Top-down and bottom-up factors in tidepool communities

Recent studies suggest that nutrient variation influences rocky intertidal community structure, however empirical evidence is rare. In the Gulf of Maine, tidepools that occur on seagull feeding roosts are potentially subjected to regular nutrient loading from seagull guano. The results of a survey conducted on Swan's Island, ME show that roost tidepools have very low macroinvertebrate and macroalgal diversity as well as very high phytoplankton biomass compared to non-roost tidepools. An experiment presented here tested basic food chain hypotheses in tidepool communities. These basic food chain models predict that in a tidepool with one trophic level (phytoplankton only), phytoplankton biomass will increase when nutrients are enriched. In contrast, these models predict that in two trophic level tidepools (phytoplankton and mussels) herbivory will prevent an increase in phytoplankton biomass when nutrients are enriched. A short term 2×2 factorially designed field experiment was used to test this basic conceptual model using herbivory by mussels and enrichment with nitrogen as the main effects. The results of this investigation are consistent with the predictions of basic food chain models, and indicate that over the short time interval of a few days, herbivory by mussels is sufficient to maintain low phytoplankton levels following enrichment with nitrogen. Experimental enrichment with phosphorus in this study had no effect on phytoplankton biomass. The results of this study suggest that periodic pulses of nitrogen into tidepools will have little effect on phytoplankton biomass when mussels are present and that longer-term chronic nitrogen influxes may be driving the patterns of community structure in tidepools occurring on roosts.     

Divergent succession and implications for alternative states on rocky intertidal shores

It has been hypothesized that rockweed stands and mussel beds in sheltered bays in the Gulf of Maine, USA, are alternative community states. As a test of this hypothesis, experimental clearings of different sizes were established in stands of the rockweed Ascophyllum nodosum (L.) Le Jolis to determine if successional changes in large clearings developed species assemblages distinctly different from the surrounding A. nodosum stands. Clearings ranging from 1 to 8 m in diameter were created at 12 sites in 4 bays on Swan's Island, Maine, in 1996 to mimic the effects of ice scour, and abundances of gastropods, barnacles, mussels and fucoid algae were monitored until 2002. ANOVAs and MDS showed strong effects of clearing size and divergent successional changes in large clearings. Large clearings were quickly filled in and remained dominated by the alga Fucus vesiculosus L. and the barnacle Semibalanus balanoides (L). There was no evidence for site-specific effects, and Mantel tests showed clearing size was a better predictor of species composition than geographic distances among sites. Results suggest that large pulse disturbances using clearings of 8 m in diameter can initiate divergent successional pathways and have a protracted effect on species composition. Results are also consistent with the hypothesis that mussel beds and rockweed stands in sheltered bays may be alternative community states.     

Generalizing from experiments: is predation strong or weak in the New England rocky intertidal?

Summary Petraitis (1990) recently critized previous generalizations regarding the effects of predation in the New England rocky intertidal region (e.g., Menge 1976; Lubchenco and Menge 1978). Contrary to Lubchenco's and my conclusions, Petraitis concluded that (1) barnacles and not mussels are the favored prey of dogwhelks and (2) barnacles and not dogwhelks control mussel abundances in the mid and low rocky intertidal region. I provide evidence that these criticisms are unwarranted. First, Lubchenco and I never claimed that diet composition reflected prey preference. Moreover, predators can influence prey abundance without preferring the prey. Hence, claims regarding preferences have no bearing on our conclusions. Second, Petraitis' experiments do not invalidate Lubchenco's and my experimental results. Reanalyses of our experimental data support the earlier conclusion that at wave-sheltered sites, whelks reduce the abundance of mussels independently of barnacle abundances. Further, at all but one of Lubchenco's and my study sites, predator densities were higher than at Petraitis' site. Thus, the absence of a predator effect in Petraitis's study was most likely due to low predator density rather than a lack of generality of our earlier results. This reevaluation therefore suggests that within a broader conceptual framework, Petraitis' apparently divergent results are actually consistent with ours.     

