Heredity 64 (1990) 233—238 The Genetical Society of Great Britain Received 25August1989

Clonal structure in the , americanum Brid.

Thomas R. Meagher* and * Departmentof Biological Sciences, Rutgers University, P.O. Box 1059, Jonathan Shawt Piscataway, NJ 08855-1059. 1 Department of Biology, Ithaca College, Ithaca, NY 14850.

The present study focussed on the moss species, Climacium americanum Brid., which typically grows in the form of mats or clumps of compactly spaced gametophores. In order to determine the clonal structure of these clumps, the distribution of genetic variation was analysed by an electrophoretic survey. Ten individual gametophores from each of ten clumps were sampled in two locations in the Piedmont of North Carolina. These samples were assayed for allelic variation at six electrophoretically distinguishable loci. For each clump, an explicit probability of a monomorphic versus a polymorphic sample was estimated based on allele frequencies in our overall sample. Although allelic variation was detected within clumps at both localities, our statistical results showed that the level of polymorphism within clumps was less than would be expected if genetic variation were distributed at random within the overall population.

INTRODUCTION approximately 50 per cent of moss species that are hermaphroditic, self-fertilization of bisexual Ithas been recognized that in many taxa asexual gametophytes results in completely homozygous propagation as the predominant means of repro- sporophytes. Dispersal of such genetically iden- duction does not necessarily lead to genetic tical is functionally equivalent to asexual monomorphism (Bell, 1982; Silander, 1979; reproduction. Vrijenhoek, 1984). Indeed, localized patterns of are believed to be relatively depauper- allozymic variation have been useful tools for iden- ate in genetic variation (Crum, 1973), a circum- tifying the spatial extent of individual clones within stance which may reflect the predominance of populations (Silander, 1979) and for recog- asexual propagation as the most successful life nizing different clones in unisexual fish popula- history mode. Many bryophyte systematists believe tions (Vrijenhoek et a!., 1978). In order to under- that mosses evolve relatively slowly (Anderson, stand how genetic variation is maintained in 1963; Khanna, 1964), a view based largely on asexual populations, it is first necessary to deter- biogeographic observations. The geographic distri- mine the distribution of such genetic variation and butions of most species are relatively broad, and relate that distribution to the overall breeding intercontinentalrangedisjunctions without morphological differentiation are common. While system. Most species of mosses are believed to repro- some bryologists have assumed that almost all duce both sexually by formation, and moss reproduction is by means other than sexual asexually by a variety of methods (Longton, 1976; spores (e.g., Anderson, 1963), extremely little Wyatt, 1982; Wyatt and Anderson, 1984). Asexual information is available about how mosses actually reproduction in this group can occur by vegetative do reproduce under natural conditions. fragmentation, in which pieces of gametophytic Bryophytes in general, and mosses in par- are broken off and reestablish elsewhere, ticular, have been increasingly recognized as useful by the formation and dispersal of specialized organisms for studying a variety of concepts in reproductive structures such as axillary gemmae population biology (During and van Tooren, 1987; or rhizoidal tubers (Correns, 1899), or by the Wyatt, 1982). The reasons for this recent increase growth of localized clones. Moreover, within the in awareness of mosses as research tools are 234 T. R. MEAGHER AND J. SHAW several-fold. These plants exhibit unusual features creek crosses Old Erwin Road. The ecological for terrestrial plants, such as a conspicuous alterna- features of these sites were similar to those tion of generations with the haploid form perennial described for nearby populations of C. americanum and freeliving. Since the predominant growth form characterized in Shaw et al. (1987). In each site, is haploid, population genetic studies are greatly ten individual gametophores were sampled from simplified. Also, this group has been relatively each of ten distinct clumps. Clumps ranged in size overlooked so that virtually any population bio- from 13 cm to 100 cm in diameter, and logy study represents a novelty. gametophores were sampled evenly over a 13 cm Ironically, the fact that mosses have been rela- diameter portion of each clump. These tively understudied is a weakness as well as a gametophores were placed in plastic vials and strength. Even basic information about population stored at —70°C for several weeks until they were structure and dynamics are poorly understood for subjected to electrophoretic analysis. many species, hindering detailed investigations. Enzyme extraction was performed by grinding Many moss species occur as compact clumps of each sample in liquid nitrogen and then suspend- gametophores. The genotypic composition of such ing them in an extraction buffer of 0.1 M tris-HCI clumps, however, is generally not known. Each (pH 7.2) buffer with 0.5 per cent mercaptoethanol. clump could represent a collection of clonal ramets Extracts were absorbed onto filter paper wicks of the same genotype (as is commonly assumed in which were loaded directly into 12 per cent many bryological studies) or, alternatively, they horizontal starch gels. Following electrophoresis, could represent an aggregation of co-occurring enzyme assays were conducted for six systems that distinct genotypes. The differences between these had been found in our previous investigation to possibilitieshave significantand obvious be polymorphic: alcohol dehydrogenase (ADH), ramifications for microevolutionary processes in acid phosphatase (ACP), fluorescent esterase moss populations. (EST5), peroxidase (PER), phosphoglucomutase The present study was focussed on the moss (PGM), and superoxide dismutase (SOD). The species, Climacium americanum Brid., commonly electrophoretic and staining protocols for C. amer- known as the "tree moss" because of its dendroid icanum are outlined in detail in Shaw eta!. (1987). growth form. Sexual reproduction, as evidenced For each clump, an explicit probability of by sporophyte formation, is extremely rare in C. observing a monomorphic versus a polymorphic americanum (Shaw et a!., 1987). The species sample was estimated based on allele frequencies occupies a broad geographic range in the eastern in our overall sample. For a haploid organism, United States, from Nova Scotia and Wisconsin these probabilities for a given sample size can be south to Florida and Kansas (Crum and Anderson, calculated by use of a binomial expansion series 1981). The typical growth habit for this species is (table 1; Feller, 1968). Thus for a sample of n the formation of discreet mats or clumps along gametophores, and a diallelic locus with allele stream margins and on flood plains. These clumps frequencies Pi and P2, the probability of are generally small (Ca. <60 cm2), but are occasion- monomorphism would be ally more extensive. In order to determine the genetic structure of these clumps in C. americanum, P, (monomorphism) =(p'+p°) the distribution of genetic variation was analysed for the ith locus. A combined probability across by an electrophoretic survey of two natural popula- multiple loci can be obtained by multiplying tions.

