Fiddler crab–vegetation interactions in hypersaline habitats

Fiddler crab–vegetation interactions in hypersaline habitats

Journal of Experimental Marine Biology and Ecology, 225 (1998) 53–68 L Fiddler crab–vegetation interactions in hypersaline habitats Benjamin E. Noma...

518KB Sizes 0 Downloads 6 Views

Journal of Experimental Marine Biology and Ecology, 225 (1998) 53–68

L

Fiddler crab–vegetation interactions in hypersaline habitats Benjamin E. Nomann, Steven C. Pennings* University of Georgia Marine Institute, Sapelo Island, GA 31327, USA Received 9 April 1997; received in revised form 20 July 1997; accepted 8 August 1997

Abstract Abiotic conditions often change ecological interactions. Studies in areas with low to moderate soil salinities have demonstrated a facultative mutualism between fiddler crabs and salt marsh vegetation. In these habitats, fiddler crab burrowing increases plant growth, and plant roots help support the walls of crab burrows. We looked for these interactions in hypersaline soils bordering unvegetated salt pans in a Georgia salt marsh. Crab burrows and vegetation cover were positively associated. Neither crab removals nor burrow additions demonstrated a positive effect of crabs on vegetation. However, both vegetation removals and the addition of an artificial canopy (suspended shadecloth) demonstrated a strong positive effect of vegetation on crab burrows. In contrast to previous studies, we found no evidence that plants supported burrow walls. Instead, crabs likely associate with vegetation to avoid predators. Our results caution against extrapolating experimental results between habitats with different abiotic conditions.  1998 Elsevier Science B.V. Keywords: Abiotic stress; Facilitation; Fiddler crab; Mutualism; Salt marsh; Uca

1. Introduction Ecological interactions may change as a function of abiotic conditions (Dunson and Travis, 1991; Travis, 1996). Tansley (1917) provided one of the first examples of this dependence: competitive dominance hierarchies of plant species were reversed depending upon whether the soil was acidic or basic. Abiotic conditions have since been found to mediate the outcome of ecological interactions as varied as competition (Moloney, 1990), plant parasitism (Pennings and Callaway, 1996), and a plant–fungal mutualism (Kelrick and Nomann, unpublished data). Facilitative interactions may be common and important in structuring communities

*Corresponding author. Tel.: 1 1 912 4852293; fax: 1 1 912 4852133; e-mail: [email protected] 0022-0981 / 98 / $19.00  1998 Elsevier Science B.V. All rights reserved. PII S0022-0981( 97 )00209-8

54

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

(Bertness and Callaway, 1994; Callaway, 1995), however, species may shift between competitive and facilitative interactions depending upon abiotic conditions. For example, Callaway and King (1996) demonstrated that aerenchymous plants may either facilitate or compete with neighbors depending on temperature. In New England salt marshes, either competition or facilitation may dominate plant interactions depending on soil salinities (Bertness and Shumway, 1993; Bertness and Hacker, 1994). Fiddler crabs (Uca spp.) are ubiquitous in salt marshes along the western Atlantic coast (Teal, 1958). Facilitative interactions between crabs and low to middle marsh vegetation (primarily Spartina alterniflora) have been described by Bertness (1985) and Montague (1980; 1982). Crab burrowing increases plant production through moderating soil conditions (e.g. increasing soil aeration, oxidation-reduction potential, and in situ decomposition of belowground plant debris). In soft sediments, plants appear to facilitate crab burrowing by stabilizing the substratum, but in areas where vegetation forms dense root mats, crab burrowing is effectively prevented. Upper zones of southern salt marshes can become hypersaline, sometimes to the extent that vegetation cannot exist, and unvegetated salt pans form (Wiegert and Freeman, 1990). Fiddler crab–vegetation interactions have not been previously examined in this habitat, but may differ from those described in low- to moderate-salinity habitats. In hypersaline soils, plant productivity might be affected by salinity more than by aeration or decomposition, and sediments might be stable enough that support of burrows from plant roots was unimportant. We conducted a series of experiments to determine if crabs and vegetation facilitated each other in hypersaline marsh habitats, and if so, whether the mechanisms were identical to those previously described in lower-salinity marsh habitats.

