CC BY
7
Вестник Томского государственного университета. Биология. 2019. № 46. С. 48-63
БОТАНИКА
UDC 581.524.2(470.56) doi: 10.17223/19988591/46/3
Olga I. Sumina, Elena M. Koptseva
Saint-Petersburg State University, Saint-Petersburg, Russian Federation
Seed distribution drivers at an early stage of vegetation development in a sand quarry
Distribution of viable seeds on the bare ground slope in a sand quarry located in the vicinity of Saint-Petersburg was studied in relation to the relief of the ground surface, species composition of pioneer communities and surrounding vegetation, as well as the timing of a growing season. Seeds were found even on bare ground and poorly vegetated sites. Ground contains three times more seeds in autumn than in early summer. Most of seeds belonged to perennial grasses (mainly anemochorous, apophytes, and weeds), with a few coming from other life-forms (from annuals to trees), which were spread in the quarry or in the surrounding vegetation. Seeds concentrated on the slope, their number was the lowest in the quarry bottom in both observation periods, since conditions favored to plant germination there. The microrelief of the ground surface can also influence seed distribution: in June, viable seeds mostly located in depressions. We suppose that in dissemination period, seeds of anemochorous plants concentrate on hillocks which trap them; however, more data is needed to conclusively prove this hypothesis.
The paper contains 3 Tables, 3 Figures and 49 References.
Keywords: seed bank; natural recovery; pioneer vegetation; viable seeds; seasonal dynamics; relief; safe sites.
Introduction
Seed bank, consisting of all viable seeds present in soil [1, 2] is not only a "common characteristic of many plant species, which allow storage of genetic diversity" [3:29] or "plant regeneration strategy, buffering environmental variation to allow species persistence" [4:201], but also one of the mechanisms of plant communities' self-regulation [5] and the source of their hidden biodiversity.
Seed banks have been studied extensively since the 19-th century (the first publication in 1825 belongs to Dureau de la Malle [6]), and the obtained results showed that the role of the below-ground plant community is much more significant than just the above-ground part we observe. Most research focused on comparison of plant community and seed bank species composition or assessment of its quantifying changes, but "the processes that govern assembly of the soil seed
bank following severe habitat disturbance are poorly understood" [7:58]. Little is known about details of seed bank formation process during the early plant community development. E. Chaideftou and co-authors [8] noted that especially little is known about the seed bank inter-annual temporal turnover, and they studied the seed bank composition in permanent plots from year to year. However, the seed bank dynamics during a single year period also has evident ecological importance and must be investigated.
Nowadays, large areas of land are impacted by mechanical disturbances, which destroy their soil and vegetation covers completely. The natural recovery of these bare grounds and the set of colonist-species depend on local flora resources [9-12]. Most of pioneer species are anemochorous [13], so the seed bank should form via "seed rain". Seeds inflow and outflow from the soil bank (via germination, death, water erosion, etc.) happen constantly, so the seed bank is a very dynamic unit.
The species diversity of pioneer communities does not only depend on the success of seed distribution, but also on plant germination, growth, and endurance [14]. The reserve of propagules is always more than the number of seedlings and thriving plants, and the quality of substrate plays an important role in the safety of pioneer species [15]. For example, the high ground density causes a reduction in germination [16] or lack of moisture leads to the death of seedlings [17]. It is well known that natural vegetation recovery is strongly influenced by substrate properties [18, 19]. Open grounds of disturbed sites are unfavorable to plant sprouts because of intensive insolation, extreme fluctuations of soil temperatures, rapid drying of substrate surface, etc. [20]. Plant individuals distribution depends on micro-topography of the ground surface [21, 22], even small depressions contributing to seed accumulation [23, 24]. Seeds gathering in depressions keep their viability better [25], thanks to accumulation of fine material and snow, leading to better moistening. The speed of colonization and pioneer species richness depend on proportion of fine (clay) fractions in the substrate [12, 26-28], due to the fact that the high content of these fractions increases water holding capacity of soil [23]. At the initial stage of primary succession, moistening plays an extremely important role, since plants suffer from nutrient deficit [26, 29], so the colonization process always depends on ground moisture [30, 31].
The territory disturbed by bulldozing or other techniques is often a patchwork of microhabitats (or "microsites"), which differ sharply in their properties (temperature, moisture, nutrients, etc.), including their suitability for settlement, survival, and growth of plants [1, 32]. The favorable microhabitats are referred to as "safe sites" [1]. Natural vegetation recovery is controlled by various abiotic factors, which can be unfavorable to plants (dryness, high temperature of ground surface, excessive solar lighting, strong wind, etc.), and as a result, species are able to survive within safe sites only [24].
The trend of regenerative succession depends on habitat local conditions, according to one point of view [18, 29, 33], or on plant diasporas' distribution at the
beginning of colonization, according to another [15, 34]. One of the interpretations of the Egler's Initial Floristic Composition model [35] suggests that seeds from all seral stages are present in the initial vegetation [36], but another scenario (gradual seed bank formation) is possible, as well.
