Populations of Primula Veris
Total Page：16
File Type：pdf, Size：1020Kb
Heredity 71 (1993) 252—258 Received 8 December 1992 Genetical Society of Great Britain Genetics of colonizing and established populations of Primula veris SUSAN ANTROBUS* & ANDREW J. LACK School of Biological and Molecular Sciences, Oxford Brookes University, Headington, Oxford OX3 OBP, U.K. Primulaveris, a long-lived iteroparous herb which has declined in abundance in southern England since the 1940s, has recently been observed colonizing disturbed sites. The genetic structure of P. veris populations in the Oxford region was investigated to determine whether colonizing popula- tions differ from those longer established. Nine enzyme systems, revealing 19 presumptive loci, were screened in 11 established and seven colonizing populations. Levels of variation were low in all populations, and all populations except one approximated panmixia. The proportion of the total variation among populations was small (FST =0.039) as were genetic distances between popula- tions. No significant differences in genetic structure were recorded between established and colon- izing populations. The lack of differentiation within and between populations is surprising for a long-lived, insect pollinated species with gravity dispersed seeds, for which gene flow would be expected to be spatially restricted. Keywords:colonization,dispersal, gene flow, genetic structure, isozyme, Primula veris. Introduction Yeo, 1973; Wedderburn & Richards, 1990). Pollen dispersal is normally up to 12 m and seed dispersal, by Theability of plant species to colonize new areas and gravity, is normally limited to 0.5 m (Richards & the ecological and genetic consequences of this coloni- Ibrahim, 1978). In mature habitats, Tamm (1972) zation have received much study. Successful colonizing observed that recruitment of seedlings into the adult plants are likely to have a good seed dispersal mechan- population was a rare event and that regeneration was ism and rapid reproduction or vegetative growth. Self- achieved mainly by infrequent and local branching of fertility may also be an advantage. Some of the the rhizome. variation in a species may be lost in colonizing popula- The existence of colonizing populations of this tions, through founder effects, and such effects have species in many places was, therefore, surprising. The been detected at all scales of colonization from long aim of this study was to see what attributes it possessed distance to local, and among plants which are mainly which allowed it to colonize well, and what the genetic inbreeding as well as those that are mainly outbreeding consequences of this colonization might be. We are (Brown & Marshall, 1981; Barrett & Husband, 1990). concerned with colonization at a local scale, within the Generally, inbreeding plants show more differentiation species' established range. between populations and less within them than out- breeders (Loveless & Hamrick, 1984; Hamrick & Materials Godt, 1990) and many show founder effects strongly, and methods even at a local scale (Lack & Kay, 1988). Untilthe 1 940s P. veris was an abundant meadow and One species that we observed colonizing new sites in pasture plant in lowland England, with extensive Britain was the cowslip, Primula veris. This species populations, which were likely to have been long lived does not appear to have the attributes of a good colon- due to constant cutting or grazing. Since then P. veris izer. It is a long-lived iteroparous perennial (Tamm, has greatly decreased in England through the destruc- 1972) which is insect pollinated, mainly by bees, tion of many grasslands for arable use or, in those that heterostylous and an obligate outcrosser (Proctor & remain, relaxation of grazing pressure leading to a less suitable habitat. Grime et al. (1988) regarded it as a *Present address: British Antarctic Survey, High Cross, Madingley stress tolerator (with some competitive ability) Road, Cambridge, CBS OET, U.K. although they too noted that populations, some of them 252 GENETICS OF PRIM ULA VERIS 253 extensive, have become established on verges of ancient grassland sites such as West Mead or sites that motorways and in disturbed habitats. were considered ancient owing to the presence of ancient grassland plant indicators, e.g. Stonesfield Common. It is likely that these populations are well Populationsampling procedure established. Colonizing populations included those Populationswere sampled from 18 sites (Table 1, Fig. grOwing on roadsides, in a disused quarry and a dis- 1; further details, including grid references and habitat used sewage works. The age of some of these sites is are given by Antrobus, 1992). They ranged in size known, giving an upper age limit to the populations. from