2. Methods

2.1. Study site Research was conducted in three marshes on Sapelo Island, Georgia, USA (Fig. 1A,B). At each site we worked in a high marsh zone characterized by salt pans and extremely salt-tolerant vegetation. Located above the Spartina alterniflora Loisel zone and below the Juncus roemerianus Scheele / terrestrial zone (Fig. 1C), the salt pans were unvegetated flats in a patchy band parallel to the shore | 1–20 m wide. The vegetation fringing and forming patches within salt pans was dominated by Batis maritima L., Distichlis spicata (L.) Greene, Salicornia bigelovii L., and S. virginica L (taxonomy follows Radford et al., 1968). Five to 10% of high tides inundated this zone. Tidal water did not pond in the salt flats; however, soil pore-water steadily evaporated during long (15–25-day) intervals between flooding, resulting in extremely high ( . 100 ppt) concentrations of salts within the soil (Antlfinger and Dunn, 1979; Wiegert and Freeman, 1990). The mud and sand fiddler crabs, Uca pugnax (Smith 1870) and U. pugilator (Bosc 1801) respectively, comprised | 95% of the crab individuals present in the high salt marsh, with the remaining | 5% consisting primarily of Armases and Sesarma spp. (Teal, 1958; pers. obs.).

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

55

Fig. 1. (A) location of Sapelo Island, GA. (B) Location of study sites. (C) Typical marsh zonation pattern.

2.2. Crab distribution patterns To characterize crab distribution patterns in relation to vegetation cover and soil salinity, a 0.25-m 2 quadrat, divided into 100 5 3 5-cm cells, was used to sample the Marsh Landing salt pan and fringing vegetation on 19 August 1995. In each of three ‘zones’, eight replicates were haphazardly located. The vegetation bordering the upper edge of the salt pan was sampled 0.5–1.0 m into vegetation (border) and 2.0–3.0 m into vegetation (vegetation), and the salt pan was sampled 2.0–4.0 m away from the salt pan–vegetation interface. Percent cover of above-ground vegetation and feeding pellets was estimated as the amount needed to fully cover 5 3 5-cm cells (one cell 5 1%), and the number of burrows present was recorded. One core of sediment, 15 cm in depth and 1.85 cm in diameter, was collected in each quadrat to determine interstitial salinity. Cores were taken . 5 cm from the nearest crab burrow. Soils were usually too dry to obtain pore water directly. Instead, cores were weighed while wet, dried for 3 days at 608C, and weighed again to measure initial water content. A known amount of distilled water was added, and containers sealed to prevent evaporation. Samples were vigorously stirred after 1 day, and the salinity of the supernatant measured with a refractometer after 2 days. Original pore-water salinities were calculated based on the initial water content of the soil, the volume of distilled water added, and the final salinity reading. In order to explore temperature patterns that might influence fiddler crab behavior, soil

56

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

and air temperatures were measured at Marsh Landing between 1 and 2 p.m. on 23 September 1996. Temperatures were measured 2.5 m into the salt pan, and 0.5 and 2.5 m into the vegetation. Independent measurements were taken | 7 mm below (n 5 15 / zone) and | 10 mm above the sediment surface (n 5 15 / zone). These two heights were chosen to represent the burrow entrance and the approximate location of the body of a foraging crab. Temperatures were only measured at one time on one day of the year but would of course vary diurnally and seasonally. Moreover, actual temperatures of crabs would be influenced by additional factors such as body orientation and color; therefore, these readings represent only a preliminary estimate of the crabs’ thermal regime.

