Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Composition and viability of the seed bank along a successional gradient on a Baltic sea shore meadow

Composition and viability of the seed bank along a successional gradient on a Baltic sea shore... Jerling, L. 1983. Composition and viability of the seed bank along a successional gradient on a Baltic sea shore meadow. - Holarct. Ecol. 6: 150-156. The seed-bank in a Baltic sea-shore meadow was studied and soil samples were taken regularly along a 60-m transect. The transect reflected a successional gradient. The seed bank had its highest density around the mean water level for August, which was about 10 cm above the yearly mean water level. It was totally dominated by seeds of Juncwi gerardii. The depth distribution was investigated. Only at iwo of ten sampling points (i.e. around mean water level of August) could a significant peak in the second cm from the top be observed, otherwise the distribution was quite even. A comparison between present vegetation and the seed bank revealed a strong difference. L. Jerling, Dept of Ecology. Botanical Inst., Univ. of Stockholm. Litia Frescati, S-I06 9i Stockholm, Sweden. Introduction The tiumber of buried viable seeds has beeti estimated many times outside Sweden in many types of habitats (for a review see i.e. Harper 1977). Investigation of the seed bank in relation to succession was first done by Oosting and Humphrey (1940), but after them only few attempts have been made (see Livingstone and Alessio 1968. Hayashi and Numata 1971). Milton (1939) sampled and analysed buried seeds in a salt marsh of Wales. In Denmark Jensen (1969) estimated the content of buried seeds in arable soil and 0dum in archaeologically dated soil samples (1965) and in ruderal soils (1978). In Sweden, however, only a few investigations have been reported (Liljelund and Jerling 1980, Zimmergren 1980). Little is known about the formation of seedlings from the seed bank each year. Barallis (1965) found that the weed flora in a winter weed crop represented only 4% of the viable seed population in the top 10 cm of soil. Similar observations were made by Kropac (1966). However, in reeolonization of disturbed areas the seed bank is of great importance {Fries 1853, Oosting and Humphrey 1940, Ericson 1977, Liljelund 1979, Cook Accepted 30 November 1981 © HOLARCTIC ECOLOGY 1980, Zimmergren 1980). Such dispersal in time is of great interest in areas with low predictability. The seed bank ean also act as a genetic stabilizer when the number of individuals fluctuates (Epling et al. 1960, Gottlieb 1974, Templeton and Levin 1979) and thus disperse genetie information in time. This study is a description of the seed bank along a suecessional gradient in a Baltic sea-shore meadow about 60 km SSW of Stockholm. It includes a comparison between present vegetation and seed-bank. The vegetation, flora and soil conditions were described by Walientinus {1967, 1970) and the terminology of vegetation follows this work. Material and methods Study area The sea-shore meadow, surrounded by Phragmites communis reeds, was very sheltered and had a vegetational gradient related to sea-water level. The hydrolitoral vegetation consisted in grazed areas, of Scirpus maritimus vegetation. The geolitoral vegetation started with a Juncus gerardii belt which was replaced by a HOLARCnC ECOLOGY 6:2 (1983) Tab. I. The relative abundance of different species in the vegetation and in the seed bank on different points along the transect. Relative abundance in vegetation + SE (upper figure) is expressed as the mean percentage cover in siz squares. Relative abundance of seeds ± SE (lower figure) is expressed as the mean percentage in six samples. Significant differences between seed bank and vegetation ane shown by asterisks. 6m Triglochin marifinum Phragmites communis Agroslis slolonifera Pou irrigafa Fesfucu rubra 12 m 14.2+6.3 18 m _* 5.5 + 3.7 24 m _ NS _ *** 7.5 + 2.0 0.8+0.4 4.5±2.7 11.813.0** 3.812.3 1.2±0.8 13.613.5*** 1.510.8 0,3 + 0.3 12.2 + 2.8* 0.510.3 3.612.2^^ 0.3+0.2 18.2 + 3.6* 1.5+1 33.715.7* 81.1123.8 0.3±0.2 4.711.9 0.310.2 11.514.1* 1.110.3 ScirpiLs maritimus Scirpus uniglumis Carex disticha Carex nigra Juncus gerardii Sperguiaria marina Cardamine prafensis Pofenfilla anserina 3.8+1.0 0.7+0.3 2.411.2 17.314.0*** 67.8112.6 0.7 + 0.6 1.210.3 20.613.2*** 56.719.8 8.212.4 23.218.0 0.210.2 12.811.9'^^ 17.819.71 6.912.04*** 5.712.3 1.211.2 Trifolium fragiferum (iliiii.x maritima I'lantago murifima Leontodon aufumnalis Unidentified dicotyledons Total cover (%) Total number of seeds 7.213.2 0.310.3 28.816.3 0.3 + 0.3 7.410.9'^^ 26.9112.1 5.413.5*** 3.8+2.5 0.310.3 68.013.4 376 9.911.9^^ 5.111.1 30.712.8* 0.410.3 1.210.8 1.911.1 0.410.3 73.617.0 270 15.313.5''^ 15.616.7 23.313.3* 0.910.6 Festuca rubra belt. In areas influenced by freshwater the Festucu ruhra belt was replaced by a Carex nigra belt. The soil consisted of a heavy clay except in the middle of the meadow although the transect used here did not touch this area. In the field Along a 60-m transeet through the meadow, six soil samples were taken each 6th m between the 7 and 10 June 1979. The samples were taken with a brass corer (diameter 6 cm) and then put in plastic bags for transHOLARCTIC ECOLOGY 6;2 (1983) portation to the laboratory. The soil samples were taken down to a depth of 10 cm but only the top 7 cm were