SEARCH

SEARCH BY CITATION

Keywords:

  • competition;
  • consumer pressure;
  • density-dependent processes;
  • environmental stress;
  • facilitation;
  • herbivory;
  • marine;
  • plant–herbivore interactions;
  • rocky reefs;
  • stress-gradient hypothesis

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

1. The stress-gradient hypothesis predicts an increase in the importance and intensity of positive species interactions towards extreme ends of gradients generated by either physical stress or consumer pressure. However, little attention has been devoted to assessing how the co-occurrence of different gradients of stress and variations in the abundance of the benefactor can influence switches in species interactions.

2. On shallow rocky reefs, we assessed shifts in the effects of different covers of Vermetid tube-building gastropods (benefactor) on macroalgae (beneficiary), under experimental conditions generated by crossing a gradient of consumer pressure (sea urchin density) and a gradient of physical stress (sediment deposition).

3. Negative effects of Vermetids on macroalgae in the absence of herbivores switched to positive at intermediate grazing pressure, but sedimentation and benefactor cover determined their intensity. Thus, association with Vermetids provides macroalgae with a refuge from herbivores. When consumer pressure was the greatest, facilitation persisted both at natural and moderately enhanced sedimentation if the benefactor cover was reduced. When the benefactor monopolized space, facilitation was only observed at natural levels of sedimentation. Thus, the relationship between the outcome of the benefactor–beneficiary interaction (expressed as the Relative Interaction Index) and consumer pressure varied from linear to asymptotic or quadratic, according to sedimentation levels and benefactor abundance.

4.Synthesis. These results show that shifts in the direction and intensity of species interactions are regulated by the interplay of biological and physical factors. In addition, they suggest that density-dependent processes are more likely to shape species interactions at extreme ends of gradients of stress.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The formulation of the stress-gradient hypothesis (SGH) by Bertness & Callaway (1994) represents a milestone of research on the mechanisms regulating the balance between negative and positive species interactions (Brooker et al. 2008). The SGH predicts that the role of positive interactions (i.e. facilitation) will be greater at increasing levels of stress (adverse physical conditions or consumer pressure). Since its formulation, a number of studies have empirically or theoretically tested the predictions from the SGH, mostly focusing on plant–plant interactions. This large research effort, resulting in at times apparently contradictory evidence, has promoted attempts to assess how factors such as life-history traits and ontogenetic shifts of interacting species (Callaway & Walker 1997; Miriti 2006; Maestre et al. 2009), the nature and severity of the gradient of stress (Michalet et al. 2006; Kawai & Tokeshi 2007; Maestre et al. 2009) or the performance variable examined (Maestre, Valladares & Reynolds 2005) can alter predictions from the SGH.

Despite resulting advances, several gaps in our understanding of the mechanisms regulating the balance between positive and negative interactions remain to be filled in. First, little attention has been devoted to assessing how the co-occurrence of different sources of stress can influence switches in species interactions (but see Kawai & Tokeshi 2007 and le Roux & McGeoch 2010). In particular, interactions between physical gradients of stress and those due to consumer pressure (disturbance sensuGrime 1977) are almost utterly unexplored (Smit, Rietkerk & Wassen 2009), despite the recognition that natural communities are generally structured by the interplay of biological and physical factors (Menge & Sutherland 1987; Benedetti-Cecchi, Bulleri & Cinelli 2000). Empirical studies that have simultaneously taken into account biotic and physical stressors suggest that these interact to determine the balance between competition and facilitation (Crain 2008; Eskelinen 2008). Recently, Smit, Rietkerk & Wassen (2009) have developed conceptual models that convincingly show how the inclusion of biotic stress along physical stress gradients can alter predictions from the SGH. To the best of our knowledge, no study has, however, experimentally crossed biotic and physical gradients of stress in a full orthogonal fashion, as is necessary to disclose variations in thresholds of stress intensity at which shifts in species interactions occur.

A second point in need of consideration is the abundance at which the benefactor occurs (Bruno & Bertness 2001; Kikvidze, Armas & Pugnaire 2006; Irving & Bertness 2009). Density dependency of processes regulating species co-existence has long been recognized in both terrestrial and aquatic systems (Harper & White 1974; Drake 1991). Supposedly, for any given set of environmental conditions, the response of beneficiaries to the presence of the benefactor can be hypothesized to be dependent upon the relative density at which the latter is found, especially when demographic dynamics of interacting species are controlled by the same limiting resources. Thus, without explicitly controlling for the effects of the density of the facilitator, shifts in the balance between negative and positive species interactions can be difficult to interpret and generalize (Irving & Bertness 2009).

Using a shallow subtidal rocky reef assemblage as a model system, we investigated how predictions from the SGH, in its original formulation, can be altered when a gradient of physical stress is superimposed on a gradient of consumer pressure and when the density of the benefactor is allowed to vary. In benthic assemblages, competition for primary space between sessile species can be intense, regardless of the trophic level they belong to (Bulleri 2009). Nonetheless, associational defence can enable the persistence of prey species in the face of intense consumer pressure (Hay 1986; Stachowicz 2001). For example, on subtidal rocky reefs off the coast of California, the presence of the colonial anemone, Corynactis californica, facilitates the persistence of algal turfs made of filamentous and foliose forms, at intermediate densities of the sea urchins, Strongylocentrotus purpuratus and Strongylocentrotus franciscanus, by impeding their movement and thus reducing their feeding efficiency (Levenbach 2008, 2009).

