Interspecific interactions modulate bioturbation efficiency and nutrient dynamics in freshwater benthic communities

Abstract

Using the snails Filopaludina bengalensis and Gabbia orcula, along with tubificid worms and Chironomus sp., one monospecific and one combinatorial experiment were conducted in microcosms over 28 days to explore the direct and indirect interactive effects of non-predatory snails and tubificid worms on the bioturbation activity of chironomid larvae. These experiments examined how the species, whether alone or in combinations of two or three, influenced nutrient cycling in freshwater environments. On a comparative scale, the snails demonstrated higher N and P efflux than the tubificid worms and chironomid larvae, with the values normalized to biomass. At the community level, the chironomid, in combination with F. bengalensis and tubificid worms, showed a significant increase in N and P flux compared to the control. Unlike the two-species treatments, only the chironomid combined with F. bengalensis displayed higher N flux relative to the three-species treatment. Additionally, the indirect interactions of grazer snails more effectively inhibited the tube-dwelling behaviour of larvae than gallery-diffuser oligochaetes. Thus, the study reveals that species-specific functional traits and their interactions have a stronger effect on nutrient dynamics than species richness alone.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-39976-5.

Introduction

Organic nutrients originating from allochthonous and autochthonous sources are deposited in the sediment of freshwater ecosystems. Bioturbation modifies the fate of organic deposits either directly by promoting resuspension or indirectly by altering their physical state^1,2. Bioturbation by the benthic macrofauna is acknowledged as an engineered mechanism that facilitates changes in the physical state of substrates and regulates nutrient dynamics in aquatic ecosystems^3,4. Bioturbators, through burrowing, feeding, excretion, and mucus secretion, promote the organic mineralization, thereby accelerating the diffusive flux of solutes from sediment to the overlying water^5,6. The transport of solutes and water is primarily induced by the nature of the biogenic structure on the substratum. In addition, the morphology and behaviour of the macrofauna are influential factors favouring biogeochemical processes in freshwater ecosystems^7,8.

Both theoretical and empirical studies show that different types of benthic bioturbators vary in effective traits (bioturbation potential) in ways that can influence habitat modification and biogeochemical processes. These two ecological processes are not identical, with their exclusive mode of trait expression leading to different contributions to ecological processes. For example, the diverse feeding and burrowing strategies of bioturbators can affect the physical and chemical properties of the benthic habitat in various ways, thereby influencing nutrient dynamics in aquatic habitats^9,10. Species from different functional groups exhibit different yet comparable impacts on ecosystem processes. However, on a community scale, multiple species with the same or different functional traits fulfil novel ecological functions, in aggregate. The correspondence of relating the species-specific function to the biodiversity can better explain the complexity of the ecosystem¹¹. Due to current concerns regarding species extinction, many studies have focused on examining species-specific impacts^8,12–15. An increasing trend towards manipulative experiments to assess links between species composition and richness on ecosystem functioning, such as organic mineralization and nutrient cycling, is observed. Since the relationship between species richness and nutrient regeneration partly reflects species-specific traits⁷, it is crucial to highlight the contributions of functional traits of species to biogeochemical cycling.

The bioturbation efficiency varies with the morphological features and behavioural activities of the bioturbators^7,14,16–18. In the freshwater ecosystem, where the diverse assemblages of benthic fauna showed considerable variation in terms of size, shape, behaviour, and physiological activities^19,20, the outcomes of the bioturbation are expected to differ significantly^7,18,21. Among the known benthic macroinvertebrates, the tubificid worm, chironomid larvae, and the snails Filopaludina bengalensis and Gabbia orcula are recognized as dominant fauna in the freshwater habitats of India^22,23. While these species are taxonomically distinct, they are often considered significant contributors to the bioturbation process, sharing similar trophic status. While the upward conveyor tubificid worms construct galleries in the sediment and transport sediment particles upward^4,15, the downward conveyor chironomid larvae show constant undulation within the burrow tube or the nest and transport sediment particles from the surface to the deeper zone^15,17. In comparison, the snails F. bengalensis and G. orcula are considerably robust and actively graze over the sediment⁸. Owing to their burrowing, nudging, and browsing activities, the snails are found to modify the sediment stratigraphy extensively⁸. A large number of studies reveal the species-specific effects of snails^8,24–27, chironomid larvae¹³and tubificid worms^4,28, still, little effort has been made to decipher the functional importance of their co-occurrence in the freshwater habitat.

The freshwater snails, chironomid larvae, and tubificid worms are the crucial components of benthic ecosystems^8,15, though the functional traits of the species vary considerably. For instance, the biomass, mobility and life span of the snails are considerably higher than the worms and the larvae. As a consequence, the effective outcome of the bioturbation is expected to vary with the functional responses. In the present study, an attempt was made to justify the bioturbator species composition (functional traits combination) as predictors of the per-capita contribution to N and P efflux. Moving beyond species-specific observations, this study reflects how interference competition among bioturbators modulates biogeochemical processes. By quantifying these interspecific interactions, it becomes easier to apply the ecological realism of microcosm findings to natural environments, which improves the ability to predict ecosystem functions. As a result, the effect of the non-trophic interactions among the bioturbators could be inferred. To compare the effects of individual species in their natural density on the nutrient dynamics in aquatic environments and to determine whether the non-competitive interactions among them in multispecies communities affect species-specific activities, we conducted two experiments in laboratory microcosms. Therefore, the purpose of the present study was twofold. First, we compared the impacts of four benthic macrofauna occurring in their natural density on the physicochemical properties of sediment, and overlying water and on the productivity of the system. Secondly, since the interspecific interactions between non-predator species sharing the same trophic level affect their habitat utilization pattern^21,29–31, we intended to investigate how a potent bioturbator chironomid larva, alone or in combination with snails and tubificid worms having diverse biological traits, could alter the nutrient dynamics in freshwater habitat.