2.3. Manipulating crab density and plant cover To determine whether plants facilitated crabs and / or vice versa, we conducted an experiment where crab density and vegetation cover were independently manipulated. To assess the influence of burrowing crabs on above-ground productivity of plants, crabs were removed from the vegetation fringing salt pans. Exclosures (1.0 3 2.0 3 0.3 m (W 3 L 3 H )) were constructed of 0.65-cm (mesh size) Vexar plastic fencing attached to wooden posts driven into the marsh. The fencing extended | 20 cm below the soil surface to prevent burrowing crabs from entering plots. A 10-cm wide strip of 4-mil clear plastic sheeting was attached to the upper edge of the exclosure to inhibit immigration of climbing crabs. Because the exclosures cut rhizomatic connections of plants, 1 3 2-m control plots were also cut along their perimeters to a depth of | 20 cm. Observations indicated that crabs could not crawl over the fence but that very small crabs could pass through the mesh. Crabs were removed from exclosures by hand and pitfall trap. Hand-collection occurred only when a crab was visually observed so as not to disturb the soil or vegetation. A single pitfall trap consisting of a | 700-ml glass jar was placed inside each exclosure, with the jar’s mouth flush with the soil surface. Jars were emptied of crabs approximately every 2 weeks, and the total number of crabs captured by both methods was recorded. To examine the influence of plants on crab burrowing and feeding, above-ground vegetation was removed from additional 1 3 2-m plots. Vegetation was cut to within | 1 cm of the soil surface and reclipped as necessary to prevent regrowth. Crab removal, vegetation removal and control treatments were grouped in eight blocks, each with one replicate of each treatment. Six blocks were located at Marsh Landing and two at Cabretta. Within each block, plots were selected to maximize initial vegetation similarities, were located near each other ( | 0.5–2-m gaps between plots), and were randomly assigned to treatments. Because plots were so large, they encompassed parts of both the ‘border’ and ‘vegetation’ zones described in Section 2.2. Vegetation was removed on 17–22 June 1995. Crabs were intensively removed from the exclosures by hand and pitfall from 8 July 1995 to 7 August 1995; pitfall trapping continued until 28 August 1996. Plant and pellet cover and burrow data were recorded on 23–29 August 1995 and on 25 June 1996 using two 0.25-m 2 quadrats per 1 3 2-m plot. On the latter date, burrows were classified into size categories: small (0.1–0.5 cm), medium (0.6–0.9 cm), and large

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

57

( . 0.9 cm). The two quadrats were placed to provide for a | 20-cm buffer between the sampled areas and pitfall traps and / or plot edges. On 23 and 28 August 1996 the vegetation from a 0.25 3 1.0-m area within each 1 3 2-m plot was harvested to within 1 cm of the marsh surface, sorted according to species, dried at 608C, and weighed to the nearest 0.01 g. Two soil salinity cores, 2–3 cm in depth, were taken from each plant removal and control plot on 18 September 1995.

2.4. Manipulating light intensity To simulate light and cover conditions of vegetated areas in the absence of plant stems and roots, the salt pan surface was shaded. A single shaded and a single control plot were placed in the salt pan within 2–3 m of each block of crab removal, vegetation removal and control plots described above. After selecting plot locations, treatments were randomly assigned to plots. Shade cloths that reduced light by | 50% were supported 15 cm over the 90 3 90-cm plots by wooden stakes, starting 17 June 1995. Burrow numbers and feeding pellet cover within a 0.125-m 2 quadrat centered within each plot were recorded on 18 September 1995 and on 25 June 1996. Soil salinity cores (2–3 cm in depth) were taken on 18 September 1995, and 20 August 1996.