Although the focus of research on positive species interactions in subtidal habitats has been almost exclusively on associational defence, physical factors, such as sediment deposition and wave action, are key determinants of the structure of rocky benthic assemblages (Bulleri 2009). Thus, these habitats provide an ideal system in which to assess how interactions between physical and consumer pressure gradients influence shifts in species interactions.

Here, we investigated the direction and intensity of the interaction between two primary space occupiers, turf-forming macroalgae and tube-building Vermetid gastropods (Vermetus triqueter), in different conditions generated by crossing a gradient of consumer pressure (grazing by sea urchins) and a gradient of physical stress (sediment deposition). On Mediterranean rocky reefs, the distribution and abundance of algal turfs is controlled by the grazing activity of the sea urchins Paracentrotus lividus and Arbacia lixula (Benedetti-Cecchi, Bulleri & Cinelli 1998; Bulleri, Benedetti-Cecchi & Cinelli 1999). Irregular surfaces formed by tube-building gastropods could indirectly benefit turf-forming macroalgae, by constraining the movement of consumers. Thus, we predicted that negative effects of tube-building gastropods on algal turfs due to pre-emption of space would prevail in the absence of herbivores, but would shift to positive at intermediate densities of herbivores (i.e. at increased grazing pressure), to eventually switch back to negative when the grazing pressure becomes extreme (Levenbach 2009).

These predictions can be altered in a number of ways when taking into account the abundance of the benefactor and the depositional regime. The intensity of positive effects is, in fact, likely to be dependent not only on grazing pressure, but also on the amount of primary space available for macroalgae to settle within the matrix formed by the tube-building gastropods (i.e. the relative cover of tubes). In addition, Vermetids, as shown for other filter or mucus feeders, can be smothered by sediments, whereas algal turfs are largely resistant to this type of stress (Airoldi 2003). Although the majority of empirical tests of the SGH account for stress gradients that influence both interacting species or at least the beneficiary (i.e. the species that benefits from the association with the stress-tolerant species), here, it is the benefactor (Vermetus) that is more susceptible to the stress generated by high sediment deposition. Thus, when sedimentation deposition is large, algal turfs might be able to gain competitive superiority over tubes and overgrow them. A decline in the cover of Vermetids might, on the other hand, imply a loss of refuge from grazing for macroalgae.

Effects of increasingly harsher environmental conditions are not limited to the benefactor and the beneficiary, but extend to grazers. A thick layer of sediments may, in fact, prevent urchins from accessing prey items, protecting macroalgae from consumption. Alternatively, sea urchins could re-suspend sediments during their feeding bouts, influencing their grazing behaviour and/or ameliorating physical stress for tube-building gastropods. Finally, sediment layers might limit the diffusion into the water column of nutrients that sea urchins release through their excretions, indirectly favouring algal growth. In summary, switches in the effects of Vermetids on algal turfs generated by consumers are likely to be dependent upon the response of each the three interacting taxa (algal turfs, Vermetids and urchins) to enhanced levels of sediment deposition and are, therefore, difficult to predict.

Materials and methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

This study was carried out on shallow rocky reefs along an exposed coast c. 10 km south of the town of Livorno (NW Mediterranean, 43°30′N, 10°20′E) and lasted for c. 21 weeks. At depths from 3 to 5 m, stands of turf-forming algae, mostly made of filamentous forms, alternate with patches monopolized by encrusting coralline algae (Bulleri, Benedetti-Cecchi & Cinelli 1999; Bulleri, Bertocci & Micheli 2002) or by the Vermetid gastropod, Vermetus triqueter. In July 2009, 54 patches monopolized by Vermetids (cover ∼ 100%), c. 2 m2 in extension and 4–5 m apart, were randomly chosen in the proximity of urchin shelters. Eighteen patches were then randomly assigned to each of the following Vermetid cover treatments (Fig. 1): (i) natural cover (cover ∼ 100%); (ii) 50% removal (cover ∼ 50%) and (iii) total removal (cover ∼ 0%). Vermetid cover was manipulated within four 10 × 10 cm plots, randomly selected along the external margin of each patch (Fig. 1). In these plots, macroalgae could be present as epiphytes, but values of cover were generally very low (F. Bulleri, pers. obs.). Vermetids were removed using hammer and chisel, paying attention to minimize the damage to remnant individuals. In order to avoid confounding the effects of benefactor density with those of altered patchiness, 50% Vermetid removals were produced by subdividing the plot in 25 2 × 2-cm subquadrats and removing organisms from 12.5 randomly selected subquadrats. Field observations suggest that this procedure did not increase the mortality in remnant individuals.

image

Figure 1.  Diagram of the experimental design, illustrating the 27 combinations generated by crossing the factors Vermetid cover, urchin density and sediment deposition. Eighteen patches (large ovals) were assigned to each of the following Vermetid cover treatments: natural cover (cover ∼ 100%), 50% removal (cover ∼ 50%) or total removal (cover ∼ 0%) of Vermetids. The cover of Vermetids was manipulated in four replicated plots (= 4) selected along the margins of each patch. Six patches of each Vermetid cover treatment were randomly allocated to each of three different urchin density treatments: urchins absent, intermediate and high urchin density. Two patches for each combination of Vermetid cover and urchin density treatments were finally assigned to each of the following sedimentation levels: natural, moderate addition or large addition.