Table 1 Probability distribution of a sample of 10 haploid MATERIALSAND METHODS genotypes within clumps based on population allele frequencies Pt and P2 Clumpsof C. americanum were sampled in two Ratio of Ratio of locations in the Piedmont of North Carolina. The g1 to g2 Probability g1 to g2 Probability first sample site was on the flood plain of the Eno River 50 m downstream from its intersection with 10:0 p° 4:6 210pp Guess Road in the city of Durham (Durham 9:1 10p'p2 3:7 120pp County, NC). The second sample site was located 8:2 45pp 2:8 45pp in the Korstian Division of the Duke Forest 7:3 120p,p 1:9 10pp 6:4 210pp 0:10 (Orange County, NC) along New Hope Creek 5:5 approximately 200 m upstream from where the 252pp CLONAL STRUCTURE IN CLIMACIUM AMER/CANUM 235 together all of the P,assumingthe loci are variation was predominantly due to variation unlinked. A combined probability that a clump between clumps rather than to variation within will be polymorphic for at least one locus can be clumps (tables 3—4). obtained by subtracting the combined probability In our previous results (Shaw et a!., 1987) of monomorphism from 1. Within each site, demonstrated that two varieties of C.americanum explicit probabilities of monomorphism versus which differ in morphology can also be distin- polymorphism were calculated for each clump guished by their enzyme phenotypes coded by sampled using the cumulative allele frequencies several loci. In the Eno River sample, nine of observed for that site. Clumps varied in these prob- the clumps showed a PGM enzyme pattern abilities due to small differences in sample size. characteristic of Climacium americanum var. americanum and the remaining clump was clearly var. kindbergii. Within clump allelic variation RESULTS was restricted to the PER locus, although allelic variation among clumps was observed for ACP, Fourof the six enzymes assayed showed poly- PGM, and SOD as well (table 3). morphism in the Eno River collection, whereas all At New Hope Creek, the PGM banding pattern six were polymorphic in the New Hope Creek (plus morphological characteristics) showed that collection (table 2). In both populations, allelic nine clumps were var. kindbergii and the remaining clump was var. americanum. Within clump vari- Table2Cumulative allele frequencies in the current sampled ation occurred for ADH, PER and SOD, and clumps of Climacium americanum. Allele numbering overall allelic variation was also observed for ACP follows Shaw et a!., 1987. Sample sizes (N) vary due to and ESTfl, in addition to the morph difference in uneven scorability of gels the PGM (table 4). Enoriver NewHope Creek Two series of expected frequencies of monomorphic clumps were determined for each Enzyme N Alleles N Alleles population (table 5). In the first series, allelic vari- ation at all loci, including the morph differences 1 2 1 2 at the PGM locus, were used to estimate expected ADH 37 100 — 99 027 073 ACP 94 0•89 011 100 010 090 monomorphism. This analysis showed a significant ESTfl 50 — 100 98 059 041 excess of monomorphic clumps in both popula- PER 100 085 015 99 026 074 tions. In the second series, only the most frequent PGM 100 090 010 100 010 090 morph type in each sample was used, and allelic 97 020 080 SOD 100 090 010 variation at the PGM locus was not included. When

