 42) form a 'Rassenkreis' in NW Venezuela (Reig et al., 1978(Reig et al., , 1979) (Fig. I).Biochemical data are not available at the present for P. g. ochraceus and P. poliopus, but they are included here for the sake of completeness.
Eleven of the 22 loci scored showed no variation; these were Adh, Aldo,Got,Idh,Ldh.1,Nadharia,Pgi,Pgm,.The allelic frequencies of the eleven variable loci are given in Table 2.The following remarks are pertinent: (1) No population is fixed for all loci studied.
(2) The most common allele for a given locus in a (3) Among variable loci, Lap is the only one whictt shows variation in all populations.Intrapopulational genetic variation is measured by the average frequency of heterozygotes per locus (H) and by the proportion of polymorphic loci in the population (P).The first statistic (H) is the expected frequency of heterozygotes that would exist under a Hardy-Weinberg equilibrium, and is an average of all the loci sampled (Nei, 1975).The proportion of poly.morphic loci (P) is calculated using two criteria: (a) a locus is considered polymorphic if the frequency of the most common allele is not greater that 0.95; and (b) if the frequency is not greater than 0.99.Average heterozygosities and the proportion of polymorphic loci are given in Table 3.The overall mean heterozygosity for all the populations sampled is 0.0885, which is within the range reported for other rodents (Ayala, 1975).The San Carlos population has the highest H (0.1533) and La Trilla the lowest (0.0589).The overall mean proportion of polymorphic loci for all population is 18.2% and 20.5% depending respectively on whether the .95or .99criterion is used.Again San Carlos has the greatest proportion of polymorphic loci (36.4% and 45.5% respectively) while La TriUa has the least (9,1% and 9.1%).
To measure genetic divergence between populations, we used Nei's (1975) indices of 'genetic identity' (D and 'genetic distance' (D).Both statistics were calculated for all possible pairwise comparisons between populations and are given in    complex is 0.969 _+ 0.003 for all 10 pairwise comparisons; mean identity between P. urichi and the P. guab rae complex is 0.889 _+ 0.011 for all five comparisons.Thus, it is apparent that the karyomorphs of the P.
guairae complex are more closely related among themselves than this complex to P. urichi.Within the P. guairae complex, increased genetic divergence and higher chromosome number are roughly correlated.Thus, for the two 2n = 46 populations, I = 0.986; between the 2n = 46 and 2n = 48 karyomorphs I = 0971; and I = 0968 between 2n = 48 and 2n = 50, and so on.We used the unweighted pair group method to generate dendograms based on genetic identities (Fig. 2) and genetic distances (Fig. 3).It is evident that genetic divergence and chromosome number are correlated within P. guairae and that the complex of P. guairae karyomorphs are a group distinct from P. urichi.guairae (Reig et al., 1978(Reig et al., , 1979)).Isolation between the 2n = 46, 2n = 48, and 2n = 50 karyomorphs is less certain; however the following facts suggest that they are genetically isolated semispecies: (i) each form has a definite parapatric area of distribution (Fig. 1); (ii) chromosomal polymorphism within and between morphs is almost inexistent; (iii) there are no clines within any chromosomal type.In other words, there is an abrupt transition from one form to the other.This indicates that although one or two Robertsonian changes may not prevent gone flow between karyomorphs (White, 1973), in our case they are doing so.This pattern of chromosome differentiation in P. guairae is similar to what Patton & Gardner (1972) reported in populations of Proechimys in Per6.Assuming that P. guairae populations are at least semispecies, we ask, then, how much genetic differentiation there exists among them?As a paradigm of speciation, Drosophila willistoni is a nice example to start with.Genetic identity is 0.798 among semispecies ofD.willistoni (Ayala, 1975).It is clear that P. guairae karyomorphs are genetically very similar under this paradigm of speciation: in fact, ] = 0.970 for local populations of D. wiUistoni and | = 0.969 for semispecies of P. guairae.Moreover, genetic identity between the sibling species P. guairae and P. urichi is 0.880.For sibling species of D. willistoni, genetic identity is 0.563, 37% smaller.Spiny rats do not follow a pattern of speciation through increasing genetic divergence as D. willistoni; comparison with other rodents is in order.Available data on genetic differentiation and speciation in rodents can be treated separately for fosserial and non-fossorial ones.Fossorial rodents seem to follow two broad patterns of speciation (Patton & Yang, 1977): (i) species complexes with extensive chromosome differences associated with reproductive isolation and low genetic variability, and (ii) species with extensive allozymic and chromosomal differentiation and no reproductive barriers associated with karyotypic variation.Pattern (i) is well exemplified by the mole rat, Spalax ehrenbergi (Nero & Shaw, 1972), and by the pocket gophers of the Thomomys talpoides complex (Nevo et al, 1974) and Geomys bursarius group (Selander etal, 1974).Average heterozygosities for these rodents are 0.037, 0.047, and 0.040, respectively.The overall mean heterozygosity for rodents is 0.057 (Dobzhansky et al, 1977;Selander, 1976).Hence, Venezuelan Proechimys are genetically more variable than these los-serial forms, and are more heterozygous than the 'average' rodent as well.The average identity among karyomorphs of S. ehrenbergi is 0.978 and for T. taipoides | = 0.925, so mole rats and pocket gophers show little genetic divergence.The species of the Geomys bursarius group are equally very closely related, 72% to 83% of predominant alleles are shared by all species pairs of the group.Thus, chromosomal forms of Proechimys have similar levels of genetic identity as occurs in Spalax, Thomomys and Geomys.

