Search

Ecosystem engineers have large impacts on the communities in which they live, and these impacts may feed back to populations of engineers themselves. In this study, we assessed the effect of ecosystem engineering on density-dependent feedbacks for midges in Lake Mývatn, Iceland. The midge larvae reside in the sediment and build silk tubes that provide a substrate for algal growth, thereby elevating benthic primary production. Benthic algae are in turn the primary food source for the midge larvae, setting the stage for the effects of engineering to feed back to the midges themselves. Using a field mesocosm experiment manipulating larval midge densities, we found a generally positive but nonlinear relationship between density and benthic production. Furthermore, adult emergence increased with the primary production per midge larva. By combining these two relationships in a simple model, we found that the positive effect of midges on benthic production weakened negative density dependence at low to intermediate larval densities. However, this benefit disappeared at high densities when midge consumption of primary producers exceeded their positive effects on primary production through ecosystem engineering. Our results illustrate how ecosystem engineering can alter density-dependent feedbacks for engineer populations.

We created 24 mesocosms containing mixtures of herbaceous species arranged along experimentally maintained gradients of stress (external abiotic constraints limiting biomass production) and disturbance (events causing partial or total live biomass destruction) in order to determine the degree to which community assembly is deterministic or historically contingent during succession. In this 7-yr experiment, we found taxonomic divergence and functional convergence during the last 3 yr. Although communities became more functionally dissimilar as the difference in the level of stress increased, they were equally taxonomically different irrespective of the amount of difference between them in terms of stress and disturbance. In addition, comparing communities experiencing the same conditions, taxonomic community structure was more dissimilar as the levels of stress and disturbance decreased. Therefore, community assembly was largely deterministic from a functional perspective but more historically contingent from a taxonomic perspective, and the relative importance of taxonomic historical contingency decreased as the levels of stress and disturbance frequency increased.

Habitat heterogeneity is a primary mechanism influencing species richness. Despite the general expectation that increased heterogeneity should increase species richness, there is considerable variation in the observed relationship, including many studies that show negative effects of heterogeneity on species richness. One mechanism that can create such disparate results is the predicted trade-off between habitat area and heterogeneity, sometimes called the area-heterogeneity-trade-off (AHTO) hypothesis. The AHTO hypothesis predicts positive effects of heterogeneity on species richness in large habitats, but negative effects in small habitats. We examined the interplay between habitat size and habitat heterogeneity in experimental mesocosms that mimic freshwater ponds, and measured responses in a species-rich zooplankton community. We used the AHTO hypothesis and related mechanisms to make predictions about how heterogeneity would affect species richness and diversity in large compared to small habitats. We found that heterogeneity had a positive influence on species richness in large, but not small habitats, and that this likely resulted because habitat specialists were able to persist only when habitat size was sufficiently large, consistent with the predictions of the AHTO hypothesis. Our results emphasize the importance of considering context (e.g., habitat size in this case) when investigating the relative importance of ecological drivers of diversity, like heterogeneity.

Most land plants acquire nitrogen (N) through associations with arbuscular (AM) and ectomycorrhizal (ECM) fungi, but these symbionts employ contrasting strategies for N acquisition, which may lead to different stocks of soil carbon (C). We experimentally test this hypothesis with a mesocosm system where AM and ECM tree seedling roots, or their hyphae only, could access mineral soils with 13 C- and 15 N-enriched organic matter. We quantified loss of soil C and N, plant uptake of N and new inputs of plant C to soil. We found that AM, but not ECM, seedlings reduced soil C relative to controls. Soil C loss was greater in the presence of roots relative to hyphae only for both AM and ECM seedlings, but was correlated with plant N uptake for AM seedlings only. While new plant C inputs stimulated soil C loss in both symbioses, we detected plant C inputs more frequently and measured higher rates of decomposer activity in soils colonized by AM relative to ECM seedlings. Our study experimentally demonstrates how mycorrhizal strategies for N can affect soil C and C:N, even at the scale of an individual plant. Such effects may contribute to broad patterns in soil C across terrestrial ecosystems.

Identifying and quantifying the relevant properties of habitat structure that mediate predator-prey interactions remains a persistent challenge. Most previous studies investigate effects of structural density on trophic interactions and typically quantify refuge quality using one or two-dimensional metrics. Few consider spatial arrangement of components (i.e., orientation and shape) and often neglect to measure the total three-dimensional (3D) space available as refuge. This study tests whether the three-dimensionality of interstitial space, an attribute produced by the spatial arrangement of oyster (Crassostrea virginica) shells, impacts the foraging success of nektonic predators (primary blue crab, Callinectes sapidus) on mud crab prey (Eurypanopeus depressus) in field and mesocosm experiments. Interstices of 3D-printed shell mimics were manipulated by changing either their orientation (angle) or internal shape (crevice or channel). In both field and mesocosm experiments, under conditions of constant structural density, predator foraging success was influenced by 3D aspects of interstitial space. Proportional survivorship of tethered mud crabs differed significantly as 3D interstitial space varied by orientation, displaying decreasing prey survivorship as angle of orientation increased (0° = 0.76, 22.5° = 0.13, 45° = 0.0). Tethered prey survivorship was high when 3D interstitial space of mimics was modified by internal shape (crevice = 0.89, channel = 0.96) and these values did not differ significantly. In mesocosms, foraging success of blue crabs varied with 3D interstitial space as mean proportional survivorship (± SE) of mud crabs was significantly lower in 45° (0.27 ± 0.06) versus 0° (0.86 ± 0.04) orientations and for crevice (0.52 ± 0.11) versus channel shapes (0.95 ± 0.01). These results suggest that 3D aspects of interstitial space, which have direct relevance to refuge quality, can strongly influence foraging success in our oyster reef habitat. Our findings highlight the importance of spatial arrangement in mediating consumptive pathways in hard-structured habitats and demonstrate how quantifying the three-dimensionality of living space captures aspects of habitat structure that have been missing from previous empirical studies of trophic interactions and structural complexity. This article is protected by copyright. All rights reserved.