Table 3 Allelic variation within and among clumps of Climacium americanum collected along the Eco River, Durham County, NC. Genetic status of clump is indicated as monomorphic (M) or polymorphic (P)

Clump

Enzyke Allele 1 2 3 4 5 6 7 8 9 10

ADH 1 2 — — 1 6 10 10 — — 8 2 0 — — 0 0 0 0 —— 0

ACP 1 10 10 10 10 0 10 4 10 10 10 2 0 0 0 0 10 0 0 0 0 0 — — — EST 1 — 0 0 0 0 — 0 2 — 10 10 10 10 — — 10 — — PER 1 8 10 10 10 10 10 0 9 10 8 2 2 0 0 0 0 0 10 1 0 2

PGM 1 10 10 10 10 10 10 0 10 10 10 2 0 0 0 0 0 0 10 0 0 0

SOD 1 10 10 10 10 10 10 0 10 10 10 2 0 0 0 0 0 0 10 0 0 0

Genetic status P M M M M M M P M P 236 T. R. MEAGHER AND J. SHAW

Table 4Allelic variation within and among clumps of Climacium americanum collected along New Hope Creek, Orange County, NC. Genetic status of clump is indicated as monomor- phic (M) or polymorphic (P)

Clump

Enzyme Allele 1 2 3 4 5 6 7 8 9 10

ADH 1 0 0 10 7 0 0 0 10 0 0 2 10 10 0 2 10 10 10 0 10 10 ACP 1 0 0 0 0 0 0 0 0 0 10 2 10 10 10 10 10 10 10 10 10 0 EST5 1 10 10 0 10 10 0 10 0 8 0 2 0 0 10 0 0 10 0 10 0 10

PER 1 0 0 10 3 2 1 0 0 0 10 2 10 10 0 7 8 9 10 10 9 0

PGM 1 0 0 0 0 0 0 0 10 0 0 2 10 10 10 10 10 10 10 0 10 10

SOD 1 0 0 0 7 1 1 0 0 0 10 2 10 10 10 3 9 6 10 10 10 0

Genetic status M M M P P P M M M M

Table SCumulative observed and expected frequencies of monomorphic and polymorphic clumps of Climacuim americanum. Expected counts are based on overall allele frequencies in each population for the total sample (a) or for a subset that is homogeneous with respect to varietal status (b). Log-likelihood ratio (G2, Bishop et a!., 1975) test results comparing observed and expected ratios are also presented