Download figure to PowerPoint

Two replicated patches of each Vermetid cover treatment were then assigned to each of the nine combinations generated by crossing the following two factors (Fig. 1): urchin density (absent: c. 0 individuals m−2; intermediate: c. 6 individuals m−2; high: c. 12 individuals m−2) and sediment deposition (natural; moderate enhancement: addition of c. 100 g m−2 day−1; large enhancement: addition of c. 200 g m−2 day−1). Densities of urchins included in the experiment encompassed the density range at this site (Bulleri, Benedetti-Cecchi & Cinelli 1999) and were maintained manually, by adding or removing individuals as necessary, every fortnight (see Fig. S1). Handling of urchins has been previously shown not to alter their feeding behaviour (Bulleri, Tamburello & Benedetti-Cecchi 2009).

The amount of sediment to be added to create a gradient of physical stress was calculated on the basis of the natural range of sedimentation at this site (Bulleri & Benedetti-Cecchi 2008). Every fortnight, 15 and 30 g of sediment were added to each of the four plots within patches assigned to a moderate and a large enhancement of sediments, respectively (Fig. 1). Due to adverse sea conditions, the addition of sediments could not be always performed at scheduled times, but delays were never >3 days. At our study site, inputs of sediments to shallow water habitats are due to run-off from the overtopping sandstone cliff during heavy rain episodes. In order to use the same type of material, the sediment was collected up the cliff, as described in Bulleri & Benedetti-Cecchi (2008). The sediment was passed through a 2-mm mesh and the fraction not retained, composed of fine particles with a grain size within the range of those trapped by algal turfs, was used for our manipulation.

The percentage cover of macroalgae and Vermetids in experimental plots was quantified visually by means of a 10 × 10-cm plastic frame, subdivided into 25 subquadrats. A score from 0% to 4% was given to each taxon in each subquadrat and the percentage cover was obtained by summing over the entire set of subquadrats. Organisms were generally identified to the species level, except for filamentous and encrusting forms that were treated as morphological groups. Species susceptible to grazing (i.e. those composing algal turfs) were subsequently grouped as necessary to test our hypothesis.

The response of algal turfs to experimental conditions was evaluated by means of a four-factor anova, including the factors: Vermetid cover (3 levels; fixed), urchin density (3 levels; fixed and crossed with Vermetid cover), sedimentation level (3 levels; fixed and crossed with Vermetid cover and Urchin density) and patch (2 levels; random and nested in the interaction of the other three factors). The same linear model was used to assess variations in the cover of Vermetids at the end of the experiment.

The Relative Interaction Index (RII), proposed by Armas, Ordiales & Pugnaire (2004), was used to measure the relative interaction intensity between algal turfs and Vermetids, under different combinations of grazing pressure and sediment deposition. This index was chosen for its strong mathematical and statistical properties (Armas, Ordiales & Pugnaire 2004). The RII can be expressed as RII = (Bw − Bo)/( Bw + Bo), where Bw and Bo represent the biomass (percentage cover in this study) of the facilitated species (algal turfs) in the presence or absence of the facilitator (Vermetids), respectively. The RII ranges between 1 and −1 and is therefore symmetrical around zero and does not present discontinuities in its range, with identical absolute values for competition and facilitation. In this study, values of Bo were calculated as the average percentage cover of algal turfs in plots from which Vermetids were totally removed, separately for each combination of urchin density and sediment deposition. Variation in the RII was analysed by means of the same anova model previously described, but in this case only two levels were included in the factor Vermetid cover (natural vs. 50% removal). Homogeneity of variances was tested using Cochran’s test and data were transformed when necessary. Data were analysed even when homogeneity of variances could not be achieved, as the anova is robust for departure from this assumption when there are many independent replicates and sizes of samples are equal (Underwood 1997). Results were, however, interpreted with caution, by judging the significance more conservatively (α = 0.01). Student-Newman-Keuls (SNK) tests were used for a posteriori comparisons of the means.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Response of algal turfs to experimental treatments

Four months after the start of the experiment, Vermetids had negative effects on the cover of algal turfs in the absence of herbivores (Table 1, Fig. 2). Increasing Vermetid cover caused a progressive decline in the cover of algal turfs, although the difference between 50% and total Vermetid removal treatments was significant only at natural sediment deposition levels (Fig. 2A–C).

Table 1. anovas of the effects of the initial cover of the facilitator (Vermetids: natural cover, 50% removal and total removal), density of sea urchins (urchins: absent, intermediate and high), sediment deposition (sediment: natural, moderate enhancement and large enhancement) and patch on the percentage cover of algal turfs, RII values and final Vermetid cover (see text for further details) Thumbnail image of
image

Figure 2.  Mean percentage cover (+SE) of algal turfs at different combinations of grazer density (absent, intermediate, high) and Vermetid cover (total removal, 50% removal and natural cover), separately for (A) natural sediment deposition, (B) moderate and (C) large enhancement of sediment deposition; = 8. Letters above bars illustrate the results of SNK tests; different letters indicate significant differences. Note that letters above bars do not allow comparisons of algal turf cover across graphs (a posteriori comparisons among levels of sediment deposition within each combination of Vermetid cover and grazer density are reported in Table 1). Bars marked with the symbol † were indicated as significantly different from the SNK tests, despite no identification of alternatives to the null hypothesis.

Download figure to PowerPoint

In general, Vermetids promoted algal development at intermediate grazer densities, but there was large variation according to their cover and sedimentation regime. At natural sedimentation levels, the cover of algal turfs was significantly greater in 50% Vermetid removal plots than in plots where they had been totally removed or left untouched (Fig. 2A). At enhanced levels of sedimentation, although the a posteriori analysis did not detect significant differences, there was a trend for the cover of algal turfs to be greater in plots where Vermetid cover had been reduced by 50% or left at natural values (Fig. 2B,C).

