Ecotypic differentiation within Festuca rubra L. occurring in a heterogeneous coastal environment* L. J. Rhebergen & H. J. M. Nelissen** Biological Laboratory, Department of Ecology, Free University, P.O. Box 7161, 1007 MC Amsterdam, The Netherlands Keywords: Ecotype, Festuca rubra, Gene flow, Heritability, Selection

Abstract In order to study the genetic differentiation between Festuca rubra L. individuals growing in a heterogeneous environment, indices of salt tolerance, mean relative growth rates and the numbers of tillers formed by plants grown in a Hoagland solution, were determined. It was found that plants from salt marsh sites have a high index of salt tolerance, a high mean relative growth rate and numerous tillers; plants from coastal sand dunes are less tolerant, grow slowly and form few tillers; plants from the inland polder sites are rather salt sensitive, fast growing and form a high number of tillers. The heritability of the mean relative growth rate and the tiller number appeared to differ from zero. Apparently, these characters have been under recent selection and thus give a picture of the adaptations of individual plants to the different environments encountered. An indication of gene flow has been found, although the effect of gene flow seems to be small in the face of the force of selection. It was concluded that the distinction of three ecotypes within the species F. rubra is insufficient to describe the differentiation found. Considering the differences observed, it seems more reasonable to speak of ecotypic variation.

Introduction Species occupying a wide range of habitats must be adapted to varying circumstances encountered. These adaptations may be physiological or morphological and result in a genetically based differentiation. Differentiation does not require geographic isolation. It has been demonstrated that plants in a heterogeneous environment show local differentiation, if the selection pressure is strong in * Nomenclature follows Heukels & van Ooststroom (1977). ** The authors are indebted to Prof. Dr W. H. O. Ernst and Dr J. A. C. Verkleijfor helpfulsuggestionsduring the investigations and for critically reading the manuscript, to Drs T. Dueck for correcting the English text and to Miss D. Hoonhout for typing the manuscript. The investigations were supported by the Foundation for Fundamental Biological Research(BION), which is subsidized by The Netherlands Organization for the Advancement of Pure Research (zwo). Vegetatio 61, 197-202 (1985). © Dr W. Junk Publishers, Dordrecht. Printed in the Netherlands.

the face of the stabilizing influence of gene flow (Jain & Bradshaw, 1966; Caisse & Antonovics, 1978). For this reason Ehrlich & Raven (1969) concluded that local interbreeding populations are the evolutionary units of importance, selection being the most primary force influencing them. Most investigations made on genetic variation, deal with differentiation, selection and gene flow in view of tolerance to salt or heavy metals (Aston & Bradshaw, 1966; McNeilly & Bradshaw, 1968). In only some of the studies a close examination has been made of the genetic basis of micro-differentiation of species with more or less wide ecological amplitudes (Silander & Antonovics, 1979; Antlfinger, 1981). Festuca rubra is a mainly outbreeding, perennial grass species with a more or less continuous geographic distribution. Individuals of this species occur on acidic and alkaline inland soils, on more

198

Materials and methods

or less clayey salt marshes as well as in dry and sandy dunes along the coast. In The Netherlands three subspecies are distinguished on the basis of the morphology of the plants in their own habitat: the inland form Festuca rubra L. ssp. rubra, the salt marsh form Festuca rubra L. ssp. litoralis (Mey.) Auquier and the dune form Festuca rubra L. ssp. arenaria (Osb.) Richt. (Auquier, 1968; Heukels & van Ooststroom, 1977)). The differentiation within the species at the subspecies level, especially in salt tolerance has been reported repeatedly (Hannon & Bradshaw, 1968; Rozema et aL, 1978, Venables & Wilkins, 1978; Kahn & Marshall, 1981). However, most of these studies were concerned with differentiation between isolated populations. The three habitat types mentioned above are found in close proximity at a number of places along the Dutch coast, where both clinal variation and sharp boundaries occur. The aim of the present paper is to describe the local differentiation, its genetic basis and to discuss the influence of selection and gene flow on this micro-differentiation within the Festuca rubra taxa on the Frisian island ofS chiermonnikoog.

a

Study site Plants used in this investigation, were sampled on the Dutch island of Schiermonnikoog (53 ° 29' N, 6 ° 12' E), from three main habitats: a salt marsh, a dune area and a polder. In Figure la, the sampling sites are indicated. Figure 1b shows a diagrammatic longitudinal section through the beach plain, a sandy salt-marshlike habitat separated from the North Sea by an artificial sand ridge (since 1958-1959). This beach plain is inundated several times during the winter season with salt water from the Wadden Sea. The beach plain consists of a lower part with plant species such as Glaux maritima and Juncus gerardii and some embryo dunes with Agrostis stolonifera, Festuca rubra and on the top, Ammophila arenaria. A detailed description of this heterogeneous habitat is given by Rozema (1978). The sampling sites 1-14 are situated on or near the beach plain. The

North Sea

m

I

"'~

\

S

1¢

18"

-

-.

