Research Article |
Corresponding author: Anett Endrédi ( endredi.anett@ecolres.hu ) Academic editor: Peter Poschlod
© 2023 Anett Endrédi, Ármin Sőth, Dóra Ércz, Balázs Deák, Orsolya Valkó, János György Nagy.
This is an open access article distributed under the terms of the Creative Commons Attribution License (CC BY 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Citation:
Endrédi A, Sőth Á, Ércz D, Deák B, Valkó O, Nagy JG (2023) Exploring life-history traits of an endangered plant (Vicia biennis L.) to support the conservation of marginal populations. Nature Conservation 54: 121-147. https://doi.org/10.3897/natureconservation.54.105606
|
We aimed to investigate the reproduction-related traits of Vicia biennis L., an endangered and poorly known wetland species in its western marginal populations (in Hungary), and discuss the conservational and ecological implications. We measured the mass, viability, and physical dormancy of half-year-old seeds in five in-situ collected seed lots, while potential seed longevity (i.e., seed bank type) was estimated from repeatedly germinating subsamples from four ex-situ collected seed lots for 3–8 years. Plant survival, flowering, and seed production were studied in different light-, irrigation-, and competition conditions in a botanical garden experiment. We found that 1) half-year-old seeds have a high germination capacity (78–100%), 2) and high level of physical dormancy (72–100%) in all examined Hungarian populations, and 3) the seeds can preserve their germination capacity for more than five years, although their viability sharply decreases, probably falling below 10% within ten years, when they are stored at room temperature. The results of the botanical garden experiment suggested that 1) the species is annual, not biennial; 2) it shows strong sensitivity to precipitation and low competitiveness for water; and 3) it can produce hundreds of seeds even in suboptimal (dry or shady) conditions. Although the species was found to be well-adapted to a temporally heterogeneous environment, its future vulnerability can increase depending on the duration of dry seasons. Further investigation of genetic diversity and soil seed bank is needed to estimate the actual vulnerability of the species while strengthening the populations through seed sowing, and additional vegetation control in the habitats is suggested.
Fabaceae, germination, physical dormancy, seed longevity, seed production
Recently, several studies have shown that human-induced loss, fragmentation, and degradation of natural habitats, together with accelerating climate change, are among the most important drivers of current and future biodiversity crisis (
An increasing number of models were developed to predict the responses of species and communities to future environmental changes in order to prioritize conservation actions (
Life history traits of rare plant species are often poorly known, and the relative importance of these species in maintaining ecosystem functioning has been underestimated for a long time (
For the effective in-situ and ex-situ conservation of rare species, it is essential to have a basic knowledge of life-history traits related to germination and seedling establishment that can considerably affect population growth and stability (
This study aimed to examine the life strategy of a poorly known, endangered plant species (Vicia biennis L.) and its reproduction-related traits in order to reveal potential sources of the species’ vulnerability and understand its current demography and distribution.