When the grazing pressure was the greatest, the cover of algal turfs decreased dramatically at natural levels of sediment deposition, irrespective of Vermetid cover (Fig. 2A). When sediment deposition was moderately enhanced, the cover of algal turfs in plots lacking of Vermetids tended to be smaller than 50% Vermetid removal plots, but greater than natural Vermetid cover plots (Fig. 2B); differences were not, however, significant. In contrast, at the upper end of the sedimentation gradient, the presence of Vermetids, either at natural or halved cover, caused a significant decrease in the cover of algal turfs (Fig. 2C, Table 1). Indeed, following the total removal of Vermetids, at the extreme ends of the biotic and physical gradients of stress, the cover of algal increased to values matching those recorded in the absence of grazers.

Variations in the RII across combinations of factors

In the absence of grazers, RII values were all negative (Fig. 3), suggesting that the pre-emption of space by Vermetids negatively influenced the cover of algal turfs. The intensity of this negative interaction varied among combinations of Vermetid cover and sediment deposition (Table 1). RII values were significantly less negative when the cover of Vermetid was reduced by 50% than when it was left untouched, both at natural and moderately enhanced levels of sediment deposition (Fig. 3A,B, Table 1). In contrast, there was no variation between 50% removal and natural Vermetid cover plots when sedimentation stress was the greatest (Fig. 3C).

image

Figure 3.  Mean RII values (±SE), at different combinations of grazer density (absent, intermediate, high) and Vermetid cover (natural or 50% removal), separately for (A) natural sediment deposition, (B) moderate and (C) large enhancement of sediment deposition; = 8. Letters besides symbols illustrate the results of SNK tests; different letters indicate significant differences. Note that letters do not allow comparisons of RII values across graphs (a posteriori comparisons among levels of sediment deposition within each combination of the levels of Vermetid cover and density of grazers are reported in Table 1).

Download figure to PowerPoint

Relative Interaction Index values generally shifted to positive at intermediate densities of urchins, suggesting that algal turfs can achieve protection from grazers through association with Vermetids (Fig. 3). At natural levels of deposition, the value of RII was slightly negative where the cover of Vermetids was left untouched, but positive where it was reduced by 50% (Fig. 3A). In contrast, there was no variation between Vermetid cover treatments at enhanced levels of sedimentation (Fig. 3B,C).

At the upper end of the grazing pressure gradient, there was substantial variation in RII values, according to Vermetid cover and sediment deposition (Table 1). Positive RII values indicate that Vermetids, irrespective of their cover, could still facilitate the persistence of algal turfs when sediment deposition was at natural levels (Fig. 3A). In contrast, when sediment deposition was moderately increased, positive effects of Vermetids on algal turfs persisted when their cover was reduced, but not when it was left at natural values (Fig. 3B). Finally, when sediment deposition was the greatest, effects of Vermetids on algal turfs were negative, regardless of their cover (Fig. 3C).

Effects of grazer density on the facilitator

The cover of Vermetids differed according to their initial cover and the density of grazers, but was not influenced by the sedimentation regime (Table 1). Differences in Vermetid cover due to the initial manipulation were still clearly evident at the end of the experiment (SNK test: natural cover > 50% removal > total removal). The cover of Vermetids was, however, lower in plots exposed to the grazing of high densities of urchins than in those assigned to other treatments (no urchins: 51.44 ± 4.77; intermediate urchin density: 44.79 ± 4.44; high urchin density: 36.54 ± 4.24; values averaged across levels of Vermetid removal and sediment deposition treatments, = 72).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

The intensity and direction of the effects of Vermetids on algal turfs were regulated by the interplay between physical stress and consumer pressure. These effects were dependent upon the relative cover of the benefactor (Vermetids), introducing further complexity to the scenarios generated by crossing gradients of sediment deposition and density of urchins. In accordance with the findings of Kawai & Tokeshi (2007), our results confirm the importance of taking into account the co-occurrence of different gradient of stress (either physical or biotic in nature) and warn against the over-simplistic assumption of dominance of single drivers. In addition, they suggest that conceptual frameworks, developed to explain variations in plant–plant interactions along stress gradients, can be successfully extended to the case of interactions between species at different trophic levels (i.e. primary producers and mucus-filter feeders, in this case), advancing our understanding of community assembly rules (Altieri, Silliman & Bertness 2007).

In the absence of urchins, pre-emption of space by Vermetids had marked negative effects on algal turfs, irrespective of the sedimentation regime. Vermetids and algal turfs, despite belonging to different trophic levels, compete for space (Fig. 4A). Algal turfs are mostly composed of filamentous forms that, being highly tolerant to sedimentation stress and reproducing vegetatively, can rapidly colonize free space in degraded environments (Airoldi 1998). The colonization of the substratum by Vermetids, relying on the supply of larvae from the water column, is slower. However, because of the great plasticity of their shell, they can hold primary space at the expense of other sessile forms (Schiaparelli & Cattaneo-Vietti 1999).