•

ok bp Beach Plain

~

l/

c,,% Dunes

I

wooo.os.

,

~ ,

t

141312 Kobbeduinen

I

11

-

,'°

--,--

B

t

109

Beach Plain

87

6

54321 Sand ridge

Fig. 1. Location of the sampling sites on Schiermonnikoog. (a) Map of Schiermonnikoog. (b) North south cross-section through the beach plain, taken along the line A-B.

199 m a x i m u m distance between these sites is 350 m, the minimum is 2 m. Soil samples, taken in January 1983, were dried overnight (80 o C). pH (H20) was measured in water extracts (soil:water = 1:2.5 on dry weight basis). Sodium concentrations in extracts (1:10) were determined by means of flame emission photometry and chloride using a Marius micro-chlorocounter. Plant analysis Up to 15 plants together with the seed produced were sampled from each site. Prior to the experiments, plants were grown in the greenhouse under standard conditions (12 hr light, 50 W / m 2, 20 ° C day temperature, 15 o c night temperature, 70% R.H.) for 6 months. The experiments can be separated into a clone experiment and a seedling experiment. In the clone experiment 4 plants of each clone (each plant consisting of one tiller, from which the roots had been removed) were placed in culture solution for two weeks. The culture solution in all experiments was 1/4 strength Hoagland solution as modified by Johnson et al. (1957). After these two weeks most of the tillers had formed new roots. This procedure was necessary to ensure that all the plants were in the same condition at the start of the experiment. After having determined the fresh weight of the plants, two plants of each clone were grown in I / 4 strength Hoagland culture solution; the other two plants were placed in the same medium supplemented with 100 mM sodium chloride (to avoid an osmotic shock the salt level was raised to the m a x i m u m of 100 mM in steps in one week). The culture solution was changed once a week. After 8 weeks fresh weight and the number of tillers were determined. Salt tolerance was determined according to Rozema & Visser (1981), using the ratio of the mean relative growth rate ]~ (fresh weight) on culture solution with salt added and ]~ on medium without salt, as an index of salt tolerance. In the seedling experiment two-week old seedlings were placed in culture solution. After an acclimatization period of one week, fresh weight was determined and half the seedlings (with a m a x i m u m of 8) belonging to families of half-sibs were placed in a medium with 100 mM sodium chloride (added in steps in one week). After 5 weeks fresh weight and number of tillers were determined. Salt tolerance was determined as a family mean. F r o m each

sampling site four families were analyzed. To determine the degree in which differences between individuals may be contributed to the additive effects of genes, the heritability (narrow sense) of two growth characteristics (R on medium without sodium chloride and the number of tillers formed) were calculated using the intraclass correlation coefficient of half-sib families. The heritability thus obtained is the observed correlation as a proportion of the correlation expected if the differences found were completely caused by the additive effects of genes (Falconer, 1981). The statistical analyses (Bartlett's test of homogeneity of variances and the LSD calculations) were according to procedures described by Sokal & Rohlf (1969).

Results

Table 1 shows the pH(H20), and the sodium and chloride concentrations of the soil of the sampling sites. F r o m these results it can be concluded that F. rubra occurs on Schiermonnikoog on a wide range of pH (3.7-7.5) and sodium chloride values (<0.1-32.9 mmol NaCI/100 g dry weight of the soil). Table 2 shows the mean values for index of tolerance, the mean relative growth rate, the logarithm of the number of tillers formed and in which of the three main habitat types each sampling site is situated. The highest mean values for salt tolerance clearly correspond with the salt marsh sites. The plants from the polder sites have a significantly lower index of tolerance, whereas plants from the dunes show a more or less intermediate response. When comparing the various samples from the arti-

