The aim of the study is to investigate the relative importance of plant-plant interactions with regard to flooding and drought effect on perennial plant performances in wetlands. Flooding is expected to be the major driver and, accordingly, the importance of drought is hardly if ever taken into account. Focusing on five widespread species, the growth, the survival and the competitive ability of plants were monitored on permanent plots spread along two elevation gradients. Flooding duration and drought intensity were found to vary substantially along the ~ 0.5 meter range elevation gradient. Flooding and drought alternate over the hydrological year and the pin-point surveys were thus conducted over the course of one year. The data were modeled taking into account survival, recruitment and competitive growth throughout flooding and drying out periods. Flooding and drought both directly impacted the plant performances and their competitive effect, with the effect of drought being much more general among species and of higher magnitude than flooding. The importance of competition was found to be high for all species, particularly during the drying out period. It varied more along the flooding gradient than along the drought gradient. The higher flooding tolerance shown by the studied species compared to drought may be related to species specific growth timing together with efficient response traits. These results offer new insights into the filters operating over the species pools. This suggests that the drying out period and drought conditions may be even more important for species’ relative success and the importance of competition than the flooding pattern. The general applicability of this result, obtained in mild Atlantic climate and fertile wetlands, remains to be studied.
Citation: Merlin A, Bonis A, Damgaard CF, Mesléard F (2015) Competition Is a Strong Driving Factor in Wetlands, Peaking during Drying Out Periods. PLoS ONE 10(6): e0130152. https://doi.org/10.1371/journal.pone.0130152
Academic Editor: Maura (Gee) Geraldine Chapman, University of Sydney, AUSTRALIA
Received: December 5, 2014; Accepted: May 18, 2015; Published: June 15, 2015
Copyright: © 2015 Merlin et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited
Data Availability: The data file is fully available on Dryad. DOI:doi:10.5061/dryad.rq8n5
Funding: AM received a PhD grant from the Ministère de l'Enseignement Supérieur et de la Recherche Français. The Parc Naturel Régional du Marais poitevin provided the funding for maintaining the in situ grazing experimental setting. The authors thank the mairie des Magnils-Reigniers for granting permission to work in the common.
Competing interests: The authors have declared that no competing interests exist.
The functioning of wetlands has been widely investigated, taking into account the alternation of flooding and drying out, e.g. [1, 2, 3]. By contrast, recent papers investigating plant performance and biotic interactions in wetlands have only considered the flooding pattern [4, 5, 6, 7, 8, 9]. Some others papers have considered the complete hydrological cycle, but focused only on the final plant performance without distinguishing the flood and dry out periods . The suggestion by  of quantifying plant stress in relation to flooding (anoxia stress) and drying out (drought stress) thus constitutes a notable exception. It was accordingly demonstrated [12, 13] that these stress gradients were pertinent for the definition of the hydrological niche of wetlands species
Assessment of the importance of biotic interactions in wetlands must thus be conducted while taking into account the impact of flooding and drought for perennial plants. According to the Stress-Gradient Hypothesis , stress and disturbance modulate the nature and magnitude of biotic interactions: facilitation is expected in stressful habitats, whereas competition is expected to peak where stress or disturbances are milder. Several studies generally support this hypothesis (see meta-analysis by ), while species-specific and specific stress-type response patterns [16, 17, 18] rule out its use as a general pattern.
A wide range of parameters have been proposed for investigating plant-plant interaction strength. Here, we consider i) species’ competitive effect, i.e. its ability to deplete resources for others , and ii) competition importance, i.e. the reduction of the fitness of a plant species by the presence of competitors in relation to any other factor that influence plant fitness . Competition importance offers a basis for appreciating the relative importance of competition against any other factors effect (here abiotic stress) and was for this reason chosen preferentially to the competition intensity which focuses on the effect of neighbors only . Competition importance was previously found to increase along with productivity . In flooded grasslands,  found that competition importance was peaking at one side of the flooding gradient or the other, depending on the ecology of the species. A similar result was shown along salinity gradients [22, 23, 24]. Such contrasts in the responses to environmental gradients may be related to species-specific response and to non-univocal effect of the environmental gradient, which may act as stress or not depending on the level considered and its timing.
To our knowledge, no work has yet documented how plant performances, plant competitive effect and competition importance vary with aeration stress, related to flooding, and with drought stress, occurring during the drying out of the wetlands.
Competition between species has been repeatedly reported to impact plant survival, recruitment and growth [25, 26]. These life-cycle parameters are thus appropriate to measure the role and importance of competition [27, 28], notably in natural communities  and to dissect out the various processes influencing the dynamics of populations (e.g. ). A model quantifying the importance of competition along plant density and environmental gradients has been developed by  and used to explain the results of the modeled competitive interactions.
In this study, this model framework was used to investigate plant performances, plant competitive effect and competition importance on the basis of field-based vegetation surveys in dense wetland vegetation. They were investigated along elevation gradients, 0.5 m wide on average, which drives an aeration stress gradient-during the flooding period- and a drought intensity gradient –during the drying out period- as shown by .