used. The 6 and 12 m points fell within the Carex nigra belt and the 18, 24 and 30 m points in the Festuca rubra belt. The Juncus gerardii belt included the 36, 42 and 48 m points and the Scirpus maritimus belt the 54 and 60 m points. The vegetation was analysed in each of six squares, placed on every 6th m of the transect. These squares were distributed at random distances (between 0 and 10 m) from the transect on a line perpendicular to the transect i.e., parallell to the shore line. The coverage 151 30 m 36 m 2.211.1 1.510.9 42 m 4.7±2.7 2.911.6 48 m 3.1+3.4^^ 7.2+3.2 8.3+3.0 8.014.4^^ 54 m 60 m 26.013.4^^^ 19.012.1** 11.715.1 8.8±].3 32.413.8** 11.511.9 39.414.7** O.I+0.1 0.1+0.1 0.1+0.1 0.1+0.1 9.0+4.5'^^ 0.1+0.1 0.7+0,4'^^ 0.1+0.1 0.510.5^^ 2.8+1.2 4.1 + 1.7 44.8+7.1*** 84.6+14.6 0.110.1 42.7+4.3*** 91.2±t9.1 40.0+3.2*** 94.2+14.2 30.312.7** 89.719.7 13.0+0.9** 87.9+18.3 11.5+1.6" 66112 0.1+0.1 0.110.1 0.110.1 10.511.8'^^ 2.911.1 37.019.2** 0,110.1 12.8 + 2.5'*^ 10.512.1 31.213.1*** 0.110.1 0.110.1 0.110.1 55.514.0 1148 12.511.7^^ 6.2+2.8 45.0+2.8*** O.I 10.1 12.111.4^'' 5.512.1 38.915.4*** 0.110.1 7.210,9** 0.3 + 0.1 15.5 + 2.2** 6.411.8* 0.610.4 5.110.6** 0.110.1 0.1 + 0.1 62.2+1.6 2393 0.1+0.1 60.2+5.72 1509 48.718,1 817 percents were used as the measure of abundance. Each square was 0.5 X 0.5 m. In the laboratory The soil, still moist, was cut into 1 -em sliees and put in a greenhouse for germination 15 June 1979. After two months, when the germination had deereased, the soilslices were dried, crushed, watered and again put in the greenhouse for germination. The samples, watered twice a week, were covered by plastic film. The samples were supplied with extra light from fluorescent tubes 152 12 h per day. Once a month the samples were stirred so that new seeds would be exposed to the light and the soil loosen up. The seedlings that germinated were counted once a week and removed as soon as they were identified. The germination trial was terminated 15 September 1980. Statistical treatment The confidence limits of the mean indicated in Fig. 1 are calculated using Student's t-dlstribution as described by Zar {1974). Differences between abundance in vegetaHOLARCnC ECOLOGY 6;2 (198.1) well as between depths (p < 0.005). In addition a variation in depth-distribution of seeds between zones was significant (p < 0.005). The only significant differences between largest and second largest total number of seeds of different layers were found at the 42 and 48-m points. Here the second depth layer contained more seeds than the others (p < 0.05), Of vegetationai belts xhe Juncus gerardii belt and the Scirpus maritimus had more seeds in the second layer but in the other belts this was not so (Tab. 2). Regarding Ihe three most abundant species (Tab. 2) in the seed bank. Glaitx maritima had a peak in abundance Results of seeds in the third layer in the Fesfuca ruhra belt. This The distribution of seeds per square metre was not uni- was so also in the Juncus gerardii belt but here the third form along the transeet but had peaks at 36. 42 and 48 layer did not differ from the second which in turn did m, all within ihc Juncus gerardii belt (Fig. 1), The 48-m not differ from the others. The peak in the fifth layer in point corresponded with the mean water level of the the Carex nigra belt was not significant but differed year and the 42-m point with the mean water level of from all the others except the fourth layer at the p < 0.1 August and September. The latter was about 10 em level. above mean water level of the year (Wallentinus 1970). Triglochin mariiimum was evenly distributed vertiOf the 13 species that germinated (Tab. 1), Juncius cally, except in the Scirpus mariiimus belt. Here the gerardii contributed the vast majority of the total third layer contained more seeds than the first but not number of seedlings (86.8%) (Fig. 1). Two more more than the others. species contributed noticeably, namely Trigloch'm The second depth-layer in the Juncus gerardii belt, as maritimum (5.6%) and Gtaux maritima (5.9%), The well as in the Scirpus maritimus belt, contained more part of the seed bank made up of Juncus gerardii was Juncus gerardii seeds than the other layers; otherwise not uniform along the transect but had a peak around no differences could be found for Juncus gerardii. the 42-m point. The composition of the present vegetation and the Analysis of variance revealed differences in seed seed bank are shown in Tab. 1 for the six most abundant number per square metre between zones (p < 0.005) as speeies. Note that Juncus gerardii was significantly overrepresented in the seed bank in every comparison and Plantago maritima underrepresented in all but one. tion and seed bank are tested using the Mann-Whitney U-tcst outlined by Zar (1974). The variation in seed number between vegetationai belts and between depth layers was analysed using two-factor anova. The ranking of seed abundance between depth layers was done using one-factor ANOVA followed by Neuman-Keuls multiple range test Zar (1974). Levels ot significance used here are p < 0.05{*), p < 0.01(**) and p < 0.005(***). 10 seeds/m 200- 150- 100- 50- Depth (cm) seeds/iii2 4 • O - Juncus gerardii, • = Fig. 1. Avobe: The total number of seeds per m^ along the transect. Brackets show 95% confidence limits. other seeds. Below: The distribution of seeds relative to depth along the transect expressed as seeds per m^ within a depth layer. 