2.5. Additional vegetation manipulations Two other experiments were conducted in which we manipulated vegetation and / or shading to explore the relationship between vegetation and crab burrowing and feeding In one experiment, we removed vegetation and / or shaded the substrate as described above in 0.5 3 0.5-m plots (n 5 12 / treatment / zone) in three zones at Lighthouse marsh: the salt pan, an area immediately above the salt pan dominated by Batis maritima, and an area immediately below the salt pan dominated by Salicornia virginica. In the salt pan we located control and shaded plots. In the two vegetated zones we located control, vegetation removal, and vegetation removal 1 shaded plots. Treatments were initiated on 15 August 1994, and sampled with a 0.125-m 2 quadrat on 20 September 1995. In another experiment, we reciprocally removed two major components of the vegetation fringing salt pans: dominant plants (i.e., Spartina alterniflora and Borrichia frutescens) which dominate large zones of the marsh where salinities are moderate, and salt tolerators (Batis maritima, Distichlis spicata, Salicornia virginica). Plots (0.5 3 0.5 m, n 5 15 / treatment) were located in groups of three in vegetation mixtures near salt pans in the Marsh Landing and Lighthouse marshes, and randomly assigned to two removal and one control treatments. In these mixtures the ‘dominant plants’ had scattered robust stems that did not form a dense canopy, whereas the ‘salt tolerators’ had smaller stems but formed a dense canopy over the soil surface. Thus, removing dominants affected structure more than shade, and removing salt tolerators affected shade more than structure. Reciprocal removals were initiated on 30 March 1994 and maintained by periodic weeding. Burrows were counted on 4 September 1995. Results from both marshes were similar and were pooled to increase statistical power.

58

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

2.6. Burrow effects on individual Salicornia bigelovii plants To determine the effects of crab burrows on individual plants, we placed artificial burrows near individual Salicornia bigelovii plants. The experiment was conducted in Marsh Landing in S. bigelovii stands fringing a salt pan. Paired experimental and control plants were greater than 5.5 cm from the nearest neighboring plant, were within 0.5 m of each other, and were similar in size. Using a portable electric drill, a single | 1.2-cm diameter hole was drilled to a depth of 10 cm at a distance of 2 cm from the base of each experimental plant. A large diameter hole, similar in size to large burrows, was chosen to obtain a maximum effect. Holes were drilled on 14 June 1995. Plants were checked 1–3 times / week, deaths noted, and burrows redrilled as necessary until 5 September 1995, when the surviving individuals were harvested, dried at 608C, and weighed to the nearest 0.01 g. We examined the effect of adding a burrow on survival and on final mass. Only cases in which both plants of a pair survived (n 5 4) were used in the analysis of final mass.

2.7. Statistical analyses Proportions were arcsine-squareroot transformed. Data sets in which the variance increased with the mean were ln transformed. Data were analyzed with t-tests or ANOVA with Tukey means tests using Statistix (Analytical software, Tallahassee FL).

3. Results

3.1. Crab distribution patterns Plant cover differed strongly between the three zones: plants were absent in the salt pan and more than twice as abundant in the vegetation zone than in the border zone (Fig. 2A). Soil salinity was approximately two times greater in the salt pan than in the vegetation or border zones (Fig. 2B). Burrows were absent in the salt pan quadrats, present in the border zone, and common in the vegetation zone (Fig. 2C). Feeding pellets were about twice as abundant in the vegetation zone as in the salt pan or border zones (Fig. 2D). Temperature patterns on 23 September 1996 differed depending on whether readings were taken above or below ground. Although the differences were relatively minor, below-ground temperatures were highest in the salt pan, whereas above-ground temperatures were highest in the vegetation zone (Fig. 2E, ANOVA, interaction term, P , 0.0001).

3.2. Manipulating crab density and plant cover During the first 38 days of the crab removal experiment, we removed 8463 crabs, including most of the large individuals, from each exclosure. Thereafter, we no longer counted the crabs that we removed but noticed that we were observing and removing only small and medium-sized crabs. Examination of burrow densities indicated that we

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

59

Fig. 2. Plant cover, physical factors, and crab activity measures in three marsh zones (n 5 8 / zone except for temperatures, n 5 15 / location (air or soil) / zone). Soil temperatures were measured 7 mm below the soil surface, to represent the burrow enterance; air temperatures were measured 10 mm above the soil surface, to represent the body location of a foraging crab.

reduced the number of large burrows by . 75% but had no effect on the number of small or medium burrows (Fig. 3). Small crabs were able to pass through the mesh and probably grew into the medium-size class quickly enough that we could not reduce their numbers. Vegetation cover and dry mass (all species were pooled) did not differ significantly between crab removal and control treatments (Fig. 4A,B). Vegetation removal reduced crab burrow numbers by approximately 50% (Fig. 5A). Cover of feeding pellets was similar between treatment and control replicates in 1995,

60

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

Fig. 3. Burrow numbers in 1 3 2-m control and crab removal plots at Marsh Landing and Cabretta (n 5 8 / treatment). Error bars are61 SE.

but removal plots had one-third fewer pellets in 1996 (Fig. 5B). We observed no difference between treatments in soil pore-water salinity (Fig. 5C).