Vicia biennis L. is a wild legume species with a 1–3m long, branching and climbing herbaceous stem system. Originally it was described as a biennial (hemitherophyte) plant, which is preserved in its Latin name, but recently, this was questioned in the literature. Nowadays, it is occasionally referred to as annual (
Although the species has a wide distribution area (from Hungary to Kazakhstan), its known populations are often isolated and situated far from each other. They live near wetland habitats: edge of gallery forests, willow bushes, or reed-framed watersheds in the steppe zone. These habitat types are profoundly affected by land use, climatic changes, and invasive alien species (
The westernmost populations live in the Pannonian Biogeographic Region in Hungary, where the species is endangered and strictly protected by law (
Location of the ex-situ and in-situ populations in Hungary validated after 1990. Examined populations (Tiszaderzs, Püspökladány and Lakitelek) are highlighted. The map is based on the data of the Vascular Plants of Hungary online database (
The size of the found populations was highly variable, ranging from one specimen to more than 100 individuals. However, despite all conservational efforts, this shows a high within-population variation between years. Based on this and the vulnerability of wetland habitats, the species was suggested for ex-situ conservation by the Hungarian National Park Directorates (
Despite the species’ endangered conservation status, vulnerability, and its well-known crop (e.g., Vicia sativa) and weed (e.g., V. cracca) relatives, only morphological data and no functional trait data are available about V. biennis in the literature or public databases (e.g., SID (
In our research, we aimed to experimentally investigate some reproduction-related traits of the species in semi-natural conditions: Seed and germination-related traits (i.e., seed mass, potential seed longevity, dormancy, germination capability) were examined in greenhouse experiments, while plant survival, flowering, seed production and their reaction to different light exposure and precipitation were investigated in a botanical garden experiment. The main questions and hypothesis of the study were the followings:
A vast majority of species in the Fabaceae family produce seeds with some kind of dormancy – mostly physical (PY) or combinational (i.e., physical + physiological, PY+PD) dormancy (
Due to the limited number of seeds available from the endangered populations of V. biennis, we decided not to test all potential dormancy types but rather focus on the more important physical dormancy (PY). Based on the expected life form of the species (i.e., summer annual) and the seed characteristic of other Vicia species, high between-population variances in seed dormancy, weight, and viability were expected. As the examined in-situ populations are small and variable (containing ~10–50 individuals), detecting signs of inbreeding depression (e.g., low viability of seeds) was also conceivable.
The longevity of seeds (i.e., the time they can preserve their germination capacity) is a crucial seed trait affecting the survival of endangered, annual/biennial plant populations. Although this highly depends on the species’ attributes, environmental conditions affect it too (
Although we did not have the conditions (lab and a sufficient number of seeds) to conduct a proper longevity test, we wanted to estimate at least the potential seed bank type of the species by testing the viability of seeds with different ages in the short-term ex-situ seed collection. Based on the known ecological correlates, it is not clear what seed bank type to expect: E.g., species with similarly large seed sizes have a higher chance to have a short-lived, transient seed bank (
V. biennis often grows close to wetland habitats in the half-shaded edge of taller vegetation (e.g., bushes or cattails); thus, we hypothesized that the species has a relatively low light requirement and high water demand for healthy development and seed production.
Three in-situ and one ex-situ Hungarian population were selected as seed sources to study the seed traits of the species (Fig.
The Tiszaderzs population was found in 1999, near a canal connected to an oxbow of the River Tisza (Cserőközi-Holt-Tisza). At that time, the known habitats of the species were threatened by invasive weeds (e.g., Amorpha fruticosa, Vitis spp., Echinocystis lobata, Solidago gigantea) (
In the last decade, the forest habitat has become more shaded, and despite all conservational efforts (e.g., early spring weed control and some reintroduction attempts), the number of V. biennis individuals showed a significant decrease here. Meanwhile, the shift of the population towards the continuously disturbed (cut) edge of willow bushes was observed.
The Püspökladány population was found in 2009 in a ditch between fields. Here, the plant community was mainly composed of typical wetland species (e.g., Typha angustifolia, T. latifolia, Bolboschoenus maritimus s. l., Lycopus exaltatus, Butomus umbellatus) (
The Lakitelek population was recorded in 2012. A few dozen individuals occurred on a pond’s steep shore, and a similar number of plants were found nearby, on the embankment between artificial ponds (
The ex-situ population was established in 2009 in the Botanical Garden of MATE (Gödöllő). For the establishment, seeds were collected from the Tiszaderzs population in 2007 and germinated in 2009. The species’ preference for light and moist conditions was tested in the first year. The results and the conditions are presented in this paper. In the following years, the population was maintained partly by spontaneous germination and partly by additional seed sowing and seedling planting. All seeds/plants used for the sowing and plantation originated from the previous seed lots of the same ex-situ population. The number of seed-producing individuals in this population varied between 50 and 100, and about 50% of the mature seeds was collected each year.
In the germination tests, we used both in-situ and ex-situ collected seeds. In this paper, we will use the term “seed lot” for seeds collected from the same population in the same year, while “seed sample” refers to subsamples of seed lots for germination experiments or weighing.