The aim of our research is to receive data on seed distribution on the disturbed territory under influence of the surface relief, the season of the vegetation period and plant cover species composition. The study used the example of a sand quarry, because such quarries are widespread in different regions, which gives an opportunity to compare the obtained results in the future.
Materials and methods
The study was carried out in 2014, in the vicinity of Saint-Petersburg, Russia (southern taiga subzone; 60°05'30'N, 30°26'52"E). A small sand quarry (about 7500 m2) was located on a hill, previously occupied by a grass birch forest. The top of the hill and its eastern slope were completely destroyed by quarrying. In spite of sand extraction ceasing in 2006, even 8 years later there were sites of open ground within the quarry. In 2015, sand extraction was restarted, and we had no possibility to continue our study, so all obtained results were collected during one year.
The data were collected on a sandy slope (eastern exposition, approximately 30-35°, the length about 25 m) covered by patchy vegetation, so the total plant cover varied from 0 to 20% in different places. Adjacent vegetation consisted of a fragment of the birch forest, meadows with grasses and weeds, and communities dominated by Chamaenerion angustifolium (L.) Scop. or Calamagrostis epigeios (L.) Roth, occurring in spots at the quarry bottom.
We address the following questions: (1) Is the seed bank (consisting of viable seeds) present at disturbed sites with bare ground and poorly developed pioneer vegetation? (2) What is the difference between the seed bank composition in the beginning and the end of a growing season? (3) How is the seed bank composition influenced by the relief of the ground surface and the surrounding vegetation?
Ground samples were taken by core with cross section of 4^4 cm, immersed to the depth of 5 cm [37], their number was determined in accordance with the published recommendations [38] (See below). The samples were placed in craft-paper bags and kept in a dark dry place. The plan of data collection allowed us to assess such factors affecting the seed bank as:
Timing of the growing season. To reveal the dynamics of seeds in the ground, we collected data when most of viable seeds germinated at the end of June (28.06.2014), and at the end of the growing season (7.09.2014), when most of new seeds had already matured.
Location. Ground samples were collected from 4 positions in the relief: top of the slope, its middle and foot, and the bottom of the quarry (Fig. 1).
Fig. 1. Sampling scheme: 4 positions in the relief of the quarry: top of the slope (T), its middle (M) and foot (F), and the bottom of the quarry (B). At each position, 3 depressions and 3 hillocks were sampled
Microrelief of ground surface. The surface of sand was not smooth, various hillocks, small ridges as well as dimples and grooves formed an irregular microrelief. Vertical and horizontal sizes of these elements varied from 10 to 40 cm. In each of the aforementioned positions on the slope and quarry bottom, samples were collected from 3 elevated elements of the microrelief (named as "hillocks") and 3 from lower ones ("depressions"). Thus, 24 samples were studied in each from two observation periods. Therefore, the total number of samples was 48.
Vegetation of the slope. Each ground sample was collected inside of 30^30 cm plot, after all plant species were recorded. On 24 plots in June and another 24 plots in September, the total list of species, total plant cover and each species percentage cover estimations were made. As it was mentioned above, study plots were established in correlation with elements of the microrelief, and one of each three plots situated on hillocks or depressions was always located on a bare substrate.
Surrounding vegetation. The total list of species in the surrounding plant communities was made in 2014, also we used the floristic data collected in 20122013. Botanical nomenclature for plants is given by SK Cherepanov [39].
The protocol of sample treatment was established on the basis of the published recommendations [40, 41], and the results of our preliminary experimental selection of optimal methods for seed detection and their viability assessment. Each soil sample was stirred diligently and divided into 2 parts. Later, all manipulations were made sequentially with both of them. The ground was sifted through a sieve with a cell diameter of 1mm, and the material remaining in the sieve was examined to find seeds. The sifted ground was split into 2 parts again, and each half was exposed to flotation. Detected seeds were dried at room temperature. A special standard seed collection along with special guides and an atlas [42, 43] to determine the species affiliation of seeds were used.
Without conditions favorable to germination, "dormant" seeds may stay in soil keeping viability for a very long time [3]. Various special manipulations, such as stratification, phytohormones, etc. are applied to force seeds to sprout, but even in this case, we cannot be sure that all viable seeds have germinated and are
accounted for. For this reasons, to assess seed viability we used the staining by 0.1% water solution of fuchsine [44]. Seeds were exposed to fuchsine for 1 hour. Viable seeds had uncolored germ and embryo roots or faintly painted spots on roots and cotyledons. Seeds with good colored germ or bright large spots on roots and cotyledons were unviable.
Statistical data processing was carried out by StatSoft STATISTICA 10. Distribution of viable seeds in the relief of the quarry was tested by the Kolmogorov-Smirnov criterion and analyzed using non-parametric tests. So, the Kruskal-Wallis Criterion is used for identification of differences between compared groups of data. Also, we used the Mann-Whitney test (U-test) allowing us to evaluate the performance differences between comparing groups in pairs for a more precise description of the observed trends. We adopted a significance level of p < 0.05.