Eno River New Hope Creek

Genetic status Observed Expected Observed Expected

Monomorphic 7 1 7 0 Polymorphic 3 9 3 10 G2 8.2*** 13.7*** b

Eno River New Hope Creek

Genetic status Observed Expected Observed Expected Monomorphic 6 2 6 0 Polymorphic 3 7 3 9 G2 37ns 115*** *** P<0-005. differences due to different morph types were thus them to be so. Most of the electrophoretically eliminated, only the New Hope Creek sample detectable variation in the two populations we showed a significant excess of monomorphism. studied occurred among the discreet clumps, yet some variation did exist within clumps. Elec- trophoretic surveys of moss populations are still DISCUSSION limited in number, and care is often taken to avoid sampling within clumps (e.g., Daniels, 1982; and Our results indicate that clumps of Climacium Shaw et al., 1987). In the only other study of which americanum are not necessarily genetically we are aware that has explicitly examined the homogeneous, although there is a tendency for genetic structure of moss clumps, Cummins and CLONAL STRUCTURE IN CLIMACIUM AMER/CANUM 237

Wyatt (1981) obtained similar results demonstrat- clumped growth habits. Using cytological markers, ing intra-clump variability in Atrichum angustatum. Anderson and Lemmon (1974) found that an Other studies of electrophoretic variation in extensive population of the moss, Weissia con- moss populations have demonstrated surprisingly troversa, consisted of numerous genetically dis- high levels of variation comparable in several cases crete clumps, between which gene exchange to levels typical of highly outcrossed angiosperms appeared to be limited or absent under natural (Wyatt, Stoneburner and Odrzykoski, 1989). conditions. Daniels (1982, 1985a, b) has found variation in As a means of dispersal and propagation, vege- enzyme phenotype within populations of largely tative fragmentation is obviously effective for C. asexual species of Sphagnum in Great Britain and americanum, as shown by the abundance of non- Scandinavia, but it is not clear how much of the sporophyte producing populations along streams variation occurred within phsical1y discrete throughout the Piedmont of North Carolina. The clumps. Krzakowa and Szweykowski (1979) have haploid nature of the gametophytes sampled made also demonstrated enzyme polymorphism in the the determination of genetic uniformity of clumps largely asexual liverwort, Plagiochila asplenioides. quite robust. Even with a relatively small sample Zielinski (1986) found genetic evidence of cross size within clumps and only six useful marker loci, fertilization between individuals that were variable it was possible to detect statistically significant at a peroxidase locus within clumps of the liver- departures from expected levels of polymorphism wort, Pellia epiphylla. Lower but significant levels within clumps. Thus the statistical approach pres- of cross fertilization occurred in a related species, ented here should prove to be generally useful in Pellia borealis, again demonstrating that clumps of studies of genetic variation in moss populations. this liverwort are at least sometimes genetically heterogeneous. Climacium is a small genus of three or four REFERENCES dioicous species. Populations of C. dendroides, a circumboreal relative of C. americanum, tend to ANDERSON,L. EL. 1963. Modern species concepts: mosses. show a strong female Mas. In fact, most clumps Bryologist, 66, 107-119. of C. dendroides appear to be unisexual and the ANDERSON,L. EL. AND LEMMON, B. E. 1974. Gene flowdis- frequent absence of male plants directly limits tancesin the moss Weissia controversa Hedw. J. Hattori Bot. Lab., 38, 67-90. sexual reproduction (Bedford, 1938). Transplanta- BEDFORD.T. H. B. 1938.Sex distribution in colonies of tion of male plants into non-sporophyte producing Climacium dendroides W. & M. and its relation to fruit female clumps resulted in successful sporophyte bearing. NWest Naturalist, 13, 213-221. formation. BELL,G.1982. The Masterpiece of Nature: The Evolution and Comparable data on sex ratios in relation to Genetics of Sexuality. University of California Press, Berkeley. sporophyte formation are not available for C. BISHOP,Y.M. M., FEINBERG. S. EL. AND HOLLAND, pw.1975. americanum. Our field observations, however, indi- Discrete multivariate analysis: theory and practice. MIT cate that sexual reproduction is exceedingly rare Press, Cambridge, Mass. CORRENS, C. 1899. Untersuchungen uber die Vermehrung der in this species. We have observed sporophytes in Laubmoose durch Brutorgane und Stecklinge. G. Fischer, only three populations in the Piedmont of North Jena, xxiv+472 pp. Carolina, although C. americanum is extremely CRUM, H. A. 1973. Mosses of the Great Lakes Forest. Contrib. abundant throughout the region. In our previous University Michigan Herbarium 10, 1—404. electrophoretic study (Shaw et a!., 1987), one CRUM, H. A. AND ANDERSON, L. E. 1981. Mosses of Eastern North America. Columbia University Press, New York. population along a stream feeding into the Eno CUMMINS, H., AND WYATT, R. 1981. Genetic variability in River was selected for study because there were natural populations of the moss Atrichum angustatum. sporophytes present. This population, however, Bryologist, 84, 30-38. did not appear to be more variable than any of the DANIELS, R. E. 1982. Isozyme variation in British populations other populations included in our studies. of Sphagnumpulchrum(Braithw.) warnst. J. Bryol. 12, 1—11. Limited gamete dispersal distances characteris- DANIELS, R. EL. 1985a. Isozyme variation in populations of tic of mosses (Anderson and Lemmon, 1974; Sphagnum recurvum var. mucronatum from Britain and Wyatt, 1977) make cross fertilization between Finland. J. Bryol., 13, 563-570. clumps of species such as C. americanum extremely DANIELS,R. EL.1985b. Isozyme variation in Finnish and British populations of Sphagnum compactum. Ann. Bot. Penn., 22, unlikely. Consequently, low levels of genetic varia- 275-279. bility within clumps clearly limit the possibility of DURING. H. J. AND VAN TOOREN. B. F. 1987. Recent develop- genetic recombination during sexual reproduction ments in bryophyte population ecology. Trends in Ecology in C. americanum and other species with similar and Evolution, 2, 89-93. 238 T. R. MEAGHER AND J. SHAW