The fluxes of energy, matter, and organisms are important structuring forces of metaecosystems. Such ecosystem fluxes likely interact with environmental heterogeneity and differentially affect the diversity of multiple communities. In an aquatic mesocosm experiment, we tested how ecosystem flux and patch heterogeneity affected the diversity of bacteria, phytoplankton, and zooplankton metacommunities, and the structure and functioning of metaecosystems. We built metaecosystems consisting of three mesocosms that were either connected by flux of living organisms, organic material, and nutrients (alive ecosystem flux) or only by flux of organic material and nutrients (dead ecosystem flux). The three patches of each metaecosystem were either homogeneous or heterogeneous in nutrient loading. We found that the three groups of organisms responded differently to our treatments: flux of living organisms increased bacterial diversity irrespective of nutrient heterogeneity, while flux effects on phytoplankton diversity depended on nutrient heterogeneity, potentially indicating source-sink effects. Although zooplankton diversity was largely unaffected by our manipulations, subtle changes of community composition in response to ecosystem flux had strong effects on lower trophic levels, highlighting the importance of indirect flux effects via alterations in trophic interactions. Furthermore, differential effects of communities on the mean and spatial variability of local abiotic environments influenced the development of metaecosystem heterogeneity through time. Despite identical nutrient loading at the scale of the metaecosystem, abiotic conditions diverged between homogeneous and heterogeneous metaecosystems. For example, concentrations in dissolved organic carbon (DOC) were higher in homogeneous than heterogeneous metaecosystems, possibly because of differential responses of the algal community to local environmental conditions. Similarly, we found that flux effects on organisms translated into effects on DOC concentrations at the patch level, suggesting that flux-mediated changes in abundances of species can alter abiotic conditions. Our study shows that the dynamics of biotic and abiotic compartments of spatially structured ecosystems are intricately linked, highlighting the importance of integrating metacommunity and metaecosystem perspectives.

Maintaining multiple ecological functions (“multifunctionality”) is crucial to sustain viable ecosystems. To date most studies on biodiversity-ecosystem functioning (BEF) have focused on single or few ecological functions and services. However, there is a critical need to evaluate how species and species assemblages affect multiple processes at the same time, and how these functions are interconnected. Dung beetles represent excellent model organisms because they are key contributors to several ecosystem functions. Using a novel method based on the application of 15N-enriched dung in a mesocosm field experiment, we assessed the role of dung beetles in regulating multiple aspects of nutrient cycling in alpine pastures over appropriate spatial (up to a soil depth of 20 cm) and temporal (up to 1 yr after dung application) scales. 15N isotope tracing allowed the evaluation of multiple interrelated ecosystem functions responsible for the cycling of dung-derived nitrogen (DDN) in the soil and vegetation. We also resolved the role of functional group identity and the importance of interactions among co-occurring species for sustaining multiple functions by focusing on two different dung beetle nesting strategies (tunnelers and dwellers). Species interactions were studied by contrasting mixed-species to single-species assemblages, and asking whether the former performed multiple functions better than the latter. Dung beetles influenced at least seven ecological functions by facilitating dung removal, transport of DDN into the soil, microbial ammonification and nitrification processes, uptake of DDN by plants, herbage growth, and changes in botanical composition. Tunnelers and dwellers were found to be similarly efficient for most functions, with differences based on the spatial and temporal scales over which the functions operated. Although mixed-species assemblages seemed to perform better than single-species, this outcome may be dependent on the context. Most importantly though, the different functions were found to be interconnected sequentially as reveled by analyzing 15N content in dung, soil and vegetation. Taken together, our current findings offer strong support for the contention that the link between biodiversity and ecosystem functions should be examined not function by function, but in terms of understanding multiple functions and how they interact with each other.

Diel variability in nutrient concentrations is common but not universal in aquatic ecosystems. Theoretical models of photoautotrophic systems attribute the absence of diel uptake variation to nutrient scarcity, such that diel variability in nutrient uptake disappears as nutrients become limiting. We tested this prediction in a mesocosm experiment, by exposing benthic algal communities to a range of nitrogen (N) and phosphorus concentrations and recording the rates of uptake during both day and night. We found that higher concentrations of N produced diel variability in uptake and that the difference between the day and night total mass uptakes approximately equaled N demand for observed primary production as seen in other studies. At lower concentrations of N, uptake rates during the day and night were indistinguishable. These results are the first empirical evidence to imply that diel nitrate patterns in streams and rivers indicate a release from N limitation and offer a new way to assess nutrient limitation.

The diversity and composition of local communities depends strongly on the pool of species that have been able to colonize that community from elsewhere. Typically this is thought to depend on a larger regional species pool that is subject to local environmental constraints that act as "filters." Often, however, colonists arrive from multiple sources that differ in habitat conditions and have therefore already experienced distinct "prefiltering." Consequently, it is the interaction of species from these distinct pools that determine the composition of local communities. This interaction is particularly important when certain colonist pools provide keystone species with disproportionate roles on community assembly. We propose to identify these key colonist pools and their interaction with local habitat filters by quantifying community-level responses to colonist pool manipulation. We tested this framework to assess the contribution of surface and burrow sediment bacteria to bacterial communities associated with the fiddler crab, Uca panacea. In a mesocosm experiment, we combined normal and autoclaved surface and burrow sediment in a factorial experimental design, and we evaluated the community-level responses of carapace and gut microbial assemblages to sediment treatments with next-generation sequencing of the 16S rRNA gene. Results from carapace bacterial communities indicate that burrow sediments contribute most recruits, but surface sediments provide a few key colonizers that become established in the carapace community. In contrast, the composition of gut-associated microbial communities responded only to surface bacteria manipulation, despite being highly dissimilar from the community composition in both the surface and burrow source pools. These results suggest that assembly in the gut depends primarily on colonization from the surface sediment and regulation by habitat filtering. For fiddler crab-associated bacteria, we can conclude that key colonist pools and habitat filters regulate the influence of multiple colonist pools. Incorporating and distinguishing the contribution of multiple sources of species, rather than a single regional species pool, may better explain community dynamics in many systems, especially those with weak habitat filters.