Tablel. pH(H20), sodium and chloride concentrations (in mmol/100 g dry soil) in some of the sampling sites. Means of three replications. January 1983. Site

pH

Na

CI

Site

pH

Na

CI

1 2 3 4 5 9 10 12

6.9 7.0 7.0 7.1 7.2 7.4 7.2 7.0

<0.1 0.1 0.1 0.2 4.0 2.9 0,6 1,4

<0.1 <0.I <0.1 0.4 3.2 3.2 0.7 1.0

13 14 15 16 17 18 19 20

7.4 6.8 7.5 6.2 3.9 3.7 5.5 3.9

1.4 0.1 28.9 27.0 <0.1 0.1 0.1 0. I

0.9 <0.1 22.9 32.9 <0.1 <0.1 <0.1 <0. I

200 Table 2. M can values of indices of tolerance ( TO, mean relative growth rate (R in g/g/week) and the logarithm of the n u m b e r of tillers formed (log n). LS D (a = 0.05) and heritability ± standard error (h 2 5: se) are included, s: salt marsh site, d: dune site, p: polder site, s-d: top of embryo dune in the beach plain. Site

T1

Site

R

Site

log n

12s 8s-d 5s 15s 6s 9s 1 Is 10s-d 7s 16s 17d 13d 14d 2d ld 3d 18p 20d 19p 4d

92.1 90.7 84.6 82.7 81.2 80.8 76.5 75.9 75.0 69.5 67.4 63.2 52.5 47.8 47.3 43.7 41.4 35.7 34.4 32.1

9s 18p 19p ld 12s 11 s 16s 6s 7s 5s 4d 15s 10s-d 14d 20d 3d 13d 17d 8s-d 2d

0.32 0.31 0.31 0.30 0.30 0.30 0.29 0.29 0.29 0.28 0.27 0.27 0.27 0.25 0.25 0.25 0.25 0.23 0.19 0.18

ld 15s 19p 5s 16s 7s 9s 12s 6s 10s-d 1 ls 17d 20d 2d 18p 4d 13d 14d 3d 8s-d

0.95 0.86 0.85 0.83 0.81 0.79 0.77 0.74 0.72 0.68 0.66 0.65 0.62 0.60 0.59 0.55 0.53 0.52 0.49 0.42

LSD (0.05)

h 2 + se

29.7

0.07

0.29

0.35 +0.14

0.96 5:0.24

ficial sand ridge (1, 2, 3 and 4) no consistent difference can be found between these sites. The plants from the adjacent beach plain site (5) show a significantly higher index of tolerance. The same applies to the two sites from the Kobbeduinen (13 and 14) and the adjacent site in the beach plain (12). No clear-cut differences between plants from the basis and the top of the two embryo dunes can be found with respect to salt tolerance (sites 7 and 8 and sites 9 and 10). The mean relative growth rates of plants from the salt marsh sites and from the polder are the highest. Plants from the dunes generally show a lower growth rate. It is remarkable that descending from the top of the sand ridge the growth rate increases; site 4 shows a similar growth rate as the adjacent site 5 in the beach plain. Lower growth rates (similar to plants from the dunes) are found in plants on the top of the embryo dunes than in plants at the basis. The same is true for the number of tillers formed. When comparing the results for/~ and the

number of tillers formed, plants from site 15 coming from the salt marsh Oosterkwelder appear to have a rather low growth rate but a high number of tillers. In Table 2 the values of the heritability of the mean relative growth rate and the logarithm of the number of tillers are also given. The results show that a remarkable part of the differences found between individuals can be attributed to the additive effects of genes. In Table 3 the mean indices of tolerance of families of seedlings are compared with the mean index of tolerance of one parent and the site mean. The differences found between the parental values and their offspring are striking, probably caused by the enormous variation shown by plants originating from the same site. However, this difference is small for plants from the more or less isolated sites in the beach plain (site 11) and the Oosterkwelder (site 15). Offspring from site 12 shows a lower index of tolerance compared to their parents, whereas offspring f r o m plants from the sand ridge (sites 1-4) shows a higher index.