A test of the Menge-Sutherland model of community organization in a tropical rocky intertidal food web

Summary Menge and Sutherland (1976) predicted that in physically benign habitats: (1) community structure will be most strongly affected be predation, (2) the effect of predation will increase with a decrease in trophic position in the food web, (3) trophically intermediate species will be influenced by both predation and competition, and (4) competition will occur among prey species which successfully escape consumers. These predictions were tested in a tropical rocky intertidal community on the Pacific coast of Panama. The most abundant mobile species included fishes and crabs, which occupied the top trophic level, and predaceous gastropods and herbivorous molluscs, which occupied intermediate trophic levels. The most abundant sessile organisms were encrusting algae, foliose algae, barnacles, and bivalves. Diets were broad and overlapping, and 30.3% of the consumers were omnivorous. Each consumer group had strong effects on prey occurring at lower trophic levels: (1) Fishes and crabs reduced the abundance of predaceous snails, herbivorous molluscs, foliose algae, and sessile invertebrates. (2) Predaceous gastropods reduced the abundance of herbivorous molluscs and sessile invertebrates. (3) Herbivorous molluscs reduced the abundance of foliose algae and young stages of sessile invertebrates, and altered relative abundances of the encrusting algae. The encrusting algae, although normally the dominant space occupiers, proved to be inferior competitors for space with other sessile organisms when consumers were experimentally excluded. However, the crusts escaped consumers by virtue of superior anti-herbivore defenses and competed for space despite intense grazing. Observations do not support the hypothesis that the trophically intermediate species compete. Hence, with the exception of this last observation, the predictions of the Menge and Sutherland model were supported. Although further work is needed to evaluate other predictions of the model in this community, evidence from this study joins an increasing body of knowledge supporting the model. Contradictory evidence also exists, however, indicating that aspects of the model require revision.     

Trophic exploitation in grassland food chains: simple models and a field experiment

This study provides insight into the importance of top carnivores (top-down control) and nutrient inputs (bottom-up control) in structuring food chains in a terrestrial grassland system. Qualitative predictions about food chain structure are generated using 4 simple models, each differing in assumptions about some key component in the population dynamics of the herbivore trophic level. The four model systems can be classified broadly into two groups (1) those that assume plant resource intake by herbivores is limited by search rate and handling time as described by classic Lotka-Volterra models; and (2) those that assume plant resource intake by herbivores is limited externally by the supply rate of resources as described by alternatives to Lotka-Volterra formulations. The first class of models tends to ascribe greater importance to top-down control of food chain structure whereas the second class places greater weight on bottom-up control. I evaluated the model predictions using experimentally assembled grassland food chains in which I manipulated nutrient inputs and carnivore (wolf spider) abundance to determine the degree of top-down and bottom-up control of grassland plants and herbivores (grasshoppers). The experimental results were most consistent with predictions of the second class of models implying a predominance of bottom-up control of food chain structure.     

Hybrid hierarchical clustering with applications to microarray data

In this paper, we propose a hybrid clustering method that combines the strengths of bottom-up hierarchical clustering with that of top-down clustering. The first method is good at identifying small clusters but not large ones; the strengths are reversed for the second method. The hybrid method is built on the new idea of a mutual cluster: a group of points closer to each other than to any other points. Theoretical connections between mutual clusters and bottom-up clustering methods are established, aiding in their interpretation and providing an algorithm for identification of mutual clusters. We illustrate the technique on simulated and real microarray datasets.     

Using patterns of variability to test for multiple community states on rocky intertidal shores