Results of research
Vegetation in the quarry and its neighborhoods
Natural recovery in the quarry developed during 8 years. The densest vegetation occupied the bottom of the quarry, in the other part of disturbed territory plant cover was patchy and absent in some places. The list of vascular plants recorded in quarry in 2012-2014 includes 55 species: trees (Betula pendula Roth, Alnus incana (L.) Moench, Populus tremula L., Pinus sylvestris L., Sorbus aucuparia L.), shrub (Salix cinerea L.), and herbs; dwarf shrubs were absent. Trees were represented by seedlings or young growth.
In the surrounding communities with predominance of Chamaenerion angustifolium or Calamagrostis epigeios, 16 species of herbs were marked, and only two (Erigeron acris L., Campanula rotundifolia L.) supplemented the list obtained in the quarry. Obviously, the fragment of former birch forest differed highly from other communities in species composition, and 5 species more were added to the overall list: Rubus idaeus L., Arctostaphylos uva-ursi (L.) Spreng., Convallaria majalis L., Hypericum maculatum Crantz, and Pimpinella saxifraga L. Thus, in total, 62 species were noted in the quarry and the surrounding vegetation.
Total number of seeds
The total number of seeds collected in 24 samples in September was nearly double the number in June: 513 and 275 seeds, respectively. The percentage of viable seeds became higher by autumn, as well. In the beginning of September, the number of viable seeds revealed in the collected samples increased more than threefold: from 94 to 343. The number of dead seeds was almost unchanged (in June - 181, in September - 170).
Number of species
In relevés of 48 study plots, 20 species of vascular plants were noted. Seeds from 18 of them were found in soil samples. Moreover, 3 species were present as seeds only. Thus, the total number of species present on study plots as either
plants or seeds was 23 or 37% of total number of species registered in the quarry and the surrounding vegetation (Table 1). The majority of seeds (including viable and unviable) belonged to perennial grasses, but other life-forms, such as trees
(Betulapendula, Sorbus aucuparia), dwarf-shrubs (Arctostaphylos uva-ursi), and short-lived (annuals and biennials) herbs (Berteroa incana (L.) DC., Veronica verna L. et al.) were present as well.
Table 1
Frequency of plants and/or seeds (viable and unviable) on study plots, %
No Species June (24 plots) September (24 plots)
plants seeds plants seeds
1 Artemisia campestris 63 25 29 21
2 Calamagrostis epigeois 46 17 42 54
3 Festuca rubra 17 58* 33 50
4 Rumex acetosella 21 38 17 17
5 Hieracium pilosella 13 17 21 8
6 Betula pendula 4 75 4 83
7 Berteroa incana 4 42* 0 33*
8 Chamaenerion angustifolium 8 17* 0 13
9 Tussilago farfara 17 0 13 4*
10 Achillea millefolium 13 0 8 8
11 Cirsium arvense 0 17 8 13
12 Erigeron acris 4 0 17 25
13 Artemisia vulgaris 4 0 13 0
14 Linaria vulgaris 17 0 0 0
15 Veronica verna 4 0 0 8
Noted on 1 plot
16 Sorbus aucuparia 4 4 0 0
17 Hippophae rhamnoides 4 0 0 0
18 Knautia arvensis 4 0 0 0
19 Taraxacum officinale 4 0 0 4
20 Arctostaphylos uva-ursi 0 4 0 0
21 Convallaria majalis 0 0 0 4
22 Phleum pratense 0 0 0 4
23 Agrostis tenuis 0 0 4 0
Total number of species 20 18
Seeds and plants 9 10
Plants only 9 - 2 -
Seeds only - 2 - 6
Note: * Only unviable seeds were found.
Vegetation on study plots
On 32 plots where vascular plants were noted (unlike other 16 plots located on bare sand), the total plant cover varied from 5 to 20% in June, and from 1 to 10% in September. Artemisia campestris L., Calamagrostis epigeois, Festuca rubra L., Rumex acetosella L., and Hieracium pilosella L. were the most frequent in plots, but only two of them (field sagewort and chee reedgrass) were the
most widespread all over the slope. Other plants distribution can be described as irregular mosaic. There were no real dominants between species. The maximal abundance (10%) was registered very rarely and for plants with large leaves or rosettes.
There was no distinction in species composition between two elements of the microrelief. No species was exclusive to only hillocks or depressions. On hillocks the number of species per plot (excluding plots on bare ground) varied from 1-2 to 5 in both observation periods. In depressions, this parameter changed from 2-6 species in June to 3-4 species in September. The mode is equal for hillocks and depressions: 4 species in June, and 3 in September. The average number of species (3.5) was identical for different elements of the microrelief and months. The average total cover on plots on both hillocks and depressions was 12-13% in June and 7-8% in September.
Plant and seed composition on study plots
Five aforementioned species had the highest frequency in above-ground vegetation, and their seeds were frequently found in soil samples as well, but the maximum frequency (about 80%) was noted in seeds of Betulapendula, despite the fact that the young growth of birch was only observed on 2 of 48 plots. The plants of Berteroa incana were registered on 1 plot in June, and their unviable seeds were contained in 30-40% of samples in both observation periods. Viable seed absence of short-lived weeds (from Brassicaceae and Chenopodiacea, for example) is caused by 8 years of vegetation development: in 2014, pioneer species were rare in the quarry unlike grass communities belonging to the next stage of succession.