Plant surveys were conducted on permanent plots, spread out along the elevation gradient. Each permanent plot was characterized regarding the flooding and drought magnitude following .
Our aim was to study the effect of the flooding gradient and of the drought gradient, i) on the performances and on the competitive effect of five target species presenting different occurrence patterns in situ regarding elevation, ii) on the importance of competition for the success of the target species. We also aimed to compare the magnitude of the competition importance along the flooding gradient and along the drought gradient.
Study site and characterization of the environmental gradients
This study was conducted on two grazed wet grasslands situated in the Marais poitevin, on the French Atlantic coast (46°28’N; 1°13’W). We acknowledge the cooperation of the Mairie de Magnils Reigniers for granting permission to work in the common and of the Parc Naturel Régional du Marais poitevin and the Etablissement Public du Marais poitevin for their support in maintaining the experimental settings. The climate is of mild Atlantic type, with an excess of precipitation over evapotranspiration in winter of 220 mm on average and a water deficit in summer of about 300–350 mm . These wet grasslands are fertile  and show high productivity for temperate permanent grasslands, up to 800 g dry matter/m² .
Annual cycles of flooding are typical of wet grasslands in temperate zones, with flooding occurring in autumn according to the amount of rainfall, and drying out of the marshlands generally occurs in June, but may be as early as April some years, depending on the rainfall pattern . The soil elevation varies within these wetlands, ranging between 0.36 and 0.54 m. The flooding duration and the soil drought intensity vary depending of the position along the elevation gradient, respectively during the flooding and the drying out period [7, 34]. In accordance with , we quantified the soil aeration-shortage and drought, respectively, in relation to soil flooding and drying out. Following , the flooding duration was characterized as the Sum Exceedence Value above -0.19 m (thereafter aeration SEV, SEVa) and the drought intensity as the Sum Exceedence Value below the threshold of -0.42 m (thereafter drying SEV, SEVd)  (see S1 Appendix). The SEV parameters were derived from the water-table depth level, monitored at hourly intervals all year round using level logger sensors. SEVa summarizes the aeration-stress throughout the flooding period and SEVd summarizes the drought conditions throughout the drying out period.
In the two grasslands studied, the SEVa was found to vary from 0.3 and 12.8 cm.day-1, a similar range to that found in British alluvial grasslands . The SEVd varied between 28.3 and 45.4 cm.day-1, showing locally drier conditions than in the British wetlands , probably due to the highly clay-rich soil together with large rainfall deficit during the summer.
We focused on five clonal perennial species with contrasted in situ distribution along the elevation gradient . Cynosurus cristatus and Lolium perenne are mesophilous species and occur more frequently at higher elevations. Juncus gerardi, a meso-hygrophilous species, is found at intermediate elevations and Glyceria fluitans, a hygrophilous species, is mainly recorded at lower elevations. Agrostis stolonifera is distributed over the whole the elevation gradient.
Plant abundance measurements.
Seventy permanent plots, 25 x 25 cm large, were placed every 20 cm along two elevation transects used as replicates (35 plots per sequence). On each plot, vegetation relevés were performed along both diagonals using the pinpoint method: a record was made every 4 cm for a total of 17 points per plot.
Three readings were made: on the 23–29 October, 2008, on the 3–12 June, 2009 and on the 19–20 October, 2009. The studied wet grasslands are usually grazed each year from April to November. Grazing exclusion was necessary in order to survey plant performances along the flooding and drought gradients without any confusing effect due to livestock grazing which may be selective and heterogeneous . Fences were therefore established for one year in the two grasslands, a period short enough to avoid any significant change in the vegetation species composition and relative abundance.
Plant surveys consisted in measuring the vertical density and the cover of each species in each plot. The vertical density of a species i, Yi, represents the number of times that species touches each of the 17 pins in the pinpoint frame. The cover of species i, Xi, is the number of times that species is present at each point of the pinpoint frame. The vertical density is assumed to be a function of the cover of species during the growing season  as the change in vertical density is assimilated to the biomass growth within the growing season. The use of pin point field surveys and modeling following  constitutes a field approach to plant performances.
During the flooding period, i.e. from October 2008 to June 2009, the cover of species is assumed to be a function of the vertical density of species at the end of the previous growing season. Cover can be assimilated to the survival and recruitment abilities of species in response to the environment. In this way, the processes controlling the translation of biomass into cover over the flooding period and the effect of the environment on the processes controlling the biomass growth over the drying out period can be studied. More specifically, it is assumed that the vertical density of species i at time t2 is an increasing function of cover of species i, function of the cover of species j and k at time t1, and function of the environmental gradient zr. Competitive growth of species i was modeled as: (1), with r the pin-point frame; the residual process variation during the flooding period of species i across different years and pin-point frames. ai(Z) corresponds to the growth of species i directly affected by the relation between the cover and the vertical density of the species; cj(Z) correspond to competitive effects of species j affecting growth of species i. The model assumes that the competitive effect of species i on species j is equal to the competitive effect of species i on species k. While oversimplified, we consider this assumption to be acceptable, as the main purpose of this work is to untangle the competition versus environmental effect on species performances. The parameter functions ai(Z) and cj(Z) are functions of the environmental gradient as linear functions, ai,0 + ai,1 Z and cj,0 + cj,1 Z, respectively. This means that ai,1 and cj,1 measured the effect of the environmental gradient (z) on growth and the competitive effect of species j, respectively. In this study, z is the value of aeration SEV (when considering flooding effect) and is the value of drying SEV (when considering drought effect).