Five in-situ seed lots were collected in four years (2013, 2014, 2015, 2017) from the three different populations (Fig.
We performed only one seed collection in a vegetation season to minimize the disturbance of these endangered populations. The Tiszaderzs and Lakitelek populations were sampled twice in two different years, while in the Püspökladány population, only one seed collection was performed in 2015. Samplings were performed in August, in the middle of the yielding period. Following the
Pods were opened, and only mature (round, pigmented) and intact seeds were selected for storage. In total, ~46–150 seeds were collected per population during one collection. Seeds from the same lot (collected in the same year and population) were mixed. Whereas previous observations suggested that the species is a spring germinator and a preliminary study did not show a significant effect of cold winter stratification (1 month in 4 °C) on germination (
As the size of the in-situ collected seed samples was not appropriate for repeating germination tests for years, seeds from the ex-situ population were used for the longevity experiment. For the longevity experiment, we used seeds collected from the population in four different years: 2009, 2010, 2012, and 2013 (seed lots 09EX, 10EX, 12EX, and 13EX, respectively). These ex-situ seed lots were stored under the same conditions as the in-situ seeds: in the dark, within paper bags, at room temperature (23 °C), and in ambient moist conditions.
Two series of germination experiments were conducted in greenhouse conditions to determine 1) the viability (estimated via germination capability) and physical dormancy of half-year-old seeds and 2) the average seed longevity (i.e., the time until the seeds can preserve their germination capacity). Experimental setups are shown in Fig.
Origin, treatments and number of seeds used in germination experiments. Note that seed age is rounded up as the seeds were collected in summer and germinated in spring.
Experiment | Year of collection | Source population | Seed sample | Seed age (yr) | Treatment | Sample size (n) |
---|---|---|---|---|---|---|
I. (viability & dormancy) | 2013 | Tiszaderzs | 13T | 1 | Control | 2*20 |
Scarified | 2*20 | |||||
2014 | Tiszaderzs | 14T | 1 | Control | 2*20 | |
Scarified | 2*20 | |||||
Lakitelek | 14L | 1 | Control | 23 | ||
Scarified | 23 | |||||
2015 | Püspökladány | 15P | 1 | Control | 3*20 | |
Scarified | 3*20 | |||||
2017 | Lakitelek | 17L | 1 | Control | 25 | |
Scarified | 25 | |||||
II. (longevity) | 2009 | Gödöllő (EX) | 09EX | 1 | Scarified | 3*20 |
2 | 3*22 | |||||
3 | 3*22 | |||||
7 | 3*20 | |||||
8 | 3*20 | |||||
2010 | Gödöllő (EX) | 10EX | 1 | Scarified | 3*22 | |
2 | 3*22 | |||||
6 | 3*20 | |||||
7 | 3*20 | |||||
2012 | Gödöllő (EX) | 12EX | 1 | Scarified | 3*20 | |
4 | 3*20 | |||||
5 | 3*20 | |||||
2013 | Gödöllő (EX) | 13EX | 1 | Scarified | 3*20 | |
3 | 3*20 | |||||
4 | 3*20 |
Materials and experimental setups of a the germination tests and b the botanical garden experiment.
Before the experiments, the average seed mass of all seed lots was estimated by weighing three subsamples containing exactly 20–20 seeds. In the case of Lakitelek seed lots (14L and 17L) we did not have 3×20 seeds; thus, we weighed all the 46–50 seeds individually to get the average seed masses. Subsamples were measured in grams with an accuracy of 0.0001 g.
In the first experiment, we tested the viability and dormancy of the in-situ collected seeds, mainly focusing on physical dormancy (PY).