Predictions based on theory of multiple stable states suggest that larger perturbations should lead to more unpredictable patterns of succession. This prediction was tested in the Gulf of Maine using data from 60 intertidal plots of varying size that were experimentally cleared of the rockweed Ascophyllum nodosum and from 14 benchmark sites from throughout the Gulf. Rockweed was removed from the experimental clearings ranging from 1 to 8 m in diameter in 1996 and data collected in 2004 were used to test effects of clearing size and location on divergence and variability in species composition. Benchmark data were collected in 2005, and the 14 sites were from a dataset on 53 sites throughout the Gulf of Maine. The selected sites were randomly chosen from all sites with > 80% canopy cover by A. nodosum and were expected to be similar to uncleared control plots from the experiment. Experimental removal of A. nodosum resulted in clearings at 12 sites within 4 bays. Abundances of gastropods, barnacles, mussels, and fucoid algae and the percentage cover of barnacles, mussels, fucoid algae, bare space, and other species were sampled. CAP and PERMDISP analyses revealed significant differences in multivariate dispersion and variability with both clearing size and location. Variability generally increased with clearing size and location effects were related to the north-south positioning of the sites. Benchmark sites were similar to the experimental control plots but as variable as the largest clearings. Results suggest that succession in larger clearings has been more unpredictable than in small clearings. The pattern of variability in the experimental clearings is consistent with the predictions of multiple stable states. However, the large amount of variation among the benchmark sites was due to mussels and was unexpected. This unexpected variability underscores the importance of sampling benchmark sites as part of experiments.     

Response of epiphytic algae to nutrient loading and fish density in a shallow lake: a mesocosm experiment

Studies on the effect of eutrophication on the ecology of shallow lakes, usually pay scant attention to changes within the epiphytic algal community, though the contribution of this to the ecosystem dynamics is transcendental. In order to test the influence of nutrient loadings and fish densities in the structure of algal epiphyton in a shallow lake, an experiment was performed using in situ mesocosms. Nutrient additions were related to significant decreases in the total epiphyton biovolume and that of bacillariophyceans and zygnematophyceans, but with increases in the abundance of cyanobacteria. The different response of algal groups at the higher nutrient concentrations (increases or decreases in their abundance and/or biovolume) can be related to their ecophysiological constraints such as different resistance to toxicant ammonium accumulation. Plant-associated macroinvertebrates numbers were positively correlated with total numbers of epiphyton. The presence of planktivorous fish enhanced the abundance and biovolume of all algal groups, except cyanobacteria. Fishes enhanced the abundance of plant-associated animals and of total epiphyton. Fish indirect effects (e.g., nutrients release) and their dietary particularities were among the factors that together with nutrients influenced epiphyton growth. The role of indirect effects of fishes and the importance of their dietary particularities are stressed as key factors to understand the processes controlling epiphyton ecology and the food web structure of shallow lakes. Handling editor: D. Ryder.     

Regulation of boreal forest understory vegetation: the roles of resources and herbivores

To understand inter-trophic linkages between components of the boreal forest understory vegetation, three hypotheses were tested: survival, growth and abundance of grasses and legumes were controlled by (i) resource availability alone, (ii) by herbivores alone, and (iii) by both resource availability and herbivores. The hypotheses were tested using three experimental treatments – fertilization, herbivore exclusion, and fertilization plus herbivore exclusion – in three areas having different densities of resident herbivores, mostly snowshoe hares and ground squirrels. The highest density of snowshoe hares is comparable to natural levels during peaks in the snowshoe hare cycle. As the density of herbivores increased so too did the level of response by the measured variables – survival, growth of transplants and leaf area index of established vegetation. In general, fertilization resulted in a decrease in survival and growth of transplants, and fences increased survival and growth; both responses were more noticeable at higher herbivore densities. Fertilizer and herbivore exclosure fences had only negligible effects on established grass and legume abundance at all hare densities. We have shown that some hypotheses of vegetation regulation are over-simplified because different species groups (i.e., grasses and legumes) are regulated by different factors, at different life history stages, and sometimes these factors act in opposing directions. We argue that during the increase phase and peak of the snowshoe hare cycle (high herbivore density), growth and survival of establishing plants is regulated by herbivores. During the decline and low phases of the snowshoe hare cycle herbivores will have little impact on early life stages, whereas the established, mature, vegetation will be resource-regulated. Because of the variability in responses to the same manipulations we may begin to understand which plant life history stages are most vulnerable to consumer and resource regulation, the magnitudes of these sources of regulation at each of these stages, and how these vary among species groups and types of environments.     