Discussion It may be concluded that considerable local differentiation within Festuca rubra has taken place on the island of Schiermonnikoog. This differentiation concerns not only the index of tolerance but

Table 3. C o m p a r i s o n between the index of tolerance of midparents (Tlp) and of the family means (TlFam) of plants from several sites, s: salt marsh site, d: dune site, p: polder site Site 2d 4d 5s

lls 12s

13d

TIp

TIFa m

Site

T]~

T/Fa m

52.4 35.9 95.1 76.2 69.8 77.0 72.8 106.9 97.5 98.0 95.8 68.4 62.6 50. I

114.8 68.7 69.6 I 11.6 92.1 77.6 61.1 113.4 62.5 94.6 78.1 59.9 80.4 54.9

14d

95.9 51.8 78.3 99.2 77.4 78.0 64.9 20.7 44.7 58.7 57.3 17.2 38.6 38.6

80.1 48.7 73.2 133.3 92.8 95.1 88.0 23.0 0 0 40.5 43.0 54.6 59.4

15s

16s 18p

19p 20d

201 also the mean relative growth rate and the number of tillers formed. Generally speaking, plants from the salt marsh sites are rather salt tolerant, fast growing and form a high number of tillers. Plants from the dune areas are less tolerant, grow slowly and form only a few tillers. Plants from the polder are quite sensitive to increased salt levels, grow quickly and form many tillers. The differences described with respect to salt tolerance are comparable to differences found between isolated populations (Rozema et al., 1978; Kahn & Marshall, 1981). However, besides these 'expected' reactions, plants from several sites clearly show an intermediate response. One of the sites shows some unexpected results (site 1, see Table 2). One explanation may be that only one individual from this site was tested. On the other hand, we are possibly dealing with another, chromosomally distinguishable type here (2n = 56 instead of 2n = 42). Kjellqvist (1964) suggested that plants from the loose dunes of the west European coast that resemble F. rubra, belong to the species F. arenaria Osb., the two species being separated by a difference in chromosome number. According to the habitat description given by this author, this variant is only to be expected in site 1 (and eventually site 2). Further research is necessary to discover the frequencies in which the two types occur and resolve whether or not these forms are genetically isolated. As indicated by the heritabilities of the two characteristics determined, the differences found between individuals for R and the number of tillers are caused for a considerable part by the additive effect of genes. The heritabilities given, however, are no more than approximations, because of several assumptions. In the first place it is assumed that only the additive genetic variance contributes to the covariance of half-sibs (as is true for diploids). In fact, for hexaploids (like F. rubra) the dominance variance also contributes (for a small part) in this covariance. This assumption results in an overestimation of the heritabilities (for more details, see Kempthorne, 1963). The second assumption is that individuals within families are half-sibs; this implies a high outcrossing rate. F. rubra is said to be an outcrossing species (Harberd, 1961), although some authors estimated rather high levels of selfing (Auquier, 1977). The heritabilities given are overall heritabilities, which means no distinction has been made between within-site and between-site herita-

bilities. Antlfinger (1981) studied the micro-differentiation of Borrichiafrutescens in relation to salinity. This author found no significant change in a set of morphological characters, when plants were grown on two levels of sodium chloride. All withinsite heritabilities appeared to be significantly different from zero, in contrast to between-site heritahilities. In the present case we find significant differences between various sites for a number of characteristics and rather high overall heritabilities. It seems reasonable to assume that the between-site heritabilities will differ significantly from zero, at least for some sites. Further research will elucidate this point. If the differences in characteristics between plants from different sites are partially caused by the additive effects of genes, they have probably been under recent selection, which contrasts to the case described by Antlfinger (1981). The differences in characteristics found thus give an indication of the adaptations of individual plants to the various environments encountered. This differentiating effect of the selection pressure may be counteracted by gene flow. Differences in the index of tolerance were found between the parental and offspring values in plants from the sites 12, 2 and 4 (Table 3). These may be cause in part by gene flow from F. rubra populations in the Kobbeduinen into the beach plain population and from the beach plain to the sand ridge population. However, this gene flow moves against the prevailing wind (coming from the North Sea), although changes in wind direction are frequent (especially on hot summer days). A second point worth mentioning is a difference in flowering time, which can prevent gene flow. The plants in the beach plain flower about two weeks earlier on the average than plants from the dune sites, although there is some overlap (Rhebergen, unpubl, results). It seems likely, that the local differentiation found is maintained by strong selection in the face of the effects of gene flow (the more so as seedlings have only a small chance of surviving, especially in the dense tufts of the salt marsh plants). This would be in agreement with results given by Antonovics (1968) and Caisse & Antonovics (1978). Further study is needed to determine the amount and effect of gene flow in the present case. Although the soil analysis only reflects the situation at a given moment, it may be concluded that the differences in soil characteristics have given rise

202 to local adaptations within F. rubra to the environments encountered. The resulting differentiation does not lead to three distinct subspecies or ecotypes, but to a genetical gradient within the species from salt-tolerant to sensitive individuals and from fast growers to slow growers. This confirms an idea, stated by Stebbins (1979), that ecotypic variation within species is probably indiscriminate rather than a distinct separation into ecotypes if the physiological stress is not very strong.