Discussion
Extensive genetic and chromosome variation without any concomitant development of reproductive isolation has been reported for Thomomys bottae (Patton & Yang, 1977), which fits pattern ii) above.
For T. bottae, the mean heterozygosity is 0.093, approximately the same as Proechimys.Populations of T. bottae are not reproductively isolated and have a mean genetic identity of 0.87, much lower than populations of Venezuelan spiny rats.
The cotton rats Sigmodon hispidus and S. arizonae are non-fossorial sibling species that inhabit an exten.sire area in the United States and Mexico (Johnson et al., 1972).The mean allozymic heterozygosity is rather low (2%-3%), lower than in Venezuelan spiny rats.Conspecific populations show little genetic divergence, Roger's S coefficient is 0.983.Thus conspecific populations of cotton rats show similar levels of genetic identity as karyomorphs of the P. guairae complex (cf.Chakraborty & Tateno, 1976).
Kangaroo rats (Dipodomys) are non-fossorial rodents of the western United States and northern Mexico.They have less heterozygosity (H = 0.02) than Venezuelan spiny rats and they show moderate interspecific divergence (S = 0.61).Roger's coefficient for conspecific population always exceeds 0.90 (Johnson & Selander, 1971).
The data reviewed in the previous paragraphs are summarized on a qualitative basis in Table 5.All rodents in this table, with the exception of T. bottae have in common a high interpopulational genetic similarity and isolation as a consequence of chromosome divergence, irrespective of whether they are fosserial or not.On the other hand, both the fossorial and temperate gopher (T.bottae) and the tropical non fossorial spiny rat (Proechimys) have high genetic heterozygosity.Therefore, it seems that any hypothesis to explain patterns of speciation in rodents can be either an 'ad hoe' hypothesis for a particular species, or a very general one that has little predictive power.

This paper
To illustrate this, we offer some estimates of divergence times between karyomorphs of Venezuelan spiny rats (see also Nero et al., 1974 andPatton &Yang, 1977).Nei's equation (Nei, 1975) estimates approximately 50,000 years for the divergence of karyomorphs within the P. guairae complex and 200,000 years for the separation ofP.guairae from P. urichi.These ages coincide with Pleistocene glacial periods and the formation of ecological islands of 'refugia' throughout temperate and tropical America.Indeed, such changes could be very important to speelation events.In fact, they have been set forth to explain speciation in organisms as different as butterflies (Turner, 1976), birds (Haffer, 1969), and rodents (Patton & Yang, 1977).Similar explanations could be made for Proechimys, but we think it would contribute little toward explaining why speciation occurs both in fossorial and non fossorial rodents through chromosomal changes with little protein divergence.Moreover, we can offer no explanation for the correlation between chromosome number and genetic distance found in Venezuelan spiny rats.
Recently, Wilson and his collaborators (Wilson, Sarich & Maxson, 1974) have suggested that morphology, chromosome systems, and proteins evolve at different rates in mammals.They conclude that rapid chromosome rearrangements in mammals parallels their rapid morphological evolution but that proteins evolve at a much slower rate.Our data for Venezuelan spiny rats provide a counterexample to this hypothesis, since karyomorphs of Proechimys are morphologically very little differentiated.

Fig. 1 .
Fig. 1.Map of northern Venezuela, showing the distribution of the Proechimys guairae and Proechimys trinitatis superspecies, and the location of the sampled localities.$uperspeeies are indicated by square brackets.

[Fig. 3 .
Fig. 3. Dendrogram based on Nei's D x 10 -2 .Clustering is by unweighted pair group method.Chromosome numbers are given in parenthesis for each population.
Within the chromosomal Rassenkreis of P. guairae, there is a transition from 2n = 42 to 2n = 44, which involves one Robertsonian change and three pericentric inversions; a single Robertsonian change accounts for the difference between the 2n = 44 to 2n = 46 chromosomal races.Similarly, one change of the fission type accounts in each case for the difference between the 2n = 46 and 2n = 48 between the 2n = 48 and 2n = 50 karyotypes.The karyotype of the 2n --62 population can be derived from the 2n = 50 one by six Robertsonian fissions plus two pericentric inversions.Differences between P. guairae and P. urichi (2n = 62, FN = 88) have not been worked otrt.It is reasonable to assume that reproductive isolation occurs between 2n = 50 and 2n = 62 populations of P.

Table 1 Diploid
particular population is the most common one in all of them, except Hb and Lap.To and Xdh have a different most common allele in P. urichi.

Table 2
Allelic frequencies for variable loci in six populations of Proechimys.Sample sizes are given in parentheses

Table 2 (
contd) Allelic frequencies for variable loci in six populations of Proechimys.Sample sizes are given in parentheses

Table 5
Quantitative assessment of speciation patterns in rodents