Ecosystem engineering at the sediment–water interface: bioturbation and consumer-substrate interaction

In soft-bottom sediments, consumers may influence ecosystem function more via engineering that alters abiotic resources than through trophic influences. Understanding the influence of bioturbation on physical, chemical, and biological processes of the water–sediment interface requires investigating top-down (consumer) and bottom-up (resource) forces. The objective of the present study was to determine how consumer bioturbation mode and sediment properties interact to dictate the hydrologic function of experimental filtration systems clogged by the deposition of fine sediments. Three fine-grained sediments characterized by different organic matter (OM) and pollutant content were used to assess the influence of resource type: sediment of urban origin highly loaded with OM and pollutants, river sediments rich in OM, and river sediments poor in OM content. The effects of consumer bioturbation (chironomid larvae vs. tubificid worms) on sediment reworking, changes in hydraulic head and hydraulic conductivity, and water fluxes through the water–sediment interface were measured. Invertebrate influences in reducing the clogging process depended not only on the mode of bioturbation (construction of biogenic structures, burrowing and feeding activities, etc.) but also on the interaction between the bioturbation process and the sediments of the clogging layer. We present a conceptual model that highlights the importance of sediment influences on bioturbation and argues for the integration of bottom-up influence on consumer engineering activities. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

The structuring role of free-floating versus submerged plants in a subtropical shallow lake

In shallow temperate lakes many ecological processes depend on submerged macrophytes. In subtropical and tropical lakes, free-floating macrophytes may be equally or more important. We tested the hypothesis that different macrophyte growth forms would be linked with different bottom-up and top-down mechanisms in out-competing phytoplankton. We compared experimentally the effects of submerged and free-floating plants on water chemistry, phytoplankton biomass, zooplankton and fish community structure in a shallow hypertrophic lake (Lake Rodó, 34°55S 56°10W, Uruguay). Except for the retention of suspended solids, we found no other significant bottom-up process connected with either Eichhornia crassipes or Potamogeton pectinatus. Free-floating plants had a lower abundance of medium-sized zooplankton than any other microhabitat and submerged plants were apparently preferred by microcrustaceans. Fish showed a differential habitat use according to species, size-class and feeding habits. Dominant omnivore-planktivores, particularly the smallest size classes, preferred submerged plants. In contrast, omnivore-piscivores were significantly associated with free-floating plants. The density of omnivorous-planktivorous fish, by size class, significantly explained the distribution of medium-sized zooplankton, the high number of size 0 fish being the main factor. The abiotic environment and the structure of the zooplankton community explained little of the fish distribution pattern. Our results suggest that bottom-up effects of free-floating plants are weak when cover is low or intermediate. Top-down effects are complex, as effects on zooplankton and fish communities seem contradictory. The low piscivores:planktivores ratio in all microhabitats suggests, however, that cascading effects on phytoplankton through free-floating plant impacts on piscivorous fish are unlikely to be strong.     

Weak response of saltmarsh infauna to ecosystem-wide nutrient enrichment and fish predator reduction: A four-year study

We examined the effect of whole-ecosystem manipulations of predator removal and nutrient enrichment on saltmarsh macroinfauna in the Plum Island Estuary, Massachusetts. Nitrate and phosphate loading rates were increased 10× above background levels in experimental creeks, and we significantly reduced (by 60%) the abundance of the killifish, Fundulus heteroclitus, a key predator in this system. Two creek pairs were manipulated; Creek Pair 1 for three growing seasons and Creek Pair 2 for one. Infaunal responses were examined in four habitats along the inundation gradient: mudflat, creek wall, Spartina alterniflora, and S. patens habitats. Although benthic microalgae increased synergistically in our treatments, we detected no long-term, population-level numerical response by any taxon. Similarly, no long-term species diversity or community responses were observed. However, nutrient enrichment increased the population biomass of the polychaete Manayunkia aestuarina in the creek wall habitat and the oligochaete Cernosvitoviella immota in the S. alterniflora habitat. No numerical or biomass responses of infauna were detected in predator removal treatments although indirect effects associated with killifish reduction may have contributed to an ephemeral interaction between nutrient addition and predator reduction in S. patens habitat. Our data suggest that population and community responses between benthic microalgae and macroinfauna are not tightly coupled even though some species benefit from increased benthic algae biomass by achieving larger body size.     

