Species-specific patterns of litter processing by terrestrial isopods (Isopoda: Oniscidea) in high intertidal salt marshes and coastal forests

Authors


†Author to whom correspondence should be addressed. E-mail: mzimmer@zoologie.uni-kiel.de

Summary

  • 1The species-specificity of litter processing by three species of isopods at the interface between salt marsh and coastal forest habitats in the south-eastern United States was examined.
  • 2To quantify isopod performance, measurements were taken of feeding, digestion and growth of isopods fed on three litter types (Juncus roemerianus, Quercus virginiana and Pinus palustris) and on artificial diets containing one of three classes of model phenolic compounds (simple phenolics and hydrolysable and condensed tannins).
  • 3To quantify the ecosystem impact of isopods, promotion of microbial respiration, changes in detritus chemistry, and the quantity of litter processed by isopod populations were measured.
  • 4The results support three hypotheses concerning isopod–litter interactions. (i) Isopod performance on different litter types can be predicted based on chemical litter traits, e.g. phenolic concentrations and C : N ratios. (ii) Fully terrestrial isopods are better adapted to the range of phenolics found in angiosperm litter than are semiterrestrial species inhabiting the supralittoral. (iii) Isopod species differ with respect to their impact on decomposition processes due to species-specific digestive capabilities, different effects on microbial decomposition and different rates of net litter processing.
  • 5Because isopods are transitional between semiterrestrial and terrestrial habitats, unlike most other salt marsh detritivores, they are likely to play a unique role in decomposition processes and in the flux of materials between salt marsh and terrestrial habitats.

Introduction

Because salt marshes are dominated by plants of terrestrial origin (angiosperms) but contain animals representative of both marine and terrestrial origins, they provide an evolutionary and ecological link between marine and terrestrial systems. Many of these animals, including a variety of crustaceans, molluscs, insects, fish, birds and mammals (Pennings & Bertness 2001), are motile, and this, coupled with consumer–prey interactions, can move energy and nutrients between salt marshes and adjacent habitats (Kneib 1997, 2000). Historically, much attention has been paid to possible fluxes of materials between salt marshes and marine systems (Teal 1962; Odum 1978; Nixon 1981; Pozo & Colino 1992; Flindt et al. 1999), but salt marshes may also exchange materials with terrestrial habitats through flows of groundwater (Valiela et al. 2000), litterfall and movements of organisms between marsh and terrestrial habitats.

Most of the production of salt marsh angiosperms is consumed by detritivores rather than herbivores (Smalley 1960; Teal 1962; Nixon & Oviatt 1973; Valiela & Teal 1979). A variety of species, including gastropods (Bärlocher, Newell & Arsuffi 1987; Kemp, Newell & Hopkinson 1990; Newell & Bärlocher 1993; Bärlocher & Newell 1994a,b), amphipods (Covi & Kneib 1995; Kneib, Newell & Hermeno 1997; Graça, Newell & Kneib 2000) and crabs (T. L. Buck, G. A. Breed, S. C. Pennings, M. E. Chase, M. Zimmer & T. H. Carefoot, unpublished observations), mediate detrital food webs in salt marshes. Most of these detritivores, however, are of marine origin and cannot tolerate extensive drying. Thus, the detritivore community at the border between salt marsh and terrestrial habitats is likely to be increasingly dominated by species of terrestrial origin, including semiterrestrial and fully terrestrial isopods (Isopoda: Oniscidea), that may feed upon detritus derived from either salt marsh angiosperms or terrestrial plants that border the marsh (Rietsma et al. 1982; Valiela & Rietsma 1984; Zimmer, 2002b). Terrestrial isopods are model organisms for evolutionary-ecological studies because they are represented not only by prototypal species that primarily inhabit the high intertidal area (Schmalfuss 1989), but also by fully terrestrial species that have secondarily invaded the intertidal region from land (Schmalfuss 1998). The latter would be expected to be better adapted to the litter of terrestrial plants than would detritivores of marine origin (cf. Zimmer et al. 2001, 2002). Yet their role in decomposition processes in intertidal systems is virtually unknown.