Flooding occurred at the end of autumn and ended in spring of the following year: recruitment and survival abilities of species are integrative of the fitness of the species during this period . Over the period of flooding, it is assumed that the cover of species i at year y+1 (June, 2009, here) is an increasing function of the vertical density of species at year y (October, 2008) at the end of the growing season), and a decreasing function of the vertical density of species j and k at year y and together a function of the environmental gradient z. Survival and recruitment of species i at year y+1 was modeled as: (2) with the residual process variation from one season to the next of species i among years and pin-point frames.
The competition model has previously been used to successfully model the effect of herbicide and nitrogen on survival, recruitment, and competitive growth of two perennial grass species from similar time series pin-point data [36, 37]. In this study, it will monitor the result of actual interactions between plants at different life-history stages and after the long term run of the assemblages.
As this model is a ‘three species model’, the different species surveyed were aggregated into three groups: the first group corresponded to Agrostis stolonifera, the second group corresponded to, alternatively, Lolium perenne (Lp), Cynosurus cristatus (Cc), Juncus gerardi (Jg), Glyceria fluitans (Gf), and a third group (“other aggregated species” in S1 Table) comprised the rest of the higher plant species recorded on the plot.
Model parameters (a, b, c and d) were estimated using the Bayesian method: the joint posterior distribution of model parameters was calculated using the MCMC method (Metropolis-Hastings algorithm with 120000 iterations) and a multivariate distribution. The sampled chains of all parameters were inspected to check the properties of the sampling procedure. The effect of the environmental conditions on species performances (parameter a1) and on species competitive effect (parameter c1) were assessed by the 2.5%, 50% and 97.5% percentiles provided from the marginal posterior distribution of the parameters. For a1 value significantly larger than zero, species growth increases with SEV. Likewise, if c1 is larger than zero, then the competitive effect of the species increases with SEV.
Growth was compared among species pairs modeled together: when their 95% credible intervals for a1 do not overlap, then it significantly differs from one to another.
Assessment of the importance of competition
From Eqs (1) and (2), the importance of competition was quantified using the competition models (1) and (2) and the measured cover and vertical densities (i) during the flooding period, i.e. comparing surveys in October, 2008, and in June, 2009, (3) and (ii) during the summer period, comparing data in June and October, 2009 (4), following the proposition of : (3) (4) Where and represent the changes in the measured ecological success of the species i, as modeled by Eqs (1) and (2) by changing the levels of the environmental conditions z.
How do environmental conditions impact species’ performances?
The effect of the flooding duration and drought intensity on species’ performances (survival, recruitment and growth) were distinguished from the effect of biotic interaction and assessed by the percentiles of parameter a1 (Fig 1 and S1 Table). Species’ performances, regardless of the group considered, were found to be much higher during the drought period than during the flooding period.
A positive value indicates that flooding (drought) promotes growth or competitive effect while negative value indicates a detrimental effect of flooding (drought) on growth or competitive effect. Species name for which the trend is significant appear in bold and is underlined. For Agrostis stolonifera, growth and competitive effect are reported with the model run with four different species for which the initials appear in subscript (A.stoLp, A.stoCc, A.stoJg, A.stoGf).
The increase in flooding duration, approached by the aeration SEV, had a significantly positive effect on the survival and recruitment of J. gerardi and of A. stolonifera, only when modeled with L. perenne (A.stoLp). It showed neither a significant effect on recruitment nor on survival for the other species, while this was particularly unexpected for G. fluitans. The increase in flooding duration had a significantly positive effect on the performances of the three ‘aggregated species’ groups out of the 4 (S1 Table).
Higher drought diminished the performances of all 5 studied species (L. perenne, C. cristatus, J. gerardi, G. fluitans and A. stolonifera) in all groups. Drought showed a more limited effect on J. gerardi compared with the other species. Increasing drought was found to be non-significant for the performance of 2 ‘aggregated species’ groups, negative for one and positive for one out of the 4 (Fig 1 and S1 Table).
Effect of environmental conditions on species competitive effect
Increase in flooding duration showed no significant effect of the competitive effect of any of the 5 studied species (Fig 1 and S1 Table). It had only a positive effect on the competitive effect of in 3 out of the 4 “aggregated species” groups considered (S1 Table).