Results

Effects of different macrofauna species on the N and P regimes of water

On day 0, before the introduction of the organisms, the aquatic parameters of monospecific and combination experimental microcosms were estimated and are represented in Table 1. The effects of macrofauna species occurring at their natural densities on the N and P dynamics in the water column are depicted in Fig. 1. The addition of chironomid larvae, tubificid worms, and snails enhanced N and P concentrations and reduced N: P ratios in the water column. N and P concentration measurements differed significantly among the treatments (treatment effect p = < 0.0001 for N, p = < 0.0001 for P; Fig. 1; Supplementary file Table S1a, b) and over time (time effect p = < 0.0001 for N and p = 0.019 for P; Supplementary file Table S1a, b). Except for the tubificid treatment on day 21 for N concentration, all faunal treatment groups demonstrated higher N and P concentrations in the overlying water compared to the control at each sampling event. After the 28-day experiment stabilization, the N concentrations of the four treatment groups were ranked as follows: G. orcula > F. bengalensis > tubificid worms > chironomid larvae, and the P concentrations were ranked as follows: F. bengalensis > G. orcula > tubificid worms > chironomid larvae. Notably, the mean N concentration was highest in the presence of F. bengalensis, followed by G. orcula, tubificid, and chironomid treatments. The F. bengalensis treatment also maintained the highest mean P measures in the water column, followed by tubificid, G. orcula, and chironomid treatments. Effects of each species on N and P dynamics per unit biomass (per gram) were quantified and are presented in Fig. 2. At the end of the 28-day experiment stabilization, the per gram effects of fauna on N concentrations were ordered as follows: G. orcula > F. bengalensis > tubificid worms > chironomid larvae, and on P concentrations as follows: tubificid worms > F. bengalensis > chironomid larvae > G. orcula. Considering the constant biomass effect, the mean N concentrations exhibited no distinct treatment variability (p = 0.99; Fig. 2; Supplementary file Table S2a), whereas P measures (p < 0.0001) in the water columns differed significantly (Fig. 2; Supplementary file Table S2a, b) among the four treatments. The tubificid treatment maintained the highest mean P measures in the water column, followed by chironomid, F. bengalensis, and G. orcula treatments.

Table 1.

Experimental design with a detailed description of each treatment for testing the results of monospecific effects and the effects of interactions among four different species in aquatic habitats. Data from at least nine treatment replicates were considered for each experiment. Biomass values (g/column) for the snails were based on the total weight, including the shell.

Treatment	Individual abundance (individual/column)				Species Richness	Wet biomass (g/column)
	Chironomid larvae	Tubificid worms	G. orcula	F. bengalensis
Experiment I: Monospecific experiment
C	Control without fauna
CH	45	-	-	-	1	0.29 ± 0.02
TU	-	90	-	-	1	0.41 ± 0.03
GO	-	-	16	-	1	1.97 ± 0.2
FB	-	-	-	2	1	2.15 ± 0.18
Experiment II: Combination experiment
C	Control without fauna
CH	45	-	-	-	1	0.31 ± 0.03
CH + TU	45	90	-	-	2	0.75 ± 0.03
CH + GO	45	-	16	-	2	2.18 ± 0.07
CH + FB	45	-	-	2	2	2.39 ± 0.15
CH + TU+FB	45	90	-	2	3	2.96 ± 0.08

Fig. 1.

Changes of N and P concentrations in the water column of control (C) and four faunal treatment groups with chironomid larvae (CH), tubificid worms (TU), G. orcula (GO), and F. bengalensis (FB) occurring in their natural density. Values are presented as mean ± SE.

Fig. 2.

Compared to control changes in N and P concentrations in water considering per gram effect of chironomid larvae (CH), tubificid worms (TU), G. orcula (GO), and F. bengalensis (FB). The negative concentration differences indicate lower N and P measures in the fauna treatments compared to the control. Values are presented as mean ± SE.

Effects of interspecific interactions

Combination treatment effect on nutrient dynamics

The impacts of interspecific interactions on nutrient dynamics were assessed in the presence of chironomids alone and in combination with snails and tubificid worms, as illustrated in Fig. 3. Results indicated that N and P concentration measurements significantly differed among the treatments and over time (Supplementary file Table S3a, b). It was observed that chironomids induced a substantial N and P flux from sediment to water; intensities increased further in combination with F. bengalensis and tubificid worms. Chironomids in combination with F. bengalensis and tubificid worms significantly (p < 0.05; Tukey’s HSD; Supplementary file Table S3b) augmented the N and P flux processes, respectively. Additionally, it was evident that chironomids associated with tubificids exhibited higher N and P flux to the overlying water compared to those associated with snails. The presence of snails F. bengalensis and G. orcula, in combination with chironomids, reduced N: P ratios in the water compared to treatments involving only chironomids. However, an opposite trend was observed in N: P concentrations in the water column with chironomids and tubificid treatments. At the end of the experiment, the combination of chironomids and tubificid treatments exhibited a maximum increase in N: P ratios in water compared to other treatments.

Fig. 3.