image

Figure 4.  Web of interactions among algal turfs, Vermetids, sea urchins and sediments, showing direct (solid line) and indirect (dotted line) effects and positive (+) and negative (−) interactions. Vermetids and algal turfs (a) compete for primary space. (A) In the absence of sea urchins, pre-emption of space by Vermetids limits the cover algal turfs. Large sediment deposition can (b) depress the ability of Vermetids to prevent algal overgrowth, (c) indirectly favouring the development of algal turfs. (B) At intermediate densities of sea urchins, Vermetids, by (d) constraining foraging bouts of urchins, lessen (e) grazing pressure and generate (f) indirect positive effects on algal turfs (light grey arrows). When Vermetids monopolize space (i.e. Natural cover) and sedimentation deposition is at natural levels, (a) negative direct effects prevail over (f) positive indirect effects. However, the addition of large amounts of sediments, by (b) decreasing the competitive ability of Vermetids, can (c) indirectly favour the development of algal turfs. (C) At large densities of sea urchins, Vermetids, both at natural and reduced covers, can still (f) indirectly facilitate algal turfs by (d) providing refuge from grazers, at natural levels of sedimentation. Large inputs of sediments can generate (h) indirect positive effects on algal turfs also by (g) reducing the grazing efficacy of sea urchins (dark-grey arrows).

Download figure to PowerPoint

The addition of sediments released, indirectly, algal turfs from negative effects generated by the monopolization of space by Vermetids (Fig. 4A). Enhancing sediment deposition did not affect the final cover of Vermetids, but it might have decreased the number of living animals inhabiting the tubes or caused physiological alterations. We did not assess sediment-induced mortality in Vermetids and we are, therefore, unable to provide an insight into the mechanisms operating. Nonetheless, the net result of intense physical stress was a depression in the ability of Vermetids to prevent overgrowth by algal turfs. Large amounts of sediments might have interfered with the production and/or functionality of mucus threads by Vermetids, reducing their ability to entrap food particles and, ultimately, to maintain their shells free of epiphytes.

In general, algal turfs benefited from the presence of Vermetids when the grazing pressure was increased to intermediate levels (Fig. 4B). Macroalgae can escape consumption through the association with sessile invertebrates (Levenbach 2008, 2009). Likely, Vermetids can constrain, either spatially or temporally, foraging bouts of sea urchins, by increasing the heterogeneity of the substratum (Fig. 4B). Both Arbacia lixula and P. lividus are generally observed on relatively flat surfaces (i.e. bare rock or encrusting coralline barrens), but rarely on irregular biogenic surfaces, such as those provided by erect or turfing macroalgae, likely because of a limited ability to adhere to and move across architecturally complex substrata (Gianguzza et al. 2010).

At intermediate levels of consumer pressure, the intensity of the effects of Vermetids on the algal turf was also dependent upon their relative cover and local physical conditions. A reduced cover (c. 50%) of Vermetids generally promoted the cover of algal turfs. In contrast, when at a cover of c. 100%, the effects of Vermetids on algal turfs varied according to the intensity of physical stress. Indeed, at natural levels of sediment deposition, algal turfs received no benefit from Vermetids, achieving covers comparable to those recorded in the absence of the benefactor. This suggests that negative effects of pre-emption of space by Vermetids were compensated for by positive effects due to the escape from consumers, as supported by a RII score close to zero. When sediment deposition was moderately enhanced, a c. 100% cover of Vermetids facilitated algal turfs by as much as a c. 50% cover. Again, sediments likely facilitated algal turfs indirectly, by decreasing the ability of the gastropods to prevent overgrowth (Fig. 4B).

When consumer pressure was the greatest, the direction and intensity of the effects of Vermetids on the algal turf exhibited large fluctuations, according to their cover and to the intensity of physical stress. The presence of a reduced cover of Vermetids could still mitigate the negative effects of a strong consumer pressure, either at natural or moderately enhanced sediment deposition. In contrast, when sediment deposition was the greatest, the effects of Vermetids on algal turfs switched back to negative. A larger cover of Vermetids was still able to benefit algal turfs under natural environmental conditions, but even a moderate increase in sediment deposition was sufficient to remove such facilitative effects.

Previous studies have shown that there are limits to facilitation. When stress becomes extreme, loss or decline of benefits can result from the decease of the benefactor, an impairment of its functioning (Brooker et al. 2006; Smit et al. 2007; Levenbach 2009) or, in the case in which the stress is due to a shared resource (e.g. water or nutrients in plants; space for sessile invertebrates), from the prevalence of competition (Tielbörger & Kadmon 2000; Maestre & Cortina 2004). Here, the benefactor (Vermetids) suffered some damage from the grazing by large densities of urchins. Negative effects of grazers on the cover of Vermetids were, however, independent of the intensity of physical stress and they cannot, alone, explain the diminished intensity of facilitation or the switch to competition that emerged only at enhanced sedimentation levels.

Indeed, physical stress (sedimentation) modulated the effects of grazers on primary producers in the absence of Vermetids. When at intermediate densities, urchins could control algal development across the entire range of sediment deposition, suggesting that the efficacy of consumers was not depressed when environmental conditions became harsher. When urchin density was the greatest, macroalgal cover was very small at natural sediment deposition, but increased progressively moving towards the upper end of the sedimentation gradient (see Fig. 2C). Enhancing the intensity of physical stress and grazer density likely resulted in indirect positive effects on algal turfs (Fig. 4C). Thus, environmental conditions, by setting the abundance values of the beneficiary in the absence of the facilitator, determined the persistence or decrease of facilitation at the upper end of the gradient of consumer pressure. This suggests that shifts from facilitation to competition at increasing levels of stress do not always imply a decrease in the ability of the benefactor to deliver benefits to the neighbour.