During the summer period, the competitive effect of A. stolonifera (in all groups), of J. gerardi and of all “aggregated species” groups was found to be diminished by increasing drought (Fig 1 and S1 Table). Only the competitive effect exerted by C. cristatus was improved by more intense drought conditions.
Importance of competition
The importance of competition represented the proportion of change in a species’ ecological success caused by competition relative to the impact of environmental conditions. Along the gradient of flooding duration, the competition importance showed two main patterns of variation, depending of the species. (i) For L. perenne, J. gerardi, G. fluitans and A. stolonifera-considered with L. perenne-, the importance of competition increased steadily with the flooding duration (Fig 2). For J. gerardi and G. fluitans, competition determined up to 60% of the plant performance, while this proportion remained lower than 40% for the other species. (ii) For C. cristatus and for A. stolonifera -when considered with C. cristatus and J. gerardi as species pair in the model-, the importance of competition varied within a slightly more limited range (30–40% and 50–80%, respectively) and was maximized at both ends of the flooding gradient (Fig 2).
The importance of competition represents the proportion of change in ecological success caused by competition relative to environmental conditions. The SEVa was calculated following  and expressed as cm.day-1. Fig 2 reports the results for Agrostis stolonifera modeled in pair with Lolium perenne (Fig 2a), Cynosurus cristatus (Fig 2b), Juncus gerardi (Fig 2c), Glyceria fluitans (Fig 2d). Results for Lolium perenne are reported in Fig 2a’, for Cynosurus cristatus in Fig 2b’, for Juncus gerardi in Fig 2c’ and for Glyceria fluitans in Fig 2d’.
During the drying out, the importance of competition was the highest for all studied species under milder water resource conditions (Fig 3). For L. perenne, C. cristatus and G. fluitans, the competition importance remained very high (> 90%) across a very wide range of drying SEV values and only decreased under very harsh-water limited- conditions, with drying SEV> 38 cm.day-1. For all three species, competition then almost entirely determined plant performances under the milder drought conditions. A. stolonifera showed the same pattern, but the competition importance showed a stronger diminution, being limited to 40% with the harsher drought conditions for this species, except when considered with G. fluitans.
Competition importance represents the proportion of change in ecological success caused by competition relative to environmental conditions. The SEVd was calculated following  and expressed as cm.day-1. The Fig 3 (a, b, c, d) report the results for Agrostis stolonifera modeled in pair with respectively Lolium perenne, Cynosurus cristatus, Juncus gerardi, Glyceria fluitans. Results for Lolium perenne are reported in Fig 3a’, for Cynosurus cristatus in Fig 3b’, for Juncus gerardi in Fig 3c’ and for Glyceria fluitans in Fig 3d’.
This work has shown that in fertile wetlands, and with Atlantic mild climate, flooding is a poor direct driver for species performance and for species’ competitive effect. By contrast, drought impacts all five species’ performance by directly affecting their survival and growth. Drought also controls the competitive effect of 2 out of the 5 species studied. Both flooding and drought gradients drive competition importance for all species: species-specific patterns were found along the flooding gradient, whereas competition importance almost entirely determined the species’ performances and peaks as soon as drought stress was somehow relaxed. This result was obtained with grazing being temporary excluded: it may be expected that, with grazing, competition importance would have been milder as grazing diminished plant height, a good proxy for competition effect .
The importance of competition on species performances relative to flooding and drought appears high in wet grasslands, and flooding is thus not the predominant driver of the plant performance. This suggests that the general and only focus on the impact of flooding in wetland vegetation studies is not justified, as already shown by .
The impact of gradients on plant performance
Flooding duration showed no significant impact for 3 out of the 5 perennial species studied, two mesophilous and one hygrophilous species. Those results fit well with , who also found that the elevation gradient did not significantly change biotic interferences. In agreement,  also found a limited and vanishing effect of flooding on plant performance and on competitive ability and hierarchy among species. This result may, however, be restricted to perennials which were found to be less responsive to flooding variation than annual plants .
It may in fact be questioned whether flooding constitutes a stress as defined by , as this study showed that no species’ performance was impacted by the flooding pattern. Results by  also supported the view that flooding is a poor filter of the local grassland’s species pools, as it impacted species richness to a very limited extent and did not significantly impact biomass production. This suggests that long lasting plant assemblages are, in the field, far less responsive to flooding pattern than what was reported from short term experiments [6, 44], or when focusing on biomass only . It is thus strongly advised to consider long term assemblages and various stages of plant life-cycle over the whole ecological cycle to obtain a realistic picture of plants’ performances and biotic interactions across gradients in wet grasslands.
Two main explanations may be suggested for this limited impact of flooding regime:
- Species that make up the species pool, either mesophilous, meso-hygrophilous or hygrophilous, are all able to cope with aeration stress (see ) as all the wet grasslands are subjected to water saturation in the soil for several weeks every year and over centuries. Flooding tolerance may indeed be actually shared by all species in the wetlands, at all elevations, due to previous filtering process of the species regional pool.
