Colonization dynamics of river benthos in response to local changes in bed characteristics

1. The effects of small-scale variation in the nature of a river bed upon the colonization dynamics of benthic invertebrates were explored through a series of microhabitat implants placed in a small river in southern Ontario. The implants presented variations in the levels of four microhabitat features: substrate particle size, current regime, added food and light. The sixteen possible combinations of these variables were replicated five times, for a total of eighty implants.
2. After 14 days of exposure, some of the controlled variables had been modified and a number of other variables (sand, fine (FPOM) and coarse (CPOM) particulate organic matter) had been added by the river.
3. Benthic diversity differed among the implants, with sand and CPOM accounting for 15.5% of the variation. Benthic density was more uniform and was weakly related only to added food. Application of the CANOCO program to the data set suggested that, in general, the environmental variables manipulated in this study were not the key factors influencing colonization, although they may have been important in non-linear combination.
4. A number of important positive associations were evident among colonizing species, some of which are thought to reflect those of predator and prey.
5. Control of colonization appears to be through multiple factors (abiotic, biotic, autochthonous, allochthonous) that differ in nature according to species, and which underline the complexity of selection responses of individual taxa to local variations in bed structure. Researchers must acknowledge this multifactorial organization and dynamic nature if workable models of lotic community structure and development are to be forthcoming.     

Microhabitat selection by a stream-dwelling amphipod: a muitivariate analysis approach

SUMMARY. 1. Colonization of microhabitat implants by the amphipod Gammarus pseudolimmnaeus in a small southern Ontario stream was studied in order to analyses the factors controlling habitat selection. The variables substrate particle size, current speed, presence of food and light were used in an analysis of covariance, with percentage weight of organic matter of silt and percentage interstitial space occluded by silt as the covariates.
2. Greatest numbers of amphipods settled on microhabitats featuring large substrate particles, no current and presence of food. There was also a positive relationship between total numbers and the volume of silt deposited on the microhabitats by the stream; small quantities of silt had a beneficial effect on colonization but larger quantities became detrimental.
3. The change from a positive effect occurred at approximately 25% occlusion of the interstitial space in large gravel ( diameter=3.2 cm) and at approximately 55% occlusion in small gravel ( diameter = 0.34 cm).
4. Large animals (6–16.0 mm long) were found predominantly in microhabitats featuring food and large substrate. Medium-sized animals (3–6.0 mm) were most commonly associated with no current and presence of food, and were positively affected by the amount of silt but, at the same time, were negatively affected by increasing occlusion of interstitial spaces by silt. Numbers of small Gammarus (<3.0 mm) were affected only by silt and in a similar manner to medium-sized animals.
5. Amphipod biomass was greatest in microhabitats featuring food and no current. Previous data on the behaviour of this species in laboratory stream-tanks are compared with the microhabitat selections seen.     

Benthic organic matter dynamics in Texas prairie streams

Concentrations of benthic particulate organic matter (POM) in six Texas prairie streams (2nd–4th order, intermittent and perennial) were monitored over a 20 month period to determine temporal and spatial dynamics. Benthic POM mass was highly variable, having coefficients of variation (CV) in excess of 300%. Benthic POM mass in all streams was similar with the exception of the 4th order intermittent stream which had significantly higher concentrations. Benthic POM at all sites was dominated by coarse POM (CPOM), followed by fine POM (FPOM), ultrafine POM (UPOM), and medium POM (MPOM). The dominance by CPOM is especially noteworthy in the 4th order intermittent stream where it accounted for 83% of the annual POM mass. Seasonally, benthic POM was highest in summer and lowest in the fall.     