As proposed by Smit, Rietkerk & Wassen (2009), when environmental conditions become so harsh as to cause an abatement of consumer pressure, preys receive no benefit from the association with a consumer-resistant species. Our results support this mechanism (hypothesis 1 in Smit, Rietkerk & Wassen 2009) and suggest that it is likely to play a role whenever prey are less susceptible than the consumer and the benefactor to physical stress. In addition, our findings point to the need for more thought when attempting to quantify biotic stress, as behavioural or physiological responses of consumers to changes in environmental conditions are likely to introduce discontinuities in the relationship between consumer pressure and grazer density.

Experimental work is needed to identify how large densities of grazers and elevated levels of sediment deposition can result in positive effects on macroalgae. Here, we can only speculate on the mechanisms operating. Re-suspension of sediments caused by the movement of large numbers of urchins might have affected their behaviour, ultimately decreasing their grazing efficiency (Fig. 4B). It is worth noting that, at intermediate densities of urchins or sediment level, this effect was not negative enough to affect grazing behaviour.

Alternatively, a thick layer of sediments might have reduced the diffusion of nutrients excreted by urchins into the water column, ultimately favouring their exploitation by algal turfs. In marine habitats, macroalgae can benefit from nitrogen supply from fish and invertebrates (Williams & Carpenter 1988; Taylor & Rees 1998). Sea urchin excretions can represent an important source of nitrogen and phosphate for macrophytes (Williams & Carpenter 1988). For example, ammonium inputs from the sea urchin, Diadema antillarum, promoted primary productivity on tropical coral reefs, providing up to 19% of the estimated nitrogen demand of local algal turfs (Williams & Carpenter 1988). Large concentrations of ammonium and nitrates have been reported for excretions of P. lividus (Arafa, Sadok & El Abed 2006) and previous studies have reported positive effects of P. lividus on the growth of the rhizophytic seaweed, Caulerpa racemosa, suggesting that negative effects due to consumption can be overbalanced by positive effects due to fertilization (Bulleri, Tamburello & Benedetti-Cecchi 2009).

The relationship between species interactions and the severity of environmental or biological stress can be asymptotic or quadratic rather than linear, as originally predicted by the SGH (Callaway et al. 2002; Maestre & Cortina 2004; Brooker et al. 2006; Travis et al. 2006; Levenbach 2009). What mostly differentiates these relationships is their behaviour towards the upper ends of the stress gradient. Three different scenarios can be envisioned: (i) facilitative effects can persist at increasing levels of stress, reaching an asymptote at a positive value (Callaway et al. 2002); (ii) the ability of the benefactor to deliver benefits to the beneficiary can be impaired at large values of stress, with net effects on the beneficiary switching back to negative (Maestre & Cortina 2004) or (iii) asymptote to zero (neutral net effect), thus generating a hump-shaped curve (Brooker et al. 2006; Michalet et al. 2006; Smit et al. 2007; Levenbach 2009).

Superimposing a gradient of physical stress on a gradient of consumer pressure and manipulating the abundance of the benefactor, we generated all three forms of relationship. This shows that the shape of such relationships is largely context-dependent, being determined not only by the interplay of physical and biological forces, but also by the relative abundance of the benefactor. Previous studies have shown that population-level traits of benefactors, such as density and size of individuals, can modify their facilitative properties (Irving & Bertness 2009). Although variable in intensity, the effects of Vermetids on algal turfs shifted from negative to positive when moving from absence to intermediate densities of grazers, consistently among combinations of benefactor abundance and sediment deposition levels. However, the switch back to negative effects at the upper end of the gradient of consumer pressure occurred only in three of the six combinations and was the most common outcome when the abundance of the benefactor was large. It is, therefore, at the extreme ends of gradients of stress that density-dependent processes are more likely to play a key role in determining the nature of species interactions.

In conclusion, our study has shown that switches in the sign of species interactions can be regulated by interactive effects of physical stress and consumer pressure. The contrasting support received by the SGH in terrestrial systems (Gómez-Aparicio et al. 2004; Maestre, Valladares & Reynolds 2005; Lortie & Callaway 2006) could be, at least in part, due to overlooking physical or biological forces other than those generating the single gradient of stress under investigation. Our study demonstrated that the interplay of two gradients differing in nature (i.e. physical vs. biotic) drives switches in species interaction, but interactions of multiple gradients can be common in nature (Kawai & Tokeshi 2007). For instance, we suspect that repeating our experiment along coasts varying in the exposure to wave action would yield different results.

Much of the progress in our understanding of the role of facilitative interactions in regulating the structure of natural communities made in the last decade has been undeniably stimulated by the formulation of the SGH (Bertness & Callaway 1994). More recent work has identified key attributes of gradients of stress (i.e. nature of and interaction between different types of stress; Crain 2008; Bulleri 2009; Maestre et al. 2009; Stachowicz 2001; Kawai & Tokeshi 2007; Smit, Rietkerk & Wassen 2009; this study) and of interacting species (i.e. trophic level, population and species-specific traits; Bulleri 2009; Daleo & Iribarne 2009; Irving & Bertness 2009; Maestre et al. 2009; this study), paving the way towards the formulation of a unified SGH. The inclusion of multiple gradients of stress into empirical tests of the SGH will be, however, the key in enabling predictions of changes in species interactions triggered by the ongoing alteration of trophic webs (Byrnes, Reynolds & Stachowicz 2007) under different scenarios of climate change, as well as in streamlining the use of positive species interactions in conservation and restoration actions (Callaway 2007; Halpern et al. 2007).

