Advanced search
1 file | 11.11 MB

Niche theory and competitive interactions in an amphibious plant community of West-European softwater lakes

Author
Promoter
(UGent) and AJP Smolders
Organization
Abstract
In this thesis, several questions are raised that are related to the niche concept and population ecology. They are within a framework of the amphibious softwater plant community Eleocharition multicaulis Vanden Berghen 1969. This community has declined during the 20th century and several of its species are rare or endangered. The community, with typical species like Eleocharis multicaulis, Hypericum elodes and Scirpus fluitans, has not been the subject of much research as yet. Therefore, more research is needed to better understand the requirements of Eleocharition species and the mechanisms that rule community structure, if we want to know how to restore their habitat most effectively. The research questions in this thesis can be broadly summarized as 1) which environmental factors define the realized niche of the species, and what is the realized niche of these species, 2) how important are environmental gradients compared to biotic interactions in determining community structure, 3) to what extent do the realized niches of the species support theoretical models of coexistence, and 4) is relative population performance related to plant strategy in a predictive way? To document the realized niches, soil and water chemical data were collected from several softwater lakes across Western Europe, both in summer and winter. Two different approaches were used to model the niches, either using a regression framework (chapter 3; generalized linear mixed models and proportional odds models) or by the use of discriminant analysis (chapter 2). The former was applied at a West-European scale for five species, while the latter was applied at the scale of Belgium and The Netherlands for one species, Eleocharis multicaulis. Also, two different approaches of predictor selection were applied before analysis. One way (chapter 3) was to make a shortlist of mainly proximal environmental variables (resource gradients and direct gradients sensu Austin 2005). Another was to start with all information from a large environmental dataset, and to screen for potential explanatory variables using a univariate method (chapter 2). The approach that started from a reduced list of mainly proximal variables resulted in five predictors (see below) after model fitting (chapter 3). This method has the advantage of simplification in searching for relevant niche differences, as well as robustness for prediction because of model parsimony. To a large extent, the longer list of relevant variables in chapter 2 could be matched to these five predictors by correlation. Consequently one might prefer the ‘short list’. The short list was convenient to test some expected niche patterns from theoretical ecology in a regression framework. However, its scope remains limited when the aim is also to hypothesize on relevant biogeochemical processes that shape the species’ niche. In that case, the added explanatory value of several ionic ratios and macro-ionic concentrations becomes apparent, even though they are statistically correlated to other variables with explanatory power. It is therefore suggested that univariate screening techniques are a worthy alternative for variable selection in ecology. The advice to practitioners of either approach of variable selection, is to be conscious of the pros and cons of each approach, and clearly base their field design on explicit questions or hypotheses. The main groups of influential variables that came out of the analyses in chapter 2, were those that were related to 1) the redox potential (among which soil silicon and surface water ammonium/nitrate ratio), 2) acidity & base richness, 3) base cation replenishment during winter, 4) soil nitrogen availability (among which several ionic ratios), 5) soil phosphorus availability and 6) nutrient richness of the organic layer. Our hypothesis, that Eleocharis multicaulis is confined to an acid situation rich in ammonium and carbon dioxide, nutrient-poor water and soil and a mineral substrate devoid of organic material, did not hold as a whole. The five predictors obtained with the methods in chapter 3, were: four surface water variables - calcium and the resource variables ammonium, nitrate and orthophosphate - and the soil silicon content, which was related to surface water pH. Nitrate was negatively correlated with the ammonium/nitrate ratio, which was actually used in the regression models. Calcium was related to pH and alkalinity of the surface water. Four of the five investigated species showed clear patterns in relation to the finally selected gradients in chapter 3. One species (Juncus bulbosus) was very unspecific; its realized niche was not limited within the investigated environmental range and vegetation type. The two species entitled ‘characteristic’ of the vegetation alliance (Eleocharis multicaulis and Hypericum elodes) showed opposite differences with respect to nutrients and silicon. Agrostis canina, Hypericum elodes and Hydrocotyle vulgaris were similar in their reaction to these predictors but with different positions along the gradient. As Hypericum elodes (characteristic of the Eleocharition), Agrostis canina and Hydrocotyle vulgaris (non-characteristic) were positively associated, the distinction between characteristic and non-characteristic species does not hold. Eleocharis multicaulis reacted positively to nitrogen ions, as long as nitrate remained dominant over ammonium (ammonium/nitrate ratio < 1), phosphorus and potassium remained low, the water was acid and had a high redox potential. The importance of ionic ratios as potential predictors of plant performance and abundance received special attention in this thesis. Based on a literature search of habitat studies, we give an overview of ionic ratios that have been used. For most of the ratios there is a physiological explanation how they directly act upon the plant. It can be concluded that several ionic ratios have the potential of being a strong predictor for species abundance or occurrence. Modellers of plant’s niches and performance should more often consider ionic ratios instead of testing only for simple nutrient or macro-ionic gradients. In order to assess the relative importance of nutrient availability, surface water level fluctuations and species identity in the organization of Eleocharition communities, a laboratory experiment was conducted (chapter 4). Softwater lake habitat was simulated during one growing