- Growth timing pattern may both mitigate aeration stress and accentuate drought. Flooding mainly occurs when plants are passing through a period of vegetative rest, and the associated aeration stress may thus only have slight impact in contrast to drought, which occurs in more active periods. Accordingly, the growth performances modeled from the pin-point measurements (parameter a1) actually depict a much less active growth pattern from October to early June (flooding period) than from mid-June to October. The peak in N mineralization and thereafter N availability in June and July  also suggest that a major limiting factor for growth is alleviated in early summer.
Monitoring species and community growth patterns all-year-round will provide a basis for assessment of whether or not stress occurrence coincides with plant growth and whether this may be a plausible explanation for the limited effect of flooding on plant success.
The results obtained in this study also support the view that biotic interactions are key driven factors liable to explain plant communities’ and species’ relative abundance patterns along elevation gradients [31, 40], as also shown by  and . The high plant biomass and density in the studied fertile wet grasslands  is probably a major promoter for the importance of biotic interactions in the assemblages.
Species response to flooding
The two species that are significantly impacted by flooding duration, J. gerardi and A. stolonifera, actually benefitted from longer flooding with regard to their survival and recruitment. Under flooded conditions, A. stolonifera developed longer stems along which adventitious roots grew, which favored space occupancy and high re-growth ability [48, 49], together with good competitive effect . Regarding J. gerardi, its high tolerance to flooding was first recognized by . Its ability to store resources in rhizomes  may help cope with flooding and support spring growth, as shown for other species [52, 53].
The high tolerance to flooding shown by the three other species (i.e. C. critatus, L. perenne and G. fluitans) may be related to eco-physiological and morphological adaptations, as reported in abundant literature (see e.g. synthesis by [54, 55]) and may also imply clonal traits .
In contrast to flooding, soil drought was found to be effective in controlling species’ performance and species’ competitive effect for all five species. Their performances were found to increase with water availability and, consequently, their competitive effect (see [57, 58] for the relationship between plant biomass or size and competitive effect). The impact of drought was limited on J. gerardi, probably due to its early-growing pattern [59, 60].
Showing that reduced drought had a positive effect on plant performance is by no means unexpected. However, the magnitude of drought impact and its general occurrence suggests that it plays a much more important role in the vegetation pattern in wet grasslands than flooding, and this was unexpected. Vegetation patterns commonly reported along elevation gradients may accordingly be related to drought and water availability during the drying out period even more than to the flooding pattern.
At the community level, the weaker effect of flooding-related stress can be explained by the period of flooding: it occurs over the colder period, when the conditions are mainly unfavorable for plant perennial growth in temperate wetlands, and this may have thereafter limited plant-plant interactions.
Plant-plant interactions and abiotic stress
It was predicted by  that abiotic stress limits species distribution at the harsher end of the gradient, while competition drives the vegetation pattern at the milder end. Referring to the Stress-Gradient Hypothesis [14, 61] stress and disturbance are thus expected to modulate the nature and magnitude of biotic interactions. Such a negative relationship between the importance of competition and environmental stress has been found by e.g. [5, 17, 21, 24] in tidal, wet, alpine and dry grasslands, respectively.
In the present work, a similar relationship was found in 7 out of the 8 situations studied with the competition importance reaching its maximum, determining up to 90% of individual performances, as soon as the drought level slightly decreased (i.e. SEVd <36). The detrimental effect of plant-plant interactions was thus minimized under the harsher drought conditions, as seen in dry grasslands . During the flooding period, the importance of competition also varies with elevation, but was maximized at high anoxia stress levels for 6 out of 8 cases, then showing an inverse pattern with stress than during the drying out period.
To be realistic, vegetation patterns spread along elevation gradients must be investigated while considering both the increasing drought stress from low to high elevation during summer time and the increasing aeration stress from high to lower elevations during flooding. Predictions regarding the competition importance pattern along the elevation gradient will be opposed, depending on whether one or the other stress factor is considered. This argues in favor of standardized methods to characterize environmental gradients, as proposed for plant traits measurements , taking into account the complete hydrological cycle and plant critical stages.
World-wide investigations comparing plant-plant interactions between harsh and mild habitats yield conclusive results, although not unequivocally . The different natures of the various stress and resources may explain the muddled picture frequently obtained in case studies. The studied wetlands vegetation presents a very different picture than that of alpine or saltmarsh communities where positive interactions were repeatedly found. As far as the importance of competition is concerned, this study, like , concluded that plant interaction plays a strong role in wet Atlantic grasslands. This study may shed new light on the increase of the Specific Leaf Area value at the community level found by  with flooding. This trait pattern was interpreted as a trait response to increasing aeration shortage. However, as the drying out was found to be the predominant growth period, increased SLA value at low elevation may also correspond to an increase in the growth ability of the community with decreasing drought.
Both flooding and drought impact wetlands’ plant species survival and growth but drought effect was found to be much more effective for the perennial species studied, regardless of whether the plant species are mesophilous, meso-hygrophilous or hygrophilous. This picture also holds for the species competitive ability which was found to be significantly impacted by drought and not by flooding pattern.