Experimental ecology of rocky intertidal habitats: what are we learning?

Experimental analyses of causes of patterns of distribution and abundance of intertidal animals and plants on rocky shores have been a major activity for many years. In this review, some of the themes and topics that have emerged from such analyses are briefly discussed to provide an up-date for practitioners and ecologists working in other habitats. Conceptual issues include the widespread occurrence of transphyletic use of the same resources (space and food), theories and experimental analyses of intermediate disturbance in relation to numbers of species, the complex but pervasive nature of indirect interactions among species, relative importance of ‘top-down’ versus ‘bottom-up’ control of assemblages and the importance to rocky intertidal species of ‘supply-side’ influences on densities and interactions. Methodological advances include experimental designs for complex and patchy, interacting sets of species, the importance of controls in experimental manipulations and methods for analyses of hierarchical scales of patterns and processes. Finally, some contributions to social issues (pollution, biodiversity) and some scenarios for future directions are briefly considered.     

Thermal constraints on daily patterns of aggregation and density along an intertidal gradient in the periwinkle Echinolittorina peruviana

Variations in environmental temperature have both direct and indirect effects that affect organisms at levels ranging from intra-cellular physiological processes to ecological patterns. These variations are especially important for intertidal marine ectotherms such as littorinids since they alternate between periods of immersion in seawater, and must also experience long periods of emersion. In central Chile, Echinolittorina peruviana is one of the most conspicuous species on rocky intertidal shores, occurring at high tidal levels and in the splash zone. The species is known to resist direct exposure to the sun for long periods, although juveniles tend to be restricted to protected microhabitats. Adults show seasonal variations in abundance and vertical distribution and may form aggregations that have been shown to help reduce water loss and body temperature. In this study, we evaluate the relationship between daily thermal variations throughout the vertical distribution of this species and how these affect the patterns of density and aggregation. Our results suggest that one of the leading determinants of the spatio-temporal variation of density in E. peruviana may be operative temperature (T_O: the amount of stored heat resulting from the balance between heat fluxes into and out of the body, measured with taxidermic mounts mimicking heat transfer properties of the snail). T_O showed a strong negative relationship with density and a strong positive relationship with aggregation in the highest intertidal level monitored. The strength of these relationships decreased in importance at lower levels. While T_O alone cannot explain the abundance of E. peruviana throughout its range of distribution, our results show that it does have a strong influence that should be considered in addition to other ecological factors affecting the density of intertidal littorinids.     

Periphyton-macroinvertebrate interactions in light and fish manipulated enclosures in a clear and a turbid shallow lake

In a clear and a turbid freshwater lake the biomasses of phytoplankton, periphytic algae and periphytonassociated macrograzers were followed in enclosures with and without fish (Rutilus rutilus) and four light levels (100%, 55%, 7% and < 1% of incoming light), respectively. Fish and light affected the biomass of primary producers and the benthic grazers in both lakes. The biomass of primary producers was generally higher in the turbid than the clear lake, and in both lakes fish positively affected the biomass, while shading reduced it. Total biomass of benthic grazing invertebrates was higher in the clear than in the turbid lake and the lakes were dominated by snails and chironomids + ostracods, respectively. While light had no effect on the biomass of grazers in the clear lake, snail breeding was delayed in the most shaded enclosures and presence of fish reduced the number of snails and the total biomass of grazers. In the turbid lake ostracod abundance was not influenced by light, but was higher in fish-free enclosures. Density of chironomids correlated positively with periphyton biomass in summer, while fish had no effect. Generally, light-mediated regulation of primary producers was stronger in the turbid than in the clear lake, but the regulation did not nambiguously influence the primary consumers. However, regulation by fish of the benthic grazer community was stronger in the clear than in the turbid lake, and in both lakes strong top-down effects on periphyton were seen. The results indicate that if present-day climate in Denmark in the future is found in coastal areas at higher latitudes, the effect of lower light during winter in such areas will be highest in clear lakes, with typically lower fish biomass and higher invertebrate grazer density.     