Seeds of 5 species (Artemisia vulgaris L., Linaria vulgaris Mill., Hippophaë rhamnoides L., Knautia arvensis (L.) J.M. Coult., Agrostis tenuis Sibth.) which had been noted in relevés were not found in soil samples. 3 species (Phleum pratense L., Convallaria majalis, Arctostaphylos uva-ursi) were present only as a single seed, and their plants were very rare in the quarry outside the plots. It should be emphasized that Convallaria majalis and Arctostaphylos uva-ursi are typical species of forest communities, and Phleum pratense is one of a few grass species (Poaceae), which are not pioneer plants on disturbed lands.
About a half of all species noted in relevés was found as seeds in ground samples in both observation periods. In June, seeds of 9 species were absent, and 2 species were present only as seeds. In September, the situation reversed (Table 1).
The simultaneous presence of plants and seeds on the same plot was revealed for Calamagrostis epigeios and Festuca rubra (in both observation periods), and Erigeron acris (in September). The number of such plots was low and varied from 2 to 6 for the mentioned species.
Even on bare ground plots, the seeds of 2-4 species were present, and only twice on 1 plot in June, and on 1 plot in September none was found, and the both of the empty plots were located at the bottom of the quarry.
Viable seeds
In September, the total number of seeds increased, together with the number of viable ones, and practically all species had a relatively larger proportion of viable seeds than in June (Table 2).
Table 2
Proportion of viable seeds of different species in soil samples, %
No Species June (24 samples) September (24 samples)
1 Betula pendula 47 65
2 Festuca rubra 0 77
3 Artemisia campestris 43 60
4 Rumex acetosella 82 100
5 Calamagrostis epigeios 0 72
6 Hieracium pilosella 13 67
7 Chamaenerion angustifolium 0 100
8 Cirsium arvense (100) (100)
9 Arctostaphylos uva-ursi (100) 0
10 Sorbus aucuparia (100) 0
11 Erigeron acris 0 80
12 Veronica verna 0 100
13 Phleum pratense 0 (100)
14 Achillea millefolium 0 100
15 Taraxacum officinale 0 (100)
16 Convallaria majalis 0 (100)
Note: Data are shown in parentheses, if the number of seeds per 1 samp e is less than 4.
Only 16 out of 23 species noted on study plots contributed viable seeds to the seed bank. Arctostaphylos uva-ursi, Convallaria majalis, Sorbus aucuparia were represented by a single seed only, and Phleumpratense by 4 seeds. Such weeds as Taraxacum officinale F.H. Wigg. and Cirsium arvense (L.) Scop. also had a very low number (1-4) of viable seeds in the soil bank due to their high germination capacity. No viable seed of Berteroa incana and Tussilago farfara L. was found. The first species is a pioneer annual plant, which was in the quarry in 2012, but it became very rare in 2014: we marked unique specimen on 1 plot in June. The second one mainly follows the vegetative form of reproduction, and its seeds stay viable for less than one year [45, 46]. Thus, the core of the seed bank consists of seeds of 10 species (Table 2).
The viable seeds of many species (Festuca rubra, Calamagrostis epigeios, Chamaenerion angustifolium, Erigeron acris et al.) were noted in autumn, only. We can assume that all viable seeds of these species germinated in June. The main stock of seeds was formed by Betula pendula, Calamagrostis epigeios, Artemisia campestris, Festuca rubra, Rumex acetosella, which are typical colonizers of disturbed lands. In the seed bank, the proportion of viable seeds of the aforementioned plants reached 93% in June and 86% in September. These high values were determined by a large number of birch seeds (62 and 65% of the seed bank, respectively). Adjusting for 30^30 cm plot size, the maximal number
of viable seeds of Betulapendula was more than 500 in June and about 3500 in September.
Viable seeds distribution in the relief
Viable seeds of Betula pendula and Festuca rubra were widespread along the slope and at the bottom of the quarry (Table 3). Seeds of 4 species (Artemisia campestris, Calamagrostis epigeios, Cirsium arvense, Rumex acetosella) were noted in 3 out of 4 positions in the relief, and most of them had a high frequency in the vegetation of the study plots. Viable seeds of other species demonstrated more a narrow distribution in 1-2 positions.