The importance of competition has been found to be high both during the flooding and the drying out periods, while peaking during the drying out period. Under the milder drought conditions, it was found to determine over 90% of the plant performances, and again, the picture was similar for hygrophilous, meso-hygrophilous and mesophilous species. Competition importance did vary along the flooding and the drought gradients but the pattern is neither linear nor regular among species.
In contrast to the stress gradient hypothesis, competition was the dominant process throughout the stress gradient. This may be due to the homogeneity of the growth forms characterizing the plant communities studied, where 80% of the cover is made up of clonal poaceae. Rather than a negative test of the stress gradient hypothesis, we suggest that this poor growth form diversity may not be propitious to facilitation effect, due to the driving force of grazing on growth forms.
This study highlights the drying out period and drought which constitute the “hidden side” of the elevation gradient in wetlands. As a consequence, those responsible for managing these habitats for whatever biodiversity or agriculture purposes will have to take into account this novel information.
S1 Appendix. Characterization of the aeration and drought stress magnitude.
S1 Table. Estimated effect of flooding and drought on species growth (a1) and species competitive effect (c1) summarized by the marginal posterior distribution of the parameters.
If a1 is greater than zero, then growth of the species increased with SEV. Likewise, if c1 is greater than zero, then the competitive effect of the species increases with SEV. Parameters significantly deviated from zero appear in bold font. The species considered in pair in the model that are followed by similar letters showed non-significantly different growth.
We thank Jan Bernard Bouzillé for previous discussions, Valérie Riaud-Gouesbet, Olivier Jambon and Guillaume Bouger for help in the field and water table monitoring, Jean-Pierre Caudal and Nicolas Charpentier for help in preparing figures. We are indebted to Charlotte Elisabeth Kler, Sara Mullin and Mickael Paul who checked the English. We are grateful to the Mairie des Magnils Reigniers for granting permission to work within its common and to the Parc Naturel Régional du Marais poitevin and the Etablissement Public du Marais poitevin for their support in maintaining the grazing experimental setting.
Conceived and designed the experiments: AM AB CD FM. Performed the experiments: AM. Analyzed the data: CD AM. Wrote the paper: AM AB FM CD.
- 1. Feng S, Ai Z, Zheng S, Gu B, Li Y. Effects of dryout and inflow water quality on mercury methylation in a constructed wetland. Water, Air, and Soil Pollution. 2014;225: 1–11.
- 2. Erler DV, Eyre BD, Davison L. Temporal and spatial variability in the cycling of nitrogen within a constructed wetland: A whole-system stable-isotope-addition experiment, Limnol Oceanogr. 2010;55: 1172–1187.
- 3. Lucassen EC, Smolders AJP, Lamers LPM. Water table fluctuations and groundwater supply are important in preventing phosphate-eutrophication in sulphate-rich fens: consequeces for wetland restoration. Plant and Soil. 2005;269: 109–115.
- 4. Casanova MT, Brock MA How do depth, duration and frequency of flooding influence the establishment of wetland plant communities? Plant Ecol. 2000;147: 237–250.
- 5. Jung V, Mony C, Hoffmann L, Muller S. Impact of competition on plant performances along a flooding gradient: a multi-species experiment. J. Veg. Sci. 2009;20: 433–441.
- 6. Lenssen JP, Van Kleunen M, Fischer M, De Kroon H. Local adaptation of the clonal plant Ranunculus reptans to flooding along a small-scale gradient. J. Ecol. 2004; 92:696–706
- 7. Violle C, Cudennec C, Plantegenest M, Damgaard C, Le Cœur D, Bouzillé J-B et al. Indirect assessment of flooding duration as a driving factor of plant diversity in wet grasslands. In: Sivapalan M, Wagener T, Uhlenbrook S, Zehe E, Lakshmi V, Liang X, Tachikawa Y, Kumar P, editors. Predictions in Ungauged Basins: Promises and Progress. IAHS 303; 2006.pp. 334–341.
- 8. Van Eck WHJM, van de Steeg HM, Blomand CWPM, de Kroon H. Is tolerance to summer flooding correlated with distribution patterns in river floodplains? A comparative study of 20 terrestrial grassland species. Oikos. 2004;107: 393–405.
- 9. Warwick NW, Brock MA. Plant reproduction in temporary wetlands: the effects of seasonal timing, depth, and duration of flooding. Aquat. Bot. 2003;77: 153–167.
- 10. Mueller-Dombois D, Sims HP. Response of three grasses to two soils and a water table depth gradient. Ecology. 1996;47: 644–648.
- 11. Gowing DJG, Youngs EG, Gilbert JC, Spoor G. Predicting the effect of change in water regime on plant communities. In: Wheater H, Kirby C, editors. Hydrology in a Changing Environment vol. I. John Wiley, Chichester; 1998. pp. 473–483.