6 • 6 Carex belt 18 Zfl Festuca rubra belt 36 42 48 Juncus g e r a r d i i 54 60 m Scirpgs maritimus belt belt HOLARCTIC ECOLOGY 6:2 <19R3) Tab. 2. Significant differences (at the 5-% level) between soil layers for the three most abundant speciesof seeds in the seed bank in different vegetation belts of the transect. Numbers 1-5 denote different depths, where 1 is O-I cm below soil surface. Carex nigra helt Juncus gerardii NS Fesfuca ruhra belt Juncus gerardii helt Scirpus marifimus belt Glaux maritima Triglochin marilinum Total number of seeds found NS NS NS 3>1=2=4=5 NS NS NS 2>1=3=4=5 3>1=4=5, 3=2, 2=1=4=5 NS 2>1=3=4=5 NS 2>1=3=4=5 3>i, 3 = 2=4=5 2>1=3=4=5 Discussion From the ecological point of view, the interesting portion of the seed bank is that which has a reasonable chance to germinate. Therefore the accurate way to investigate the seed bank would be to simulate possible mobilizing events in the field. This procedure, however, involves many practical problems, and the germination of seeds in soil samples is usually handled in the laboratory, as in this study. This often induces a series of experimental errors such as letting the samples dry out. forcing seeds of many plants into dormancy or even causing their death (Mlienscher 1936, Isely 1944). Breaking seed dormancy ean sometimes be done by chilling (e.g. Muenseher 1936. Binet 1960), with fluctuating temperature (e.g. Binet 1965, Boueaud 1962), or by scarifying them (Boucaud 1962). Thus a bias towards easily germinated seeds is probable in the figures presented here. Germination abilities can thus partly explain why some speeies commonly found in the vegetation are totally missing or of low abundance in the seed bank. Such species are Trifolium fragiferum. Plantago marifima, Festuca rubra, Agrostis stolonifera, Phragmites communis and Scirpus uniglumis. The four former species are usually easy to germinate (see Harper and Clatworthy 1963, Arnold 1973, Johnston 1961 and Leggati 1946, respectively). However, ethylene, commonly produced under waterlogged conditions, enforces dormancy on Plantago mariiima (Olatoye and Hall 1973) and similar mechanisms for other species are nol mpos^ihle. Agrostis stoUmifera shows a germination that is proportional to light intensity in capacity but inversely proportional in speed (Leggatt 1946). Scirptis seeds are hard to germinate. These seeds often need submersion to germinate (lseiy 1944, Walientinus 1967) which could have prevented them from germination in this experiment. Seeds oi Phragmites communis sampled in autumn and stored at -2-0°C until March will germinate at a temperature of 18-20''C {Veber 1978). Thus, one would expect that at least seeds produced in 1978 would germinate, i.e., if they had not already done so and vanished. However 1 found none. Other species in the vegetation, as deseribed in Tab. ], are either too Carex nigrabelt Deptfi (cm) Festuca belt Juncus qerardiibelt Scirpus maritimusbelt '.. 50 -lO^seeds in-2 ' • I — Fig. 2. The depth distribution of the three most abundant speeies of seeds, in different zones of the gradient. Note differences in scale. JUNCUS GERARDIt HOLARCKC ECOLOGY 6:2 (1983) infrequent or flower too poorly to be well represented in the seed bank. 1 he very high standard errors for seed bank estimates in Tab. 1 indicate that variation between samples was great. 1 his is to be expected since seeds accumulate in local depressions, on the edge of tussoeks etc. and the distribution was clumped. The frequency distribution of seeds along the transect (Fig. I)) shows that the highest density of seeds in the soil was found on or just above the mean water level of August (i.e. 42 m). The seeds, or capsules, that land on the water surface will accumulate at the water's edge if they float well. Several species of the area have seeds with a rather short floating time but as many of them (e.g. Glaux maritima, Juncus gerardii, Plantago maritima, Trifolium fragiferum) commonly disperse their seeds while still in the capsule this measure is misleading. The capsules will surely float longer and thus have a greater chance to settle at the water's edge. The vast majority of seeds in the seed bank (98.3%) consisted of Juncus gerardii, Glaux maritima and Triglochin maritimum. For the latter the area with the highest number of seeds in the soil coincided with the highest cover figures, as expected. This was not the faet for the other two dominants. The explanation may be that they flower and set fruit best in ihe Juticus gerardii belt although they are less abundant there (Jerling unpubl,). The highest production of seeds dominating the seed bank thus coincided with their highest number in the soil - which was to be expected. Ihe sample-mean peaks in deeper layers on 6, 12, 18 and 24 m eould. although they are not significant, either be a remainder from the time when the water level was found higher up (i.e. contributing to seed production area), or be a result of downward movement of seeds in the soil. The depth distribution data eould be interpreted as if all depth layers, including the deeper ones, contained an equal number of seeds, with the two exceptions in the 42 and 48-m points where the second layer contained more. Downward movement of small seeds, in loose textured soils with percolating water is reported by Harper (1977). In the study area the soil, which was moist, packed and did not show any cracks or holes, contained between 40 and 60% of clay (Walientinus 1970). Percolation of water and transportation of seeds probably were slow. This was also shown by the fact that puddles were formed readily during rain and persisted for days. In addition, the area around 60 m, usually had a ground water level below soil surface only in March, April, May and June and of those months ground was usually frozen in March and early April, while May and June had greater evaporation than precipitation. (For meteorological data see Walientinus 1970). Still there was a significant number of seeds down to 5 6 cm depth — (Fig. 2). Transportation of seeds down into the soil therefore did not seem very likely and buried seeds would mainly be a remainder from times when the shore HOLARCTIC ECOLOGY 6:2 (1983) line was found higher up. If so, their age would be at least a hundred years, because the border to the terrestrial meadow (i.e. 0 m on the transect) was estimated to be the shore-line about 1840 (Walientinus 1970, Ryberg 1971). A comparison between the present vegetation and the seed bank reveals that there usually was a great difference in abundance for any certain species. This has been noted in most seed bank investigations (Harper 1977). In this study Jimcus gerardii was strongly overrepresented in the seed bank while Festuca rubra, Agrostis sfolonifera and Plantago mariiima were underrepresented. Gtaux maritima was overrepresented in the seed bank at the 6-m point but underrepresented at the 36, 48, 54 and 60-m points. The latter observations suffer probably from systematic errors as the present vegetation was measured as percentage of cover. This eaused overestimation of herbs over grasses. All herbs exeept Plantago maritima and Glaux marifima were present in low abundance in the vegetation. Seeds of Plantago maritima were almost absent from the seed bank and the difference between representation in cover and seed bank was significant. Festuca rubra, Agrostis stolonifera and Plantago maritima produced seeds and seedlings readily. Plantago maritima seedlings can be as many as about 2500 m""^ (Jerling 1981). None of these three species has any innate or induced dormancy, although sometimes enforced (sensu Harper 1977), and seeds probably germinate the year after they are spread. The three speeies that dominated the seed bank i.e., Juncus gerardii, Triglochin maritimum and Gtaux maritima have all vegetative multiplication and small seeds with innate or induced dormancy. Seedlings of Juncus species are deseribed as having low survivorship in vegetation (Lazenby 1955), as have seedlings of Glaux maritima (Jerling unpubl.). The seeds of Juncus need light and a ripening period to germinate (Lazenby 1955). Germination of Glaux mariiima seeds is induced by fluctuating temperature and light (Binet 1965). Seeds of Triglochin maritimum germinate after treatment with cold water and germination is stimulated by light (Binet I960). Consequently, seeds of these seed bank dominants seem to be more or less dependent on bare ground to germinate and to survive as seedlings. These speeies are halophytes and almost restricted to sea shores. Such habitats often have low predictability for flooding, ice erosion etc. (i.e. inducing density independent mortality). The dormancy of these seeds and their accumulation in the seed bank give these species a better ability to reeover after such disturbances, although the advantage of this character depends on the frequency of disturbances. The chance of a seed in the soil to become mobilized is lower the deeper it is buried. However, how much the chance is decreased depends on what agents are acting. In this area sea shore meadow disturbances would probably include (1) flooding and consequent erosion. but also (2) ice-erosion, both acting around the mean water level where the seed bank was most dense. To what depth these two agents act depends of course on the exposure of the shore. On a shore meadow which develops in sheltered areas the effect will probably not extend deeper than a fen centimetres. Another probable disturbance is (3) trampling, mostly by cattle but also by deer, elk etc. The trampling by eattle is not uniform throughout the gradient but heavier in certain areas, due to selective grazing (Jerling and Andersson 1981). Around or below mean water level the abundance of Phragmites communis is high (this speeies is preferred as food in presummer) and trampling here is frequent at a time when the vegetation is not fully developed. Thus the sward is easily damaged and bare ground shown. This probably mobilizes seeds from deeper layers than does erosion. However, badgers digging for food (4), a not unfrequent occurrence on the meadow, mobilize seeds from still deeper layers. Acknowledgements - I wish to thank Gisela Elmgren and P. Torstensson for help with the laboratory work, also Joanne Rowley for criticism on the manuscript and to Margareta von Lampe for typewriting. The laboratory work was financed by the Swedish natural research council. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Ecography Wiley

Composition and viability of the seed bank along a successional gradient on a Baltic sea shore meadow

Jerling, Lenn
Ecography , Volume 6 (2) – Apr 1, 1983
Download PDF
8 pages

Loading next page...