Compared to the control, changes in N and P concentrations in water as a result of different combination treatments. Here, C = control, CH = only chironomid, CH + TU = chironomid with tubificid, CH + GO = chironomid with G. orcula, CH + FB = chironomid with F. bengalensis and CH + TU+FB = chironomid with tubificid and F. bengalensis. Values are presented as mean ± SE.

Bioturbation effect size

No significant relationship was found between the species richness of the bioturbators and the availability of nitrogen (R² = 0.01; p = 0.23; Fig. 4a) and phosphorus (R² = 0.01; p = 0.33; Fig. 4b) in the overlying water, indicating species composition serves as a better predictor for N and P dynamics than species richness. Among the two-species treatments, the combination of chironomids with F. bengalensis and tubificid worms demonstrated a significant increase (p < 0.05) in N and P flux compared to the control; however, the increase in N and P concentrations in combination with G. orcula was less pronounced. Moreover, unlike the other two-species treatments, the chironomids in combination with F. bengalensis exhibited higher N flux than the three-species treatment. Lastly, bioturbation effect sizes illustrated that, although more positive effect sizes confirmed that N: P concentrations in water increased with species richness, no significant relationship was noted between species richness and bioturbation effect size (R² = 0.02; p = 0.12; Fig. 4c). This study confirmed that the functional traits of the benthic community are more effective in predicting nutrient dynamics in the water column compared to species richness.

Fig. 4.

Effects of macrofauna species richness on (a) Nitrogen N and (b) Phosphorus P concentrations of the water. Values are represented as mean ± 95% CI. Linear regressions were performed by regressing measurements of nutrients as functions of macrofauna species richness. The dashed lines represent ± 95% CI for nutrients in control columns. The relationship between bioturbation effect sizes and bioturbator species richness (c). Positive effect size values indicate a higher N to P ratio than defaunated controls. Values are presented as mean ± SE.

Impact of multi-species interaction on the burrow decay rate of chironomid larvae

The burrow permanency of chironomid larvae demonstrated distinct treatment variations (F_4,22 = 40.53; p < 0.0001), significantly (p < 0.05; Tukey’s HSD test) decreasing upon coexistence (Fig. 5). It was found that mean burrow permanency decreased with increasing species richness, with the highest burrow decay rate occurring in the three-species treatment (CH + TU+FB). Burrow decay rates considerably increased in the presence of snails. After 21 days, nearly all chironomid burrows on the sediment bed were infilled in the presence of snails (Fig. 5). The fitted exponential decay models were all highly significant (r² > 0.74, p < 0.0001; Fig. 5) with decay constants (r) differing significantly between single-species (chironomid only) and multispecies (two and three species) treatments. The chironomid associated with F. bengalensis exhibited the fastest rate of decline (decay constants: CH = 0.01 ± 0.0, CH + TU = 0.02 ± 0.005, CH + GO = 0.042 ± 0.004, CH + FB = 0.092 ± 0.009, and CH + TU+FB = 0.104 ± 0.007). The mean burrow permanency indicated that a burrow would last on average 100 days in the chironomid-only treatment, and 61.10 ± 20.03 days, 26.67 ± 4.77 days, 11.49 ± 1.11 days, and 9.99 ± 0.72 days when associated with tubificid worms, G. orcula, F. bengalensis, and both tubificid and F. bengalensis, respectively. The results of the study also indicated that quantitative differences in chironomid burrow density among treatments diminished with increasing biomass (Fig. 6a) and species richness (Fig. 6b) of the bioturbators. Regression analysis also indicated that burrow density was not a significant predictor of nutrient flux rates (Fig. 6c, d). While burrow density may serve as a proxy for mechanical bioturbation, the direct physical disturbance on the sediment bed by the gastropods and organic mineralization of sediment deposits emerges as the primary driver of nutrient flux augmentation. Depending on the nature of specific taxa involved and the magnitude of their bioturbation activities, these biological processes can supersede the influence of physical reworking (burrow counts). The species-specific nature of these interactions suggests that the extent of nutrient flux is highly dependent upon the functional identity of the bioturbators and the specific characteristics of the benthic ecosystem.

Fig. 5.

Density (means ± SE) of chironomid burrows remaining as a function of time and treatments. The fitted exponential decay models are: CH (r² = 0.77, p = < 0.0001); CH + TU (r² = 0.88, p = < 0.0001) CH + GO (r² = 0.74, p = < 0.0001) CH + FB (r² = 0.81, p = < 0.0001) CH + TU+FB (r² = 0.74, p = < 0.0001). Here, CH = only chironomid, CH + TU = chironomid with tubificid, CH + GO = chironomid with G. orcula, CH + FB = chironomid with F. bengalensis and CH + TU+FB = chironomid with tubificid and F. bengalensis. Values are presented as mean ± SE.

Fig. 6.

Changing burrow density of chironomid larvae on the sediment bed as a function of increasing (a) biomass and (b) species richness of the bioturbators. The burrow decay results explicitly connect the relationship between burrow density and (c) N and (d) P flux rate in the water columns.

Treatment effects on sediment properties and periphyton growth

Significant treatment variations were observed concerning the water holding capacity (F_5,24 = 2.702; p = 0.044; Supplementary file Table S4a) and porosity measures (F_5,24 = 2.743; p = 0.042; Fig. 7; Supplementary file Table S4b) of the sediment that received various combinations of fauna. The Chl-a levels measured from the substrate surface displayed no distinct treatment variability (Fig. 8; RM-ANOVA, treatment effect, p = 0.33). However, at the end of the experiment (on day 28), Chl-a measures were lower in all faunal treatments compared to the control columns (Fig. 8).

Fig. 7.