Table 3
Seed bank (viable seeds) distribution in connection with the position in the relief
No Species Bottom of quarry Foot of slope Middle part of slope Top of slope
June September June September June September June September
1 Betula pendula 225 394 1350 1406 900 5794 788 4894
2 Festuca rubra 0 113 0 281 0 338 0 1350
3 Artemisia campestris 113 225 0 225 731 225 0 0
4 Rumex acetosella 0 0 338 56 225 0 225 281
5 Cirsium arvense 0 0 113 56 0 169 113 0
6 Calamagrostis epigeios 0 113 0 0 0 675 0 225
7 Chamaenerion angustifolium 0 0 0 0 0 56 0 281
8 Erigeron acris 0 0 0 0 0 731 0 169
9 Veronica verna 0 0 0 0 0 113 0 169
10 Achillea millefolium 0 0 0 0 0 506 0 0
11 Taraxacum campylodes 0 0 0 0 0 56 0 0
12 Convallaria majalis 0 0 0 0 0 56 0 0
13 Hieracium pilosella 0 0 0 0 0 0 56 113
14 Arctostaphylos uva-ursi 0 0 0 0 0 0 56 0
15 Sorbus aucuparia 0 0 0 0 0 0 56 0
16 Phleum pratense 0 225 0 0 0 0 0 0
Total number of seeds 338 1 070 1 801 2 024 1 856 8 719 1 294 7 482
Number of species 2 5 3 5 3 11 6 8
Note: Number of seeds is given as recalculation on 3Qx30cm plots.
Full data from collected samples on the distribution of seeds according to locations in the relief and microrelief are shown in Figure 2.
Fig. 2. Seasonal distribution of viable and unviable seeds between elements of the microrelief across the slope profile. Bar heights are numbers of seeds; grey parts of bars are numbers of viable seeds. H - Hillocks, D - Depressions
In general, the position factor has an impact on the distribution of viable seeds (Kruskal-Wallis test: p = 0.0405). The number of seeds at the quarry bottom was the lowest in both observation periods. However, statistically significant differences between the bottom and other slope positions are revealed in the pairwise comparison by means of the Mann-Whitney test and
only in June (p
= 0.0259; p
bottom/middle
= 0.0283; p
bottom/top
= 0.0393). In
bottom/foot
September, there were no differences (p = 0.1127) in the distribution of seeds on the relief because the number of seeds in samples varied greatly (Fig. 3).
Foot Middle
Position
□ Median □ 25%-75% I Non-Outlier Range
Bottom
Fig. 3. Viable seeds distribution on slope and quarry bottom in two observation periods. The first and third quartiles (25th and 75th percentiles, respectively), the median (50th percentile) and the edges of the statistically significant sample are shown in the Box Plot
Influence of the microrelief
Analysis of the data by distribution of viable seeds on the microrelief showed that the difference between hillocks and depressions is significant in June, only (U-test p = 0.0278). The samples collected in September revealed no influence of the microrelief.
Discussion of research
23 species of vascular plants were recorded on study plots, and only 10 of them (Betula pendula, Festuca rubra, Artemisia campestris, Rumex acetosella, Calamagrostis epigeios, Hieracium pilosella, Chamaenerion angustifolium, Erigeron acris, Veronica verna, and Achillea millefolium L.) had viable seeds in soil. Among these vasculars forming the seed bank there were 1 deciduous tree and 2 annual-biennial species, but most of them were perennial herbs, which are mostly typical colonizers of open ground, sandy dunes and other habitats with infertile soil. Many of them are considered weeds, and they have biological traits (large number of seeds, light seeds, early germination etc.) that facilitate rapid colonization and survival in such habitats.
The absence of viable seeds of the annual weed Berteroa incana in soil is an example of started changes in the seed bank composition - a decrease in the number of therophytes - observed in plant communities following restoration [47, 48]. The presence of species belonging to different life-forms in the seed bank lend credibility to the suggestion that the Egler's Initial Floristic Composition model [36] works.
Among 10 aforementioned species contributing to the seed bank on study plots, only 5 species are in the majority, with Betula pendula being the most
prominent. The influence of vegetation surrounding the quarry also appeared as the seed bank enrichment by seeds of some species (forest plants Convallaria majalis and Arctostaphylos uva-ursi, for example). Even the low number of such seeds can be effective if their inflow to quarry is annual.
The conclusion of M. Kalin with coauthors [49] that compositional similarity and spatial matching between the seed bank and the standing vegetation is not high was confirmed by our data. The presence of any species in the above-ground plant community was not directly connected to its stock in the seed bank. 4 species (Artemisia campestris, Calamagrostis epigeios, Festuca rubra, and Rumex
acetosella) had relatively high abundance on study plots, but the overwhelming majority of the seed bank (more than 60%) belonged to Betula pendula, whose young growths were observed on 2 plots out of 48.
The fact that the majority of the seed bank is formed by birch, despite this species being present in the quarry only in unfruitful young growth birches, can be an indirect confirmation of the Egler's Initial Floristic Composition model.
The seed bank is influenced by many stochastic processes, habitat configuration, land use etc. [7]. The relief of the quarry influences seed distribution. In September, the greatest number of viable seeds was found on the top and in the middle part of the slope. It is easy to explain, because during the dissemination period, seeds of anemochorous plants land at first on the elevated terrain. Later, they can move gradually down the slope under the influence of rain, wind, crumbling sand, and snow. A priori assumption that seeds will always concentrate in the lowest part of the quarry was not confirmed: the minimum number of seeds was marked at the bottom of the quarry in both observation periods. There are the most favorable conditions for plants in this lowest part of the quarry [12, 25], therefore seed germination is very active there.