3.3. Manipulating light intensity Shading salt pans increased crab burrowing by | 5 times (Fig. 6A). The total number of burrows present differed strongly between sampling dates, perhaps as a function of temporal variability in soil salinity (Fig. 6C). Shaded plots had | 3 times the feeding pellet cover of controls in 1995, but no differences were observed in 1996 (Fig. 6B).

Fig. 4. Final (A) percent plant cover and (B) total above-ground dry plant biomass in 1 3 2-m control and crab removal plots at Marsh Landing and Cabretta (n 5 8 / treatment). Error bars are61 SE.

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

61

Fig. 5. Burrow numbers, feeding pellet cover, and soil salinity during fall 1995 and summer 1996 in 1 3 2-m control and vegetation removal plots at Marsh Landing and Cabretta (n 5 8 / treatment). Error bars are61 SE.

Soil salinity did not significantly differ between treatments in 1995, but shaded plots were | 35% less saline than controls in 1996 (Fig. 6C).

3.4. Additional vegetation manipulations Shading the Lighthouse salt pan increased burrow numbers 3–4 times (Fig. 7A). Shade plots had 30–55% greater feeding pellet cover than control plots (Fig. 7B), and soil salinities were 20–50% lower in shaded than control plots (Fig. 7C). In vegetated zones, vegetation removal 1 shade plots consistently had more burrows than vegetation removal plots, with control plots intermediate (Fig. 8A). Feeding pellets were most abundant in vegetation removal plots, and least abundant in control plots (Fig. 8B).

62

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

Fig. 6. Burrow numbers, feeding pellet cover and soil salinity in 90 3 90-cm shade and control plots at Marsh Landing and Cabretta (n 5 8 / treatment). Error bars are61 SE.

Salinity was 15–35% lower in vegetation removal 1 shade plots than in the other treatments (Fig. 8C). Removing the more robust ‘dominant’ salt marsh species from vegetation mixtures had little effect on burrow numbers, but removing the more abundant ‘salt tolerators’ significantly reduced burrow numbers (Fig. 9).

3.5. Burrow effects on individual Salicornia bigelovii plants Adding artificial burrows near individual S. bigelovii plants had no effect on survival rates (Fig. 10A) but significantly reduced final mass of survivors (Fig. 10B).

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

63

Fig. 7. Burrow numbers, feeding pellet cover, and salinities in 0.5 3 0.5-m shade and control plots at Lighthouse (n 5 12 / treatment). Error bars are61 SE.

4. Discussion In contrast to Montague (1982) and Bertness (1985), our study found no evidence that crab burrowing facilitated plant growth. In fact, our burrow additions appeared to reduce plant mass, although the sample size for this test was small. In previous studies, the positive impact of crabs on marsh plants was likely mediated through burrow effects on soil oxygenation, decomposition rates, and nutrient concentrations (Montague, 1982; Bertness, 1985). We suggest that these impacts are largely irrelevant to vegetation surrounding salt pans, where extremely high soil water salinities (Antlfinger and Dunn, 1979; Wiegert and Freeman, 1990; this paper) likely exert the primary control on vegetative production. There remains the possibility that we failed to remove enough crabs to see an effect; however, burrows of large crabs (which probably have the greatest impact on sediment chemistry because of their disproportionate size) were reduced by

64

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

Fig. 8. Burrow numbers, feeding pellet cover, and soil salinities taken in 0.5 3 0.5-m control, vegetation removal 1 shade, and vegetation removal plots at Lighthouse (n 5 12 / treatment). Error bars are61 SE.