References Antlfinger, A. E., 1981. The genetic basis of microdifferentiation in natural and experimental populations of Borrichia frutescens in relation to salinity. Evolution 35: 1056-1068. Antonovics, J., 1968. Evolution in closely adjacent plant populations. VI. Manifold effects of gene flow. Heredity 23: 507-524. Aston, J. L. & Bradshaw, A. D., 1966. Evolution in closely adjacent plant populations. 11. Agrostis stolonifera in maritime habitats. Heredity 21: 649-664. Auquier, P., 1968. Festuca rubra L. subsp, litoralis (G. F. W. Mey.) Auquier: morphologie, 6cologie, taxonomie. Bull. Jard. Bot. Nat. Belg. 38: 181-192. Auquier, P., 1977. Biologic de la r~production dans le genre Festuca L. (Poacea). 1. Syst~mes de pollinisation. Bull. Soc. R. Bot. Belg. 110: 129-150. Caisse, M. & Antonovics, J., 1978. Evolution in closely adjacent plant populations. 1X. Evolution of reproductive isolation in clinal populations. Heredity 40: 371-384. Ehrlich, P. R. & Raven, P. H., 1969. Differentiation of populations. Science 165:1228 1232. Falconer, D. S., 1981. Introduction to Quantitative Genetics. Longman, New York. Hannon, N. & Bradshaw, A. D., 1968. Evolution of salt tolerance in two coexisting species of grass. Nature 220: 1342-1343.

Harberd, D. J., 1961. Observations on population structure and longevity of Festuca rubra L. New Phytol. 60: 184-206. Heukels, H. & Ooststroom, S. J. van, 1977. Flora van Nederland. Wolters-Noordhoff, Groningen. Jain, S. K & Bradshaw, A. D., 1966. Evolutionary divergence among adjacent plant populations. 1. The evidence and its theoretical analysis. Heredity 21:407-411. Johnson, C. M., Stout, P. R., Broyer, T. C. & Carlton, A. B., 1957. Comparative chlorine requirements of different plant species. Plant Soil 8:337 353. Kahn, A. H. & Marshall, C., 198 I. Salt tolerance within populations of chewing fescue (Festuca rubra L.). Commun. Soil Sci. Plant Anal. 12:1271 1281. Kempthorne, O., 1963. An Introduction to Genetic Statistics. Wiley & Sons, New York. Kjellqvist, E., 1964. Festuca arenaria Osb. A misinterpreted species. Bot. Not. 117:389 396. McNeilly, T. & Bradshaw, A. D., 1968. Evolutionary processes in populations of copper tolerant Agrostis tenuis Sibth. Evolution 22:108-118. Rozema, J., 1978. On the ecology of some halophytes from a beach plain in the Netherlands. Thesis, Amsterdam. Rozema, J., Rozema-Dijst, E., Freijsen, A. H. J. & Huber, J. J. L., 1978. Population differentiation within Festuca rubra L. with regard to soil salinity and soil water. Oecologia 34: 329 341. Rozema, J. & Visser, M., 1981. The applicability of the rooting technique measuring salt resistance in populations of Festuca rubra and Juncus species. Plant Soil 62:479 485. Silander, J. A. & Antonovics, J., 1979. The genetic basis of the ecological amplitude of Spartina patens. 1. Morphometric and physiological traits. Evolution 33:1114 1127~ Sokal, R. R. & Rohlf, F. J., 1969. Biometry. Freeman, San Fransisco. Stebbins, G. L., 1979. Fifty years of plant evolution. In: O. T. Solbrig, S. Jain, G. B. Johnson & P. H. Raven (eds.), Topics in Plant Population Biology, pp. 18 41. MacMillan Press, London. Venables, A. V. & Wilkins, D. A., 1978. Salt tolerance in pasture grasses. New Phytol. 80:613 622. Accepted 1.3.1984.