A novel method reveals how channel retentiveness and stocks of detritus (CPOM) vary among streams differing in bed roughness

1. Coarse particulate organic matter (CPOM) is a fundamental resource in freshwater streams, providing food, shelter and habitat for diverse invertebrate taxa and playing a key role in metabolism in low-order streams. Benthic CPOM stocks are determined by rates of supply and breakdown of detritus and by channel retentiveness (i.e. the capacity for the channel to trap and retain CPOM). We focussed on factors affecting the retentiveness of channels, which theoretically differs among streams with different sediment sizes and concomitant channel morphology.
2. We developed a new, rapid method to measure retentiveness using line-intercept surveys along transects. With this rapid approach, we surveyed 32 sites from three types of streams (smooth sandy channels, n = 10; gravel channels of intermediate roughness, n = 12; rough cobble channels, n = 10) in Victoria, south-eastern Australia, and tested the simple hypotheses that: (1) retentiveness increases in channels with increasing channel roughness (i.e. sandy versus gravel versus cobble-bed streams); (2) different types of channel features (e.g. log jams, cobbles, depositional areas) differ in the efficiency with which they retain CPOM. The line-intercept survey method was readily adapted to measure retentiveness as m of retentive structure per m of transect (i.e. the Linear Coverage Index) and trapping efficiency as m of CPOM per m of retentive element, for 10 different types of retentive elements.
3. Unexpectedly, the retentiveness of channels did not increase with channel roughness. This occurred because channels with different roughness were dominated by different types of retentive structure. Retentive structure in cobble sites was dominated by cobbles themselves, which were highly retentive in other studies but poorly retentive in our system. Gravel and sand sites had more log jams and depositional areas, such as pools and backwaters, and these features were more effective at trapping CPOM. Thus, retention of CPOM was highest in gravel and sand sites.
4. Our method provides a new tool for investigators testing hypotheses about CPOM retention in streams. The method is rapid, requires a minimum of equipment and personnel, and may be applied in any wadeable stream. Retentiveness is calculated in intuitive units that are directly comparable among sites and may have utility as variables in models of CPOM dynamics. We hope this method will open up new avenues for research that may shed light on how CPOM stocks vary among streams, with implications for diversity of aquatic fauna and ecosystem functions such as decomposition.

     

Reach-scale geomorphology affects organic matter and consumer δ13C in a forested Piedmont stream

1. We investigated the spatial (longitudinal position and reach geomorphology) and seasonal (spring and autumn) influences on the variation of δ¹³C among organic matter sources and consumers in a forested Piedmont river, South Carolina, U.S.A. 2. Six sites were sampled along a continuum and varied in basin area from approximately 30 to 300 km². Sites fell into two geomorphic categories (i) high‐gradient, rock bed (‘rock’) or (ii) low‐gradient, sand bed (‘sand’) sites. 3. Variation in δ¹³C was more strongly related to reach geomorphology than longitudinal position. δ¹³C of biofilm and consumers was consistently enriched at rock sites. Leaf litter (i.e. coarse particulate organic matter, CPOM) δ¹³C did not vary with bed type. There was significant δ¹³C enrichment at rock sites for biofilm, seston, fine benthic organic matter (FBOM), and eight of nine consumer trophic guilds (e.g. grazing invertebrates, insectivorous fishes). δ¹³C of biofilm and four trophic guilds was also positively correlated with drainage area, but the magnitude of enrichment was less than between bed types. 4. δ¹³C was generally enriched in spring, but this varied among organic matter types, consumers, and by bed type. CPOM and seston were enriched in spring, FBOM was enriched in autumn, and biofilm showed no trend. Five consumer guilds were enriched in spring, and only one fish guild, generalised carnivores, showed enrichment of muscle tissue in autumn. 5. Consumer δ¹³C enrichment at rock sites suggests greater reliance on algal carbon than for consumers at sand sites, but we also found δ¹³C enrichment of biofilm at rock sites. Thus, differences in consumer δ¹³C between bed types could be related to (i) increased consumption of biofilm at rock compared with sand sites, or (ii) consumption of biofilm at rock sites that is enriched relative to biofilm at sand sites or (iii) both mechanisms. 6. δ¹³C signatures in local food webs appear to respond to processes operating at multiple spatial scales. Overall downstream enrichment of biofilm and consumers was disrupted by strong local effects related to bed morphology. These results suggest that human alteration of channel habitat will have corresponding effects on stream food webs, as assessed by changes in δ¹³C.     