Some plant functional groups such as nonnative invasive and nitrogen (N)-fixing plants are widely thought to have consistent, coordinated differences in their functional traits relative to other groups such as native and non -N-fixing plants. Recent evidence suggests that these trait differences between groups can be context dependent, varying with environmental factors such as resource availability and disturbance. However, many previous comparisons among plant groups differing in invasion status have not standardized growth form between groups or have compared species that do not co-occur, which could result in invasion status per se being confounded with other factors. We determined growth and leaf functional trait responses of 20 co-occurring woody species, that is, five species within each of four functional groups (native N-fixers, native non -N-fixers, nonnative [invasive] N-fixers and nonnative [invasive] non-N-fixers), to factorial combinations of soil fertility and defoliation treatments in a mesocosm experiment to test each of two hypotheses. First, we hypothesized that nonnative invasive and N-fixing species will have functional traits associated with rapid resource acquisition whereas natives and non -N-fixing species will have traits linked to resource conservation. Second, we hypothesized that plant growth and leaf traits of nonnative and N-fixing species will be more strongly influenced by environmental factors (i.e., soil fertility and disturbance) than will natives and non-N-fixers. Plant growth, foliar nutrients, and leaf structural traits varied among plant functional groups in a manner consistent with our first hypothesis. Support for our second hypothesis was mixed; origin (native vs. nonnative) and soil fertility rarely interacted to determine plant growth or variation in leaf traits whereas interactions involving N-fixing ability and soil fertility were common. Further, there were no consistent interactive effects between plant groupings and disturbance. Our results demonstrate that variation in growth and functional traits among plant species were driven primarily by the relatively large responses of nonnative N-fixers to soil fertility, rather than by consistent differences between other plant functional groups. These findings highlight the importance of resource availability in determining trait or performance differences among plant functional groups, and provide insights into the assembly of plant functional traits in novel communities of co-occurring native and nonnative species.

Consumers can have strong effects on the biotic and abiotic dynamics of spatially-structured ecosystems. In metacommunities, dispersing consumers can alter local assembly dynamics either directly through trophic interactions or indirectly by modifying local environmental conditions. In aquatic systems, very little is known about how key grazers, such as Daphnia, structure the microbial diversity of metacommunities and influence bacterial-mediated ecosystem functions. In an outdoor mesocosm experiment with replicate metacommunities (two 300L mesocosms), we tested how the presence and absence of Daphnia and the initial density of the microbial community (manipulated via dilution) influenced the diversity and community structure of algae and bacteria, and several ecosystem properties (e.g. pH, dissolved substances) and functions (e.g. enzyme activity, respiration). We found that Daphnia strongly affected the local and regional diversity of both phytoplankton and bacteria, the taxonomic composition of bacterial communities, the biomass of algae, and ecosystem metabolism (i.e. respiration). Diluting the microbial inoculum (0.2-5μm size fraction) to the metacommunities increased local phytoplankton diversity, decreased bacteria beta-diversity, and changed the relative abundance of bacterial classes. Changes in the rank abundance of different bacterial groups exhibited phylogenetic signal, implying that closely related bacteria species might share similar responses to the presence of Daphnia. This article is protected by copyright. All rights reserved.

The stress-gradient hypothesis predicts a switch from competition to facilitation, under increasing environmental stress. However, it is unclear how important is the change in competition-facilitation balance (i.e., the net outcome of plant-plant interactions) along the stress gradient in the regulation of community temporal stability (i.e., the inverse of temporal variability in total biomass). Increasing environmental stress may enhance community temporal stability by reduced competition or eventually by leading to facilitative interactions between the dominant and subordinate species. Here, we present the results of a 5-yr mesocosm experiment that demonstrates the effects of interspecific interactions on the temporal stability of a riparian community across different drought-stress scenarios. We constructed artificial communities of dominant species (Carex elata) and three subordinate species to simulate the independent effects of environmental stress and interspecific interactions. Using removal of the dominant species, we evaluated the interplay of various mechanisms regulating the temporal stability of the subordinate species (competition-facilitation balance, species asynchrony, and dominant species stability). By simultaneous testing of these stabilizing mechanisms, we show their importance differs depending on environmental variability and harshness. The predominant role is taken by species asynchrony in a seasonally dry environment, whereas in a permanently dry environment, the importance of reduced competition increases. Reduced competition was stabilizing, in particular through increased total community biomass, whereas species asynchrony increased total community biomass and decreased biomass variation. These results suggest experiments and simulations that exclude interspecific interactions may not offer realistic predictions of the effects of changing hydrological regimes on ecosystem functioning.

Most taxa show consistent individual differences in behavior, a phenomenon often referred to as animal "personalities." While the links between individual personality and fitness have received considerable attention, little information is available on how animal personality impacts higher-order ecological processes, such as community dynamics. Using a mesocosm experiment, we subjected a representative community of alfalfa pests to different compositions of personality types of the wolf spider Pardosa milvina. We show that subtle variation in the personality composition of P. milvina populations generate wildly different prey communities, where a mixture of both active and sedentary individuals performs best at suppressing prey abundance. Our results provide the first experimental evidence that predator personality types can generate contrasting prey communities. Moreover, our results suggest that manipulating the representation of predator personality types may be a profitable avenue by which one can maximize the biocontrol potential of predator populations.

Ecological stoichiometry (ES) uses elemental ratios and mass balance to explain organismal growth, an important parameter in ecological systems. In this study, we tested quantitative predictions of the ES "minimal model" for the growth rates of two tadpole species (wood frogs, Lithobates sylvaticus and American toads, Anaxyrus americanus), by manipulating light and the quality of a leaf litter mixture in a seminatural mesocosm experiment. We predicted that wood frogs, which consume leaf litter as a resource, would respond more strongly to leaf litter quality than toads, which forage on periphyton and algae. The ES minimal model, parameterized from literature values, provided strikingly accurate quantitative predictions of nonlinear wood frog growth patterns across gradients of leaf litter quality, both in this experiment and when applied to previously published data on wood frog growth responses to various leaf litter species. In contrast, toad growth was best explained by the biomass of periphyton, which was driven primarily by light availability and only indirectly influenced by litter-derived soluble polyphenols. This study demonstrates the power of ES to predict organism growth rates, and highlights potential applications of this theory to predicting population- and community-level responses to changing forest environments.