In comparison with seaweeds, the litter of terrestrial plants often has a relatively high content of lignocellulose and/or of diverse phenolic compounds which require to be degraded either hydrolytically or oxidatively, depending on their structure. Thus, a lack of necessary digestive adaptations could have presented an obstacle to the colonization of land by isopods (Zimmer & Topp 1998a; Zimmer et al. 2001). An initial test of this hypothesis, however, using several isopod species from the rocky intertidal coast of western Canada, suggested that at least some of the necessary digestive adaptations are already present in marine isopods (Zimmer et al. 2001, 2002). Thus, both marine and semiterrestrial species were to some extent able to digest cellulose and to tolerate, digest and oxidize phenolics in their diet, and these abilities generally increased from marine to semiterrestrial to fully terrestrial species. Although suggestive, these studies only examined responses to phenolics derived from phloroglucinol, which are common in brown seaweeds and can be digested hydrolytically. Since terrestrial plants contain a variety of other phenolic classes, including some that can be digested only oxidatively (Hagerman & Butler 1991), a similar comparison of semiterrestrial and fully terrestrial isopods that feed upon angiosperm litter might provide further information on nutritional adaptations to terrestrial food sources (Zimmer et al. 2002).

Isopod communities often include several sympatric species with overlapping habitat distributions. Sympatric species utilizing similar food sources may differ in nutritional requirements and digestive capabilities, either for historical reasons or as a result of interspecific competition (e.g. Zimmer & Topp 2000; Zimmer 2002b). If species differ, they may play different roles in community processes and, if so, the increase in diversity will affect ecosystem function and/or stability (e.g. Chapin et al. 1997; Ives, Klug & Gross 2000; Petchey 2000; Wellnitz & Poff 2001). Thus, the degree to which species differ is of interest not only to evolutionary biologists examining adaptive radiation, but also to community ecologists.

To examine species-specific digestive abilities and ecosystem effects of isopods, we compared three species that inhabit supralittoral salt marshes and adjacent forests of Live Oak and Longleaf Pine. We focused on feeding and decomposition as a function of diets differing in type and concentrations of phenolic compounds. In addition to feeding experiments with natural food sources (leaf litter), we performed feeding experiments with artificial diets containing commercially available phenolic model compounds to represent different classes of phenolics. Three categories of isopod response (feeding, digestion and growth) and three categories of ecosystem impact (promotion of microbial respiration, changes in detritus chemistry and the quantity of litter processed by isopod populations) provided insight into the species-specificity of isopod-mediated decomposition processes. We tested three specific hypotheses: (1) isopod performance on different litter types can be predicted based on phenolic concentrations and C : N ratios (cf. Zimmer & Topp 2000), (2) fully terrestrial isopods are better adapted to the range of phenolics found in angiosperm litter than are semiterrestrial species, which are eco-physiologically bound to the supralittoral, and (3) isopod species differ with respect to their impact on decomposition processes owing to species-specific digestive capabilities, different effects on microbial decomposition and different rates of net litter processing.

Materials and Methods

Experiments were conducted at the Marine Institute of the University of Georgia, at Sapelo Island, GA, USA (31°27′ N; 81°15′ W), during March and April 2000. For our experiments, combinations of litter and isopods were used that were typical of the supralittoral zone of salt marshes around the island (Table 1). Juncus roemerianus Scheele (Black Needlerush) is a dominant high-marsh plant in south-eastern US salt marshes (Wiegert & Freeman 1990). Leaves initially decompose while standing (Newell 2001), but eventually fall to the marsh surface and may form mats that provide habitat for a variety of invertebrates (M. Will & S. C. Pennings, unpublished data). Quercus virginiana Mill. (Live Oak) and Pinus palustris Mill. (Longleaf Pine) are common in coastal forests in the south-eastern USA. Both species shed senescent leaves that fall to the forest floor or, if plants are growing adjacent to the salt marsh, into high-marsh habitats.

Table 1.  Isopod species studied
SpeciesSynonymsFamilyHabitatDensity (ind. m−2)
Littorophiloscia (Halophiloscia) vittata (Say, 1818)aPhiloscia vittata Say 1818aHalophilosciidaeUpper marshb, under logs<20c
Porcellionides (Metoponorthus) virgatus (Budde-Lund, 1885) PorcellionidaeCoastal forest, pine litter15–45c
Venezillo (Armadillo) parvus (Budde-Lund, 1885)Venezillo evergladensis Schultz 1963dArmadillidaeCoastal forests, oak or pine litter20–100c

Three isopod species were found in high-marsh habitats (Table 1). The isopod Littorophiloscia vittata (Say) (Philosciidae) is a semiterrestrial species found in salt marshes along the eastern coast of the USA (Rietsma et al. 1982). Members of the genus Littorophiloscia are common inhabitants of high-marsh areas (Schmalfuss 1998). The isopods Porcellionides virgatus (Budde-Lund) (Porcellionidae) and Venezillo parvus (Budde-Lund) (Armadillidae) are members of fully terrestrial families in which some species have secondarily colonized supra-littoral habitats. On Sapelo Island, P. virgatus and V. parvus occurred both in coastal forests and in the supralittoral salt marsh (M. Zimmer, S. C. Pennings, T. L. Buck & T. H. Carefoot, unpublished observations). All species are referred to generically hereafter.