Microhabitat specialization in a species-rich palm community in Amazonian Ecuador

1 Relationships between microhabitat variables (altitude, inclination, topographic position, drainage, canopy height) and the distribution and abundance of palms and palm-like plants in 50 ha of old-growth terra firme rain forest in the Yasuní National Park, lowland Amazonian Ecuador, were examined using 118 20 × 20 m plots laid out in a stratified random design.
2 If microhabitat niche differentiation is important for maintaining the species richness of the community, then (i) the distribution of the palms will be strongly influenced by microhabitat heterogeneity and (ii) palms of similar growth form will show antagonistic microhabitat relationships.
3 Mantel and cluster analyses showed that palm species distributions were strongly structured by topography. The main difference in species composition was between plots in the bottomland and plots on the upper slopes and hill tops.
4 Logistic and logit analyses showed that 20 of the 31 palm and palm-like taxa analysed had distributions that were significantly related to the microhabitat variables measured, mainly to topography but also to drainage and canopy height.
5 Spatial autocorrelation in the overall community structure was not explained by the microhabitat variables. Analyses of distributions or abundances of single species showed neighbourhood effects for seven taxa.
6 Antagonistic patterns of microhabitat preferences were recognizable among some species pairs of small palms, medium-sized palms and palm-like plants, but not among canopy palms.
7 It is concluded that microhabitat specialization is an important factor in maintaining the diversity of this palm community, while mass effects might also be important.     

Effects of buried leaf litter and vertical hydrologic exchange on hyporheic water chemistry and fauna in a gravel-bed river in northern New South Wales, Australia

1. Large amounts of coarse particulate organic matter (CPOM) are buried in the sand and gravel beds of many rivers during spates. The effects of these patchily distributed resources on hyporheic invertebrates and water chemistry are poorly understood. Buried CPOM may provide local ‘hot-spots’ of food for hyporheic detritivores and their predators, alter nutrient supply to nearby sediment biofilms, and generate habitat for some invertebrates. 2. To examine potential short-term effects on hyporheic water chemistry, nutrient concentrations and invertebrate assemblage composition, leaf packs were buried in downwelling (surface water infiltrating the hyporheic zone) and upwelling (hyporheic water emerging to the surface) zones at two sites along a gravel-bed river in northern New South Wales. At one site, pits were excavated to simulate leaf burial (procedural control) and plastic ‘leaves’ were buried to test whether invertebrates might respond to leaves as refuges rather than food. Hyporheic CPOM, sediment size fractions, and interstitial silt content were also quantified at these sites. 3. Dry weights of naturally buried CPOM (leaf litter and wood fragments) varied substantially (0.6–71.7 g L^–1 sediment). Amounts of CPOM did not differ between up- vs. downwelling zones or between sites. Hyporheic dissolved oxygen saturation was generally high (> 75%), and was lower in upwelling zones. The hyporheos was dominated taxonomically by water mites (≈ 20 species), whereas small oligochaetes were most abundant (40% of total abundance). Tiny instars of elmid beetle larvae and leptophlebiid mayfly nymphs were also common. Before experimental manipulation, faunal composition differed between up- and downwelling zones. In upwelling zones, bathynellaceans and blind peracarids were found, whereas small individuals of the surface benthos were common in samples from downwelling zones. This validated stratification of the experiment across zones of hydrologic exchange. 4. Twenty days after leaf burial, there was no effect of the treatments at either site on changes in most variables, including mean numbers of taxa and individuals per sample. Similarly, changes in faunal composition of the hyporheos in the treatments paralleled those in the controls except for a weak response in the buried leaves treatment in the upwelling zone at site 1. Artificially buried leaf litter does not seem to influence hyporheic water chemistry or fauna at these two sites. It is probable that naturally buried leaf litter is swiftly processed soon after entrainment and that repeating this experiment immediately after a flood may yield different results.     