 
/lp/wiley/composition-and-viability-of-the-seed-bank-along-a-successional-SatYf60Rm7

References (34)

Publisher
Wiley
Copyright
Copyright © 1983 Wiley Subscription Services, Inc., A Wiley Company
ISSN
0906-7590
eISSN
1600-0587
DOI
10.1111/j.1600-0587.1983.tb01076.x
Publisher site
See Article on Publisher Site

Abstract

Jerling, L. 1983. Composition and viability of the seed bank along a successional gradient on a Baltic sea shore meadow. - Holarct. Ecol. 6: 150-156. The seed-bank in a Baltic sea-shore meadow was studied and soil samples were taken regularly along a 60-m transect. The transect reflected a successional gradient. The seed bank had its highest density around the mean water level for August, which was about 10 cm above the yearly mean water level. It was totally dominated by seeds of Juncwi gerardii. The depth distribution was investigated. Only at iwo of ten sampling points (i.e. around mean water level of August) could a significant peak in the second cm from the top be observed, otherwise the distribution was quite even. A comparison between present vegetation and the seed bank revealed a strong difference. L. Jerling, Dept of Ecology. Botanical Inst., Univ. of Stockholm. Litia Frescati, S-I06 9i Stockholm, Sweden. Introduction The tiumber of buried viable seeds has beeti estimated many times outside Sweden in many types of habitats (for a review see i.e. Harper 1977). Investigation of the seed bank in relation to succession was first done by Oosting and Humphrey (1940), but after them only few attempts have been made (see Livingstone and Alessio 1968. Hayashi and Numata 1971). Milton (1939) sampled and analysed buried seeds in a salt marsh of Wales. In Denmark Jensen (1969) estimated the content of buried seeds in arable soil and 0dum in archaeologically dated soil samples (1965) and in ruderal soils (1978). In Sweden, however, only a few investigations have been reported (Liljelund and Jerling 1980, Zimmergren 1980). Little is known about the formation of seedlings from the seed bank each year. Barallis (1965) found that the weed flora in a winter weed crop represented only 4% of the viable seed population in the top 10 cm of soil. Similar observations were made by Kropac (1966). However, in reeolonization of disturbed areas the seed bank is of great importance {Fries 1853, Oosting and Humphrey 1940, Ericson 1977, Liljelund 1979, Cook Accepted 30 November 1981 © HOLARCTIC ECOLOGY 1980, Zimmergren 1980). Such dispersal in time is of great interest in areas with low predictability. The seed bank ean also act as a genetic stabilizer when the number of individuals fluctuates (Epling et al. 1960, Gottlieb 1974, Templeton and Levin 1979) and thus disperse genetie information in time. This study is a description of the seed bank along a suecessional gradient in a Baltic sea-shore meadow about 60 km SSW of Stockholm. It includes a comparison between present vegetation and seed-bank. The vegetation, flora and soil conditions were described by Walientinus {1967, 1970) and the terminology of vegetation follows this work. Material and methods Study area The sea-shore meadow, surrounded by Phragmites communis reeds, was very sheltered and had a vegetational gradient related to sea-water level. The hydrolitoral vegetation consisted in grazed areas, of Scirpus maritimus vegetation. The geolitoral vegetation started with a Juncus gerardii belt which was replaced by a HOLARCnC ECOLOGY 6:2 (1983) Tab. I. The relative abundance of different species in the vegetation and in the seed bank on different points along the transect. Relative abundance in vegetation + SE (upper figure) is expressed as the mean percentage cover in siz squares. Relative abundance of seeds ± SE (lower figure) is expressed as the mean percentage in six samples. Significant differences between seed bank and vegetation ane shown by asterisks. 6m Triglochin marifinum Phragmites communis Agroslis slolonifera Pou irrigafa Fesfucu rubra 12 m 14.2+6.3 18 m _* 5.5 + 3.7 24 m _ NS _ *** 7.5 + 2.0 0.8+0.4 4.5±2.7 11.813.0** 3.812.3 1.2±0.8 13.613.5*** 1.510.8 0,3 + 0.3 12.2 + 2.8* 0.510.3 3.612.2^^ 0.3+0.2 18.2 + 3.6* 1.5+1 33.715.7* 81.1123.8 0.3±0.2 4.711.9 0.310.2 11.514.1* 1.110.3 ScirpiLs maritimus Scirpus uniglumis Carex disticha Carex nigra Juncus gerardii Sperguiaria marina Cardamine prafensis Pofenfilla anserina 3.8+1.0 0.7+0.3 2.411.2 17.314.0*** 67.8112.6 0.7 + 0.6 1.210.3 20.613.2*** 56.719.8 8.212.4 23.218.0 0.210.2 12.811.9'^^ 17.819.71 6.912.04*** 5.712.3 1.211.2 Trifolium fragiferum (iliiii.x maritima I'lantago murifima Leontodon aufumnalis Unidentified dicotyledons Total cover (%) Total number of seeds 7.213.2 0.310.3 28.816.3 0.3 + 0.3 7.410.9'^^ 26.9112.1 5.413.5*** 3.8+2.5 0.310.3 68.013.4 376 9.911.9^^ 5.111.1 30.712.8* 0.410.3 1.210.8 1.911.1 0.410.3 73.617.0 270 15.313.5''^ 15.616.7 23.313.3* 0.910.6 Festuca rubra belt. In areas influenced by freshwater the Festucu ruhra belt was replaced by a Carex nigra belt. The soil consisted of a heavy clay except in the middle of the meadow although the transect used here did not touch this area. In the field Along a 60-m transeet through the meadow, six soil samples were taken each 6th m between the 7 and 10 June 1979. The samples were taken with a brass corer (diameter 6 cm) and then put in plastic bags for transHOLARCTIC ECOLOGY 6;2 (1983) portation to