Adult isopods were housed in 100-ml plastic jars (mesocosm experiments) or small Petri dishes (artificial diet feeding experiments) in the laboratory under ambient temperature (20 ± 2 °C) and light conditions (about 13 h light : 11 h dark). Isopod biomass was determined on a fresh mass basis and converted to dry mass through fresh mass : dry mass ratios (N = 11 per species).

Mesocosm Experiments

Isopods were kept in pairs (Littorophiloscia), separately (Porcellionides) or in groups of five to eight individuals (Venezillo), so that each replicate had about 50 mg (fresh mass) of isopod biomass. Each mesocosm contained isopods, plant litter and some sea water to maintain sufficient humidity. Every 2 days, 0·5 ml distilled water was added; faeces were not removed during the course of the experiment. Plant litter was collected in the field in March 2000. Fallen litter that showed little decay upon visual inspection was handpicked. In the laboratory, plant litter was immediately used for mesocosm experiments under ambient light and temperature conditions (see above). Known amounts (about 200 mg fresh mass) of each litter were added to separate mesocosm-jars (N = 9 for each consumer–litter combination). Litter dry mass was determined based on fresh mass : dry mass ratios (N = 20/species). With three species of isopods, three types of litter and 10 replicates of each, there were 90 separate mesocosms.

After 4 weeks, isopods were removed from mesocosms in order to compare microbial respiration with values from animal-free control mesocosms. Microbial respiration (CO2 production) was determined in mesocosms using a flow-through infrared gas analyser (ADC LCA-4), and served as a measure of the impact of isopod activity on microbial activity of the litter. A promotion factor for microbial activity was calculated by dividing respiration rates of experimental mesocosms by those of control mesocosms (random pairing).

After respiration rates were measured, litter remnants and isopod faeces were separated by hand. Litter remnants, faeces and isopods were dried (24 h, 60 °C) and weighed. Litter ingestion by isopods was calculated as mesocosm litter input minus litter output after termination of the experiment. Similarly, the difference between isopod dry mass at the beginning and end of the experiment was used to estimate growth. From these values, consumption rates [(mg food ingested)/(day × mg animal)] and relative growth rates [(mg dry mass change)/(day × mg animals)] of isopods were calculated for the 4-week duration of the experiment.

Differences in chemical composition (phenolic, C and N content) of freshly collected and remnant litter provided information on the extent to which isopods mediated changes in litter chemistry. These effects could occur either through selective consumption of litter or through influences of isopods on microbiota. To estimate effects of isopod digestive processes (including complexing of phenolic litter compounds) and subsequent decomposition of isopod faeces on the chemistry of detritus from three litter types, the phenolic, C and N contents of isopod faeces and litter were compared. Changes in litter chemistry in isopod-free control mesocosms served as a reference for changes in litter chemistry without impacts of isopods. Carbon and nitrogen were analysed in a Carlo Erba NA-1500 NCS Analyzer. Phenolics were determined as ferulic acid equivalents (‘simple phenolics’), tannic acid equivalents (‘hydrolysable tannins’) and quebracho equivalents (‘condensed tannins’) as described in Zimmer (2002a), according to the standard Folin-Ciocalteu assay (for ‘simple phenolics’), to the Prussian blue assay (Price & Butler 1977; modified after Barbehenn & Martin 1992) (for hydrolysable tannins), and to the vanillin assay (Price, Van Scoyoc & Butler 1978; for condensed tannins).