The effect of two scales of habitat architecture on benthic grazing in a river

1. This experiment studied the effects of differing levels of the complexity of substratum architecture at two spatial scales on the distribution and abundance of benthic algae and invertebrates, and the strength of the trophic interaction between invertebrate grazers and algae. Some estimates of the effects on invertebrate colonization rates were also made. 2. Four levels of microhabitat architectural complexity were created using artificial substrata (clay tiles) and placed in Mountain River, Tasmania, in two riffle types (bedrock and boulder-cobble) of differing large-scale substratum complexity. After a colonization period, invertebrate grazers were removed from half the tiles to measure the effects of grazing. Invertebrates on the tiles were also counted and identified. At the end of the experiment, algae were removed from the tiles and analysed for chlorophyll a. 3. Invertebrate grazers did not reduce algal biomass during the experiment, and microhabitat-scale architecture influenced algal biomass more strongly than riffle-scale architecture. Highly complex microhabitat architecture increased algal biomass by providing more surface area, but once standardized for surface area, algal biomass decreased as the complexity of microhabitat architecture increased. 4. Microhabitat-scale architecture was also predominant in determining invertebrate density and the identity of the dominant grazer species. In contrast to algal biomass, invertebrate densities and species density increased with the complexity of microhabitat architecture, suggesting that refuges from flow (and possibly predation) were as important to river invertebrates as the distribution of their food source. 5. Riffle-scale architecture had some effect on the colonization of two slow-moving grazer taxa, but, overall, the colonization processes of slow-moving grazers were determined mostly by the complexity of microhabitat-scale architecture.     

Patch selection by mottled sculpin (Pisces: Cottidae) in a southern Appalachian stream

1. We examined the effects of prey abundance on patch selection by a benthic fish, the mottled sculpin ( Cottus bairdi ), in a fourth order, southern Appalachian stream (North Carolina, U.S.A.). This habitat is a mosaic of small (i.e. < 0.5 m²), relatively discrete patches.
2. Patches were characterized in terms of physical habitat variables, detritus (coarse particulate organic matter, CPOM) and macroinvertebrate abundance (number and biomass). We quantified patch selection by comparing the characteristics of patches utilized by sculpin with those of locally available patches. Locally available patches were selected using a constrained random sampling design (i.e. randomly selected within a 2 m radius from each fish). We also examined the relationship between macroinvertebrate abundance, CPOM and the physical characteristics of available patches.
3. Patches selected by sculpin contained significantly higher macroinvertebrate abundances (both number and biomass) than locally available patches in five out of six seasonal samples. Sculpin also occupied patches with significantly higher amounts of CPOM in three out of five seasonal samples. Patches utilized by sculpin, however, could not be consistently differentiated from locally available patches on the basis of physical variables. In addition, macroinvertebrate abundance was not consistently related to physical habitat variables or CPOM during the course of the study.
4. Our results suggest that sculpin are able to assess patch quality on the basis of prey abundance and select patches that potentially maximize energy gain. This behaviour may produce an increase in individual fitness, especially when prey distributions are not consistently related to habitat variables. Quantifying patch use in relation to prey abundance may help elucidate the causal factors determining habitat use by benthic fishes in other lotic systems.     

Patch selection by mottled sculpin (Pisces: Cottidae) in a southern Appalachian stream

1. We examined the effects of prey abundance on patch selection by a benthic fish, the mottled sculpin ( Cottus bairdi ), in a fourth order, southern Appalachian stream (North Carolina, U.S.A.). This habitat is a mosaic of small (i.e. < 0.5 m²), relatively discrete patches.
2. Patches were characterized in terms of physical habitat variables, detritus (coarse particulate organic matter, CPOM) and macroinvertebrate abundance (number and biomass). We quantified patch selection by comparing the characteristics of patches utilized by sculpin with those of locally available patches. Locally available patches were selected using a constrained random sampling design (i.e. randomly selected within a 2 m radius from each fish). We also examined the relationship between macroinvertebrate abundance, CPOM and the physical characteristics of available patches.
3. Patches selected by sculpin contained significantly higher macroinvertebrate abundances (both number and biomass) than locally available patches in five out of six seasonal samples. Sculpin also occupied patches with significantly higher amounts of CPOM in three out of five seasonal samples. Patches utilized by sculpin, however, could not be consistently differentiated from locally available patches on the basis of physical variables. In addition, macroinvertebrate abundance was not consistently related to physical habitat variables or CPOM during the course of the study.
4. Our results suggest that sculpin are able to assess patch quality on the basis of prey abundance and select patches that potentially maximize energy gain. This behaviour may produce an increase in individual fitness, especially when prey distributions are not consistently related to habitat variables. Quantifying patch use in relation to prey abundance may help elucidate the causal factors determining habitat use by benthic fishes in other lotic systems.     