Different factors (rain, wind, erosion etc.) cause seed redistribution and change viable seed pattern, which has no direct reflection in the above-ground vegetation. The influence of the microrelief was statistically significant in one of two observation periods.
The suggestion that in September seeds of anemochorous species primarily filled the hillocks and afterwards moved to depressions seems most likely, and we hypothesize that in dissemination period, seeds of anemochorous plants concentrate not at "safe sites" (depressions), but on hillocks which trap them. In the future, more data collection would be necessary to prove this hypothesis with a statistically significant result. In June, the seed bank was significantly more present in depressions.
The obtained data do not completely confirm the idea that species can survive only within "safe sites" [1, 24]. Although in June, viable seeds concentrate in depressions, we observed neither germination as active as that at the bottom of the quarry, nor a difference between poor vegetation in either element of the microrelief.
Conclusion
Our results suggest that in the sand quarry the seed bank as a pool of viable seeds exists even at sites with bare ground or less developed plant cover. The bank includes seeds of species, which are spread in the quarry and the surrounding vegetation, and consists of seeds belonging to plants of different life-forms. The seed bank has temporal and spatial dynamics during a single year period. The number of seeds varied greatly in the beginning and the end of summer. The seed bank also changed in terms of its species diversity during the growing season. It also demonstrated heterogeneity in seed distribution with regards to the quarry relief and microrelief. A "paradoxical" result is that at the bottom of the quarry the number of viable seeds is always minimal (because of favorable conditions for germination there). Our results do not contradict the Egler's Initial Floristic Composition model, but do not support the hypothesis of plants survival at "safe sites".
The seed bank comprising viable seeds with different time and strategy of development is a complex composition. Obviously, in the field research we could not consider all its specific and diverse features. Our goal, to reveal the influence of the relief and season on seed distribution, was achieved, and the obtained data show how complex in space and time the seed bank is, as well as allow us to clarify some previously published information. We believe that our research of the seed bank short-term dynamics adds new data to study seed redistribution patterns.
Acknowledgments: The Authors thank their students Ya Dmitrakova and Yu Fedorova (Saint-Petersburg State University, Saint-Petersburg, Russia) for help during the field period, Dr. PJ Webber, Emeritus Professor (Michigan State University, East Lansing, state of Michigan, USA) for valuable critical comments on our manuscript during its preparation.
References
1. Harper JL. Population biology of plants. London, New York: Acad. Press; 1977, 892 p.
2. Grubb PJ. The uncoupling of disturbance and recruitment, two kinds of seed banks and
persistence of plant populations. Ann. Zool. Fenn. 1988;25(1):23-26.
3. Koopmann B, Müller J, Tellier A, Zivkovic D. Fisher-Wright model with deterministic seed
bank and selection. Theor. Popul. Biol. 2017;114:29-39. doi: 10.1016/j.tpb.2016.11.005
4. Plue J, Colas F, Auffret AG, Cousins SAO, Bekker R. Methodological bias in the seed bank
flora holds significant implications for understanding seed bank community functions. Plant Biol. 2017;19(2):201-210. doi: 10.1111/plb.12516
5. Bossuyt B, Honnay O. Can the seed bank be used for ecological restoration? An overview
of seed bank characteristics in European communities. J. Veg. Sci. 2008;19:875-884. doi: 10.3170/2008-8-18462
6. Dureau de la Malle, AJCA. Mémoire sur l'alternance ou sur ce problème: la succession
alternative dans la reproduction des espèces végétales vivant en sociéte est-elle une loi générale de la nature. Annales des sciences naturelles.1825;5:353-381. In France
7. Helsen K, Hermy M, Honnay O. Changes in the species and functional trait composition of
the seed bank during semi-natural grassland assembly: seed bank disassembly or ecological palimpsest? J. Veg. Sci. 2015;26(1):58-67. doi: 10.1111/jvs.12210
8. Chaideftou E, Kallimanis AS, Bergmeier E, Dimopoulos P. How does plant species composition change from year to year? A case study from the herbaceous layer of a submediterranean oak woodland. Community Ecol. 2012;13(1):88-96. doi: 10.1556/ comec.13.2012.1.11
9. Hoekstra JM, Boucher TM, Ricketts TH, Roberts C. Confronting a biome crisis: global
disparities of habitat loss and protection. Ecol. Lett. 2005;8:23-29. doi: 10.1111/j.1461-0248.2004.00686.x
10. Rehounkova K, Prach K. Spontaneous vegetation succession in gravel-sand pits: a potential for restoration. Restor. Ecol. 2008;16(2):305-312. doi: 10.1111/j.1654-1103.2006.tb02482.x
11. Sumina OI. Sravneniye floristicheskogo sostava rastitel'nosti kar'yerov, raspolozhennykh v raznykh rayonakh Kraynego Severa Rossii [Comparison of the floristic composition of the vegetation of quarries located in different regions of the Far North of Russia]. Botanicheskiy Zhurnal = Botanical Journal. 2010;95(3):368-380. In Russian
12. Sumina OI. Formirovanie rastitel'nosti na tekhnogennykh mestoobitaniyakh Kraynego Severa Rossii [Development of vegetation in technogenic habitats of the Russian Far North]. Saint-Petersburg: Inform-Navigator Publ.; 2013. 340 p. In Russian
13. Pakeman RJ. Consistency of plant species and trait responses to grazing along a productivity gradient: a multi-site analysis. Ecology. 2004;92:893-905. doi: 10.1111/j.0022-0477.2004.00928.x
14. van der Valk AG. Establishment, colonization and persistence. In: Plant succession. Theory and prediction. Glenn-Lewin DC, editor. London, Glasgow: New York Chapman and Hall; 1992. pp. 60-102.