Artificial Diet Feeding Experiments

Since the results of most quantitative analyses of phenolic compounds depend upon the specific structure of the respective compound (cf. Hagerman & Butler 1991; Waterman & Mole 1994), comparison of phenolics in different food sources must be interpreted with caution. For this reason, feeding experiments were performed with artificial diets containing commercially available phenolic model compounds, in addition to feeding experiments with natural food sources, so that the ability of the isopods to digest different phenolic compounds could be determined more precisely. Three model phenolics (2% dry mass) were added as powder to a chemically defined, cellulose-based and nutrient-rich artificial diet with a water content of 30% (cf. Zimmer & Topp 1998a; Zimmer 1999) and fed to the isopods for 5 days. Ferulic acid served as a model for simple phenolic compounds that are found in a wide variety of plants. Tannic acid served as a model for hydrolysable tannins, and quebracho, a natural condensed tannin, as a model for complex condensed tannins. Both hydrolysable and condensed tannins are characteristic of the litter of dicotyledonous plants (cf. Hagerman & Butler 1991). Each model phenolic was added separately to a nutrient-rich cellulose-based diet (see above) that was fed to individual (Porcellionides) or groups of five isopods (Littorophiloscia or Venezillo, respectively) in small Petri dishes for 5 days (N = 9 replicates per isopod species per model phenolic). Lids of Petri dishes were lined with moist filter paper to maintain high humidity levels. Faeces were collected every 12 h and lyophilized. Phenolics in diet and faeces were measured as described above (Zimmer 2002a). Consumption of artificial diets was calculated on a dry mass basis, and overall phenol digestion was estimated through quantitative comparison of the total amount of phenolics in ingested food vs egested faeces. Phenol digestion through oxidative breakdown (‘phenol oxidation’) was estimated by determining the amount of phenol oxidation products (measured as brown pigments) in acetone extracts of isopod faeces (Barbehenn, Martin & Hagerman 1996). Values for the diet were subtracted from values for the faeces to calculate digestive oxidation. In animal-free control assays (N = 9), no significant differences were detected in the phenol or brown pigment content of the artificial diet after 5 days (P > 0·3; data not shown). Thus, autogenic changes were considered to be negligible.

Results

Traits of litter

The three litter types differed in content of nutrients and phenolics (Table 2). Juncus litter had significantly lower phenolic levels than the other two litter types. Quercus litter had significantly higher nitrogen levels and a lower C : N ratio than the other two litter types. Pinus litter had the lowest nitrogen content, and the highest C : N ratio. Based on these traits, one would predict that Juncus litter would be the best diet (lowest phenolics, relatively low C : N ratio) and Pinus the worst (high phenolics, highest C : N ratio).

Table 2.  Carbon, nitrogen, and phenolic contents of leaf litter. Data are means ± 1 SD, N = 5. Shared letters indicate no significant differences for each trait
 Juncus roemerianusQuercus virginianaPinus palustris
Carbon, mg g−1 396 ± 12a 418 ± 11b 447 ± 11c
Nitrogen, mg g−111·6 ± 0·3a13·2 ± 0·2b10·3 ± 0·5c
C/N ratio343243
Simple phenolics, mg g−1 (ferulic acid equivalents) 2·5 ± 0·1a15·9 ± 0·5b15·8 ± 0·3b
Hydrolysable tannins, mg g−1 (tannic acid equivalents)  16 ± 2a 115 ± 3b 116 ± 3b
Condensed tannins, mg g−1 (quebracho equivalents)   0 ± 0a  42 ± 9b 112 ± 8c

Isopod Performance

In the mesocosms, Littorophiloscia consumed two to three times more Pinus litter than Juncus litter, and three times as much as Quercus litter (Fig. 1). However, animals gained mass only when feeding on Juncus (Fig. 2). Porcellionides ingested twice as much Juncus as Pinus, and three times more Juncus than Quercus (Fig. 1). The pattern of mass change was similar to that for Littorophiloscia, but with Juncus and Quercus sustaining only maintenance levels of growth. Venezillo consumed four times more Juncus litter than Quercus litter, and 10 times more than Pinus litter (Fig. 1). The pattern of mass change in Venezillo (Fig. 2) was similar to that in Porcellionides.

Figure 1.

Relative consumption rates (RCR) of isopods feeding on three litter types. Data are means ± 1 SD; N = 9 per litter type per isopod species; shared letters (lower case for Littorophiloscia, upper case for Porcellionides, Greek for Venezillo) indicate no significant differences among litter types (post-anova Tukey test; α = 0·05).

Figure 2.

Relative growth rates (RGR) of isopods feeding on three litter types. Data are means ± 1 SD; N = 9 per litter type per isopod species; shared letters (lower case for Littorophiloscia, upper case for Porcellionides, Greek for Venezillo) indicate no significant differences among litter types (post-anova Tukey test; α = 0·05).