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Comments and constructive criticism by E. Maggi, C. Samsa and two anonymous referees greatly improved this study and are sincerely acknowledged. This work was supported by the University of Pisa through the project Bioclima and by the MIUR through the project BIORES. Many thanks to G. Fissore (Assonautica Livorno) for assisting with boating operations.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • Airoldi, L. (1998) Roles of disturbance, sediment stress, and substratum retention on spatial dominance in algal turf. Ecology, 79, 27592770.
  • Airoldi, L. (2003) The effects of sedimentation on rocky coast assemblages. Oceanography and Marine Biology: An Annual Review, 41, 161236.
  • Altieri, A.H., Silliman, B.R. & Bertness, M.D. (2007) Hierarchical organization via a facilitation cascade in intertidal cordgrass bed communities. American Naturalist, 169, 195206.
  • Arafa, S., Sadok, S. & El Abed, A. (2006) Variation in nitrogenous compounds and gonad index in fed and starved sea urchins (Paracentrotus lividus) during live storage. Aquaculture, 257, 525533.
  • Armas, C., Ordiales, R. & Pugnaire, F. (2004) Measuring plant interaction: a new comparative index. Ecology, 85, 26822686.
  • Benedetti-Cecchi, L., Bulleri, F. & Cinelli, F. (1998) Density dependent foraging in two species of sea urchins in shallow subtidal reefs on the west coast of Italy (Western Mediterranean). Marine Ecology Progress Series, 163, 203211.
  • Benedetti-Cecchi, L., Bulleri, F. & Cinelli, F. (2000) The interplay of physical and biological factors in maintaining mid-shore and low-shore assemblages on rocky coasts in the north-west Mediterranean. Oecologia, 123, 406417.
  • Bertness, M. & Callaway, R.M. (1994) Positive interactions in communities. Trends in Ecology and Evolution, 9, 191193.
  • Brooker, R.W., Scott, D., Palmer, S.C.F. & Swaine, E. (2006) Transient facilitative effects of heather on Scots pine along a grazing disturbance gradient in Scottishmoorland. Journal of Ecology, 94, 637645.
  • Brooker, R.W., Maestre, F.T., Callaway, R.M., Lortie, C.L., Cavieres, L.A., Kunstler, G. et al. (2008) Facilitation in plant communities: the past, the present, and the future. Journal of Ecology, 96, 1834.
  • Bruno, J.F. & Bertness, M.D. (2001) Habitat modification and facilitation in benthic marine communities. Marine Community Ecology (eds M.D.Bertness, S.D.Gaines & M.E.Hay), pp. 201218. Sinauer, Sunderland, MA.
  • Bulleri, F. (2009) Facilitation research in marine systems: state of the art, emerging patterns and insights for future developments. Journal of Ecology, 97, 11211130.
  • Bulleri, F. & Benedetti-Cecchi, L. (2008) Facilitation of the introduced green alga, Caulerpa racemosa, by resident algal turfs: experimental evaluation of underlying mechanisms. Marine Ecology Progress Series, 364, 7786.
  • Bulleri, F., Benedetti-Cecchi, L. & Cinelli, F. (1999) Grazing by the sea urchins Arbacia lixula L. and Paracentrotus lividus Lam. in the Northwest Mediterranean. Journal of Experimental Marine Biology and Ecology, 241, 8195.
  • Bulleri, F., Bertocci, I. & Micheli, F. (2002) The interplay of coralline algae and sea urchins in maintaining alternative habitats. Marine Ecology Progress Series, 243, 101109.
  • Bulleri, F., Tamburello, L. & Benedetti-Cecchi, L. (2009) Loss of consumers alters the effects of resident assemblages on the local spread of an introduced macroalga. Oikos, 118, 269279.
  • Byrnes, J.E., Reynolds, P.L. & Stachowicz, J.J. (2007) Invasions and extinctions reshape coastal marine food webs. PLoS ONE, 2(3), e295.
  • Callaway, R.M. (2007) Positive Interactions and Interdependence in Plant Communities. Springer, Dordrecht, The Netherlands.
  • Callaway, R.M. & Walker, L.R. (1997) Competition and facilitation: a synthetic approach to interactions in plant communities. Ecology, 78, 19581965.
  • Callaway, R.M., Brooker, R., Choler, P., Kikvidze, Z., Lortie, C.J., Michalet, R. et al. (2002) Positive interactions among alpine plants increase with stress. Nature, 417, 844847.
  • Crain, C.M. (2008) Interactions between marsh plant species vary in direction and strength depending on environmental and consumer context. Journal of Ecology, 96, 166173.
  • Daleo, P. & Iribarne, O. (2009) Beyond competition: the stress gradient hypothesis tested in plant-herbivore interactions. Ecology, 90, 23682374.
  • Drake, J.A. (1991) Community-assembly mechanics and the structure of an experimental species ensemble. American Naturalist, 137, 126.
  • Eskelinen, A. (2008) Herbivore and neighbour effects on tundra plants depend on species identity, nutrient availability and local environmental conditions. Journal of Ecology, 96, 155165.
  • Gianguzza, P., Bonaviri, C., Milisenda, G., Barcellona, A., Agnetta, D., Vega Fernández, T. et al. (2010) Macroalgal assemblage type affects predation pressure on sea urchins by altering adhesion strength. Marine Environmental Research, 70, 8286.
  • Gómez-Aparicio, L., Zamora, R., Gómez, J.M., Hódar, J.A., Castro, J. & Baraza, E. (2004) Applying plant positive interactions to reforestation in Mediterranean mountains: a meta-analysis of the use of shrubs as nurse plants. Ecological Applications, 14, 11281138.