season, mimicking water level fluctuation, photoperiod and temperature. Artificial communities consisted of small populations of four softwater macrophyte species: Luronium natans, Baldellia ranunculoides ssp. repens, Eleocharis multicaulis and Hydrocotyle vulgaris. These communities were subjected to two levels of aqueous carbon dioxide and ammonium. Additionally, monocultures of Baldellia ranunculoides ssp. repens and Eleocharis multicaulis were grown at the higher nutrient level combination in order to measure their competitive response in community. During one growing season it was observed that water level drawdown was very influential on relative species performances and ranking order, most likely through its effect on the availability of aerial carbon dioxide to the plants. However, the water level drop acted as a disturbance to Baldellia ranunculoides ssp. repens at least. Species performance generally was not influenced by the ammonium or carbon dioxide nutrient treatments. This stressed the importance of intrinsic properties of the species (their identity) and as such, their presence, to the community outcome. Furthermore, a large competitive response was observed in both Eleocharis multicaulis and Baldellia ranunculoides ssp. repens, depending on performance measure. Hence, at least in the amphibious Eleocharition and within a certain water level range, it is possible that the interactions between the occurring species are more important than the precise physicochemical characteristics, as long as species are able to survive and grow in the given environment. Put otherwise, this means that it could be that their fundamental niche is a lot larger than their realized niche, which then greatly depends on the presence and densities of species at sites. However, much more long-term experimentation on fundamental niches and interactions is needed to fully test this hypothesis. At the local scale, niches are a convenient instrument to describe conditions under which species survive, reproduce and coexist in a community. The concept is especially used to visualize how biotic interactions shape the realized niche of a species, as opposed to its physiological capabilities, also known as its fundamental niche. Why species can be the ‘winner’ only in a part of their fundamental niche, is supposed to be linked to a trade-off. For example, a high ability to quickly take advantage of new light and nutrient supplies would not be compatible with a high ability to compete in a more resource-limited environment. This trade-off is made more explicit in the competitive hierarchy model of Keddy (1989). Particularly, there is growing evidence that species are able to coexist in communities through niche separation, in the case of resource gradients often referred to as resource partitioning. At a much larger spatial scale, the realized niche concept has been broadened in order to apply to the niches that are modelled by species distribution models at biogeographical scale (also referred to as habitat suitability models or ecological niche models). At this scale, one might expect that in very different areas, mechanisms that account for the realized niche may differ. This would result in realized niches that are closer to the fundamental niches than when they were measured at a small geographical scale. Moreover, disturbances and stochastic demography may induce a reduced – or delayed – competition intensity, potentially leading to a larger realized niche. However, in this study, considerable niche differentiation between species was still found at the investigated West-European scale (chapter 3). Some other researchers tested for mechanisms of regional coexistence of species in a metacommunity, and also found evidence for niche differentiation. Results from the current study supported the expectations from classical coexistence theory by niche differentiation, at least on the regional scale, as subplot environmental heterogeneity was not assessed. If niche separation through competition is to be responsible for coexistence at the subplot scale, fine-scale environmental heterogeneity is needed to accommodate the species. Furthermore, we must accept that competition is not the only possible biotic factor that shapes the investigated plant community and leads to coexistence. For example, also facilitation might be important. Moreover, stochastic processes of species arrival at a site and subsequent (temporary) establishment may be superposed on deterministic processes of abiotic and biotic interactions, allowing more possible patterns of coexistence. The functional (dis)similarities between the five species coincided with their observed niche overlaps and differences (chapter 3). The relative species ranking could not be inferred from their strategies, however. Neither was this the case in the community experiment in chapter 4, in which partly different species were used. Here it was attributed both to the mixed strategies of all species and the high morphological plasticity of some. According to the paradigm of coexistence through niche differentiation, fitness trade-offs underpin niche separation. The match between niche patterns and plant strategy provided one indication of this, while another was that dominance hierarchies could reverse depending on the resource gradient considered. This reversal of dominance hierarchy might be the consequence of a physiological trade-off in accessing one versus another resource, and hence in fundamental niche width along these resources. The pattern along each single resource gradient was in support of the competitive hierarchy model. The restoration of the abiotic environment of target species and the enhancement of their successful colonization and establishment in a spatial context, are important keys to success in nature restoration. This thesis contributes to the knowledge that is needed to restore the Eleocharition multicaulis, with: 1) predictive niche models of five species, 2) a biogeochemical and ecosystem-wide interpretation of the requirements of the characteristic species Eleocharis multicaulis, and 3) the insight that plant-plant interactions are important to relative population performance, at least in the short term, thereby forcing the conservation biologist to consider aspects of colonization (in a spatial and temporal context), and not only environmental constraints. The thesis concludes with opportunities for further research on the Eleocharition multicaulis, suggestions in order to improve niche modelling habits by ecologists, and challenges and obstacles in the testing of ecological theory.
