WaterSubject

The shredder guild plays an outstanding role in the functioning of headwater stream ecosystems by processing allochthonous leaf litter. Traditionally, the abiotic habitat template is regarded as the major determinant of its organization, and only a limited number of studies support the importance of biotic interactions. The aim of the present study was to examine whether competition plays a significant role in organizing the shredder guild of caddisflies in a riffle. Null-model based co-occurrence, co-existence and guild variation analyses were used in the study of guild organization. In addition, the traditional variance to mean ratio was applied for measuring the intraspecific aggregation of guild members. The non-significant metric values of co-occurrence and co-existence analyses predicted that competition was of limited importance in structuring the spatial organization of the shredder guild. The observed aggregated spatial distribution of species, suggests that besides stochastic events, deterministic forces should also contribute to the organization of the shredder guild of caddisflies.The study was carried out in the northern part of Hungary near Bernecebaráti, where the Bernecei stream flows through the northern part of the Börzsöny Mountains. The stream meanders through an oak-hornbeam woodland (Querco-Carpinetum) with riparian vegetation dominated by alder (Alnus glutinosa). The canopy provides extensive shading and leaves little light for instream primary production. Hence, matter production is predominantly based on allochthonous leaf litter from the surrounding riparian vegetation. The stream has a well-developed riffle-pool morphology with cobble and gravel as the dominant substratum, and is representative of the natural submountane streams of the Carpathian region in Central Europe.

Based upon field studies a riffle (length 13 m, average width 4.1 m) was selected in a natural second-order reach of the Bernecei stream (47°58?03?N, 18°55?02?E, 188 m a.s.l.). The water depth within the riffle varied between 3 and 8 cm. The dominant substrate was cobble (70%) with gravel (25%), the size of the particles varied between 4 and 8 cm. Caddisfly larvae were collected on 18 May 2004 using a standardized Surber sampler (area: 0.09 m², mesh size: 500 ?m). In a 10-day period before the sampling date, there was no rainfall that would influence the rate of flow of the stream. Although the riffle was well shaded, there was not any significant accumulation of coarse particulate organic matter that would have favored the aggregation of shredders. Consequently, the riffle was considered to present a relative homogeneous environment for stream dwelling caddisflies, and allowed performing distribution data-based null models for detecting competition, to be used. The riffle was embedded between pool patches, which differed in their caddisfly assemblages. Hence, we hypothesized that the caddisfly assemblage in the riffle formed a relatively independent (i.e. well segregated) community. Before sampling, the riffle was mapped and the position of each of the 62 Surber samples was chosen randomly. The sampling began downstream to avoid disturbance to the upstream sector of the riffle. To dislodge caddisflies from the substrate and allow the streamflow to carry them into the Surber net, the substrate within the quadrat frame was agitated by hand to a depth of about 5 cm, followed by careful examination of all cobble and gravel clasts. Each sample was preserved in 5% formalin, labeled and returned to the laboratory for examination. Caddisflies were identified to species level. The whole list of species is available. Shredder species were used in the further analyses. However, most of the caddisflies use a variety of feeding strategies. Thus, only species with high affinity for shredding coarse particulate organic matter were defined as shredder.The present study shows that the shredder guild members of caddisflies were assembled randomly in the riffle: neither co-occurrence, nor co-existence (CI_guild) analyses revealed any deviation from random processes. Consequently, competition (or any interaction between populations) does not influence the spatial organization of the guild, and this suggests that the incidence of one guild member does not influence the incidence of another. As a contribution to the debate within community ecology as to the usefulness of distributional data for detecting competition, the competitive effect did not appear to influence the spatial organization of the shredder guild of caddisflies. However, our results cannot be regarded as evidence of missing interactions among guild members. For example, several studies suggest that competition between caddisflies might affect fitness parameters, such as weight or development time, although not necessarily affecting the spatial distribution of the interacting individuals.

However, the aggregated spatial distribution of the individual populations seemed to support the view that non-random processes might also contribute to the organization of the guild. According to the aggregation model of co-existence, competition between guild members should force populations of the same guild into an intraspecifically aggregated spatial distribution. In the present study, the non-significant co-occurrence and co-existence indices (CI_guild) predicted that competition could not be responsible for the intraspecific aggregation pattern observed.

Besides competition, several explanations have been developed to explain intraspecific spatial aggregation of stream macroinvertebrates. First, macroinvertebrate eggs are laid in egg-masses, thus larval macroinvertebrates belonging to the same species have an a priori aggregated spatial distribution. However, this distribution should have been later diluted because of movement activity of the larvae thus egg-laying cannot be responsible for the aggregated spatial distribution. Another explanation focuses on habitat heterogeneity. Several experimental results suggest that microhabitat characteristics are of major importance in determining macroinvertebrate community organization. According to this explanation, considering the study site as a homogeneous habitat was incorrect. Applying this theory to the patterns observed, microhabitat patches should favor the incidence of each species equally, but in an intraspecifically aggregated way. This description is in agreement with the flow refugia hypothesis according to which random species aggregations are assembled in patches less affected by streamflow.

Although the present study was focused and designed for detecting competition between guild members, the habitat heterogeneity hypothesis fits best the observed patterns. Thus, it is very likely that microhabitat heterogeneity shapes guild organization of shredder caddisflies.

Guild variation analysis showed that the observed frequencies did not differ from a chance event. Although the primary aim of guild variance analysis was to gather an insight into the possible constraints affecting species assembly, it could not reveal the consequence of any biotic or abiotic factor that would favor a non-random distribution of guild members. Only a single significantly positive association (between Halesus digitatus and H. tesselatus) out of 21 analyses suggested that species abundances were non-independent from each other. However, by adjusting the P-values of the 21 comparisons to an experiment-wise error rate of 0.05, all comparisons showed independent (random) species abundance distribution. Applying this finding to the microhabitat heterogeneity hypothesis, a microhabitat patch should equally favor or not the occurrence of each guild member (for instance, the effect of presence or absence of leaf packs on shredder caddisflies. Moreover, this finding is in accordance with previous studies stating that significant interspecific aggregation among taxa is not common and mostly positive.

Considering the experimental design applied, several factors restrict the generalization of our results. First, instead of the whole macroinvertebrate community, only caddisflies were studied. Although caddisflies have an important role in leaf breakdown in forest streams, other invertebrate groups (e.g. Amphipoda, Plecoptera, etc.) with shredding activity are present in the Bernecei stream. However, the presence of other invertebrates was sporadic in the riffle during the sampling (D. Schmera, pers. obs.). Second, organization of the shredder guild of caddisflies was studied only in a single riffle and only by one sampling occasion. Nonetheless, this restriction allowed our analyses to be performed independently from the spatial and temporal variations observed in stream ecosystems.

In summary, this study demonstrates that within a riffle habitat, the shredder guild of caddisflies is randomly organized, whereas the spatial distribution of its individual populations is aggregated. It is likely that microhabitat heterogeneity within the riffle was responsible for the aggregated spatial distribution of shredder guild members of caddisflies. Deterministic forces of the stream habitat template are the most important factors structuring the shredder guild of caddisflies and the effect of competition on the spatial organization seems to be limited.