From the largest seed lot (15P), 120 seeds were randomly selected for the experiment, while only 80–80 seeds were available from the Tiszaderzs seed lots (13T and 14T), and all the 46–50 seeds of the Lakitelek seed lot (14L and 17L) were used in the experiment (Table
We recorded the number of imbibed seeds daily to study the degree of physical dormancy. Physically dormant seeds could not intake water; thus, their appearance (size and colour) did not change, while seeds with broken hard seed coats visibly swollen and their colour faded. Based on this, the physical dormancy of the seed lots was estimated as follows:
PY i= level of physical dormancy in population i
NDi= ratio of non-dormant seeds in population i
IMBc= number of imbibed seeds in the control group
UNc= number or unchanged seeds in the control group
Germination (root/shoot emergence) was monitored daily to investigate seed viability. As scarification was a reliable method to break physical dormancy (see in the Results section), the viability of the in-situ collected seed lots was estimated based on the germination of the scarified, imbibed seeds as physical dormancy did not prevent the germination of viable seeds in this group. However, the possible occurrence of viable seeds with combinational dormancy (PY+PD) “hiding” among the non-germinated, imbibed seeds was also considered in interpreting the results. Thus, seed viability of the in-situ collected seed lots was calculated as:
SVi= seed viability in population i
GERMSC= number of germinated seeds in the scarified group
NGSC= number of non-germinated seeds in the scarified group
In the second experiment series between 2009 and 2018, four ex-situ collected seed lots (09EX, 10EX, 12EX, and 13EX) were sampled and germinated repeatedly for 3–8 years to record the temporal dynamics of their germination capability (Table
All germination experiments were performed in a greenhouse from the beginning of March when the temperature in the greenhouse varied between 10 °C and 20 °C. All seeds were placed onto wet filter papers in Petri dishes, and irrigation was automatic by diffusion with the help of 2cm × 20cm filter paper stripes. Imbibition and root/shoot emergence were monitored daily for 30 days.
Survival, development, and reproduction success were studied in 2009 in the newly established ex-situ population (Fig.
For establishing the ex-situ population, seeds of Tiszaderzs population were collected in 2007 and germinated in March 2009. In June, the well-developed (longer than 30 cm) specimens were planted on a 15 × 2m plot with sandy, fertilized, and homogeneously mixed soil and naturally diverse light conditions (
Before planting, we classified the columns according to their light conditions (Fig.
Measurements: We measured the length of the stems weekly. The number of inflorescences per individual, flowers per inflorescence, green pods per inflorescence, seeds per green pods, and the number of mature legumes were counted weekly, too. Mature pods were collected, and seeds were counted as well.
As flowering and seed production is continuous during the summer, note that with monitoring once a week, particular flowers could be counted more than once, and we could not collect all of the mature legumes before the escape of some seeds. Consequently, we cannot estimate the exact number of produced flowers or seeds, but the standard sampling makes it possible to compare the seed production of the individuals grown under different conditions.
Data visualization and statistics were performed using R statistical software (R Core Team, 2019). Figures were drawn by the ‘ggpubr’ (
Germination capabilities and dormancy of seeds in the different seed samples, as well as the effect of seed weights on seed germination probability, were compared by logistic regressions (glm function;
Regarding the garden experiment, the probability of mortality after the plantation was analysed within 6-week periods (i.e., the probability of survival in the first six weeks, between 6–12 weeks, 12–18 weeks, and 18–24 weeks after plantation). The effect of microsites (i.e., light and precipitation conditions) and root competition (marginal/inner individuals) on this probability was analysed by logistic regression. The maximum heights of plants growing in different microsites and within different root competition conditions were compared with linear models. Post hoc pairwise comparisons were made using the glht function of the ‘multcomp’ package, which automatically gives adjusted p-values (
Experiment 1 – Viability and dormancy tests
The results of the first experiment are summarized in Fig.
Results of the first germination experiment: seed viability (a), dormancy (a, b), seed weight (c) and their relationship (germination probability/relative frequency of scarified seeds with different weights) (d) in the case of half-year-old, in-situ collected seeds. Different characters in panel c indicate significantly different seed lots (linear model, Tukey’s contrasts). In panel d, the heights of the bars represent the relative frequencies of germinated (hanging bars) and non-germinated (standing bars) seeds in the different seed weight classes. The sum of the heights of all bars is equal to one. E.g., 4.78% of the examined scarified seeds weighed between 5-10 mg and showed no germination (second standing bar), while another 20.21% of all tested seeds fell into the same weight category but did germinate in the experiment (second hanging bar). The fitted line shows the relationship between the seed weight and the germination capability, predicted by the logistic regression model.