Keywords
competitive response, community assembly, hypericum elodes, species distribution model, amphibious, softwater, eleocharis multicaulis, population, niche, luronium natans, competition, proportional odds model, ionic ratios, discriminant analysis

Downloads

    • full text
    • |
    • UGent only
    • |
    • PDF
    • |
    • 11.11 MB

Citation

Please use this url to cite or link to this publication:

Chicago
Vanderhaeghe, Floris. 2012. “Niche Theory and Competitive Interactions in an Amphibious Plant Community of West-European Softwater Lakes.” PhD Theses of the Research Institute for Nature and Forest. Brussels, Belgium: Research Institute for Nature and Forest (INBO).
APA
Vanderhaeghe, F. (2012). Niche theory and competitive interactions in an amphibious plant community of West-European softwater lakes. PhD theses of the Research Institute for Nature and Forest. Research Institute for Nature and Forest (INBO), Brussels, Belgium.
Vancouver
1.
Vanderhaeghe F. Niche theory and competitive interactions in an amphibious plant community of West-European softwater lakes. PhD theses of the Research Institute for Nature and Forest. [Brussels, Belgium]: Research Institute for Nature and Forest (INBO); 2012.
MLA
Vanderhaeghe, Floris. “Niche Theory and Competitive Interactions in an Amphibious Plant Community of West-European Softwater Lakes.” PhD theses of the Research Institute for Nature and Forest 2012 : n. pag. Print.
@phdthesis{3059903,
  abstract     = {In this thesis, several questions are raised that are related to the niche concept and population ecology. They are within a framework of the amphibious softwater plant community Eleocharition multicaulis Vanden Berghen 1969. This community has declined during the 20th century and several of its species are rare or endangered. The community, with typical species like Eleocharis multicaulis, Hypericum elodes and Scirpus fluitans, has not been the subject of much research as yet. Therefore, more research is needed to better understand the requirements of Eleocharition species and the mechanisms that rule community structure, if we want to know how to restore their habitat most effectively.
The research questions in this thesis can be broadly summarized as 1) which environmental factors define the realized niche of the species, and what is the realized niche of these species, 2) how important are environmental gradients compared to biotic interactions in determining community structure, 3) to what extent do the realized niches of the species support theoretical models of coexistence, and 4) is relative population performance related to plant strategy in a predictive way?
To document the realized niches, soil and water chemical data were collected from several softwater lakes across Western Europe, both in summer and winter. Two different approaches were used to model the niches, either using a regression framework (chapter 3; generalized linear mixed models and proportional odds models) or by the use of discriminant analysis (chapter 2). The former was applied at a West-European scale for five species, while the latter was applied at the scale of Belgium and The Netherlands for one species, Eleocharis multicaulis. Also, two different approaches of predictor selection were applied before analysis. One way (chapter 3) was to make a shortlist of mainly proximal environmental variables (resource gradients and direct gradients sensu Austin 2005). Another was to start with all information from a large environmental dataset, and to screen for potential explanatory variables using a univariate method (chapter 2). The approach that started from a reduced list of mainly proximal variables resulted in five predictors (see below) after model fitting (chapter 3). This method has the advantage of simplification in searching for relevant niche differences, as well as robustness for prediction because of model parsimony. To a large extent, the longer list of relevant variables in chapter 2 could be matched to these five predictors by correlation. Consequently one might prefer the {\textquoteleft}short list{\textquoteright}. The short list was convenient to test some expected niche patterns from theoretical ecology in a regression framework. However, its scope remains limited when the aim is also to hypothesize on relevant biogeochemical processes that shape the species{\textquoteright} niche. In that case, the added explanatory value of several ionic ratios and macro-ionic concentrations becomes apparent, even though they are statistically correlated to other variables with explanatory power. It is therefore suggested that univariate screening techniques are a worthy alternative for variable selection in ecology. The advice to practitioners of either approach of variable selection, is to be conscious of the pros and cons of each approach, and clearly base their field design on explicit questions or hypotheses.