Prediction of benthic macroinvertebrate composition using microhabitat characteristics derived from stereo photography

1. Benthic macroinvertebrate community and habitat features varying at the microscale ( 0.09 m ² ) were measured on one sampling occasion in the Thredbo River, Kosciusko National Park, NSW, Australia.
2. Most of the substratum habitat variables were measured in three dimensions using stereo photography. This is the first time that this method has been used so extensively for this purpose in freshwater ecology.
3. Microhabitat variables most related to benthic macroinvertebrate distribution and abundance were selected with multivariate analyses included rock length, height, area and water velocity. Individual variables alone could not account for macroinvertebrate variation, indicating the importance of interactions among variables.
4. Nine selected habitat variables were used to predict macroinvertebrate taxonomic content of additional sites. Predictions were 87% accurate for taxa with a > 50% chance of occurrence and 93% accuracy for taxa with a > 70% chance of occurrence. Variability observed in macroinvertebrate assemblages at the time of sampling was largely physically controlled and highly deterministic.     

Leaf Litter Dynamics and Its Use by Invertebrates in a High‐Altitude Tropical Andean Stream

This study evaluates litter dynamics and its relation to macroinvertebrate communities (assemblages and feeding habits) at Piburja stream (3300 m, Ecuador). Annual litterfall (216 g AFDM/m²) was not related to rainfall, but differed significantly between months. Transport of Coarse Particulate Organic Matter (CPOM) did not differ between seasons, but retention was higher during the dry season. Thus, CPOM standing stock was higher in the dry (125.28 g AFDM/m²) compared to the wet season (12.27 g AFDM/m²). Macroinvertebrate richness and diversity were positively related to Coarse Benthic Organic Matter (CBOM) especially during the wet season. Gut content analysis revealed that, regardless of season, CPOM and Fine Particulate Organic Matter (FPOM) were the most important food items in the diet of most invertebrate taxa, including those that are not considered shredders or gathering collectors in the literature. (© 2009 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Oligochaetes and eutrophication; an experiment over four years in outdoor mesocosms

Eight experimental ditch mesocosms were used to study the effect of eutrophication over four years. The experimental ditches had a sand or clay bottom. The ditches were treated with additions of phosphorus, phosphorus and nitrogen, or without additions (controls). Oligochaetes were sampled by deploying trays with substratum for colonization over twenty weeks. Both the important variables phosphorus, nitrogen and oxygen as well as the oligochaete species and numbers are presented. The effects of nutrient additions on phosphorus, nitrogen and oxygen concentrations were described together with changes in oligochaete species composition and numbers. The results were further analyzed by redundancy analysis (RDA). In the clay-lined ditches nutrient addition coincided with fluctuation in oxygen concentration. The higher the nutrient addition levels the longer the period of oxygen depletion became. During oxygen depletion the number of oligochaetes was strongly reduced or even became zero. The low nutrient status of the sandy bed in the sand-lined ditches slowed down the rate of colonization. Only a few tubificids were collected. Eutrophication effects were only observed at the highest nutrient addition level. Considerable variation is attributed to stochastic factors in the sand-lined ditches. Whether oligochaete species were present was related to the length of the colonization period. The substratum composition and food together with oxygen regime decided whether they become more or less abundant in ditches. Large-scale mesocosm experiments require time to develop. Only after the first colonization period variables of species presences and abundances can be employed to detect changes associated with eutrophication. Oligochaetes can be used to measure colonization as well as eutrophication processes.     