less than a hundred individuals to many thou- Sixty or 90 individuals were sampled from each sands. All sites were defined by a natural boundary and population. To prevent sampling the same clone more were less than 1 km in diameter; most were less than than once, individuals were normally sampled 1 m 100 m diameter: Three sites, Lower Seeds, The Dell apart. Where this was not possible, in small popula- and Upper Seeds were less than 100 m apart but were tions, genetic individuals were usually clear-cut and separated by areas with no P. veris plants. each sample came from a different genet. Populations were classified as 'established' (popula- tions 1—11), or 'colonizing' (populations 12—18). The established populations were growing on documented Electrophoresis Youngleaves were collected from each individual between November, when the leaves first appeared Table 1 Location and size of P. veris populations; % above ground, and March before flower heads polymorphic loci (F); mean number of alleles per locus (A ); appeared. At other times of the year poor or no observed heterozygosity (H0); and Nei's mean diversity enzyme activity was detected. Leaf samples were index (H) stored at 5°C and were used within 48 h of collection. Milligram quantities of leaf tissue were homogenized Approx. with 150 ul of a buffer of 0.5 M phosphate pH 7.0 con- Population numbersPA H0 Ft Established: 1. Blenheim 700 101.160.0210.021 Palace 2. Chilswell 400 101.110.0110.013 Okm Valley 3. Aston l000s 211.260.0230.024 Rowant 4. PulpitHill l000s 101.110.0080.008 5. Minster 200 101.110.0200.020 Wood 6. Stonesfield 1500 151.260.0150.017 Common 7. West Mead l000s 151.260.0220.022 8. Lower Seeds 150 151.210.0220.021 9. The Dell 75 101.160.0250.024 10. Hill End 600 151.260.0340.031 (Triangle) 11. HillEnd 400 151.260.0310.029 (Medieval) 4 —. Colonizing: 12. Showells 200 151.160.0310.031 Roadside 13. Quarry 300 101.160.0040.004 14. Lewknor 250 101.110.0080.008 Roadside 15. Cassington 500 101.110.0080.008 Lagoon 16. TarTrack 80 151.160.0170.017 17. Cumnor l000s 211.260.0250.022 18. Upper Seeds 200 101.160.0090.009 Fig. 1 Location of the populations studied, in Oxfordshire (and pop.4 in Buckinghamshire), England. 254 S. ANTROBUS & A. J. LACK taming 0.5 per cent mercaptoethanol and a few grains \C- SN of sand. 000',r N- Extracts were absorbed onto Whatman no.3 filter d0c'oo 0000000—C: 0 paper wicks and loaded onto starch gels. Horizontal C\0 0r)5r-4 NN OON starch gel electrophoresis (10.8 per cent starch) was 0 r SN performed at 60 mA for 5 h (Ferguson, 1980). A dis- 000000 00000 continuous buffer consisting of 0.002 M histidine in the rS sm N 00 gel and 0.4 M citrate in the electrode, both at pH 7.0 000qcc0000-4000 was employed throughout. Nine enzyme systems pro— N C\ r 5 duced bands of consistent clarity for routine screening: Ir phosphoglucose isomerase, PGI (two loci), phospho— -4 0000000-4000 glucomutase, PGM (two loci), shikimate dehydro- 1I N 00 0 SN 0 genase SKD (two loci), malate dehydrogenase, MDH (one -4 locus), alcohol dehydrogenase, ADH (one locus), glu- 0000 000 S e N 00 tamate oxaloacetate transaminase, GOT (five loci), 0 C 0 6-phosphogluconate dehydrogenase, 6PGD (two loci), 0000000-4 0 aldolase, ALDO (two loci), isocitrate dehydrogenase, N00 S n S IDH (one locus) and malic enzyme, ME (two loci). N '-400 C': Staining recipes were derived from recipes given by -4 doo 0 '-4000 Shaw & Prasad (1970). '.cr s-s SN 00N After recording genotype frequencies, the data were 00I,t C0 -4 analysed using the programme BIOSY5- 1 (Swofford & 00000000—0 N 00 Selander, 1989) to calculate a range of genetic variabi- 00S C\SNNCCCNN S N lity estimates. Wright's (1951) FST statistic was used to cC '-4 -00000000—000 estimate average gene flow, Nm (number of immigrants SC per generation) (Slatkin & Barton, 1989). 00000000—0 — 0 r 5 Heterostyly 00 -C0 00 0 '-4C C':c Theproportion of long- and short-styled morphs was 000000—000 assessed in three established populations, 7, 9 and 10, N r t1N00 ( 05 -C NC 0 C( 4(— and two coloniziiig populations, 17 and 18 in 1991. All 0 cC':c flowering plants in populations 9 and 18 and 100 0 00 d0000000 (0 SOCc"( 0000 li')LI') individuals in populations 7, 10 and 17 (selected using B N 0 00000 SN a grid and random numbers) were scored. C d0000o00-000 LI') LI') 0) SN LI')00 LI') Results 0'.i0c -u 0000000-4000 Nineteenloci were identified from the nine enzyme B 00.) r 5 00 U LI")4/') systems, four of which were polymorphic in at least one 0 (0 C':Q U 000 population, ADH, SKD-2, 6PGD-1 and 6PGD-2 (Table 4- 0) S C) 5r) sr')LI')SoC 2). PGI and PGM were polymorphic in plants from 4- .0 0') '-3CNC0 0 E Hill End that were used for developing protocols, but E B C': o':O_ B 000000000 were not polymorphic in any populations during 0 (') S sampling.