The main groups of influential variables that came out of the analyses in chapter 2, were those that were related to 1) the redox potential (among which soil silicon and surface water ammonium/nitrate ratio), 2) acidity \& base richness, 3) base cation replenishment during winter, 4) soil nitrogen availability (among which several ionic ratios), 5) soil phosphorus availability and 6) nutrient richness of the organic layer. Our hypothesis, that Eleocharis multicaulis is confined to an acid situation rich in ammonium and carbon dioxide, nutrient-poor water and soil and a mineral substrate devoid of organic material, did not hold as a whole. The five predictors obtained with the methods in chapter 3, were: four surface water variables - calcium and the resource variables ammonium, nitrate and orthophosphate - and the soil silicon content, which was related to surface water pH. Nitrate was negatively correlated with the ammonium/nitrate ratio, which was actually used in the regression models. Calcium was related to pH and alkalinity of the surface water.
Four of the five investigated species showed clear patterns in relation to the finally selected gradients in chapter 3. One species (Juncus bulbosus) was very unspecific; its realized niche was not limited within the investigated environmental range and vegetation type. The two species entitled {\textquoteleft}characteristic{\textquoteright} of the vegetation alliance (Eleocharis multicaulis and Hypericum elodes) showed opposite differences with respect to nutrients and silicon. Agrostis canina, Hypericum elodes and Hydrocotyle vulgaris were similar in their reaction to these predictors but with different positions along the gradient. As Hypericum elodes (characteristic of the Eleocharition), Agrostis canina and Hydrocotyle vulgaris (non-characteristic) were positively associated, the distinction between characteristic and non-characteristic species does not hold. Eleocharis multicaulis reacted positively to nitrogen ions, as long as nitrate remained dominant over ammonium (ammonium/nitrate ratio {\textlangle} 1), phosphorus and potassium remained low, the water was acid and had a high redox potential.
The importance of ionic ratios as potential predictors of plant performance and abundance received special attention in this thesis. Based on a literature search of habitat studies, we give an overview of ionic ratios that have been used. For most of the ratios there is a physiological explanation how they directly act upon the plant. It can be concluded that several ionic ratios have the potential of being a strong predictor for species abundance or occurrence. Modellers of plant{\textquoteright}s niches and performance should more often consider ionic ratios instead of testing only for simple nutrient or macro-ionic gradients.
In order to assess the relative importance of nutrient availability, surface water level fluctuations and species identity in the organization of Eleocharition communities, a laboratory experiment was conducted (chapter 4). Softwater lake habitat was simulated during one growing season, mimicking water level fluctuation, photoperiod and temperature. Artificial communities consisted of small populations of four softwater macrophyte species: Luronium natans, Baldellia ranunculoides ssp. repens, Eleocharis multicaulis and Hydrocotyle vulgaris. These communities were subjected to two levels of aqueous carbon dioxide and ammonium. Additionally, monocultures of Baldellia ranunculoides ssp. repens and Eleocharis multicaulis were grown at the higher nutrient level combination in order to measure their competitive response in community. During one growing season it was observed that water level drawdown was very influential on relative species performances and ranking order, most likely through its effect on the availability of aerial carbon dioxide to the plants. However, the water level drop acted as a disturbance to Baldellia ranunculoides ssp. repens at least. Species performance generally was not influenced by the ammonium or carbon dioxide nutrient treatments. This stressed the importance of intrinsic properties of the species (their identity) and as such, their presence, to the community outcome. Furthermore, a large competitive response was observed in both Eleocharis multicaulis and Baldellia ranunculoides ssp. repens, depending on performance measure. Hence, at least in the amphibious Eleocharition and within a certain water level range, it is possible that the interactions between the occurring species are more important than the precise physicochemical characteristics, as long as species are able to survive and grow in the given environment. Put otherwise, this means that it could be that their fundamental niche is a lot larger than their realized niche, which then greatly depends on the presence and densities of species at sites. However, much more long-term experimentation on fundamental niches and interactions is needed to fully test this hypothesis.