Variation in porosity measures in surface sediment of five faunal treatment groups (n = 3) measured at the end of 28 days of the experiment. The horizontal line represented the porosity value of the control columns. Here, CH = only chironomid, CH + TU = chironomid with tubificid, CH + GO = chironomid with G. orcula, CH + FB = chironomid with F. bengalensis and CH + TU+FB = chironomid with tubificid and F. bengalensis. Values are presented as mean ± SE.

Fig. 8.

Comparison of Chl-a measures of periphyton from the microcosms of six treatment groups. Here, C = control, CH = only chironomid, CH + TU = chironomid with tubificid, CH + GO = chironomid with G. orcula, CH + FB = chironomid with F. bengalensis and CH + TU+FB = chironomid with tubificid and F. bengalensis. Values are presented as mean ± SE.

The DA results also showed distinct treatment variations in relation to the predictor variables (Fig. 9a, b). These differences between the treatment groups with distinct species combinations further validate the considerable effectiveness of all the bioturbator species used in this experiment on the freshwater ecosystem functioning. This discriminant model signifies the contribution of the variables against the four discriminant functions (F1, F2, F3 and F4; Fig. 9b). Among these the two dominant discriminant functions (F1 and F2) explained 97.74% of the variance (the F1 explaining 90.08% and the F2 explaining 7.66%) (Fig. 9b). For most of the species combinations, the Mahalanobis distances were found to be significant (p < 0.001) (Fig. 9b). The ordination of five treatment groups in two axes along the biplot represents sufficient discrimination of various species combinations based on the functions of several variables that describe the differences between the groups and confirm the influence of all variables in the separation of groups.

Fig. 9.

Results of discriminant analysis considering aquatic parameters (N= Nitrogen, P= Phosphorus, N: P= Nitrogen to Phosphorus ratio), burrow density, sediment porosity and water holding capacity (WHC), and periphytic Chl-a contents as explanatory variables against five treatment groups with various species combinations (CH= only chironomid, CH + FB= chironomid + F. bengalensis CH + TU= chironomid + tubificid, CH + GO= chironomid + G. orcula, and CH + TU+FB= chironomid + tubificid + F. bengalensis.) as response variable. (a) A biplot representation with the ordination of various species combinations, the response variables (i), and the various bioturbation parameters, the explanatory variables (ii), (b) statistical summary of the results showing Eigenvalues, canonical discriminant function coefficients, and Mahalanobis distance matrix for response variables (significant values are bold at p < 0.001). The significant value of the Wilk’s λ = 0.006; F_28,59 = 6.782; P < 0.0001, supports the application of the multivariate analysis.

Discussion

The present study revealed how the species-specific bioturbation activities of four dominant benthic macrofauna and the functional species diversity (species combinations) influence the ecological processes and ecosystem functions, i.e., the movement and intensity of bioturbation and nutrient dynamics in freshwater habitats. Monitoring the effects of the interspecific interactions among the non-predators that share the same trophic level is a novel approach to ascertain the importance of benthic communities and the sustainable management of freshwater ecosystems. The results demonstrated that the macrofaunal effects on the ecosystem processes and functions are species-specific, indicating a significant influence of interspecific interactions on the species-specific effects. In our study, the interspecific interactions negatively affected the burrowing abilities of chironomid larvae and therefore influenced their habitat utilization pattern.

Assessment of individual species performances

Perhaps the colonized populations of the chironomid larvae and the tubificid worms are accountable for the equivalent pattern in the alterations of the nutrient quality, similar to the snails, despite their size and biomass differences; however, it would be better explaining the discrepancy in their ability to transport nutrients transport^32–34from sediment to the water. The sizes of the snails are considerably large, which is why sheer biomass is responsible for the turbulence and propulsion of the sediment, facilitating the upwelling of nutrients to the water column^8,35. Species functional traits contribute a significant effect on the diffusive nutrient fluxes from the sediment to the overlying water¹¹. The downward conveyors chironomid larvae construct J or U-shaped tubes on the superficial sediment layer¹⁵. The upward conveyors tubificid worms dig galleries on the sediment bed and deposit faecal pellets at the sediment surface¹⁵. Whereas, gliding and grazing activities of the surface bulldozers operculate snails F. bengalensis and G. orcula alter the natural texture of the surface sediment⁸. Tube-dwelling behaviour of chironomid larvae and the gallery-diffuser worms facilitates the bio-irrigation process and circulates water into the sediments, which may accelerate the solute transport processes more efficiently than the molluscs, whose activities are restricted to the superficial sediment layer^8,11. Bio-irrigation by chironomid larvae facilitates the transportation of oxygenated water deep into the sediment, thereby promoting the nitrification process. Concurrently, gallery diffusion on the sediment bed promoted by tubificid worms increases the transport of pore water to deeper sediment and the release of dissolved ammonium into the water column^2,13. All these processes enhance the increased N flux from the sediment to the overlying water. In comparison, browsing and bulldozing activities by freshwater snails on the sediment bed destabilize the oxidized microzone, thereby promoting the release of P into the overlying water⁸. Our study findings support the concept as proposed by some previous researches^7,36, that functional traits are also important in manipulating ecosystem functioning.