over 75% in exclosures relative to controls late in the experiment, and probably more so during earlier periods of intensive crab removal. This removal rate compares favorably with the 30–50% reduction achieved by Bertness (1985), who found large effects of crabs on vegetation. Superficially, our result that plant cover facilitated crab burrowing appears similar to that of other studies (Bertness and Miller, 1984; Bertness, 1985), but the mechanism was probably different. Previous studies focused on the importance of vegetation in providing structure for burrows, but vegetation might also affect crabs through several additional mechanisms including altering salinity and temperature through shading, modifying intraspecific territorial interactions by limiting display distance, or supplying

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

65

Fig. 9. Burrow numbers in 0.5 3 0.5-m dominant species removal, salt tolerator removal and control plots at Lighthouse (n 5 30 / treatment). Error bars are61 SE.

cover to hide from predators. Although we did not experimentally address all of these potential mechanisms, our results and observations suggest that vegetation surrounding salt pans acts primarily as refugia for Uca from predators. We found no evidence that Uca uses vegetation in the salt pan region to provide burrow support. Burrows in the salt pan region were not located immediately along plant stems (pers. obs.). We found greater burrow numbers in shaded versus control salt pan plots despite a lack of structure that could support burrows in either plot type. Moreover, clipping above-ground vegetation resulted in a rapid drop in burrow number, despite the

Fig. 10. Proportion of Salicornia bigelovii surviving (initial n 5 40 / treatment) and above-ground dry weight of surviving S. bigelovii control–experimental pairs (n 5 4 pairs survived) in the burrow-addition experiment at Marsh Landing. Error bars are61 SE.

66

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

fact that underground stems and large roots which could have provided burrow support remained intact for months (pers. obs.). Our initial sampling suggested that burrow numbers were correlated with soil water salinity; however, our experiments altered burrow numbers without consistently affecting soil water salinity. In general, shading by vegetation or shadecloths would be expected to reduce evaporation and moderate soil salinities (Bertness, 1991; Bertness et al., 1992; Bertness and Hacker, 1994). These effects do occur at our sites (Pennings, unpubl. data) but were not always obvious in these experiments, perhaps because many of our shade and clearing plots were quite small (0.5 3 0.5 m) and / or because the vegetation was often quite scrubby and did not heavily shade the substratum. It is possible that vegetation and shading might alter burrow numbers by mediating the temperature regime. Air temperatures were lower in the salt pan than in vegetation, probably because vegetation reduced airflow and absorbed radiation; however, air humidity was probably lower and desiccation stress higher in the salt pan. Soil surface temperatures were slightly lower in vegetation than in the salt pan, probably due to vegetation shading the soil. Temperatures were measured on only one date and were not measured at depth; however, it is possible that differences in burrow temperature might constrain fiddler crabs to vegetated areas. Examining this hypothesis would require extensive monitoring of temperature and humidity regimes experienced by actual crabs. Some workers have suggested that Uca densities are limited by territorial behavior (Zucker, 1981; Mueller, 1983). If so, removing vegetation might have facilitated intraspecific interactions, resulting in enlarged territories and lower burrow densities. However, this hypothesis would predict no effect of shade cloth on burrow numbers. Since we observed a large effect of shade cloth on burrow numbers, we believe that territoriality was not the primary mechanism controlling burrow density in our experiments. Vegetative structure has repeatedly been shown to protect prey from predators in a variety of systems (Stoner, 1979; Coen et al., 1981; Savino and Stein, 1982, 1989; Werner et al., 1983; Anderson, 1984; Ryer, 1988; Pennings, 1990). Fiddler crabs often feed in salt pans—as evidenced by the presence of feeding pellets in unshaded salt pan plots in our study—but flee when approached, suggesting avoidance of predators. Predation pressure on fiddler crabs is probably intense. Raccoons frequent these marshes and their faeces regularly contain the remains of Uca spp. (pers. obs.). Channel bass, Sciaenops ocellata, blue crabs, Callinectes sapidus, and mud crabs, Eurytium limosum, also prey on fiddler crabs (Shanholtzer (1973) cited in Montague, 1980; Kneib and Weeks, 1990). Further, certain birds feed heavily on fiddler crabs (Petit and Bildstein, 1987; Watts, 1988; Frix et al., 1991; Ens et al., 1993), and focus their activities in salt pans, possibly due to the ease of walking and attacking in unvegetated habitats. During 8 days of observations over 2 years totaling 336 foraging white ibis, Eudocimus albus (L.), we noted that . 85% of the birds were feeding in salt pans or in low-lying, sparse vegetation. Birds almost never fed in tall, dense vegetation when unvegetated areas were available, even though tall, dense vegetation comprised the majority of the marsh area. Similarly, we have consistently observed a variety of gull species feeding in salt pans, but not in dense vegetation. These observations suggest that a primary reason crabs burrow in vegetated areas of Sapelo Island marshes is to reduce their risk of predation.