Negative co-occurrence patterns are intriguing because they may reflect the outcome of interspecific interactions and therefore signal how competition shapes communities. However, other factors also contribute to these patterns. For example, theoretical studies as well as two survey-based studies have all suggested that dispersal may also impact these patterns. While natural communities commonly have nonrandom patterns of negative co-occurrence, understanding how different processes drive these patterns requires further research. We tested the influence of dispersal on co-occurrence patterns using a zooplankton mesocosm experiment with four different dispersal treatments varying in the number of dispersers delivered into mesocosms on regular intervals. Our dispersal treatments were intended to adjust the relative importance of dispersal and competition experienced within mesocosms (i.e., high dispersal results in a relatively low influence of competition on species composition and vice versa). Higher dispersal translated into increased zooplankton species richness and inter-mesocosm compositional similarity, and also changed species occupancy patterns such that species occurrences were more even across mesocosms in higher-dispersal treatments. Dispersal treatments also differed markedly in species co-occurrence patterns. Negative co-occurrence patterns were significant for all but the lowest-dispersal treatment, peaked in the intermediate-dispersal treatments, and declined in the highest-dispersal treatment. Stability analyses illustrate that co-occurrence differences are robust to the exclusion of any single mesocosm in null model analyses. Dispersal treatments did not significantly differ with respect to abiotic variation, which has been recognized as a potential driver of negative co-occurrence patterns. These results suggest that not only can dispersal influence patterns of negative co-occurrence via changes to species richness and distribution (occupancy patterns among mesocosms), but the degree to which they do so varies nonlinearly with the strength of dispersal. Critically, because negative co-occurrence patterns were nonsignificant when the contribution of dispersal was lowest, it is possible that dispersal contributes strongly to many observed patterns of negative co-occurrence. Consequently, great care should be taken prior to interpreting significant co-occurrence tests as a product of species interactions.

North American Atlantic salt marshes are generally considered to be nitrogen (N) limited systems, and plants within these marshes were historically thought to use only inorganic forms of N, such as NH4+. Recent research has suggested that Spartina alterniflora may take up some organic nitrogen compounds directly. To determine the availability of dissolved organic nitrogen (DON) to S. alterniflora in marshes along the North American Atlantic coast, porewater was sampled at eight field sites along the North American Atlantic Coast, ranging from Maine (44° N) to Florida (30° N). To determine if a latitudinal gradient existed in the assimilation of either DON or NH4+, replicated mesocosm experiments were conducted with S. alterniflora plants at three field sites (Massachusetts, Virginia, South Carolina), which represented three distinct ecotypes in the latitudinal range. To determine how microbial activity may influence DON availability, we also assessed rates of microbial activity. We found that porewater DO...

Stability contributes to the persistence of ecological communities, yet the interactions among different stabilizing forces are poorly understood. We assembled mesocosms with an algal resource and one to eight different clones of the consumer Daphnia ambigua and tracked algal and Daphnia abundances through time. We then fitted coupled ordinary differential equations (ODEs) to the consumer-resource time series. We show that variation in different components of stability (local stability and the magnitude of population fluctuations) across mesocosms arises through variation in life history traits and the functional processes represented by ODE model parameters. Local stability was enhanced by increased algal growth rate and Daphnia mortality and foraging rate. Population fluctuations were dampened by high Daphnia conversion efficiency and lower interaction strengths, low algal growth rate, high Daphnia death rate, and low Daphnia foraging. These results indicate that (1) stability in consumer-resource systems may arise through the net effect of multiple related stabilizing pathways and (2) different aspects of stability can vary independently and may respond in opposite directions to the same forces.

Leaf litter provides an important nutrient subsidy to headwater streams, but little is known about how tree genetics influence energy pathways from litter to higher trophic levels. Despite the charge to quantify carbon (C) and nitrogen (N) pathways from decomposing litter, the relationship between litter decomposition and aquatic consumers remains unresolved. We measured litter preference (attachments to litter), C and N assimilation rates, and growth rates of a shredding caddisfly (Hesperophylax magnus, Limnephilidae) in response to leaf litter of different chemical and physical phenotypes using Populus cross types (P. fremontii, P. angustifolia, and F1 hybrids) and genotypes within P. angustifolia. We combined laboratory mesocosm studies using litter from a common garden with a field study using doubly labeled litter (13 C and 15 N) grown in a greenhouse and incubated in Oak Creek, Arizona, USA. We found that, in the lab, shredders initially chose relatively labile (low lignin and condensed tannin concentrations, rapidly decomposing) cross type litter, but preference changed within 4 d to relatively recalcitrant (high lignin and condensed tannin concentrations, slowly decomposing) litter types. Additionally, in the lab, shredder growth rates were higher on relatively recalcitrant compared to labile cross type litter. Over the course of a three-week field experiment, shredders also assimilated more C and N from relatively recalcitrant compared to labile cross type litter. Finally, among P. angustifolia genotypes, N assimilation by shredders was positively related to litter lignin and C:N, but negatively related to condensed tannins and decomposition rate. C assimilation was likewise positively related to litter C:N, and also to litter %N. C assimilation was not associated with condensed tannins or lignin. Collectively, these findings suggest that relatively recalcitrant litter of Populus cross types provides more nutritional benefit, in terms of N fluxes and growth, than labile litter, but among P. angustifolia genotypes the specific trait of litter recalcitrance (lignin or tannins) determines effects on C or N assimilation. As shredders provide nutrients and energy to higher trophic levels, the influence of these genetically based plant decomposition pathways on shredder preference and performance may affect community and food web structure.