  • Grime, J.P. (1977) Evidence for existence of 3 primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist, 111, 11691194.
  • Halpern, B.S., Silliman, B.R., Olden, J.D., Bruno, J.P. & Bertness, M.D. (2007) Incorporating positive interactions in aquatic restoration and conservation. Frontiers in Ecology and the Environment, 5, 153160.
  • Harper, J.L. & White, J. (1974) The demography of plants. Annual Review of Ecology and Systematics, 5, 419463.
  • Hay, M.E. (1986) Associational plant defenses and the maintenance of species diversity: turning competitors into accomplices. American Naturalist, 128, 617641.
  • Irving, A.D. & Bertness, M.D. (2009) Trait-dependent modification of facilitation on cobble beaches. Ecology, 90, 30423050.
  • Kawai, T. & Tokeshi, M. (2007) Testing the facilitation–competition paradigm under the stress-gradient hypothesis: decoupling multiple stress factors. Proceedings of the Royal Society of London, Series B: Biological Sciences, 274, 25032508.
  • Kikvidze, Z., Armas, C. & Pugnaire, F.I. (2006) The effect of initial biomass in manipulative experiments on plants. Functional Ecology, 20, 13.
  • Levenbach, S. (2008) Behavioral mechanism for an associational refuge for macroalgae on temperate reefs. Marine Ecology Progress Series, 370, 4552.
  • Levenbach, S. (2009) Grazing intensity influences the strength of an associational refuge on temperate reefs. Oecologia, 159, 181190.
  • Lortie, C.J. & Callaway, R.M. (2006) Re-analysis of meta-analysis: support for the stress-gradient hypothesis. Journal of Ecology, 94, 716.
  • Maestre, F.T. & Cortina, J. (2004) Do positive interactions increase with abiotic stress? – a test from a semi-arid steppe. Proceedings of the Royal Society of London, Series B: Biological Sciences, 271, S331S333.
  • Maestre, F.T., Valladares, F. & Reynolds, J.F. (2005) Is the change of plant–plant interactions with abiotic stress predictable? A meta-analysis of field results in arid environments. Journal of Ecology, 93, 748757.
  • Maestre, F.T., Callaway, R.M., Valladares, F. & Lortie, C.J. (2009) Refining the stress-gradient hypothesis for competition and facilitation in plant communities. Journal of Ecology, 97, 199205.
  • Menge, B.A. & Sutherland, J.P. (1987) Community regulation: variation in disturbance, competition, and predation in relation to gradients of environmental stress and recruitment. American Naturalist, 130, 730757.
  • Michalet, R., Brooker, R.W., Cavieres, L.A., Kikvidze, Z., Lortie, C.J., Pugnaire, F.I. et al. (2006) Do biotic interactions shape both sides of the humped-back model of species richness in plant communities? Ecology Letters, 9, 767773.
  • Miriti, M.N. (2006) Ontogenetic shift from facilitation to competition in a desert shrub. Journal of Ecology, 94, 973979.
  • le Roux, P.C. & McGeoch, M.A. (2010) Interaction intensity and importance along two stress gradients: adding shape to the stress-gradient hypothesis. Oecologia, 162, 733745.
  • Schiaparelli, S. & Cattaneo-Vietti, R. (1999) Functional morphology of vermetid feeding-tubes. Lethaia, 32, 4146.
  • Smit, C., Rietkerk, M. & Wassen, M.J. (2009) Inclusion of biotic stress (consumer pressure) alters predictions from the stress gradient hypothesis. Journal of Ecology, 97, 12151219.
  • Smit, C., Vandenberghe, C., Den Ouden, J. & Mueller-Schaerer, H. (2007) Nurse plants, tree saplings and grazing pressure: changes in facilitation along a biotic environmental gradient. Oecologia, 152, 265273.
  • Stachowicz, J.J. (2001) Mutualism, facilitation, and the structure of ecological communities. BioScience, 51, 235246.
  • Taylor, R.B. & Rees, T.A.V. (1998) Excretory products of mobile epifauna as a nitrogen source for seaweeds. Limnology and Oceanography, 43, 600606.
  • Tielbörger, K. & Kadmon, R. (2000) Temporal environmental variation tips the balance between facilitation and interference in desert plants. Ecology, 81, 15441553.
  • Travis, J.M.J., Brooker, R.W., Clark, E.J. & Dytham, C. (2006) The distribution of positive and negative species interactions across environmental gradients on a dual-lattice model. Journal of Theoretical Biology, 241, 896902.
  • Underwood, A.J. (1997) Experiments in Ecology: Their Local Design and Interpretation Using Analysis of Variance. Cambridge University Press, Cambridge.
  • Williams, S.L. & Carpenter, R.C. (1988) Nitrogen-limited primary productivity of coral reef algal turfs: potential contribution of ammonium excreted by Diadema antillarum. Marine Ecology Progress Series, 47, 145152.

Supporting Information

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  9. Supporting Information

Figure S1. Mean density of sea urchins (+SE) in experimental areas.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials may be re-organized for online delivery, but are not copy-edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

FilenameFormatSizeDescription
JEC_1750_sm_FigS1.doc86KSupporting info item

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.