15. Walker LR. Patterns and processes in primary succession. In: Ecosystems of Disturbed Ground. Ecosystems of the World. Vol. 16. Walker LR, editor. Amsterdam, New York, Oxford: Elsevier Publ.; 1999. pp. 585-610.
16. Sheldon JC. Behavior of seeds in soil. 3. Influence of seed morphology and behavior of seedlings on establishment of plants from surface-lying seeds. J. Ecol. 1974;62(1):47-66.
17. Chapin DM, Bliss LC. Seedling growth, physiology, and survivorship in a subalpine, volcanic environment. Ecology. 1989;70(5):1325-1334.
18. Matthews JA. The ecology of recently-deglaciated terrain. A geological approach to glacier forelands and primary succession. Cambridge, New York: Cambridge Univ. Press; 1992. 386 p.
19. Kapelkina LP, Sumina OI, Lavrinenko IA, Lavrineneko OV, Tikhmenev EA, Mironova SI. Samozarastaniye narushennykh zemel' Severa [Natural revegetation on disturbed lands of the North]. St. Petersburg: VVM Press; 2014. 204 p. In Russian
20. Chapin FS. III. Physiological controls over plant establishment in primary succession. In: Primary succession on land. Miles J and Walson DWH, editors. Oxford, London: Blackwell Scientific Publications; 1993. pp. 161-178.
21. Matthews JA, Whittaker RJ. Vegetation succession on the Storbreen glacier foreland, Jotunheimen, Norway. Arct. Alp. Res. 1987;19(4):385-395.
22. Chambers JC, MacMahon JA, Brown RW. Alpine seedling establishment: the influence of disturbance type. Ecology. 1990;71(4):1323-1341.
23. Jumpponen A, Vare H, Mattson KG, Ohtonen R, Trappe JM. Characterization of "safe sites" for pioneers in primary succession on recently deglaciated terrain. J. Ecol. 1999;87:98-105.
24. Walker LR, Bellingham PJ, Peltzer DA. Plant characteristics are poor predictors of microsite colonization during the first two years of primary succession. J. Veg. Sci. 2006;17(3):397-406. doi: 10.1111/j.1654-1103.2006.tb02460.x
25. Enright NJ, Lamont BB. Seed banks, fire season, safe sites and seedling recruitment in five co-occurring Banksia species. J. Ecol. 1989;77:1111-1112.
26. Borgegard SO. Vegetation development in abandoned gravel pits: effects for surrounding vegetation, substrate, and region. J. Veg. Sci. 1990;1:675-682.
27. Harper KA, Kershaw GP. Soil characteristics of 48-year-old borrow pits and vehicle tracks in shrub tundra along the canol No.1 Pipeline corridor, Northwest Territories, Canada. Arct. Alp. Res. 1997;29:105-111.
28.CannoneN,GerdolR.Vegetation as anecologicalindicatorofsurfaceinstabilityinRockGlaciers. Arct, Antarct., Alp. Res. 2003;35(3):384-390. doi: 10.1657/1523-0430(2003)035[0384:VA AEIO]2.0.CO;2
29. Grubb PJ. The ecology of establishment. In: Ecology and design in landscape. Bradshaw AD, Goode DA and Thorp EHS, editors. Oxford: Blackwell Sci. Publ.; 1986. pp. 83-98.
30. Komarkova V, Wielgolaski FE. Stress and disturbance in cold region ecosystems. In: Ecosystems of disturbed ground. Ecosystems of the World. Vol. 16. Walker LR, editor. Amsterdam, New York, Oxford: Elsevier Publ.; 1999. pp. 39-122.
31. Jorgenson MT, Kidd JG, Carter TC, Bishop S, Racine CH. Long-term evaluation of methods for rehabilitation of lands disturbed by industrial development in the Arctic. In: Social and environmental impacts in the North. NATO Science Series (Series: IV: Earth and Environmental Sciences). Vol. 31. Rasmussen RO and Koroleva NE, editors. NATO Science Series (Series: IV: Earth and Environmental Sciences). Dordrecht, Boston, London: Kluwer Acad. Publ.; 2003. pp. 173-190. doi: 10.1007/978-94-007-1054-2_13
32. Reader R J, Buck J. Topographic variation in the abundance of Hieracium floribundum: relative importance of differential seed dispersal, seedling establishment, plant-survival and reproduction. J. Ecol. 1986;74(3):815-822.