the laboratory. The soil samples were taken down to a depth of 10 cm but only the top 7 cm were used. The 6 and 12 m points fell within the Carex nigra belt and the 18, 24 and 30 m points in the Festuca rubra belt. The Juncus gerardii belt included the 36, 42 and 48 m points and the Scirpus maritimus belt the 54 and 60 m points. The vegetation was analysed in each of six squares, placed on every 6th m of the transect. These squares were distributed at random distances (between 0 and 10 m) from the transect on a line perpendicular to the transect i.e., parallell to the shore line. The coverage 151 30 m 36 m 2.211.1 1.510.9 42 m 4.7±2.7 2.911.6 48 m 3.1+3.4^^ 7.2+3.2 8.3+3.0 8.014.4^^ 54 m 60 m 26.013.4^^^ 19.012.1** 11.715.1 8.8±].3 32.413.8** 11.511.9 39.414.7** O.I+0.1 0.1+0.1 0.1+0.1 0.1+0.1 9.0+4.5'^^ 0.1+0.1 0.7+0,4'^^ 0.1+0.1 0.510.5^^ 2.8+1.2 4.1 + 1.7 44.8+7.1*** 84.6+14.6 0.110.1 42.7+4.3*** 91.2±t9.1 40.0+3.2*** 94.2+14.2 30.312.7** 89.719.7 13.0+0.9** 87.9+18.3 11.5+1.6" 66112 0.1+0.1 0.110.1 0.110.1 10.511.8'^^ 2.911.1 37.019.2** 0,110.1 12.8 + 2.5'*^ 10.512.1 31.213.1*** 0.110.1 0.110.1 0.110.1 55.514.0 1148 12.511.7^^ 6.2+2.8 45.0+2.8*** O.I 10.1 12.111.4^'' 5.512.1 38.915.4*** 0.110.1 7.210,9** 0.3 + 0.1 15.5 + 2.2** 6.411.8* 0.610.4 5.110.6** 0.110.1 0.1 + 0.1 62.2+1.6 2393 0.1+0.1 60.2+5.72 1509 48.718,1 817 percents were used as the measure of abundance. Each square was 0.5 X 0.5 m. In the laboratory The soil, still moist, was cut into 1 -em sliees and put in a greenhouse for germination 15 June 1979. After two months, when the germination had deereased, the soilslices were dried, crushed, watered and again put in the greenhouse for germination. The samples, watered twice a week, were covered by plastic film. The samples were supplied with extra light from fluorescent tubes 152 12 h per day. Once a month the samples were stirred so that new seeds would be exposed to the light and the soil loosen up. The seedlings that germinated were counted once a week and removed as soon as they were identified. The germination trial was terminated 15 September 1980. Statistical treatment The confidence limits of the mean indicated in Fig. 1 are calculated using Student's t-dlstribution as described by Zar {1974). Differences between abundance in vegetaHOLARCnC ECOLOGY 6;2 (198.1) well as between depths (p < 0.005). In addition a variation in depth-distribution of seeds between zones was significant (p < 0.005). The only significant differences between largest and second largest total number of seeds of different layers were found at the 42 and 48-m points. Here the second depth layer contained more seeds than the others (p < 0.05), Of vegetationai belts xhe Juncus gerardii belt and the Scirpus maritimus had more seeds in the second layer but in the other belts this was not so (Tab. 2). Regarding Ihe three most abundant species (Tab. 2) in the seed bank. Glaitx maritima had a peak in abundance Results of seeds in the third layer in the Fesfuca ruhra belt. This The distribution of seeds per square metre was not uni- was so also in the Juncus gerardii belt but here the third form along the transeet but had peaks at 36. 42 and 48 layer did not differ from the second which in turn did m, all within ihc Juncus gerardii belt (Fig. 1), The 48-m not differ from the others. The peak in the fifth layer in point corresponded with the mean water level of the the Carex nigra belt was not significant but differed year and the 42-m point with the mean water level of from all the others except the fourth layer at the p < 0.1 August and September. The latter was about 10 em level. above mean water level of the year (Wallentinus 1970). Triglochin mariiimum was evenly distributed vertiOf the 13 species that germinated (Tab. 1), Juncius cally, except in the Scirpus mariiimus belt. Here the gerardii contributed the vast majority of the total third layer contained more seeds than the first but not number of seedlings (86.8%) (Fig. 1). Two more more than the others. species contributed noticeably, namely Trigloch'm The second depth-layer in the Juncus gerardii belt, as maritimum (5.6%) and Gtaux maritima (5.9%), The well as in the Scirpus maritimus belt, contained more part of the seed bank made up of Juncus gerardii was Juncus gerardii seeds than the other layers; otherwise not uniform along the transect but had a peak around no differences could be found for Juncus gerardii. the 42-m point. The composition of the present vegetation and the Analysis of variance revealed differences in seed seed bank are shown in Tab. 1 for the six most abundant number per square metre between zones (p < 0.005) as speeies. Note that Juncus gerardii was significantly overrepresented in the seed bank in every comparison and Plantago maritima underrepresented in all but one. tion and seed bank are tested using the Mann-Whitney U-tcst outlined by Zar (1974). The variation in seed number between vegetationai belts and between depth layers was analysed using two-factor anova. The ranking of seed abundance between depth layers was done using one-factor ANOVA followed by Neuman-Keuls multiple range test Zar (1974). Levels ot significance used here are p < 0.05{*), p < 0.01(**) and p < 0.005(***). 