Ecological theory on biological invasions attempts to characterize the predictors of invasion success and the relative importance of the different drivers of population establishment. An outstanding question is how propagule pressure determines the probability of population establishment, where propagule pressure is the number of individuals of a species introduced into a specific location (propagule size) and their frequency of introduction (propagule number). Here, we used large-scale replicated mesocosm ponds over three reproductive seasons to identify how propagule size and number predict the probability of establishment of one of world's most invasive fish, Pseudorasbora parva, as well as its effect on the somatic growth of individuals during establishment. We demonstrated that, although a threshold of 11 introduced pairs of fish (a pair is 1 male, 1 female) was required for establishment probability to exceed 95%, establishment also occurred at low propagule size (1-5 pairs). Although single introduction events were as effective as multiple events at enabling establishment, the propagule sizes used in the multiple introductions were above the detected threshold for establishment. After three reproductive seasons, population abundance was also a function of propagule size, with rapid increases in abundance only apparent when propagule size exceeded 25 pairs. This was initially assisted by adapted biological traits, including rapid individual somatic growth that helped to overcome demographic bottlenecks.

Predator-prey relationships in poikilotherms are often size dependent, such as when adults of two interacting species are capable of eating juveniles of the other species. Such bi-directional predation can be important during the establishment and spread of an invading species, but its role remains poorly understood. Using a combination of laboratory and mesocosm experiments and field introductions, we demonstrate that guppies, Poecilia reticulata, prey on juvenile killifish, Rivulus hartii, and thereby facilitate their establishment in the habitat of a potential predator. Laboratory studies found that mature guppies can consume larval Rivulus, and experimental stream studies showed that guppies reduced the number of Rivulus surviving from eggs. Growth trials found that interspecific competition, while significant, cannot account for the declines in the survival of juvenile Rivulus seen in field surveys. Finally, a field experiment, in which guppies were introduced into previously guppy-free stream reaches, resulted in a marked reduction in the abundance of juvenile Rivulus relative to guppy-free controls. Together, these results indicate that reducing the native Rivulus population represents an important.mechanism promoting guppy invasion success.

Spatial self-organization can occur in many ecosystems with important effects on food web dynamics and the maintenance of biodiversity. The consumer-resource interaction is known to generate spatial patterning, but only a few empirical studies have investigated the effect of the consumer on resource distribution. Here we report results from a large aquatic mesocosm experiment used to investigate the effect of the consumer Daphnia magna on the distribution of its resource, the green algae Chlorella vulgaris. We maintained large tanks with capacity for 26 ,000 L with either algae or both algae and Daphnia in different temperature conditions. We found that the presence of D. magna inhibited spatial structure in algal distribution that arose as a consequence of increasing temperature. We conjecture that this homogenization effect might be caused by a combination of high mobility combined with high rates of algal consumption by Daphnia. Our study emphasizes the importance of both local constraints on growth and behavioral responses in either promoting or suppressing spatial self-organization in natural populations.

Polymorphisms play critical roles in allowing organisms to adapt to novel environments while enabling ecological speciation under divergent selection. Ambystoma maculatum, the spotted salamander, exhibits a unique polymorphism in the structure and appearance of its egg masses with two common morphs, white and clear. Amphibian egg jelly layers mediate interactions between embryos and the environment and are more responsive to ecological pressures of natural selection than other egg coat components. The A. maculatum egg mass polymorphism was hypothesized to be adaptive with regard to varying dissolved nutrient levels in ponds. We conducted two mesocosm experiments, collected field data, and constructed a population projection model to determine how dissolved nutrient levels affect embryonic and larval development and relate to the distribution of the morphs in natural ponds. We found that upon hatching there was an interaction between nutrient level and egg mass morph wherein individuals from white morphs were larger in low nutrient habitats. This interaction persisted throughout the larval stage, and along with the higher abundance of white morphs in ponds with low conductivity, we demonstrate that the white morph is advantageous in low nutrient environments. Our findings provide evidence for the role of environmental heterogeneity in enabling the persistence of a structural egg mass polymorphism, with maintenance occurring across multiple scales and persistence across its range. This indicates that polymorphisms can maximize performance in heterogeneous environments, while persisting over long timescales without leading to sympatric speciation.

It has been proposed that plant biodiversity may increase the erosion resistance of soils, yet direct evidence for any such relationship is lacking. We conducted a mesocosm experiment with eight species of riparian herbaceous plants, and found evidence that plant biodiversity significantly reduced fluvial erosion rates, with the eight-species polyculture decreasing erosion by 23% relative to monocultures. Species richness effects were largest at low levels of species richness, with little increase between four and eight species. Our results suggest that plant biodiversity reduced erosion rates indirectly through positive effects on root length and number of root tips, and that interactions between legumes and non-legumes were particularly important in producing biodiversity effects. Presumably, legumes increased root production of non-legumes by increasing soil nitrogen availability due to their ability to fix atmospheric nitrogen. Our data suggest that a restoration project using species from different functional groups might provide the best insurance to maintain long-term erosion resistance.

Three mesocosm experiments were performed in an outdoor facility to quantify the responses of five mangrove species grown from seedling to sapling stage to increasing rates of dissolved iron supply. Stem extension and biomass of mangroves were measured in the first two experiments, and in the third experiment, rates of microbial iron reduction were measured in relation to stem extension of two mangrove species. In all experiments, mangrove growth was enhanced by increasing iron supply, although some species showed iron toxicity at the higher supply rates. In the first two experiments, stem extension rates of Rhizophora apiculata, Bruguiera gymnorrhiza, and Xylocarpus moluccensis best fit Gaussian curves with maximal growth at supply rates of 50-60 mmol Fe x m(-2) x d(-1), whereas growth of Avicennia marina and Ceriops tagal increased to the highest rate (100 mmol Fe x m(-2) x d(-1)) of iron supply. Changes in leaf chlorophyll concentrations and iron content of roots mirrored the growth responses. In the third experiment, rates of microbial iron reduction were greater with R. apiculata and A. marina than in controls without plants; for both species, there was a positive relationship between stem extension and iron reduction. The rates of iron reduction and rates of iron supplied to the plants were well within the range of interstitial iron concentrations and rates of iron reduction found in the natural mangrove soils from which the seedlings were obtained. The responses of these species show that mangroves growing from seedling to sapling stage have a strong nutritional requirement for iron, and that there is a close relationship between plant roots and the activities of iron-reducing bacteria. These results suggest that mangrove growth may be limited in some natural forests by the rate at which iron is solubilized by iron-reducing bacteria. Such biogeochemical conditions have significant implications for successful recruitment, establishment, and early growth of mangroves.