33. Grubb PJ. Some generalizing ideas about colonization and succession in green plants and fungi. In: Colonization, succession and stability. Gray AJ, Crawley MJ and Edwards PJ, editors. Oxford: Blackwell Sci. Publ.; 1987. pp. 81-102.
34. Martineau Y, Saugier B. A process-based model of old field succession linking ecosystem and community ecology. Ecol. Model. 2007;204(3-4):399-419. doi: 10.1016/j. ecolmodel.2007.01.023
35. Egler FE. Vegetation science concepts I. Initial floristic composition, a factor in old-field vegetation development with 2 figs. Vegetatio. 1954;4(6):412-417.
36. Wilson JB. Does the Initial Floristic Composition model of succession really work? J. Veg. Sci. 2014;25(1):4-5.
37. Forcella F, Webster T, Cardina J. Protocols for weed seed bank determination in agro-ecosystems. In: Weed management for developing countries. Addendum 1. Labrada R, editor. FAO Plant Production and Protection Paper. 2003;120(Add.1):3-18.
38. Dessaint F, Barralis G, Caixinhas ML, Major J-P, Recasens J, Zanin G. Precision of soil seedbank sampling: How many soil cores? Weed Research. 1996;36:143-151.
39. Cherepanov SK. Sosudistye rasteniya Rossii i sopredel'nyh gosudarstv (v predelah byvshego SSSR) [Vascular plants of Russia and adjacent states (the former USSR)]. St. Petersburg: Mir & Sem'ya-95 Publ.; 1995. 991 p. In Russian
40 Kemeny G, Nagy Z, Tuba Z. Application of nested samples to study the soil seed bank in semiarid sandy grassland. Acta Bot. Hung. 2003;45(1-2):127-137.
41. Gonzalez S, Ghermandi L. Comparison of methods to estimate soil seed banks: The role of seed size and mass. Community Ecol. 2012;13(2):238-242. doi: 10.1556/ ComEc.13.2012.2.14
42. Maysuryan NA, Atabekova AI. Opredelitel semyan i plodov sornykh rasteniy [Guide of seeds and fruits of weeds]. Moscow: Kolos Publ.; 1978. 288 p. In Russian
43. Bojnansky V, Fargasova A. Atlas of seeds and fruits of Central and East-European flora. Netherlands: Springer Publ.; 2007. 1046 p. doi: 10.1007/s10933-009-9319-6
44. Mesgaran MB, Mashhadi HR, Zand E, Alizadeh HM. Comparison of three methodologies for efficient seed extraction in studies of soil weed seed banks. Weed Research. 2007;47(6):472-478.
45. Gubanov IA. Mat'-i-machekha obyknovennaya. Biologicheskaya flora Moskovskoy oblasti [Coltsfoot. Biological flora of Moscow Region. Vol. 1]. Moscow: MGU Publ.; 1974. pp. 169-181. In Russian
46. Sheptukhov VN, Gafurov RM, Papaskiri TV. Atlas osnovnykh vidov sornykh rasteniy Rossii [Atlas of the main species of weed plants in Russia]. Moscow: Kolos Publ.; 2009. 192 p. In Russian
47. Bossuyt B, Butaye J, Honney O. Seed bank composition of open and overgrown calcareous grassland soils - a case study from Southern Belgium. J. Environmental Management. 2006;79:364-371. doi: 10.1016/j.jenvman.2005.08.005
48. Fagan KC, Pywell RF, Bullock JM, Marrs RH. The seed banks of English lowland calcareous grasslands along restoration chronosequence. PlantEcol. 2010;208:199-211. doi: 10.1007/ s11258-009-9698-9
49. Mary Kalin T, Arroyo MTK, Cavieres LA, Castor C, Humana AM. Persistent soil seed bank and standing vegetation at a high alpine site in the central Chilean Andes. Oecol. 1999;119:126-132. doi: 10.1007/s004420050768
Received 11 January 2019; Revised 28 March 2019; Accepted 25 April 2019; Published 27 June 2019
Author info:
Sumina Olga I, Dr. Sci. (Biol.), Professor, Department of Vegetation Science and Plant Ecology, Saint-Petersburg State University, 7-9 Universitetskaya emb., Saint-Petersburg 199034, Russian Federation. ORCID iD:: https://orcid.org/0000-0001-6191-228X E-mail: o.sumina@spbu.ru
Koptseva Elena M, Cand. Sci. (Biol.), Assistant, Department of Vegetation Science and Plant Ecology, Saint-Petersburg State University, 7-9 Universitetskaya emb., Saint-Petersburg 199034, Russian Federation. ORCID iD: https://orcid.org/0000-0003-4387-9521 E-mail: e.koptseva@spbu.ru
For citation: Sumina OI, Koptseva EM. Seed distribution drivers at an early stage of vegetation development in a sand quarry. Vestnik Tomskogo gosudarstvennogo universiteta. Biologiya = Tomsk State University Journal of Biology. 2019;46:48-63. doi: 10.17223/19988591/46/3