10 seeds/m 200- 150- 100- 50- Depth (cm) seeds/iii2 4 • O - Juncus gerardii, • = Fig. 1. Avobe: The total number of seeds per m^ along the transect. Brackets show 95% confidence limits. other seeds. Below: The distribution of seeds relative to depth along the transect expressed as seeds per m^ within a depth layer. 6 • 6 Carex belt 18 Zfl Festuca rubra belt 36 42 48 Juncus g e r a r d i i 54 60 m Scirpgs maritimus belt belt HOLARCTIC ECOLOGY 6:2 <19R3) Tab. 2. Significant differences (at the 5-% level) between soil layers for the three most abundant speciesof seeds in the seed bank in different vegetation belts of the transect. Numbers 1-5 denote different depths, where 1 is O-I cm below soil surface. Carex nigra helt Juncus gerardii NS Fesfuca ruhra belt Juncus gerardii helt Scirpus marifimus belt Glaux maritima Triglochin marilinum Total number of seeds found NS NS NS 3>1=2=4=5 NS NS NS 2>1=3=4=5 3>1=4=5, 3=2, 2=1=4=5 NS 2>1=3=4=5 NS 2>1=3=4=5 3>i, 3 = 2=4=5 2>1=3=4=5 Discussion From the ecological point of view, the interesting portion of the seed bank is that which has a reasonable chance to germinate. Therefore the accurate way to investigate the seed bank would be to simulate possible mobilizing events in the field. This procedure, however, involves many practical problems, and the germination of seeds in soil samples is usually handled in the laboratory, as in this study. This often induces a series of experimental errors such as letting the samples dry out. forcing seeds of many plants into dormancy or even causing their death (Mlienscher 1936, Isely 1944). Breaking seed dormancy ean sometimes be done by chilling (e.g. Muenseher 1936. Binet 1960), with fluctuating temperature (e.g. Binet 1965, Boueaud 1962), or by scarifying them (Boucaud 1962). Thus a bias towards easily germinated seeds is probable in the figures presented here. Germination abilities can thus partly explain why some speeies commonly found in the vegetation are totally missing or of low abundance in the seed bank. Such species are Trifolium fragiferum. Plantago marifima, Festuca rubra, Agrostis stolonifera, Phragmites communis and Scirpus uniglumis. The four former species are usually easy to germinate (see Harper and Clatworthy 1963, Arnold 1973, Johnston 1961 and Leggati 1946, respectively). However, ethylene, commonly produced under waterlogged conditions, enforces dormancy on Plantago mariiima (Olatoye and Hall 1973) and similar mechanisms for other species are nol mpos^ihle. Agrostis stoUmifera shows a germination that is proportional to light intensity in capacity but inversely proportional in speed (Leggatt 1946). Scirptis seeds are hard to germinate. These seeds often need submersion to germinate (lseiy 1944, Walientinus 1967) which could have prevented them from germination in this experiment. Seeds oi Phragmites communis sampled in autumn and stored at -2-0°C until March will germinate at a temperature of 18-20''C {Veber 1978). Thus, one would expect that at least seeds produced in 1978 would germinate, i.e., if they had not already done so and vanished. However 1 found none. Other species in the vegetation, as deseribed in Tab. ], are either too Carex nigrabelt Deptfi (cm) Festuca belt Juncus qerardiibelt Scirpus maritimusbelt '.. 50 -lO^seeds in-2 ' • I — Fig. 2. The depth distribution of the three most abundant speeies of seeds, in different zones of the gradient. Note differences in scale. JUNCUS GERARDIt HOLARCKC ECOLOGY 6:2 (1983) infrequent or flower too poorly to be well represented in the seed bank. 1 he very high standard errors for seed bank estimates in Tab. 1 indicate that variation between samples was great. 1 his is to be expected since seeds accumulate in local depressions, on the edge of tussoeks etc. and the distribution was clumped. The frequency distribution of seeds along the transect (Fig. I)) shows that the highest density of seeds in the soil was found on or just above the mean water level of August (i.e. 42 m). The seeds, or capsules, that land on the water surface will accumulate at the water's edge if they float well. Several species of the area have seeds with a rather short floating time but as many of them (e.g. Glaux maritima, Juncus gerardii, Plantago maritima, Trifolium fragiferum) commonly disperse their seeds while still in the capsule this measure is misleading. The capsules will surely float longer and thus have a greater chance to settle at the water's edge. The vast majority of seeds in the seed bank (98.3%) consisted of Juncus gerardii, Glaux maritima and Triglochin maritimum. For the latter the area with the highest number of seeds in the soil coincided with the highest cover figures, as expected. This was not the faet for the other two dominants. The explanation may be that they flower and set fruit best in ihe Juticus gerardii belt although they are less abundant there (Jerling unpubl,). The highest production of seeds dominating the seed bank thus coincided with their highest number in the soil - which was to be expected. Ihe sample-mean peaks in deeper layers on 6, 12, 18 and 24 m eould. although they are not significant, either be a remainder from the time when the water level was found higher up (i.e. contributing to seed production area), or be a result of downward movement of seeds in the soil. The depth distribution data eould be interpreted as if all depth layers, including the deeper ones, contained an equal number of seeds, with the two exceptions in the 42 and 48-m points