In feeding experiments with artificial diets, patterns of consumption and digestion of model phenolics differed among isopod species (Fig. 3a, Table 3). Effects of phenolics on feeding were variable depending upon species and type of phenolic. Feeding by Littorophiloscia was stimulated by ferulic and tannic acid. Feeding by Porcellionides was stimulated by tannic acid. Feeding by Venezillo was slightly, but significantly, reduced by quebracho. The effect of phenolics on digestion of the diet also differed among phenolics and species (Fig. 3b, Table 3). None of the phenolics affected the digestibility of the artificial diet by Littorophiloscia, but ferulic acid increased digestibility by Porcellionides, as did both ferulic acid and tannic acid in Venezillo. Phenol oxidation (Fig. 3c) was weak in all isopod species, but some striking differences were apparent in how well different isopods oxidatively degraded different phenolics (Table 3). In particular, Littorophiloscia oxidized ferulic acid and quebracho much more readily than tannic acid, and Porcellionides oxidized ferulic acid much more readily than tannic acid or quebracho. Only Venezillo oxidized non-trivial amounts of tannic acid.

Figure 3.

Consumption (a) and digestive processes – (b) digestibility of diet; (c) phenol oxidation; (d) phenol digestion – of isopods feeding on artificial diets containing one of three model phenolics. Data are means ± 1 SD; N = 9 per litter type/isopod species; shared letters indicate no significant differences among litter types (anova followed by Tukey test).

Table 3. anova tables for consumption and digestive processes of isopods feeding on artificial diets containing three model phenolics: (a) two-way anovas; (b) individual anova tables for each isopod species
 dfConsumptionDigestibilityPhenol oxidationdfPhenol digestion
 SSFPSSFPSSFPSSFP
(a)
Phenolics  30·9328·8<0·001 1 374 3·9 0·011 91275·2<0·001 2   702 1·3 0·280
Isopods  21·4466·9<0·001 5 58624·1<0·001 19924·7<0·001 2 1 302 2·4 0·098
Interaction  60·58 8·9<0·001 6 492 9·3<0·001 43818·1<0·001 4 6 276 5·8<0·001
Error 961·03  11 140   388  7219 500  
Total1073·98  24 600  1938  8027 780  
(b)
Littorophiloscia
Phenolics  30·8050·8<0·001 2 349 4·9 0·006 89234·9<0·001  2 81415·4<0·001
Error 320·17   5 088   273    2 200  
Total 350·97   7 437  1164    5 014  
Porcellionides
Phenolics  30·6858·5<0·001 3 339 8·1<0·001 39858·5<0·001  1 746 2·8 0·078
Error 320·78   4 408    73    7 368  
Total 351·46   7 747   471    9 114  
Venezillo
Phenolics  30·0213·6 0·040 2 17814·1<0·001  6115·1<0·001  2 418 2·9 0·073
Error 320·08   1 648    43    9 936  
Total 350·10   3 826   104   12 350  

Overall changes in the contents of phenolics through various digestive processes (henceforth called ‘digestion’) also differed in a species-specific manner (Fig. 3d, Table 3). Littorophiloscia digested ferulic acid better than tannic acid or quebracho, but Porcellionides and Venezillo exhibited non-significant differences in digestion of phenolics. Similar patterns occurred during digestion of natural litter diets (Fig. 4, Table 4). Hydrolysable and condensed tannins (tannic acid and quebracho equivalents, respectively) were always enriched in faeces of Littorophiloscia, while the content of simple phenolics (ferulic acid equivalents) usually did not change, suggesting that the former were only weakly digested while the latter were digested along with other elements of the diet. Porcellionides exhibited similar changes, except that simple phenolics were enriched and condensed tannins unchanged in the faeces of isopods that fed on Pinus litter. These results suggest the digestion of simple phenolics in Juncus and Quercus litter, and of condensed tannins (by oxidation) in Pinus litter. In contrast, Venezillo digested hydrolysable tannins from Pinus, but did not strongly digest simple phenolics from any diet or hydrolysable tannins from Quercus.

Figure 4.

Change in content of three classes of phenolics due to digestive processes by isopods feeding on three litter types (a, Juncus; b, Quercus; c, Pinus). C, control; L, Littorophiloscia; P, Porcellionides; V, Venezillo. Initial values for litter are indicated by dashed lines at 100%; control means that differ from initial values are indicated with an asterisk above bars. Data are means ± 1 SD; N = 9 per litter type per isopod species; shared letters indicate no significant differences among treatments for each phenolic class within each litter type (anova followed by Tukey test).