Effects of global environmental changes on parasitoid–host food webs and biological control

Global environmental changes threaten biodiversity and the interactions between species, and food-web approaches are being used increasingly to measure their community-wide impacts. Here we review how parasitoid–host food webs affect biological control, and how their structure responds to environmental change. We find that land-use intensification tends to produce webs with low complexity and uneven interaction strengths. Dispersal, spatial arrangement of habitats, the species pool and community differences across habitats have all been found to determine how webs respond to landscape structure, though clear effects of landscape complexity on web structure remain elusive. The invasibility of web structures and response of food webs to invasion have been the subject of theoretical and empirical work respectively, and nutrient enrichment has been widely studied in the food-web literature, potentially driving dynamic instability and altering biomass ratios of different trophic levels. Combined with food-web changes observed under climate change, these responses of food webs could signal changes to biological control, though there have been surprisingly few studies linking food-web structure to pest control, and these have produced mixed results. However, there is strong potential for food-web approaches to add value to biological control research, as parasitoid–host webs have been used to predict indirect effects among hosts that share enemies, to study non-target effects of biological control agents and to quantify the use of alternative prey resources by enemies. Future work is needed to link food-web interactions with evolutionary responses to the environment and predator–prey interactions, while incorporating recent advances in predator biodiversity research. This holistic understanding of agroecosystem responses and functioning, made possible by food-web approaches, may hold the key to better management of biological control in changing environments.     

Effects of consumers and enrichment on abundance and diversity of benthic algae in a rocky intertidal community

Human alteration of nutrient cycling and the densities of important consumers have intensified the importance of understanding how nutrients and consumers influence the structure of ecological systems. We examined the effects of both grazing and nutrient enrichment on algal abundance and diversity in a high-intertidal limpet-macroalgal community on the South Island of New Zealand, a relatively nutrient-poor environment. We used a fully factorial design with three levels each of grazing (manipulations of limpet and snail densities) and nutrients (nutrient-diffusers attached to the rock). Top-down control by grazers appears to be the driving organizing mechanism for algal communities in this system, with strong negative effects of grazing on algal diversity and abundance across all levels of nutrient enrichment. However, in contrast to the conclusions drawn from the analysis of the whole algal community, there was an interactive effect of grazing and enrichment on foliose algae, an important component of the algal system. When herbivory was reduced to very low levels, enrichment generated increases in the abundance and biomass of foliose algae. As expected, top-down control was the primary determinant of algal community structure in this system, controlling abundance and diversity of macrophytes on the upper shore. Contrary to expectations, however, increased nutrients had no community-wide effects, although foliose algal abundance increases were greatest with high nutrients and reduced grazing. It seems likely that most of the corticated algal species have limited capacity to respond to nutrient pulses in this nutrient-poor environment.     

Community stability: effects of limpet removal and reintroduction in a rocky intertidal community

Summary The stability of a high rocky intertidal community was assessed in a controlled field experiment in which the most common consumers, limpets, were temporarily removed. Compared to the unmanipulated plots, the exclusion plots developed greater algal abundance and altered species composition of both algae and barnacles. The community was not perturbed beyond its capacity to recover, since the community structure of the limpet-removal plots converged on the structure of the unmanipulated plots following limpet reintroduction. Different components of the community recovered at different rates, depending on whether or not the species had a size-related escape from the limpets. Algae had no size-related escape from limpets. The difference in algal abundance between manipulated and unmanipulated plots lasted less than six months after limpet reintroduction. Barnacles, however, had a size-related escape from limpets and therefore recovered more slowly. The difference in barnacle species composition between the perturbed and unperturbed plots lasted for 17 months after limpet reintroduction. The length of the limpet removal period (16 or 28 months) did not appear to affect the rate of community recovery.