Species-specific traits in multispecies benthic communities predict the nutrient dynamics

The effects of multi-specific treatments were more complex than the single-species treatment. The burrows constructed by the chironomid larvae were reduced in number in the presence of snails and tubificid (Fig. 5), reflecting their struggle for habitat utilization and overall habitat modification^37,38. Our experiment also confirmed that the snails and oligochaete worms affected the burrow permanency of chironomid larvae and their bioturbation activities. Grossly, the study outcomes ascertain the fact that the competi effects among the species influence the nutrient transport process as the nitrogen and phosphorus concentrations in the water column are altered with the addition of species (Fig. 4a, b). The study also illustrated the reduction rate of chironomid burrows best fitted with the exponential decay model. The decay rate was maximum when larvae were combined with snails than alone or in combination with tubificid worms (Fig. 5). While heavily grazing on fine detritus over the sediment³⁹, snails destroyed the larval burrows, complementarities in space occupation^30,40may be the reason that could reduce the interspecific interaction between tubificid worms and chironomid larvae. Likewise, in the absence of the snails, as the galleries and burrow channels become more complex with time, the chironomid and tubificid treatments were found to enhance diffusive flux from the deeper sediment layer. This resulted in a proportionate increase in the N to P ratio in the water, consistent with the observations of the previous study¹¹. Our data indicate interference competition between gastropods and chironomids, where surface bulldozing by the gastropod physically destroys or collapses the sedentary tube structures built by chironomid larvae. Conversely, interactions between distinct conveyor taxa are characterized as non-antagonistic. As the physical structures, like galleries and tubes, made by tubificid worms and chironomid larvae are vertically separated, they avoid spatial overlap. Instead of interfering, their activities thereby promote functional synergy in nutrient dynamics rather than negative interference. Therefore, it can be stated that the particular species composition in a community better predicts the nutrient dynamics in the freshwater system. An earlier study also demonstrated that the grazing activities on the sediment bed allow for the reduction of the spatial variability, resulting in the relocation of some other bioturbators having different functional traits, thereby reducing their bioturbation intensity⁴¹.

The magnitude of the net flux of solutes to the water is directly controlled by the biogenic activities of macrofauna that regulate sediment oxygenation¹¹. The suspension feeder snails F. bengalensis and G. orcula heavily graze on fine detritus of the substratum³⁹. While the detritivore tubificid worms act as upward conveyors, they are moving within the burrows in a restricted fashion, often able to propel a smaller amount of water. The conveyor belt organism, chironomid larvae, acts as the downward conveyor. The chironomid larvae show constant body undulation within the burrow tube or the nest^15,37. Disturbances due to the activities of non-predatory snails and tubificid worms on chironomid burrows can be considered as interference competition¹⁵that induced negative effects on the burrow permanency of the larvae. The study observations suggest that the snails, tubificid worms, and chironomid larvae represented the same functional feeding group, and shared the same habitats while exploiting spaces and resources differently¹⁰. Several empirical studies have already shown that the co-occurrence of non-predators often reduces the growth and survival rate^12,42, decreases burrowing ability³⁷, interrupts the feeding success¹⁰of co-occurring species, and influences the abundance of the producer community⁴². In corroboration with these studies, the present experiments also demonstrate that the coexistence could alter the species-specific activity of chironomid larvae.

Species interactions underpin many ecosystem processes, like nutrient dynamics in aquatic ecosystems¹⁶. As an extension, the present study contributes to the understanding of how species interactions specifically drive nutrient dynamics and their implications for broader ecosystem processes. A linear regression represented no statistical significance (P > 0.05) of both the N and P dynamics (Fig. 4a, b) as a function of the species richness. Lack of substantial correlation between overall measures of nutrient (N and P) dynamics across different species combination treatments apparently reflects that the effects on fundamental ecosystem processes are species-specific and depend on behaviour and non-competitive interactions of the individual organisms present.

However, the burrow density was significantly correlated with the species richness, indicating the differences among the species in the creation of burrows (Fig. 6). The DA results (Fig. 9a, b) validate the effectiveness of all the bioturbator species used in this experiment on the freshwater ecosystem functioning, varying with the species composition (species richness). As shown in the ordination of the various combinations of bioturbators (response variables) as characterized by the explanatory variables (the bioturbation indicators like N, P, Chl-a, burrow density and porosity), a difference (significant values of the Mahalonobis distance; Fig. 9) was prominent in terms of the species combinations. A considerable extent of differences is prominent among the various species compositions, reflecting that the bioturbation indicators have varied significantly. Perhaps the differences in the burrow-building capacity and other matters may have contributed to these differences among the species combinations.

The results reveal that the N and P fluxes show distinct treatment variability (Table 3a). However, post hoc comparisons (Table 3b) justified that not all the treatments differed significantly from the control. Only the chironomid in combination with F. bengalensis and tubificid worms significantly augmented the process of N and P flux compared to the control. Increased functional diversity might show enhanced nutrient flux. However, high functional variation among the co-occurring species significantly affects important ecosystem functions compared with different combinations of fauna. Earlier studies^16,43reported that the effects of species richness on ecosystem functions are not additive, but a consequence of interspecies interactions. Such observation was prominent in the present context, where the chironomid in combination with F. bengalensis showed higher N flux compared to the three-species treatment.

These observations support the idea that interspecies interactions are better predictors of species-specific activity, which could be crucial for ecosystem functions in bioturbated sediments. In summary, the observations of the present study prove beneficial for selecting insects, snails, and oligochaete worms in the bioturbation of the freshwater ecosystem. In the current context, where nutrient management in eutrophicated freshwater bodies is prioritised, investigating different species combinations will be valuable for understanding the role of bioturbators and their applications in sustaining freshwater ecosystems. As illustrated in the present study, the various combinations of snails, chironomid larvae, and tubificid worms may facilitate eutrophication management of the freshwater ecosystems like ponds, lakes and reservoirs. Since the observations were made in microcosm level, a probable scaling effect is warranted when the snails, chironomids, and tubificids are used for the management of the larger water bodies. An exploration of the scaling effects using the bioturbators may be carried out to precisely predict the enhancement of water quality and nutrient parameters of the water bodies.