Table 4. anova tables for change in content of three classes of phenolics due to digestive processes by isopods feeding on three litter types
 dfFerulic acidTannic acidQuebracho
 SSFPSSFPSSFP
Litter 2178 80038·7<0·001  436 10079·6<0·001 98 28040·1<0·001
Isopods 2166 00035·9<0·001  356 90065·1<0·001 70 49028·8<0·001
Interaction 4 26 330 2·9 0·030  447 30040·8<0·001 94 56038·6<0·001
Error72166 300    197 400   78 450  
Total80537 500  1 438 000  341 800  

Decomposition Processes

The impacts of isopods on microbial respiration varied with food type and isopod species (Fig. 5, Table 5). Both Littorophiloscia and Porcellionides significantly reduced microbial respiration on Juncus but increased respiration of Quercus and Pinus litter by 50–100%. In contrast, Venezillo doubled microbial respiration on every litter type.

Figure 5.

Promotion of microbial respiration by isopods feeding on three litter types. Values for control mesocosms are indicated by dashed lines at 100%; means that differ from controls are indicated with an asterisk above bars. Data are means ± 1 SD; N = 9 per litter type per isopod species; shared letters (lower case for Littorophiloscia, upper case for Porcellionides, Greek for Venezillo) indicate no significant differences among litter types (post-anova Tukey test; α = 0·05).

Table 5. anova tables for promotion of microbial respiration by isopods feeding on three litter types: (a) two-way anovas; (b) individual anova tables for each isopod species
(a)dfSSFP
Litter 213·0435·1<0·001
Isopods 210·6428·7<0·001
Interaction 4 4·00 5·4 0·001
Error7213·36  
Total8041·04  
(b)dfLittorophilosciaPorcellionidesVenezillo
SSFPSSFPSSFP
Litter 27·0247·9<0·001 9·7823·7<0·0010·240·40·653
Error241·76   4·96  6·64  
Total268·78  14·74  6·88  

Effects of isopods on nitrogen and carbon content of litter were less obvious. The three litter types initially contained 1–1·3% N and 40–45% C (Table 2). Microbial decomposition in isopod-free controls increased the C content of the litter by about 5%, but the N content dropped by 10–30% (Fig. 6). The effects of isopods on N and C content were similar in magnitude to the effect of microbial decomposition in isopod-free controls, but were not consistent across all litter types (Table 6). Littorophiloscia tended to increase litter C content, but this effect was significant only for Juncus and Pinus litter. Porcellionides tended to reduce litter N content, but this effect was significant only for Quercus and Pinus litter. In contrast, Venezillo tended to increase litter N content, but this effect was significant only for Juncus and Pinus litter.

Figure 6.

Change in content of carbon and nitrogen in three litter types (a, Juncus; b, Quercus; c, Pinus) during incubation in mesocosms. C, control; L, Littorophiloscia; P, Porcellionides; V, Venezillo. Initial values for litter are indicated by dashed lines at 100%; control means that differ from initial values are indicated with an asterisk above bars. Data are means ± 1 SD; N = 9 per litter type per isopod species; shared letters indicate no significant differences among treatments for each element within each litter type (anova followed by Tukey test).

Table 6. anova tables for change in content of carbon and nitrogen in three litter types during incubation in mesocosms
 dfCarbonNitrogen
 SSFPSSFP
Litter 2 705 88·2<0·001   755 3·6 0·031
Isopods 21221152·6<0·001 6 62431·9<0·001
Interaction 4 494 30·9<0·001   795 1·9 0·117
Error72 288   7 472  
Total802708  15 650  

Microbial decomposition in isopod-free controls did not significantly affect the phenolic content of Juncus litter, but all other values were reduced after 4 weeks of microbial decomposition, sometimes by up to 80% (Fig. 7). Isopod activity often had strong effects on phenolic content of litter, but effects were highly species-specific (Table 7). Isopod activity sometimes strongly reduced phenolic content of litter (e.g. the effect of Porcellionides on simple phenolics in Quercus and Pinus and hydrolysable tannins in Pinus) and sometimes strongly increased it (e.g. the effect of Venezillo on simple phenolics and condensed tannins in Pinus). These effects were not consistent across phenolic classes or litter types and, further, differences between specific phenolics from different litter types have to be taken into account.

Figure 7.

Change in content of three classes of phenolics in three litter types (a, Juncus; b, Quercus; c, Pinus) during incubation in mesocosms. C, control; L, Littorophiloscia; P, Porcellionides; V, Venezillo. Initial values for litter are indicated by dashed lines at 100%; control means that differ from initial values are indicated with an asterisk above bars. Data are means ± 1 SD; N = 9 per litter type per isopod species; shared letters indicate no significant differences among treatments for each phenolic class within each litter type (anova followed by Tukey test).