We examined the relationship between the isolation of experimental aquatic mesocosms and the abundance of an aquatic insect colonist, Notonecta irrorata, over two years. We used a curve-fitting approach to assess whether linear or quadratic models better describe the relationship between isolation and abundance. For two measures of mesocosm isolation, distance to nearest source and distance to the largest source population, there was a significant quadratic relationship between isolation and abundance. Abundance of colonizing N. irrorata was not found to be significantly related to a third measure of isolation, mesocosm connectivity. These results indicate that the relationship between habitat isolation and colonist abundance may not be a monotonic decline across all spatial scales, a finding that contradicts the usual assumption incorporated in measures of habitat connectivity. Our results suggest that under some circumstances individuals that have undertaken dispersal may bypass patches they encounter early in this process and preferentially settle in patches encountered later. This behavioral preference in conjunction with decreased numbers of potential colonists at sites far from the source environment could lead to the "hump-shaped" colonist abundance by habitat isolation relationship we observed in this study. We suggest that simple assumptions about the relationship between habitat isolation and the probability a site is colonized need to be reexamined and alternative possible forms of this relationship tested.

The temporal stability of aggregate community and ecosystem properties is influenced by the variability of component populations, the interactions among populations, and the influence of environmental fluctuations on populations. Environmental fluctuations that enhance population variability are generally expected to destabilize community and ecosystem properties, but this will depend on the degree to which populations are synchronized in their dynamics. Here we use seminatural experimental ponds to show that reduced synchrony among zooplankton taxa increases the temporal stability of zooplankton density, abundance, and ecosystem productivity in fluctuating environments. However, asynchrony only occurs at long timescales (∼80-day periods) and under recurring environmental perturbations. At shorter timescales (∼10-day periods) and in constant environments, synchronous dynamics dominate. Our findings support recent theory indicating that compensatory dynamics can stabilize communities and ecosystems. They further indicate that environmental fluctuations can enhance the likelihood of long-period asynchrony and thus stabilize community and ecosystem properties despite their short term destabilizing effects.

Most plant diversity-function studies have been conducted in terrestrial ecosystems and have focused on plant productivity and nutrient uptake/retention, with a notable lack of attention paid to belowground processes (e.g., root dynamics, decomposition, trace gas fluxes). Here we present results from a mesocosm experiment in which we assessed how the richness of emergent macrophyte functional groups influences aboveground and belowground plant growth and microbial-mediated functions related to carbon and nitrogen cycling, with an emphasis on methane (CH4) efflux and potential denitrification rates. We found that an increase in the richness of wetland plant functional groups enhanced belowground plant biomass, altered rooting patterns, and decreased methane efflux, while having no effect on aboveground plant production or denitrification potential. We hypothesize that the greater root production and increased rooting depth in the highest diversity treatments enhanced CH4 oxidation to a relatively greater degree than methane production, leading to an overall decrease in CH4 efflux across our plant functional group richness gradient.

Periodical cicadas emerge from below ground every 13 or 17 years in North American forests, with individual broods representing the synchronous movement of trillions of individuals across geographic regions. Due to predator satiation, most individuals escape predation, die, and become deposited as detritus. Some of this emergent biomass falls into woodland aquatic habitats (small streams and woodland ponds) and serves as a high-quality allochthonous detritus pulse in early summer. We present results of a two-part study in which we (1) quantified deposition of Brood X periodical cicada detritus into woodland ponds and low-order streams in southwestern Ohio, and (2) conducted an outdoor mesocosm experiment in which we examined the effects of deposition of different amounts of cicada detritus on food webs characteristic of forest ponds. In the mesocosm experiment, we manipulated the amount of cicada detritus input to examine if food web dynamics and stability varied with the magnitude of this allochthonous resource subsidy, as predicted by numerous theoretical models. Deposition data indicate that, during years of periodical cicada emergence, cicada carcasses can represent a sizable pulse of allochthonous detritus to forest aquatic ecosystems. In the mesocosm experiment, cicada carcass deposition rapidly affected food webs, leading to substantial increases in nutrients and organism biomass, with the magnitude of increase dependent upon the amount of cicada detritus. Deposition of cicada detritus impacted the stability of organism functional groups and populations by affecting the temporal variability and biomass minima. However, contrary to theory, stability measures were not consistently related to the size of the allochthonous pulse (i.e., the amount of cicada detritus). Our study underscores the need for theory to further explore consequences of pulsed allochthonous subsidies for food web stability.

The stoichiometric light:nutrient hypothesis (LNH) links the relative supplies of key resources with the nutrient content of tissues of producers. This resource-driven variation in producer stoichiometry, in turn, can mediate the efficiency of grazing. Typically, discussions of the LNH attribute this resource-stoichiometry link to bottom-up effects of light and phosphorus, which are mediated through producer physiology. Emphasis on bottom-up effects implies that grazers must consume food of quality solely determined by resource supply to ecosystems (i.e., they eat what they are served). Here, we expand upon this largely bottom-up interpretation with evidence from pond surveys, a mesocosm experiment, and a model. Data from shallow ponds showed the "LNH pattern" (positive correlation of an index of light : phosphorus supply with algal carbon : phosphorus content). However, algal carbon : phosphorus content also declined as zooplankton biomass increased in the ponds. The experiment and model confirmed that this latter correlation was partially caused by the various bottom-up and top-down roles of grazers: the LNH pattern emerged only in treatments with crustacean grazers, not those without them. Furthermore, model and experiment clarified that another bottom-up factor, natural covariation of nitrogen : phosphorus ratios with light : phosphorus supply (as seen in ponds), does not likely contribute to the LNH pattern. Finally, the experiment produced correlations between shifts in species composition of algae, partially driven by grazing effects of crustaceans, and algal stoichiometry. These shifts in species composition might shape stoichiometric response of producer assemblages to resource supply and grazing, but their consequences remain largely unexplored. Thus, this study accentuated the importance of grazing for the LNH; de-emphasized a potentially confounding, bottom-up factor (covarying nitrogen : phosphorus supply); and highlighted an avenue for future research for the LNH (grazer-mediated shifts in producer composition).