Generally, the seeds of the species showed high viability (on average, 78–88% of the scarified seeds germinated), and there were no significant differences between the different in-situ seed lots (adjusted p>0.08, logistic regression) (Fig.
Almost all seed samples were different in average seed mass (p<0.001, linear model) (Fig.
The results of the second experiment series are shown in Fig.
Despite the highly protective seed coat (i.e., the generally high initial physical dormancy) and high initial viability, all examined ex-situ collected seed samples showed a sharp decrease in germination capability in the first 5–6 years (p<0.001, logistic regression) (Fig.
Summary statistics (mean and SD) of traits (lifespan, height, pod, and seed production) measured on the different microsites are summarized in Table
Summary statistics (mean and sd) of the traits of plants grown in the botanical garden in different light and precipitation conditions (Shady= always shaded weekly irrigated microsite (n=17), HS= half-shaded, weekly irrigated microsite (n=30), Sunny= always sunny, weekly irrigated microsite (n=30), HS+dry= half-shaded, biweekly irrigated microsite (n=23)).
Microsite | Lifespan (weeks) | Max. height (cm) | Number of pods/individual | Number of seeds/individual | Number of seeds/pods | |||||
---|---|---|---|---|---|---|---|---|---|---|
mean | sd | mean | sd | mean | sd | mean | sd | mean | sd | |
Shady | 21.00 | 3.78 | 201.37 | 41.42 | 80.43 | 39.79 | 264.43 | 111.56 | 3.39 | 0.63 |
HS | 18.33 | 5.96 | 276.91 | 50.82 | 148.07 | 76.67 | 529.25 | 256.75 | 3.62 | 0.40 |
Sunny | 15.30 | 8.82 | 209.00 | 54.43 | 184.50 | 130.95 | 679.91 | 507.77 | 3.69 | 0.36 |
HS + dry | 14.39 | 5.98 | 161.47 | 49.52 | 53.45 | 51.44 | 183.05 | 152.20 | 3.67 | 0.49 |
All ex-situ planted individuals (n=98) germinated, established and finished their life cycle within one vegetation season (from spring to autumn) as summer annual plants. Within the 24-week-long field experiment, the highest average lifespan (21 weeks after plantation) was found in shady conditions, and a decreasing trend was observed along the light gradient (Half-shady (HS)=18.3 weeks, and Sunny=15.3 weeks, Fig.
Survival and growing in different light conditions and competition: lifespan after plantation (mean + SE) (a), temporal patterns in survival (b), maximum height (mean + SE) (c) and temporal patterns in growing (d). Lighter boxplots show individuals with no competitors within 40cm, while darker boxplots indicate plants with neighbouring grass in 20cm distance. Dots and bars on the boxplots show mean and SE values. Grey and white stripes on panels b and d indicate 6-week-long periods.
The plants reached their maximum height 12–16 weeks after plantation (Fig.
Flowering started around week 28, reached the peak around week 33, and ended around week 39 (Fig.
Flowering and yielding in different light conditions and competition: first week of flowering after plantation (a), temporal patterns in flowering (b), number of matured pods/individual (c) and temporal patterns in producing pods (d). Lighter boxplots show individuals with no competitors within 40cm, while darker boxplots indicate plants with neighbours within 20cm distance. Dots and bars on the boxplots show mean and SE values. Grey and white stripes on panels b) and d) indicate 6-week-long periods.
The first pods appeared two weeks after the start of flowering. The number of pods showed a similar temporal pattern to the flower’s (Fig.
To summarize our results, Vicia biennis seeds showed high seed viability and physical dormancy in all the studied Hungarian populations, while seed mass was highly variable across years and source populations. The germination results suggest that the species has a long-term persistent seed bank, but seed viability sharply decreases below 10% within ten years in dry storage and room temperature.