where the second layer contained more. Downward movement of small seeds, in loose textured soils with percolating water is reported by Harper (1977). In the study area the soil, which was moist, packed and did not show any cracks or holes, contained between 40 and 60% of clay (Walientinus 1970). Percolation of water and transportation of seeds probably were slow. This was also shown by the fact that puddles were formed readily during rain and persisted for days. In addition, the area around 60 m, usually had a ground water level below soil surface only in March, April, May and June and of those months ground was usually frozen in March and early April, while May and June had greater evaporation than precipitation. (For meteorological data see Walientinus 1970). Still there was a significant number of seeds down to 5 6 cm depth — (Fig. 2). Transportation of seeds down into the soil therefore did not seem very likely and buried seeds would mainly be a remainder from times when the shore HOLARCTIC ECOLOGY 6:2 (1983) line was found higher up. If so, their age would be at least a hundred years, because the border to the terrestrial meadow (i.e. 0 m on the transect) was estimated to be the shore-line about 1840 (Walientinus 1970, Ryberg 1971). A comparison between the present vegetation and the seed bank reveals that there usually was a great difference in abundance for any certain species. This has been noted in most seed bank investigations (Harper 1977). In this study Jimcus gerardii was strongly overrepresented in the seed bank while Festuca rubra, Agrostis sfolonifera and Plantago mariiima were underrepresented. Gtaux maritima was overrepresented in the seed bank at the 6-m point but underrepresented at the 36, 48, 54 and 60-m points. The latter observations suffer probably from systematic errors as the present vegetation was measured as percentage of cover. This eaused overestimation of herbs over grasses. All herbs exeept Plantago maritima and Glaux marifima were present in low abundance in the vegetation. Seeds of Plantago maritima were almost absent from the seed bank and the difference between representation in cover and seed bank was significant. Festuca rubra, Agrostis stolonifera and Plantago maritima produced seeds and seedlings readily. Plantago maritima seedlings can be as many as about 2500 m""^ (Jerling 1981). None of these three species has any innate or induced dormancy, although sometimes enforced (sensu Harper 1977), and seeds probably germinate the year after they are spread. The three speeies that dominated the seed bank i.e., Juncus gerardii, Triglochin maritimum and Gtaux maritima have all vegetative multiplication and small seeds with innate or induced dormancy. Seedlings of Juncus species are deseribed as having low survivorship in vegetation (Lazenby 1955), as have seedlings of Glaux maritima (Jerling unpubl.). The seeds of Juncus need light and a ripening period to germinate (Lazenby 1955). Germination of Glaux mariiima seeds is induced by fluctuating temperature and light (Binet 1965). Seeds of Triglochin maritimum germinate after treatment with cold water and germination is stimulated by light (Binet I960). Consequently, seeds of these seed bank dominants seem to be more or less dependent on bare ground to germinate and to survive as seedlings. These speeies are halophytes and almost restricted to sea shores. Such habitats often have low predictability for flooding, ice erosion etc. (i.e. inducing density independent mortality). The dormancy of these seeds and their accumulation in the seed bank give these species a better ability to reeover after such disturbances, although the advantage of this character depends on the frequency of disturbances. The chance of a seed in the soil to become mobilized is lower the deeper it is buried. However, how much the chance is decreased depends on what agents are acting. In this area sea shore meadow disturbances would probably include (1) flooding and consequent erosion. but also (2) ice-erosion, both acting around the mean water level where the seed bank was most dense. To what depth these two agents act depends of course on the exposure of the shore. On a shore meadow which develops in sheltered areas the effect will probably not extend deeper than a fen centimetres. Another probable disturbance is (3) trampling, mostly by cattle but also by deer, elk etc. The trampling by eattle is not uniform throughout the gradient but heavier in certain areas, due to selective grazing (Jerling and Andersson 1981). Around or below mean water level the abundance of Phragmites communis is high (this speeies is preferred as food in presummer) and trampling here is frequent at a time when the vegetation is not fully developed. Thus the sward is easily damaged and bare ground shown. This probably mobilizes seeds from deeper layers than does erosion. However, badgers digging for food (4), a not unfrequent occurrence on the meadow, mobilize seeds from still deeper layers. Acknowledgements - I wish to thank Gisela Elmgren and P. Torstensson for help with the laboratory work, also Joanne Rowley for criticism on the manuscript and to Margareta von Lampe for typewriting. The laboratory work was financed by the Swedish natural research council.

Journal

Ecography – Wiley

Published: Apr 1, 1983

There are no references for this article.