Compensatory dynamics are an important suite of mechanisms that can stabilize community and ecosystem attributes in systems subject to environmental fluctuations. However, few experimental investigations of compensatory dynamics have addressed these mechanisms in systems of real-world complexity, and existing evidence relies heavily on correlative analyses, retrospective examination, and experiments in simple systems. We investigated the potential for compensatory dynamics to stabilize plankton communities in plankton mesocosm systems of real-world complexity. We employed four types of perturbations including two types of nutrient pulses, shading, and acidification. To quantify how communities responded to these perturbations, we used a measure of community-wide synchrony combined with spectral analysis that allowed us to assess timescale-specific community dynamics, for example, whether dynamics were synchronous at some timescales but compensatory at others. The 150-d experiment produced 32-point time series of all zooplankton taxa in the mesocosms. We then used those time series to evaluate total zooplankton biomass as an aggregate property and to evaluate community dynamics. For three of our four perturbation types, total zooplankton biomass was significantly less variable in systems with environmental variation than in constant environments. For the same three perturbation types, community-wide synchrony was much lower in fluctuating environments than in the constant environment, particularly at longer timescales (periods ≈ 60 d). Additionally, there were strong negative correlations between population temporal variances and the level of community-wide synchrony. Taken together, these results strongly imply that compensatory interactions between species stabilized total biomass in response to perturbations. Diversity did not differ significantly across either treatments or perturbation types, thus ruling out several classes of mechanisms driven by changes in diversity. We also used several pieces of secondary evidence to evaluate the particular mechanism behind compensatory responses since a wide variety of mechanisms are hypothesized to produce compensatory dynamics. We concluded that fluctuation dependent endogenous cycles that occur as a consequence of consumer-resource interactions in competitive communities were the most likely explanation for the compensatory dynamics observed in our experiment. As with our previous work, scale-dependent dynamics were also a key to understanding compensatory dynamics in these experimental communities.

Species and trophic richness often increase with habitat size. Although many ecological processes have been evoked to explain both patterns, the environmental stress associated with small habitats has rarely been considered. We propose that larger habitats may be species rich simply because their environmental conditions are within the fundamental niche of more species; larger habitats may also have more trophic levels if traits of predators render them vulnerable to environmental stress. We test this hypothesis using the aquatic insect larvae in water-filled bromeliads. In bromeliads, the probability of desiccation is greatest in small plants. For the 10 most common bromeliad insect taxa, we ask whether differences in drought tolerance and regional abundances between taxa predict community and trophic composition over a gradient of bromeliad size. First, we used bromeliad survey data to calculate the mean habitat size of occurrence of each taxon. Comparing the observed mean habitat size of occurrence to that expected from random species assembly based on differences in their regional abundances allowed us to obtain habitat size sensitivity indices (as Z scores) for the various insect taxa. Second, we obtained drought sensitivity indices by subjecting individual insects to drought and measuring the effects on relative growth rates in a mesocosm experiment. We found that drought sensitivity strongly, predicts habitat size sensitivity in bromeliad insects. However, an increase in trophic richness with habitat size could not be explained by an increased sensitivity of predators to drought, but rather by sampling effects, as predators were rare compared to lower trophic levels. This finding suggests that physiological tolerance to environmental stress can be relevant in explaining the universal increase in species with habitat size.

Invasive plant fecundity underlies propagule pressure and ultimately range expansion. Predicting fecundity across large spatial scales, from regions to landscapes, is critical for understanding invasion dynamics and optimizing management. However, to accurately predict fecundity and other demographic processes, improved models that scale individual plant responses to abiotic drivers across heterogeneous environments are needed. Here we combine two experimental data sets to predict fecundity of a widespread and problematic invasive grass over large spatial scales. First, we analyzed seed production as a function of plant biomass in a small-scale mesocosm experiment with manipulated light levels. Then, in a field introduction experiment, we tracked plant performance across 21 common garden sites that differed widely in available light and other factors. We jointly analyzed these data using a Bayesian hierarchical model (BHM) framework to predict fecundity as a function of light in the field. Our analysis reveals that the invasive species is likely to produce sufficient seed to overwhelm establishment resistance, even in deeply shaded environments, and is likely seed-limited across much of its range. Finally, we extend this framework to address the general problem of how to scale up plant demographic processes and analyze the factors that control plant distribution and abundance at large scales.

An ongoing debate in ecology revolves around how species composition and ecosystem function are related. To address the mechanistic controls of this relationship, we manipulated the composition of dissolved organic matter (DOM) fed to aquatic bacteria to determine effects on both bacterial activity and community composition. Sites along terrestrial to aquatic flow paths were chosen to simulate movement of DOM through catchments, and DOM was fed to downslope and control bacterial communities. Bacterial production was measured, and DOM chemistry and bacterial community composition (using denaturing gradient gel electrophoresis of 16S rRNA genes) were characterized following incubations. Bacterial production, dissolved organic carbon (DOC)-specific bacterial production, and DOC consumption were greatest in mesocosms fed soil water DOM; soil water DOM enhanced lake and stream bacterial production by 320-670% relative to lake and stream controls. Stream DOM added to lake bacteria depressed bacterial production relative to lake controls in the early season (-78%) but not the mid-season experiment. Addition of upslope DOM to stream and lake bacterial communities resulted in significant changes in bacterial community composition relative to controls. In four of five DOM treatments, the bacterial community composition converged to the DOM source community regardless of the initial inoculum. These results demonstrate that shifts in the supply of natural DOM were followed by changes in both bacterial production and community composition, suggesting that changes in function are likely predicated on at least an initial change in the community composition. The results indicate that variation in DOM composition of soil and surface waters influences bacterial community dynamics and controls rates of carbon processing in set patterns across the landscape.