It is important to note that in this additive experimental design (Experiment II), biomass varied greatly with increasing species richness of the bioturbators. Therefore, it is relevant to mention that the impacts of bioturbation in freshwater systems may be influenced by both the functional identity of the species and the total biomass of the organisms. This study design does not allow us to statistically separate the richness effects from the effects of density and biomass, and therefore reflects the collective impact of increasing community complexity. Future studies considering biomass as a fixed factor or applying multi-metric indices such as Rao’s Quadratic Entropy (Rao’s Q) would be necessary to clarify the specific contribution of functional richness relative to the total faunal biomass.

Multispecies benthic communities regulate the producer growth

The periphytic Chl-a measures changed over time in all treatment groups. At the end of the experiment, all the treatment microcosms with various species combinations showed lower Chl-a content compared to the control. This result can be attributed to the differences in the trophic status and the feeding activities of the bioturbators. Different empirical studies confirm that the snails can stimulate algal growth by increasing the release of nutrients from the sediment^24,26. Simultaneous grazing activity of snails also reduces periphytic algal biomass⁸. Furthermore, bioturbation activities can lead to an increase in the turbidity of the water, which may restrict periphyton growth with the alterations in light penetration and physicochemical water conditions^6,8,13. Therefore, the variation in periphytic algal Chl-a deposition among the treatment cores in the presence or absence of bioturbators was the consequence of these contrasting forces. Extending this observation in the context of biomanipulation, the regulation of the algal growth can be achieved by augmenting the snail abundance along with the chironomid larvae and tubificid worms. An amplified population of the bioturbators would potentially release nutrients that may promote the periphyton growth. Our findings highlight a critical biogeochemical trade-off in the use of gastropods for biomanipulation. While snails effectively suppress periphytic biofilms through top-down grazing, their bioturbation activities simultaneously augment the internal loading of N and P. These contrasting effects highlight the importance of the strategic assembly of bioturbator taxa during restoration. Therefore, the introduction of snails must be balanced with primary producers to prevent excessive nutrient load in the freshwater ecosystem. An appropriate combination of bioturbators is therefore required, else the bioturbation effects may promote the eutrophication process and hinder successful management. Apparently, the species combination of bioturbators and the relative abundance of the constituent species are significant in facilitating biomanipulation and reducing eutrophication risks in a scale-dependent manner. Inferring these microcosm findings to natural ecosystems requires recognising specific factors like sediment heterogeneity, predator presence, and advective transport that might change at larger scales and modulate the observed interactions. In larger-scale systems, spatial complexity and hydrodynamics are likely to reduce the direct non-predatory interference between snails and conveyors. Furthermore, the presence of a producer community and diverse trophic interactions would probably influence the net biogeochemical output of benthic communities.

Among the various macroinvertebrates recognized for the bioturbation activity in freshwater, the body size, life span and behavioural traits become prominent contributors to enhance the outputs of the processes. Thus, freshwater snails F. bengalensis and G. orcula, chironomid larvae and tubificid worms carry in those aspects of life history traits considerably. In terms of biomass, the snails are multiple times bigger than the larvae and worms, and the life spans are also considerably different. While the larvae and the worms continuously filter sediment materials, the snails bulldoze the sediment materials, thereby varying in material output in their respective bioturbation ability. In natural conditions, a combination of these animals is obvious, which would represent their interaction ability. The combination of the species appeared to be relevant in the bioturbation output measured in terms of changes in N, P, and Chl-a measures and sediment architecture, and then the species richness, particularly due to the differences in the functional traits, contributed to the bioturbation output. The bioturbation ability of the macroinvertebrates contributed significantly to the habitat structures. For instance, the chironomid larvae created burrows in the sediment, while the tubificid worms remained within the sediment. In contrast, the benthic snails created varied architecture on the sediment surface, which varied with the size. Considering the variations in the traits and activities, coupled with the observed results on bioturbation, a variation in the rate of the bioturbation process was obvious. Perhaps, the composition of the bioturbators will therefore be more important in bringing the desired changes in the bioturbation processes.

Conclusion

The results of the study reflect that the functional traits of the bioturbators, the freshwater snails F. bengalensis and G. orcula, the tubificid worms, and the larvae of Chironomus sp. determine the bioturbation ability with considerable differences in species-specific performances. Presumably, the species-specific disparity in the biomass, behavioural activity, and adaptation may have affected the efficacy in modifying benthic habitat structure and nutrient flux in freshwater habitats. The combined effects of the functional traits of the species may be responsible for the turbulence and propulsion of the sediment, facilitating the upwelling of nutrients through the water column. Consequently, the bioturbators significantly affected the diffusive nutrient fluxes from the sediment to the overlying water. The resultant effect may prove augmentative for eutrophication, while the decreasing amount of chl-a substantiates that the grazing by the snails may be helpful in biomanipulation of the freshwater system. The study results also highlighted the interspecific interactions in the benthic habitat, influencing both the habitat utilization patterns and the functional traits. Since the observations were made at a lower scale (microcosm study) than what can be observed in the natural conditions, like wetlands, ponds and even riverine systems, possible deviation in the N, P, Chl-a dynamics is warranted. Our experiment showed that the snails affected the burrow permanency of chironomid larvae more extensively than the oligochaete worms. Besides, the presence of F. bengalensis and tubificid worms in combination with chironomids positively influenced the nutrient efflux from the sediment to the water. This empirical data substantiates the fact that community composition exerts a more profound influence on N and P cycles than species richness, identifying functional identity as the governing driver of nutrient flux. In all instances, the combination of the snails, chironomids, and tubificid worms acted in harmony to enhance water quality and nutrient parameters, which, however, may vary in larger freshwater systems, reflecting the scaling effect and non-trophic interactions.