All specimens in our experiment behaved as annual (not biennial), and their mortality, growth, and reproduction showed a strong sensitivity for irrigation/soil moisture content: on sandy but fertilized soil and in half-shady conditions, the plants performed significantly better (lived longer, grew higher and produced more flowers and seeds) when they were irrigated weekly and not only every second week. When getting the same amount of water, the plants had longer stems and more flowers in the half-shady area than in shady or sunny conditions, but the pod and seed production were similar or even higher (when competition for water was low) on the sunny part of the plot. However, higher competition for the limited water resulted in a significant decrease in seed production.
Annual plant species are typically associated with highly disturbed and temporally variable habitats, where producing seeds with different dormancy levels and high longevity can efficiently “spread the risk” caused by the unpredictable environment (
V. biennis is mainly connected to the edge of temporal or persistent wetland habitats, and we found that the adult plants show a strong sensitivity to water availability. Based on this, a possible explanation for the generally observed high seed dormancy across populations and years can be the widespread insufficient precipitation or high inter-annual variation in precipitation in their habitats. The long-term climate data of Hungary support this: in the last decades, there was a significant decrease in spring precipitation and the number of wet days (with more than 1mm rainfall), and an increasing trend in the duration of dry seasons, while a large inter-annual variation can be observed for all the three parameters (
Seed mass is also important in coping with environmental stress. At an individual level, larger seeds support a higher survival rate for seedlings (
In this study, we found significant differences in the seed weights between the different populations and years. In general, we found significantly larger seeds in the larger populations (Tiszaderzs and Lakitelek in 2017) containing more than 20 individuals, while the 10–15 individuals living in the smallest populations (Püspökladány and Lakitelek in 2014) produced the smallest seeds. Although seed mass usually correlates with seed viability (i.e., seeds developed in unfavourable conditions are often less viable), we found high seed viability in all populations. A possible explanation for this is that even these populations with small seeds have reached an average seed weight of 10 mg, which – according to our logistic model – can be enough to maintain high seed viability (i.e., an expected germination capacity higher than 0.8). These results suggest that the small numbers of individuals in the populations have not yet led to a significant deterioration in seed quality.
At the species level, seed longevity shows some correlation with other seed traits, like seed mass: species with smaller average seed mass are more likely to have long-term persistent seed banks (i.e., seeds that are still viable after five years in the seed bank) (
From a conservation perspective, the most crucial question is how threatened the species actually is by climate change or habitat change/loss. According to the climate models of Hungary (
The species’ long-term survival will depend on how it can adapt to these changes or ‘escape’ them and find more suitable areas.
Adaptive capacity is based on the adequate genetic diversity of populations. In small, highly fluctuating populations, genetic diversity is more likely to be reduced by inbreeding and genetic drift, and these populations may become more sensitive to environmental change (
The probability and level of inbreeding depression depend on the species’ mating system. Although self-compatibility of flowers is common in the Fabaceae family (e.g.,
The primary mechanism to escape from the negative effects of climate and habitat change is dispersal. Similarly to other Vicia species (e.g.,
Based on the above, the following conservation priorities are suggested:
The authors have declared that no competing interests exist.
No ethical statement was reported.
The work of BD was supported by the NKFI FK 135329 project, while the work of OV was supported by the NKFI KKP 144096 project.
AE and JGYN designed the experiments and collected the seed material. AE, ÁS, and DÉ did the germination experiments, AE conducted the botanical garden experiment, analyzed the data, and wrote the first version of the manuscript. JGYN, BD, and OV contributed to the final version of the manuscript.
Anett Endrédi https://orcid.org/0000-0001-6572-6468
Ármin Sőth https://orcid.org/0000-0002-6417-3168
Balázs Deák https://orcid.org/0000-0001-6938-1997
Orsolya Valkó https://orcid.org/0000-0001-7919-6293
János György Nagy https://orcid.org/0009-0005-3854-0840
All of the data that support the findings of this study are available in the main text.