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

67

The nature of ecological interactions can be strongly affected by abiotic factors (Dunson and Travis, 1991; Stephens and Bertness, 1991; Callaway and King, 1996; Travis, 1996). Our results suggest that crab–vegetation interactions in salt marshes are not the same in all habitat types. Working in higher, more hypersaline habitats than previous workers, we found crab facilitation of vegetation to be absent, and vegetative facilitation of crabs to be strong but driven by different mechanisms than suggested by previous workers. Our findings caution against the over-extrapolation of results from single experiments to other situations where abiotic factors may differ.

Acknowledgements We thank M. Grant, T. Page, C. Pullen, and C. Richards for field assistance, C. Richards for use of her Lighthouse plots, and D. Casey for Fig. 1. Ten percent ($1000) of research was funded by the US Department of energy’s (DOE) National Institute for Global Environmental Change (NIGEC) through the NIGEC Western Regional Center at the University of California, Davis (DOE Cooperative Agreement No. DE-FCO390ER61010). Financial support does not constitute an endorsement by DOE of the views expressed in this article. Support for Benjamin Nomann in the summer of 1995 was provided by the Sapelo Foundation through the University of Georgia Marine Institute Student Intern Program. This is contribution number 800 from the University of Georgia Marine Institute.

References Anderson, O., 1984. Optimal foraging by largemouth bass in structured environments. Ecology 65, 851–861. Antlfinger, A.E., Dunn, E.L., 1979. Seasonal patterns of CO 2 and water vapor exchange of three salt marsh succulents. Oecologia 43, 249–260. Bertness, M.D., 1985. Fiddler crab regulation of Spartina alterniflora production on a New England salt marsh. Ecology 66, 1042–1055. Bertness, M.D., 1991. Interspecific interactions among high marsh perennials in a New England salt marsh. Ecology 72, 125–137. Bertness, M.D., Callaway, R., 1994. Positive interactions in communities. Trends Ecol. Evolution 9, 191–193. Bertness, M.D., Hacker, S.D., 1994. Physical stress and positive associations among marsh plants. The American Naturalist 144, 363–372. Bertness, M.D., Miller, T., 1984. The distribution dynamics of Uca pugnax (Smith) burrows in a New England salt marsh. J. Exp. Mar. Biol. Ecol. 83, 211–237. Bertness, M.D., Shumway, S.W., 1993. Competition and facilitation in marsh plants. Am. Naturalist 142, 718–724. Bertness, M.D., Gough, L., Shumway, S.W., 1992. Salt tolerances and the distribution of fugitive salt marsh plants. Ecology 73, 1842–1851. Callaway, R.M., 1995. Positive interactions among plants. Bot. Rev. 61, 306–349. Callaway, R.M., King, L., 1996. Temperature-driven variation in substrate oxygenation and the balance of competition and facilitation. Ecology 77, 1189–1195. Coen, K.L., Heck, Jr. K.L., Abele, L.G., 1981. Experiments on competition and predation among shrimps of seagrass meadows. Ecology 62, 1484–1493.