A current principle of ecological stoichiometry states that the nitrogen to phosphorus ratio (N:P) of primary producers should closely match that from environmental nutrient supplies. This hypothesis was tested using data from ponds in Michigan, USA, a freshwater mesocosm experiment, a synthesis of studies from diverse systems (cultures, lakes, streams, and marine and terrestrial environments), and simple dynamic models of producer growth and nutrient content. Unlike prior laboratory studies, the N:P stoichiometry of phytoplankton in Michigan ponds clustered around and below the Redfield ratio (7.2:1 by mass), despite wide variation in N:P supply ratios (2:1-63:1 by mass) and the presence of grazers. In a mesocosm experiment, the N:P stoichiometry of phytoplankton cells again deviated from a nearly 1:1 relationship with N:P supply. Phytoplankton seston exhibited lower N:P content than expected at high N:P supply ratios, and often higher N:P content than anticipated at low N:P supply ratios, regardless of herbivore presence. Similar devi- ations consistently occur in the N:P stoichiometry of algae and plants in the other diverse systems. The models predicted that both high loss rates (sinking, grazing) and physiological limits to nutrient storage capacity could attenuate producer stoichiometry. In the future, research should evaluate how limits to elemental plasticity of producers can influence the role of stoichiometry in structuring communities and ecosystem processes.

While experiments are vital for understanding how ecological systems operate, different philosophies exist concerning how experiments should be conducted (e.g., Petranka 1989, Dunham and Beaupre 1998, Resetarits and Fauth 1998, Skelly and Kiesecker 2001; also see the special features in Herpetologica [1989; 45:111– 128] and Ecology [1996; 77:663–705, see Dahler and Strong 1996]). Recently, Skelly (2002) asked how experimental venue (i.e., cattle tanks set up as mesocosms in a field setting vs. screened enclosures placed into natural ponds) influences competitive interactions between two species of larval anurans (Pseudacris crucifer and Rana sylvatica) and how results from the two venues match a standard of realism. He observed that density affected competitive interactions among larval anurans in mesocosms but not in enclosures and concluded that enclosures were more realistic because the observed size of tadpoles measured in the field was more similar to the size of tadpoles predicted by the enclosure experiment than by the mesocosm experiment. Although an empirical examination of venue is valid, we believe that this study has serious flaws and claims differences between venues that erroneously devalue the use of mesocosms. Our goal is to reinterpret the results from Skelly (2002) in light of its design, point out methodological/statistical issues associated with his study, and argue that both venues can make meaningful contributions to the field of ecology if they are designed correctly with regard to the questions being asked and the specific population of interest.

Tadpole density dependence has repeatedly been studied in laboratories and mesocosms, but rarely in natural ponds. The present study manipulated density in three ponds; each pond was divided into one low density section and one high density section. The experiment was run for eight years, switching density treatments within ponds among years. This experimental design cannot definitively separate the experimental density effects from certain lag effects. This problem is discussed. Nonetheless, in all ponds tadpole survival was affected by the density manipulation. The reduced survival of larvae at high density was most pronounced in the pond with the overall highest density. In this pond, the density-dependent mortality almost offset the original, experimental, density difference between sections. Effects on tadpole and metamorph size and development rate were found in those cases in which density-dependent mortality was weak and treatment effects on density persisted to the later stages of tadpole development. However, even if significant, these were generally of small magnitude compared to the variation among ponds and years. 1 suggest that different modes and degrees of regulation among populations of one species of frog are to be expected, not only due to variation in pond quality but also due to variation in the surrounding terrestrial habitat. This may determine the balance between tadpole and adult population regulation.

Ecological theory suggests that community structure determines how eco- systems respond to perturbation. However, this idea is not well-tested empirically, in part because it is difficult to define the appropriate reference state for evaluating post-pertur- bation dynamics in field experiments. In this study, we (1) evaluated two different methods of determining the reference state, and (2) tested hypotheses derived from theory and previous experiments about the effects of herbivore community structure on autotroph responses to nutrient pulses in freshwater plankton communities. We hypothesized that (1) phytoplankton are less sensitive to a nutrient pulse with increasing zooplankton biomass, size, and Daphnia abundance, and (2) phytoplankton recover more quickly from the nutrient pulse with increased zooplankton biomass and increased abundance of taxa with rapid numerical responses, especially small-bodied cladocerans. We tested these hypotheses in field mesocosm experiments, which crossed four zooplankton community structures with two nutrient pulse treatments in a fully randomized factorial design. We found that phy- toplankton sensitivity decreased with increased zooplankton biomass, size, and Daphnia biomass regardless of the method used to define the reference state, supporting our first hypothesis and suggesting that results for sensitivity are robust to methodology. This result, together with previous work, clearly demonstrates that zooplankton have a strong effect on phytoplankton sensitivity. Further, as predicted, return times were shorter with increased zooplankton biomass and biomass of small cladocerans. However, contrary to expectations, return times also declined with Daphnia and copepod biomass, suggesting that the effects of zooplankton on return time are more complex than we hypothesized. Moreover, con- clusions for return time depended somewhat on methodology, indicating that we may need to reevaluate whether the concept of return time is relevant for field experiments.

Diel vertical migration (DVM) is a dynamic behavioral pattern found extensively in the world's oceans and lakes, yet the role of food and temperature distribution on DVM is still unclear. While DVM has been mostly studied in systems with surface food maxima, deep-water food maxima are quite common in lakes and oceans. In such ecosystems, optimal conditions of temperature and food are uncoupled. In a Swiss high-mountain lake (Oberer Arosasee) with a deep-water food maximum, we found that Daphnia galeata adults and juveniles exhibit DVM behavior almost throughout the year and migrated upwards, out of food-rich environments, at night. In a large indoor mesocosm experiment, we were able to show that Daphnia respond to fish-mediated cues by migrating into deeper water layers. In the presence of fish and using natural vertical food and temperature distributions in the mesocosms, we could reproduce the vertical distribution of adult Daphnia observed in the field. The indoor experiments show that food and temperature modulate the actual depth at which the animals stop migrating, whereas fish and ultraviolet radiation likely determine the timing of migration (its synchronization with dawn and dusk). Overall, our results show that the nighttime movement into the surface waters contributes to the fitness of Daphnia, given the costs and benefits associated with the trade-off between food and temperature. [KEYWORDS: deep-water chlorophyll maximum; diel vertical migration; enclosure experiment; fish kairomones; fish stocking; food–temperature trade-off; high-mountain lakes; phototaxis; plankton towers; zooplankton]