Table 7. anova tables for change in content of three classes of phenolics in three litter types during incubation in mesocosms
 dfFerulic acidTannic acidQuebracho
 SSFPSSFPSSF P
Litter 2104 50073·2<0·001181 400113·0<0·00134 660108·1<0·001
Isopods 2  6 414 4·5 0·015  5 462  3·4 0·039 8 769 13·7<0·001
Interaction 4 65 64023·1<0·001 15 990  4·9 0·001 8 103 12·6<0·001
Error72 51 360   57 810  15 390  
Total80227 900  260 600  66 920  

Discussion

Our results demonstrate that, among the tested litter types, food quality for isopods (measured in terms of isopod growth) was readily predicted based on phenolic content and C : N ratio (cf. Zimmer & Topp 2000). Moreover, the ability of isopods to process phenolic compounds accorded with the general prediction that the ability of isopods to process phenolics should increase from marine to semiterrestrial to terrestrial species (cf. Zimmer et al. 2001, 2002). Finally, our data on decomposition processes indicate that these isopod species are not redundant members of the community, but have species-specific effects on decomposition of leaf litter. Below, we discuss these points in turn.

Quality of Different Litter Types

Based on its overall low phenolic content and low C : N ratio, Juncus litter should have been the best of the three foods for isopods. Conversely, Pinus litter, with its high phenolic content and high C : N ratio, should have been of low quality. As expected, growth of all three isopods was best on Juncus and least on Pinus. Consumption of Juncus was not always higher than that of other litter types, but this probably reflected animals compensating for poor quality food through increased consumption (e.g. Dallinger & Wieser 1977; Rushton & Hassall 1983; Cruz-Rivera & Hay 2000).

Species-Specific Digestive Capabilities of Isopods

Based on their evolutionary history, we expected the ability to digest different phenolic compounds to differ between the supralittoral Littorophiloscia and the terrestrial Porcellionides and Venezillo. In particular, we predicted that the terrestrial species would be increasingly capable of processing larger and more complex phenolic compounds. This prediction was generally supported by our results. In contrast to some other salt marsh detritivores (cf. Valiela & Rietsma 1984; Bärlocher & Newell 1994b), model phenolics did not deter feeding on artificial diets by isopods. Further, these phenolics did not clearly affect total digestibility in Littorophiloscia, but in three out of six cases increased digestibility of artificial diets in the two terrestrial species. Because phenolics are typically thought of as digestibility-reducing defences, such increases in digestibility would appear to be counterintuitive; however, similar effects of phenolics and degradation products of tannic acid have been observed in the terrestrial isopod Porcellio scaber (Zimmer 1999). In P. scaber, a tannin-rich diet induced increased numbers of bacterial endosymbionts (Zimmer 1999) that are thought to contribute to digestive processes (Zimmer & Topp 1998a,b; Zimmer 1999; Zimmer et al. 2001, 2002). Further, owing to surfactants present in the gut lumen (Zimmer 1997, in P. scaber), isopods are thought to reduce protein precipitation by ingested phenolics. Neither the presence of surfactants nor the role of endosymbionts in digestion by the isopods studied here is known, but the present results suggest that the two terrestrial species are better adapted to phenolics than is the semiterrestrial species.

The ability to oxidize phenolics is believed to be one of a plethora of important adaptations to a terrestrial lifestyle, because lignins and many non-hydrolysable phenolics, both of which are common in terrestrial food sources, can be digested only through oxidation (Zimmer et al. 2001, 2002). Phenol oxidation differed between isopod species and model phenolics, but did not show a pattern of increasing oxidative ability in terrestrial species. On the other hand, patterns of digestion of different model phenolics offered in artificial diets reflected the accepted evolutionary model for isopods. Although the total ability to digest model phenolics did not differ among isopod species, Littorophiloscia digested a higher percentage of the monomer ferulic acid than of the complex polymers tannic acid or quebracho. In contrast, Venezillo and Porcellionides digested all three model phenolics equally effectively. Similarly, when feeding on natural diets, Littorophiloscia typically was the best of the isopods at digestion of simple phenolics but intermediate or least effective at digesting hydrolysable or condensed tannins.

Although differences between the semiterrestrial and fully terrestrial species in these attributes are subtle, they accord with previous comparative studies (Zimmer et al. 2001, 2002). These results suggest that the ability to tolerate and even digest phenolics is present to some degree in marine and semiterrestrial isopods. Many physiological and digestive adaptations may have been necessary for isopods to colonize land; however, our results suggest that the presence of phenolics in terrestrial food sources would not have presented an obstacle to this evolutionary step.