- 12. Silvertown J, Dodd ME, Gowing DJ, Mountford JO. Hydrologically defined niches reveal a basis for species richness in plant communities. Nature. 1999;400: 61–63.
- 13. Silvertown J. Plant coexistence and the niche. Trends Ecol Evol. 2004;19: 605–611.
- 14. Bertness MD, Callaway R. The role of positive forces in natural communities: a post-cold war perspective. Trends Ecol Evol. 1994; 9: 191–193. pmid:21236818
- 15. He Q, Bertness MD, Altieri AH. Global shifts towards positive species interactions with increasing environmental stress. Ecol Lett. 2013;16: 684–706
- 16. Maalouf JP, Le Bagousse-Pinguet Y, Marchand L, Touzard B, Michalet R. The interplay of stress and mowing disturbance for the intensity and importance of plant interactions in dry calcareous grasslands. Ann. Bot. 2012;110: 821–828. pmid:22782241
- 17. Brooker RW, Callaghan TV. The balance between positive and negative plant interactions and its relationship to environmental gradients: a model. Oikos. 1988;81: 196–207.
- 18. Guo H, Pennings SC. Mechanisms mediating plant distributions across estuarine landscapes in a low-latitude tidal estuary. Ecology. 2012;93: 90–100. pmid:22486090
- 19. Goldberg DE. Components of resource competition in plant communities. In: Grace JB, Tilman D, editors. Perspectives on plant competition. Academic Press, San Diego; 1990. pp. 27–49
- 20. Welden CW, Slauson WL. The intensity of competition versus its importance: an overlooked distinction and some implications. Q Rev Biol. 1986;61: 23–44. pmid:3961095
- 21. Gaucherand S, Liancourt P, Lavorel S. Importance and intensity of competition along a fertility gradient and across species. J. Veg. Sci. 2006;17: 455–464.
- 22. Barbour MG. The effect of competition and salinity on the growth of a salt marsh plant species. Oecologia. 1978;37: 93–99
- 23. Mesléard F, Lepart J., Tan Ham L. Impact of grazing on the vegetation dynamics in former ricefields. J. Veg. Sci. 1995;6: 683–690
- 24. Greiner La Peyre MK, Grace JB, Hahn E, Mendelssohn IA. The importance of competition in regulating plant species abundance along a salinity gradient. Ecology.2001; 82: 62–69.
- 25. Howard TG, Goldberg DE. Competitive response hierarchies for germination, growth, and survival and their influence on abundance. Ecology. 2001;82: 979–990.
- 26. Fayolle A, Violle C, Navas ML. Differential impacts of plant interactions on herbaceous species recruitment disentangling factors controlling emergence, survival and growth seedlings. Oecologia. 2009;159: 817–825. pmid:19132404
- 27. Brooker RW, Kikividze Z. Importance: an overlooked concept in plant interaction research. J Ecol. 2008;96: 703–708.
- 28. Frekleton RP, Watkinson AR, Rees M. Measuring the importance of competition in plant communities. J Ecol. 2009;97: 379–384.
- 29. Aarssen LW, Keogh T. Conundrums of competitive ability in plants: what to measure? Oikos. 2002;96: 531–542.
- 30. Damgaard CF, Fayolle A. Measuring the importance of competition: a new formulation of the problem. J Ecol. 2010;98: 1–6.
- 31. Amiaud B, Bouzille JB, Tournade F, Bonis A. Spatial patterns of soil salinities in old embanked marshlands in western France. Wetlands. 1998;18: 482–494.
- 32. Gardarin A, Kazakou E, Garnier E, Carrère P, Cruz P, Donato A et al. Plant trait-digestibility relationships across management and climate gradients in permanent grasslands. J Appl Ecol. 2014;51: 1207–1217.
- 33. Dumont B, Rossignol N, Loucougaray G, Carrère P, Chadoeuf J, Fleurance G et al. When does grazing generate stable vegetation patterns in temperate pastures? Agric Ecosyst Environ. 2012;153: 50–56.
- 34. Violle C, Bonis A, Plantegenest M, Cudennec C, Damgaard C, Marion B et al. Plant traits capture species diversity and coexistence mechanisms along a disturbance gradient. Oikos. 2010; 120: 389–398
- 35. Damgaard C, Riis-Nielsen Y, Schmidt IK. Estimating plant competition coefficients and predicting community dynamics from non-destructive pin-point data: a case study with Calluna vulgaris and Deschampsia flexuosa. Plant Ecol. 2009;201: 687–697.
- 36. Damgaard CF, Strandberg B, Mathiassen SK, Kudsk P. The combined effect of nitrogen and glyphosate on the competitive growth, survival and establishment of Festuca ovina and Agrostis capillaris. Agric Ecosyst Environ. 2011;142: 374–381.