At the local scale, niches are a convenient instrument to describe conditions under which species survive, reproduce and coexist in a community. The concept is especially used to visualize how biotic interactions shape the realized niche of a species, as opposed to its physiological capabilities, also known as its fundamental niche. Why species can be the {\textquoteleft}winner{\textquoteright} only in a part of their fundamental niche, is supposed to be linked to a trade-off. For example, a high ability to quickly take advantage of new light and nutrient supplies would not be compatible with a high ability to compete in a more resource-limited environment. This trade-off is made more explicit in the competitive hierarchy model of Keddy (1989). Particularly, there is growing evidence that species are able to coexist in communities through niche separation, in the case of resource gradients often referred to as resource partitioning.
At a much larger spatial scale, the realized niche concept has been broadened in order to apply to the niches that are modelled by species distribution models at biogeographical scale (also referred to as habitat suitability models or ecological niche models). At this scale, one might expect that in very different areas, mechanisms that account for the realized niche may differ. This would result in realized niches that are closer to the fundamental niches than when they were measured at a small geographical scale. Moreover, disturbances and stochastic demography may induce a reduced -- or delayed -- competition intensity, potentially leading to a larger realized niche.
However, in this study, considerable niche differentiation between species was still found at the investigated West-European scale (chapter 3). Some other researchers tested for mechanisms of regional coexistence of species in a metacommunity, and also found evidence for niche differentiation. Results from the current study supported the expectations from classical coexistence theory by niche differentiation, at least on the regional scale, as subplot environmental heterogeneity was not assessed. If niche separation through competition is to be responsible for coexistence at the subplot scale, fine-scale environmental heterogeneity is needed to accommodate the species. Furthermore, we must accept that competition is not the only possible biotic factor that shapes the investigated plant community and leads to coexistence. For example, also facilitation might be important. Moreover, stochastic processes of species arrival at a site and subsequent (temporary) establishment may be superposed on deterministic processes of abiotic and biotic interactions, allowing more possible patterns of coexistence.
The functional (dis)similarities between the five species coincided with their observed niche overlaps and differences (chapter 3). The relative species ranking could not be inferred from their strategies, however. Neither was this the case in the community experiment in chapter 4, in which partly different species were used. Here it was attributed both to the mixed strategies of all species and the high morphological plasticity of some. According to the paradigm of coexistence through niche differentiation, fitness trade-offs underpin niche separation. The match between niche patterns and plant strategy provided one indication of this, while another was that dominance hierarchies could reverse depending on the resource gradient considered. This reversal of dominance hierarchy might be the consequence of a physiological trade-off in accessing one versus another resource, and hence in fundamental niche width along these resources. The pattern along each single resource gradient was in support of the competitive hierarchy model.
The restoration of the abiotic environment of target species and the enhancement of their successful colonization and establishment in a spatial context, are important keys to success in nature restoration. This thesis contributes to the knowledge that is needed to restore the Eleocharition multicaulis, with: 1) predictive niche models of five species, 2) a biogeochemical and ecosystem-wide interpretation of the requirements of the characteristic species Eleocharis multicaulis, and 3) the insight that plant-plant interactions are important to relative population performance, at least in the short term, thereby forcing the conservation biologist to consider aspects of colonization (in a spatial and temporal context), and not only environmental constraints.
The thesis concludes with opportunities for further research on the Eleocharition multicaulis, suggestions in order to improve niche modelling habits by ecologists, and challenges and obstacles in the testing of ecological theory.},
  author       = {Vanderhaeghe, Floris},
  isbn         = {9789040303289},
  keyword      = {competitive response,community assembly,hypericum elodes,species distribution model,amphibious,softwater,eleocharis multicaulis,population,niche,luronium natans,competition,proportional odds model,ionic ratios,discriminant analysis},
  language     = {eng},
  pages        = {VII, 226},
  publisher    = {Research Institute for Nature and Forest (INBO)},
  school       = {Ghent University},
  series       = {PhD theses of the Research Institute for Nature and Forest},
  title        = {Niche theory and competitive interactions in an amphibious plant community of West-European softwater lakes},
  volume       = {INBO.T.2012.1},
  year         = {2012},
}