FELLER, w. 1968. An Introduction to Probability Theory and Its VEIJENHOEK, R. C., ANGUS, R. A. AND SCHULTZ, R. .i. 1978. Applications. Wiley, New York, Variation and clonal structure in a unisexual fish. American KHANNA, K. R. 1964. Differential evolutionary activity in Naturalist, 112, 41—55. bryophytes. Evolution 18, 642-670. WYATT, IL 1977. Spatial pattern and gamete dispersal distances KRZAKOWA, M. AND SZWEYKOWSKI,J. 1979. Isozyme poly- in Atrichum angustalum, a dioicous moss. Bryologist, 80, morphism in natural populations of a liverwort, Plagiochila 284-29 1. asplenioides. Genetics 93, 711—719. WYATT, R. 1982. Population ecology of bryophytes. J. Hattori LONGTON, R. E. 1976. Reproductive biology and evolutionary Bot. Lab., 52, 179-198. potential in bryophytes. I Hattori Bot. Lab., 41, 205-223. WYATT, R. AND ANDERSON, L. E. 1984. Breeding systems in SHAW, J., MEAGHER, T. R. AND HARLEY, P. 1987. Elec- bryophytes. In Dyer, A. F. and Duckett, J. G. (eds) The trophoretic evidence of reproductive isolation between two Experimental Biology of Bryophytes, Academic Press, New varieties of the moss, Climacium americanum. Heredity, 59, York, pp. 39-64. 337—343. WYATT, R., STONEBURNER, A. AND ODRZYKOSKI, 1. 3. 1989. SILANDER,J. A. 1979. Microevolution and clone structure in Bryophyte isozymes: systematic and evolutionary implica- Spartina patens. Science, 203, 658—660. tions. In Soltis, D. E. and Soltis, P. S. (eds) Plant Isozymes, VRIJENH0EK, R. C. 1984. Ecological differentiation among Dioscorides Press, Portland, Oregon. clones: the frozen niche variation model. In Wohrmann, ZIELINSKI R. 1986. Cross-fertilization in the monoecious Pellia K. and Loeschcke, V. (eds) Population Biology and Evol- borealis, n =18,and spatial distribution of two peroxidase ution, Springer-Verlag, Berlin, pp. 217—23 1. genotypes. Heredity, 56, 299-304.