Materials and methods

Collection of sediment and macrofauna

The sediment sample was obtained from a local pond in Howrah (22°35′12.95″N, 88°18′33.51″E), West Bengal, India. The sun-dried sample was sieved through a 500 μm mesh⁷. Chironomid larvae and Tubificid worms were collected from a local ornamental fish market in Galiff Street (22°36′19.86″N, 88°22′22.31″E), Kolkata, West Bengal, India. Living larvae and worms were placed into polythene bags with a small amount of water and secured with rubber bands to trap air. The bags were kept inside an insulated box⁴⁴and transported to the laboratory for rearing and identification. In the laboratory, the freshly collected tubificid worms were placed in a plastic tray and maintained under slowly running water until they were transferred to the experimental microcosms⁴⁵. The chironomid larvae were transferred to a plastic tray containing deionised water and autoclaved sediment, with a continuous air supply provided^46,47. The freshwater snails G. orcula and F. bengalensis were collected from various local ponds and wetlands in the Kolkata metropolitan area (22° 32′ 27.96″ N, 88° 20′ 16.08″ E), West Bengal, India. The snails were collected either by dragging an aquatic hand net (mesh size 200 μm) along the bottom region of the littoral zone or hand-picked directly from the bottom sediment^48,49.

Experimental design

Chironomid larvae, tubificid worms, and operculate freshwater snails G. orcula and F. bengalensis, which possess different functional traits, were selected for the study. The identification of freshwater snails was aided by the available literature^50,51. Based on morphological characteristics, the genus-level identification of chironomid larvae was conducted using appropriate keys^52,53. The morphological features of tubificid worms were identified by following the available literature^54,55. The microcosm experiments were carried out in cylindrical glass columns (0.4 m long with an internal diameter of 0.075 m) filled with a 0.05 m thick sediment bed overlaid by a 0.35 m water column. Before filling the water, several plastic tapes were affixed to the inner surface of the water column’s wall, which would serve as a substratum for periphytic chlorophyll-a development⁵⁶. The microcosms were stabilized for 2 weeks (pre-incubation period) before introducing fauna into the systems⁷. The day of introduction of organisms to the microcosms was considered as day-0 of the experiment. The entire experiment was conducted in an outdoor environment under natural aeration and sunlight for an additional 28 days following the introduction of fauna.

Monospecific experiments comprised five treatment groups (See Experiment I of Table 2). Given that the snails, insect larvae, and oligochaete worms differ in terms of biomass, behavioural activity, and ability to move surface sediment, it was difficult to directly compare species-specific effects based on fixed density or biomass. Therefore, the study employed two different snail species, the operculate benthic F. bengalensis and G. orcula, as model freshwater snails (suspension feeders), alongside the chironomid larvae and tubificid worms (known bioturbation agents), using their natural density (relative abundance in the field condition) instead of a fixed density⁵⁷. Microcosms were inoculated with snail species, chironomid larvae, and tubificid worms at natural densities, as detailed in Table 1. The relative abundances of F. bengalensis, G. orcula, chironomid larvae, and tubificid worms aligned with previously reported values^8,13,15,22. These densities were further validated by field surveys conducted across various lentic systems within the Kolkata metropolitan area, West Bengal, India, between May and July 2016. Owing to the obvious differences in the features of the macroinvertebrate bioturbators, the estimated treatment effects, namely, N, P and Chl-a, were normalized with the biomass. Thus, the study results were represented as per gram (wet biomass) effects of the macrofauna as the response variables. The combination experiment involved six treatments (See Experiment II of Table 2) to illustrate the consequences of interspecific interaction on nutrient dynamics.

Table 2.

The initial values of water parameters and sediment properties in the microcosms of monospecific and combination experiments.

Parameters	Values
Monospecific experiment
Nitrogen (N) content	1.554 ± 0.186 mg.L− 1
Phosphate (P) content	0.043 ± 0.004 mg.L− 1
Dissolved Oxygen (DO)	6.8 ± 0.2 mg.L− 1
pH	7.2 ± 0.5
Combination experiment
Nitrogen (N) content	2.238 ± 0.267mg.L− 1
Phosphate (P) content	0.028 ± 0.001 mg.L− 1
Dissolved Oxygen (DO)	7.3 ± 0.4 mg.L− 1
pH	7.1 ± 0.3
Sediment properties at the beginning of the experiment the end (day-28) of the experiment for eight treatments
Percent of organic carbon carbon	1.12 ± 0.07
Available Phosphorus (ppm)	0.87 ± 0.02
Total Nitrogen (g/Kg)	1.37 ± 0.14