- 37. Damgaard CF, Strandberg B, Mathiassen SK, Kudsk P. The effect of glyphosate on the growth and competitive effect of perennial grass species in semi-natural grasslands. J Environ Sci Health. 2014; B49: 897–908. pmid:25310805
- 38. Westoby M, Falster DS, Moles AT, Vesk PA, Wright IJ. Plant ecological strategies: some leading dimensions of variation between species. Annu Rev Ecol Evol Syst. 2002;33: 125–159
- 39. Bartelheimer M, Gowing D, Silvertown J. Explaining hydrological niches: the decisive role of below-ground competition in two closely related Senecio species. J Ecol. 2010;98: 126–136.
- 40. Shipley B, Keddy PA & Lefkovitch LP. Mechanisms producing plant zonation along a water depth gradient: a comparison with the exposure gradient. Can. J. Bot. 1991;69: 1420–1424.
- 41. Lenssen JP, van de Steeg HM, de Kroon H. Does disturbance favor weak competitors? Mechanisms of changing plant abundance after flooding. J. Veg. Sci. 2004;15: 305–314.
- 42. Baldwin AH, Egnotovich MS, Clarke E. Hydrologic change and vegetation of tidal freshwater marshes: greenhouse, and seed-bank experiments. Wetlands. 2001;21: 519–531.
- 43. Grime JP. Plant Strategies and Vegetation Processes. Wiley, Chichester; 1979.
- 44. Mony C, Mercier E., Bonis A., Bouzillé JB. Reproductive strategies may explain plant tolerance to flooding: a mesocosm experiment using six marsh species. Aquat. Bot. 2010;92: 99–104. pmid:21144492
- 45. Rascio N. The underwater life secondarily aquatic plants: some problems and solutions. Crit Rev Plant Sci. 2002;21: 401–427
- 46. Rossignol N, Bonis A, Bouzillé J-B. Relationships between grazing pattern and vegetation structure on the spatial variations of net N mineralisation in a wet grassland. Appl Soil Ecol. 2006;31: 62–72.
- 47. Brose U, Tielborger K. Subtle differences in environmental stress along a flooding gradient affect the importance of inter-specific competition in an annual plant community. Plant Ecol. 2005;178: 51–59.
- 48. Soukupova L. Allocation plasticity and modularity in clonal graminoids in response to waterlogging. Folia Geobot. Phytotax. Praha. 1994;29: 227–236.
- 49. Benot ML, Bonis A, Mony C. Do spatial patterns of clonal fragments and architecture responses to defoliation depend on the structural blue-print? An experimental test for two rhizomatous Cyperacea. Evol. Ecol. 2010;24: 1475–1487.
- 50. Cooper AC. The effects of salinity and waterlogging on the growth and cation uptake of salt marsh plants. New Phytol. 1982;90: 263–275.
- 51. Dong M, de Kroon H. Plasticity in morphology and biomass allocation in Cynodon dactylon, a grass species forming stolons and rhizomes. Oikos. 1994;70:99–106.
- 52. Lenssen JPM, Ten Dolle GE, Blom C. The effect of flooding on the recruitment of reed marsh and tall forb plant species. Plant Ecol. 1998;139: 13–23.
- 53. Clevering OA, Van Vierssen W, Blom CWPM. Growth, photosynthesis and carbohydrate utilization in submerged Scirpus maritimus during spring growth. New Phytol. 1995;130: 105–116.
- 54. Volder A, Bonis A, Grillas P. Effects of drought and flooding on the reproduction of an amphibious plant, Ranunculus peltatus. Aquat. Bot. 1997;58: 113–120.
- 55. De Wilde M, Sebei N, Puijalon S, Bornette G. Responses of macrophytes to dewatering: effects of phylogeny and phenotypic plasticity on species performance. Evol. Ecol. 2014;28: 1155–1167.
- 56. Benot ML, Bonis A, Rossignol N, Mony C. Spatial patterns in defoliation and the expression of clonal traits in grazed meadows. Botany. 2011;89: 43–54.
- 57. Peltzer DA, Köchy M. Competitive effects of grasses and woody plants in mixed-grass prairie. J Ecol. 2001;89: 519–527.
- 58. Fraser LH, Miletti TE. Effect of minor water depth treatments on competitive effect and response of eight wetland plants. Plant Ecol. 2008;195, 33–43.
- 59. Bertness MD, Ellison AM. Determinants of pattern in a New England salt marsh plant community. Ecol Monogr. 1987;57: 129–147.
- 60. Bouzille J B, Bonis A, Clement B, Godeau M. Growth pattern of Juncus gerardi clonal population in a coastal habitat. Plant Ecol. 1997;132: 39–48.
- 61. Wilson SD, Keddy PA. Species competitive ability and position along a natural stress/disturbance gradient. Ecology. 1986;67: 1236–1242.
- 62. Cornelissen J H, Lavorel S, Garnier E, Díaz S, Buchmann N, Gurvich DE, et al. A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Aust. J. Bot. 2003;51: 335–380.
- 63. Callaway RM, Brooker RW, Choler P, Kikvidze Z., Lortie CJ, Michalet R, et al. Positive interactions among alpine plants increase with stress. Nature. 2002;417: 844–848. pmid:12075350