68

B.E. Nomann, S.C. Pennings / J. Exp. Mar. Biol. Ecol. 225 (1998) 53 – 68

Dunson, W.A., Travis, J., 1991. The role of abiotic factors in community organization. Am. Naturalist 138, 1067–1091. Ens, B.J., Klaassen, M., Zwarts, L., 1993. Flocking and feeding in the fiddler crab (Uca tangeri): prey availability as risk-taking behavior. Neth. J. Sea Res. 31, 477–494. Frix, M.S., Hostetler, M.E., Bildstein, K.L., 1991. Intra- and interspecies differences in responses of Atlantic sand (Uca pugilator) and Atlantic marsh (U. pugnax) fiddler crabs to simulated avian predators. J. Crustacean Biol. 11, 523–529. Kneib, R.T., Weeks, C.A., 1990. Intertidal distribution and feeding habits of the mud crab Eurytium limosum. Estuaries 13, 462–468. Moloney, K.A., 1990. Shifting demographic control of perennial bunchgrass along a natural habitat gradient. Ecology 71, 1133–1143. Montague, C.L., 1980. A natural history of temperate Western Atlantic fiddler crabs (Genus Uca) with reference to their impact on the salt marsh. Contrib. Mar. Sci. 23, 25–55. Montague, C.L., 1982. The influence of fiddler crab burrows and burrowing on metabolic processes in salt marsh sediments. In: Kennedy, V.S. (Ed.), Estuarine Comparisons. Academic Press, San Francisco, pp. 283–301. Mueller, V.K., 1983. Untersuchungen zur Populationsbiologie, Aktivitatsrhythmik und geographischen Verbreitung von Uca tangeri (Decapoda, Brachyura). Zool. Jb. Syst. 110, 221–266. Pennings, S.C., 1990. Predator-prey interactions in opistobranch gastropods: effects of prey body size and habitat complexity. Mar. Ecol. Prog. Ser. 62, 95–101. Pennings, S.C., Callaway, R.M., 1996. Impact of a parasitic plant on the structure and dynamics of salt marsh vegetation. Ecology 77, 1410–1419. Petit, D.R., Bildstein, K.L., 1987. Effect of group size and location within the group on the foraging behavior of white ibises. Condor 89, 602–609. Radford, A.E., Ahles, H.E., Bell, C.R., 1968. Manual of the Vascular Flora of the Carolinas. UNC Press, Chapel Hill, NC. Ryer, C.H., 1988. Pipefish foraging: effects of fish size, prey size and altered habitat complexity. Mar. Ecol. Prog. Ser. 48, 37–45. Savino, J.F., Stein, R.A., 1982. Predator-prey interaction between largemouth bass and bluegills as influenced by simulated, submerged vegetation. Trans. Am. Fisheries Soc. 111, 255–266. Savino, J.F., Stein, R.A., 1989. Behavioural interactions between fish predators and their prey: effects of plant density. Anim. Behav. 37, 311–321. Stephens, E.G., Bertness, M.D., 1991. Mussel facilitation of barnacle survival in a sheltered bay habitat. J. Exp. Mar. Biol. Ecol. 145, 33–48. Stoner, M.A., 1979. Species-specific predation on amphipod crustacea by the pinfish Lagodon rhomboides: mediation by macrophyte standing crop. Mar. Biol. 55, 201–207. Tansley, A.G., 1917. On competition between Galium saxatile L. (G. hercynium Weig) and Galium sylvestre Poll. (G. asperum Shreb) on different types of soil. J. Ecol. 5, 173–179. Teal, J.M., 1958. Distribution of fiddler crabs in Georgia salt marshes. Ecology 39, 185–193. Travis, J., 1996. The significance of geographical variation in species interactions. Am. Naturalist 148, S1–S8. Watts, B.D., 1988. Foraging implications of food usage patterns in yellow-crowned night-herons. Condor 90, 860–865. Werner, E.E., Gilliam, J.F., Hall, D.J., Mittelbach, G.G., 1983. An experimental test of the effects of predation risk on habitat use in fish. Ecology 64, 1540–1548. Wiegert, R.G., Freeman, B.J., 1990. Tidal salt marshes of the southeast Atlantic coast: a community profile. US Department of the Interior Biological Report 85 (7.29). Zucker, N., 1981. The role of hood-building in defining territories and limiting combat in fiddler crabs. Anim. Behav. 29, 387–395.