Food,microhabitat, or both? Macroinvertebrate use of leaf accumulations in a montane stream

1. Leaf packs in streams may serve as food and substrate to many macroinvertebrates, but the relative importance of these two functions has not been disentangled. To test the hypothesis that leaf packs are colonized primarily for their food value rather than as microhabitat, the colonization of leaf packs of red alder and of polyester cloth in a natural stream was compared. 2. Species of shredders showed large differences in the colonization of the two types of leaf pack with almost no use of artificial leaf packs. Non-shredders were also more abundant on natural leaf packs, however, they colonized artificial leaf packs six to eighty times proportionally more than shredders. 3. The effect of different leaf types was virtually eliminated for non-shredders when the amount of fine parriculate detritus (food for many non-shredders) trapped in the leaf pack was added to the analysis as a covariate. Therefore, to non-shredders, leaf types differed only in the retention of fine particulate organic matter. 4. Comparisons of the use of both kinds of leaf packs in riffles versus pools revealed that significantly fewer animals colonized pool leaf packs. 5. These results suggest that food value, and not microhabitat, is the primary determinant of leaf pack use for most shredders and non-shredders.     

Temporal and Spatial Patterns in Inputs and Stock of Organic Matter in Savannah Streams of Central Brazil

Coarse particulate organic matter (CPOM) inputs from riparian vegetation into streams and CPOM benthic stock vary naturally in space and time, but most studies in the tropical savannah (Cerrado) have been done over a small temporal scale (<1 year), which does not allow for the determination of inter-annual patterns. We found that CPOM collected over two years differed temporally and spatially, whereas there was no significant variation between years for the benthic stock, which indicates high stability in the energy balance of streams. The largest monthly inputs occurred between August and October, at the end of the dry season and the onset of the rainy season, which was partially explained by precipitation. Other factors such as photoperiod, which was not studied, could also have important roles in this pattern. Spatial differences in CPOM between streams were attributed to topography and channel morphology. The plant density was lowest in the stream with a more irregular topography and a deeper channel, which results in drier riparian soil. The benthic stock was highest in the stream with a flat channel, where the lower water speed facilitates the accumulation of CPOM in the stream bed. Inter-annual differences in CPOM were attributed more to the differences in the beginning of the dry and wet periods between years than to the average values of precipitation. Longer-term studies are needed to clarify this temporal pattern.     

Macroinvertebrate community structure in relation to environmental variables in a Swiss glacial stream

1. Benthic macroinvertebrate distribution was examined in relation to channel characteristics (including stability), substratum, hydraulic variables, primary production (chlorophyll a ) and coarse particular organic matter (CPOM) in an alpine glacial stream, the Mutt (Upper Rhône valley, Switzerland). Co-inertia analysis and canonical correspondence analysis were used to identify the major environmental gradients influencing community variations.
2. The Mutt (length: 3.6 km, altitudinal range: 1800–3099 m a.s.l.) exhibited typical characteristics of a kryal stream. Average summer temperature remained below 2 °C immediately downstream from the snout but was on average 5 °C higher 1700 m downstream. Seasonal variations in water sources were evidenced by the high late-summer (September) contribution of groundwater with increased conductivity.
3. Sixty-six taxa were recorded from the five reaches sampled at three periods (snowmelt, ice melt and low water in late summer) in 1996 and 1997, of which 29 were Chironomidae. Three taxa of Diamesinae were the first colonizers of the stream below the glacier, but 16 taxa, including Ephemeroptera, Plecoptera and Trichoptera, were already recorded 200 m downstream. Water depth, channel slope and Pfankuch's Index of channel stability were strongly correlated with the longitudinal faunal gradient. Maximum temperature, current velocity and water conductivity were also correlated, but to a lesser extent.
4. The rapid incorporation of non-chironomid taxa into the stream community represented a departure from Milner & Petts's (1994) conceptual model of invertebrate succession downstream of glacial margins. The results confirmed that glacial stream communities are primarily driven by physical determinants.     

Temporal and spatial macroinvertebrate variance compared: crucial role of CPOM in a headwater stream

The aim of the present study was to estimate the extent to which macroinvertebrate assemblages in a headwater stream are determined by environmental conditions and temporal dynamics. Six mesohabitats defined by substratum were sampled monthly throughout one year, environmental conditions at each sampling point being precisely described. Environmental variables could be truncated into two main gradients related to (a) the availability of food and space resources (CPOM) and (b) hydraulic conditions. The response of the macroinvertebrate assemblage to the environmental gradients and temporal dynamics were analyzed using GAM and (p)RDA. Twice as high portion of variance in the faunal data was attributed to environmental gradients compared to temporal dynamics. Total abundance, as well as the abundance of almost all feeding groups, was dependent on the availability of food and space resources (CPOM), while their proportions were determined by the quality of food resources driven by hydraulic conditions. Temporal dynamics was of lower importance. Our results suggest the role of CPOM be crucial in woodland headwater streams, as it greatly enhances habitat quality, serving both as a food source to dominant shredders and a space source to most macroinvertebrates.     