Species-Specific Impacts of Isopods on Decomposition Processes

Different species of detritivores may have very different impacts on decomposition processes. Feeding by detritivores, for instance, may either promote or reduce microbial activity (e.g. Teuben & Roelofsma 1990; Daniel & Anderson 1992; Kandeler et al. 1994; Graça et al. 2000; Kautz & Topp 2000). In the present study, only Venezillo promoted microbial respiration in all three litter types, while the effects of the other isopods depended on the litter type. The underlying mechanisms are unknown, since we did not directly monitor changes in bacterial or fungal communities. Nonetheless, these results clearly suggest that the impacts of these isopods on decomposition processes are species- and litter-specific. Species-specific effects of isopods on decomposition processes were evident throughout this study: in addition to differences in phenolic digestion and effects on microbial respiration, isopods also differed in consumption of different litter types and in impacts on litter chemistry (phenolic, C and N contents). Similar conclusions were drawn from field data on isopods and microbiota in a floodplain forest (Zimmer & Topp 1999) where two species of terrestrial isopods appeared to have differential effects on biomass and activity of litter and soil microbiota.

The species-specific impacts on decomposition that we documented in the laboratory will be altered in the field because of differences in abundance of the three species. Further, the presence of more than one food source in the field, along with corresponding feeding preferences of consumers, will influence feeding rates as compared to experimental single-food diets. Based on our data for litter consumption, we can only roughly estimate the contribution of each isopod species to decomposition processes in the high intertidal salt marsh and coastal forests in this region of Georgia and at this time of year. Owing to its relatively high density and large size (isopod biomass of 0·1–0·3 g dry mass m−2; M. Zimmer, unpublished data), Porcellionides would consume roughly 7–20, 3–9 or 4–12 g m−2 of Juncus, Quercus or Pinus litter, respectively, every year. In contrast, the less frequent and smaller species Littorophiloscia would only weakly contribute to the decomposition of Juncus and Quercus litter by ingesting less than 2 g m−2 per year. Assuming that our results on consumption of Pinus litter by this isopod are not just artefacts of a single-food diet resulting in compensatory ingestion, Littorophiloscia could consume an annual amount of up to 5 g m−2 of Pinus litter. In contrast, Venezillo would only consume 0·1–0·6 or 0·3–1·5 g m−2 of Pinus or Quercus litter, respectively, annually, and only 1–6 g m−2 of Juncus litter per year.

Most studies of decomposition in salt marshes have focused on lower intertidal levels than we consider here (Pennings & Bertness 2001). At these lower elevations, a variety of invertebrates, including the snails Littoraria irrorata and Melampus bidentatus and the crab Armases cinereum contribute to decomposition by consuming detritus (Bärlocher et al. 1987; Kemp et al. 1990; Newell & Bärlocher 1993; Bärlocher & Newell 1994a,b; Covi & Kneib 1995; Kneib et al. 1997; Graça et al. 2000). These species are often present at higher biomass levels than are isopods, and consequently will contribute more strongly than isopods to detrital processing. With the exception of Armases, however, none of these species can tolerate extensive drying. Moreover, although some salt marsh detritivores are deterred from feeding on Spartina litter by ferulic acid (Valiela & Rietsma 1984; Bärlocher & Newell 1994b), isopods were not and were even capable of extensively digesting several phenolics (this study). Thus, isopods may play an important role in decomposition at the terrestrial fringe of salt marsh habitats. Because isopods are also consumed directly by Armases (T. L. Buck, G. A. Breed, S. C. Pennings, M. E. Chase, M. Zimmer & T. H. Carefoot, unpublished observations), which is an omnivore that migrates between littoral salt marshes and coastal forests, they may also play a role in the fluxes of energy and nutrients between terrestrial and salt marsh habitats.

Acknowledgements

We thank Steve Newell, University of Georgia Marine Institute, for valuable comments on our study and the manuscript. Two anonymous referees were helpful in clearing up some of our interpretations. We are grateful for financial support provided by the University of Georgia Marine Institute Visiting Scientist Program (MZ), the Georgia Coastal Ecosystems LTER (OCE 99–82133, TLB, SCP) and the Natural Sciences and Engineering Research Council of Canada (THC). This is contribution 2000–2 of the International Isopod Research Group (IIRG), and contribution no. 905 of the University of Georgia Marine Institute.

Ancillary

Advertisement