Monitoring changes in water and sediment features and periphytic Chl-a content

To assess the single-species and multi-species treatment effects on the water and sediment properties, as well as the growth of the producer component, scheduled sampling and measurements were performed (Supplementary file Fig. S1) focusing on nitrogen (NH₄⁺-N, NO₂ˉ-N, NO₃ˉ-N) and phosphate (PO₄³⁻-P) contents in the water, sediment porosity, and periphytic chlorophyll-a (Chl-a) content. The volume of 50 ml of water taken from each experimental microcosms in course of the estimations of nutrients was renewed immediately with the same amount of deionized water. Although volume replacement with deionized water represents a potential confounder, it was implemented to regulate the chemical matrix and impede the introduction of exogenous ions or trace impurities. The sediment samples were sliced from the upper 2 cm of three replicated microcosms from experiment II. The Chl-a concentration of the periphyton developed on plastic tape was estimated by removing a single tape from each microcosm of experiment II and scraping off the deposited algal film using a brush and razor. NH₄⁺-N in water was measured at 640 nm using the phenate method⁵⁸. NO₂ˉ-N concentration was measured at 543 nm using a sensitive diazotization method⁵⁸. The Brucine method⁵⁹estimated NO₃ˉ-N concentration. PO₄³⁻-P levels in water were determined using the ascorbic acid method⁵⁸. The sediment porosity was analysed by estimating the ratio of the volume of pore space to the total volume of sediment⁶⁰, and water holding capacity (WHC) was estimated from the ratio of the mass of the water contained in the saturated sediment to the mass of the saturated sediment⁶¹. The Chl-a content was determined using a 90% acetone extraction procedure⁶².

Bioturbation effect size

The natural logarithm (ln) of each treatment means (for 1, 2, and 3 species treatments) divided by the control mean, known as the log response ratio, was calculated to represent the bioturbation effect sizes⁶³. This ratio is widely applied in estimating the relative differences between treatments. At each level of species richness, the overall effect sizes were aggregated to obtain the effect sizes for that level³. The bioturbation effect sizes were subsequently compiled using a regression equation as a function of species richness to examine the effects of bioturbator species richness on the nitrogen to phosphorus regime of the water column.

Burrow permanency

By monitoring the burrow permanency of chironomid larvae in the sediment, it was possible to estimate the rate of burrow decay. The chironomid larvae were examined within various species combinations to calculate this decay rate. Each microcosm contained 45 Chironomid larvae (10,400 individuals/m²), and best represented five distinct combination treatments with four different macroinvertebrates (See experiment II of Table 1). To track the change in burrow numbers over time, the microcosms subjected to various fauna treatments were observed weekly. In this study, the average number of intact chironomid burrows recorded immediately after the introduction of larvae into the only chironomid treatment columns was considered the initial burrow count. The intact burrows of chironomid larvae were identified by the presence of small vertical tubes sticking out of the surface sediment. All microcosms were inspected weekly for 28 days. During each visit, the presence or absence of burrows was noted, and the total count of intact burrows within each microcosm was recorded to assess changes in burrow density over time. Burrow decay rates for each treatment were estimated by counting the remaining chironomid burrow numbers over time and fitting the exponential decay model (Bt = B₀ e^–rt)^64,65. The decay constant (r) was then utilised to calculate the mean burrow permanency (B_pr = 1/r) for each of the five faunal treatments⁶⁴.

Data analysis

Data were tested for normality and homoscedasticity using the Shapiro-Wilk test and Levene’s test, respectively, followed by an assessment of the sphericity of the covariance using Mauchly’s test. The data on the porosity measures, burrow maintenance rates, and Chl-a contents met the distributional assumptions; however, the data for N and P concentrations did not. As the data transformation (log or square-root transformed) on N and P concentrations failed to meet any distributional assumptions, treatment variations were evaluated using mixed model ANOVA, with time (days) and treatments as fixed factors, and replicates as random factors. Following the ANOVA, multiple comparisons among the treatment means were evaluated by performing Tukey’s HSD post-hoc test, with the family-wise error rate maintained at an alpha level of 0.05. to identify specific differences between treatment means. To determine the treatment variations on the porosity measures, water holding capacity and burrow maintenance rates, one-way ANOVA was performed using treatments as the main factor. A linear regression model was implemented to validate the relationship between bioturbation effect sizes and species richness. One-way RM-ANOVA was conducted to ascertain significant treatment variation in mean values over time concerning periphytic Chl-a measures. Furthermore, to determine if the seven test parameters (N, P, N: P, burrow density, porosity, WHC, and Chl-a) could distinguish the five treatment groups, data from these groups with various species combinations were analysed using discriminant analysis (DA). Discriminant function analysis (DA) is a multivariate statistical method that enables discrimination among target groups based on certain quantitative explanatory variables^66,67. In this study, several aquatic parameters (N, P, N: P), burrow density, sediment porosity, WHC, and periphytic Chl-a contents were considered as predictor variables to differentiate the treatment groups with five distinct species combinations. The outcome would facilitate an understanding of the differences between treatment groups concerning various predictor variables. All statistical analyses were performed following Zar⁶⁸and using XLSTAT software (Addinsoft 2010)⁶⁹.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Anupam Chakraborty contributed to the execution of the experiments and the data collection and analysis and preliminary and final draft preparation; Goutam Kumar Saha contributed to the planning and the supervision of the experiments and methodological details; Gautam Aditya conceptualized and planned the experimental design, supervised statistical analysis and data presentation and the preliminary and final draft preparation.

Funding

The first author AC acknowledges the financial assistance from UGC, India, through University of Calcutta in accomplishing this compilation (Sanction No.UGC/487/Fellow (Univ) dated 04.07.2017: University of Calcutta).

Data availability

The data that support the findings of this study are available from the first author Dr. Anupam Chakraborty, upon authentic and reasonable request.

Declarations

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Ethical approval and consent to participate

All authors have read, understood, and have complied as applicable with the statement on “Ethical responsibilities of Authors” as found in the Instructions for Authors”.