The effects of food presentation and microhabitat upon resource monopoly in a ground-foraging ant (Hymenoptera: Formicidae) community

In Neotropical wet forests several species of omnivorous, resource-defending ants, live and forage in close proximity to one another. Although the forest floor is heterogeneous in microhabitat and food quantity, little is known about the impact of microhabitat and food variation upon resource monopoly among ants. We investigated how food type and microhabitat influence food monopoly in resource-defending ants in old-growth tropical wet forest in the Caribbean lowlands of Costa Rica. We measured several microhabitat characteristics at 66 points in a 0.5 hectare plot, and baited each point with two categories of tuna bait. These baits were presented in "split" and "clumped" arrangements. We measured the frequency of bait monopoly by a single species, as well as the number of recruited ant foragers at a bait. Out of five common species, two (Wasmannia auropunctata and Pheidole simonsi) more frequently monopolized one bait type over the other, and one (P. simonsi) recruited more ants to the split baits. We then considered the recruitment response by all ant species in the community. We found that the frequency of monopoly, sharing, and the absence of ants at a given point in the rainforest differed with bait type. The frequency of monopoly was associated with microhabitat type in two out of eight microhabitat variables (leaf litter depth and palms); variation in two other types (canopy tree distance and leafcutter ant trails) was associated with changes in forager number. In at least two ant species, food presentation affected monopoly at baits; among all resource-defending ants, the microhabitats where ants foraged for food and the type of food located determined in part the frequency of monopoly and the number of foragers at the food item. These results suggest that the location and presentation of food items determines in part which ant species will utilize the resource.     

Seasonality and composition of benthic coarse particulate organic matter in two coastal tropical streams with different land uses

This work studies benthic CPOM in two streams of Ecuador: the Atacames stream, located in a developed watershed, and the Súa stream, located in a rural watershed and used as a reference. It is tested whether the amount, composition and timing of benthic CPOM will differ between them as a function of watershed and riparian land uses. Benthic CPOM was collected at five study sites on each stream with a Surber net and classified into four categories: leaves, twigs and bark, flowers and fruits and debris. Leaves were further identified to genus or species. There were no significant differences in the amount, composition and timing of benthic CPOM between the streams. CPOM storage showed strong seasonality linked to seasonal rainfall and a weak relation with land uses, channel width and stream order. Diversity of the benthic CPOM was high and 30 species contributed to the benthic leaf pool. Presence or absence of Ficus species with heavy leaves that are easily retained in the streambed explained the spatial distribution of benthic CPOM, so spatial differences in the composition of the riparian vegetation in these tropical streams seem to be more important to explain CPOM distribution than in their temperate counterparts.

     

Predicting the structure and diversity of young-of-the-year fish assemblages in large rivers

1. Interactions between environmental variables and 0+ fish assemblages in the upper River Garonne (France) were quantified during late August 1995.
2. The abundance and diversity of the fish assemblages in floodplain channels were modelled using Artificial Neural Network (ANN) analysis and nine variables: the abundance of the six dominant species, fish specific richness, overall abundance of 0+ fish and the Shannon index of diversity. Multiple regression analysis was also used to assess ANN performance.
3. Using 596 samples, correlation coefficients ( r adjusted) between observed and estimated values of the nine dependent parameters were all highly significant ( P < 0.01). Expected values from the tested data were significantly related to the observed values. The correlation coefficient between observed and estimated values ( r ) varied from 0.70 to 0.85.
4. The ANN provided a high quality prediction, despite the complex nature of the relationship between microhabitat composition and fish abundance.
5. Garson's algorithm was used to provide the explanatory power needed in ecology when using black-box models. Parameters contained in the models (i.e. weighting) were used to determine the relative contributions of explanatory variables and thus to ascertain the structure of fish communities.