Influence of peat substrate on the distribution and behaviour patterns of sand shrimp, Crangon septemspinosa, under experimental conditions

Introduction

Peat bogs occupy 12% of the Canadian landscape (Keys, 1992) and represent 90% of Canadian wetlands (Price and Waddington, 2000). Peat moss is most commonly used for horticultural purposes as a soil conditioner. In eastern New Brunswick, peat mining is particularly important (Daigle and Lamarche, 1999), representing 36% of the national production and valued at over 98 million dollars in 2000 (Thibault, 2001).

Expansion of the peat industry has raised major concerns regarding its influence on the quality of the coastal and estuarine environments (Osborne, 1982; Sallantaus and Pätilä, 1983; Olsson and Näslund, 1985; Selin et al., 1985; Shotyk, 1986; Lavoie, 1995). This is because of the presence of dissolved organic matter (Surrette et al., 2002), nutrients (Sallantaus and Pätilä, 1985; Surrette et al., 2002); heavy metals (Glooschenko et al., 1985; Di Giulio and Ryan, 1987) and peat fibers (Korpijaakko and Pheeney, 1976; Selin and Nyrönen, 1985) that are released in runoff from mined peatlands and transported to fresh and marine waters.

In some locations in Eastern New Brunswick, extensive peat mining has resulted in the accumulation of large quantities of peat in rivers, estuaries and along coastal areas (Lavoie, 1995; Ouellette, 2002). The introduction of large amounts of peat deposits over a period of time could significantly modify the composition of the bottom sediment, changing it from sand or mud to an organically-rich fibrous substrate. The extent to which peat particles are mixed with other substances (sand, mud, etc.) ultimately determines the characteristics of the bottom sediment. A thick deposit of peat fibers, in varying states of decomposition, creates mats of organic debris that can also retain large quantities of water (Williams, 1958).

Anthropogenic changes resulting from the introduction of excessive peat into the sediment could alter a previously established association between a species and its microhabitat. Sediment, like other abiotic factors (e.g. salinity, temperature, oxygen), strongly influences species distribution throughout various environmental niches. This is particularly relevant as it has been demonstrated that species establish intricate animal–sediment relationships (Sanders, 1958; Gray, 1974; Rhoads, 1974). Consequently, changes in substrate type could trigger modifications in population dynamics of individual groups of organisms and ultimately in the interactions of species that form part of existing food webs (Bonsdorff et al., 1995; Mannino and Montagna, 1997).

Aquatic habitats are characterized by a wide variety of bottoms that range from hard substrates such as cobbles and gravel to softer-based substrates, including silt, mud and fine sand. Numerous studies have described the close associations that exist between certain species and the softer variety of substrates (Sanders, 1958; Ruello, 1973; Moller and Jones, 1975; Rulifson, 1981; Aziz and Greenwood, 1982; Minello et al., 1987; Pinn and Ansell, 1993; Mannino and Montagna, 1997). While peat can also be classified as a ‘soft’ substrate, it differs from others because of its high organic content and angular-shaped fibers. Fewer studies, however, (Williams, 1958; Olsson and Näslund, 1985; Olsson and Persson, 1986; Growns and Richardson, 1988; Daigle et al., 1995; Strychar, 1997; Strychar and MacDonald, 1999) have considered the interaction of aquatic organisms and peat substrate.

Preliminary work (Ouellette et al., 1997) conducted in an estuarine habitat containing a peat deposit revealed the presence of the sand shrimp, Crangon septemspinosa. This euryhaline species is widely distributed along the northwestern Atlantic coast, ranging from Newfoundland to eastern Florida (Squires, 1990). It plays an important role in the food web, both as predator (Olmi and Lipcius, 1991; Bertram and Leggett, 1994; Kamermans and Huitema, 1994) and prey (Bulak and Heidinger, 1980; Friedland et al., 1988; Blackwell et al., 1995). It has also served as an indicator to gauge stress and toxicity in its environment, as demonstrated by various in vivo and in situ experiments (McLeese and Metcalfe, 1979; McLeese et al., 1980; Burridge and Haya, 1993).

The sand shrimp is a burrowing species that utilizes the sediment for shelter (Haefner, 1979) as opposed to other non-burrowing pelagic species. Unlike sedentary benthic species, it can relocate as the habitat becomes less suitable. Because of its close association with the sediment, it was felt that this epibenthic species could serve as a useful indicator of potential environmental stress resulting from modification to substrate by the accumulation of peat.

Studies have shown that certain species of Penaeid (Rulifson, 1981) and Crangonid (Pinn and Ansell, 1993) shrimp are not distributed in their habitats either uniformly or randomly but display specific preferences for particular substrates. Recent investigations (Ouellette, 2002) demonstrated that there were significantly fewer sand shrimp (C. septemspinosa) found on substrates with high peat concentration (>66% of substrate composition) than on sediments consisting of higher sand and mud content. To shed further light on the possible influence of peat on the distribution and behaviour patterns of sand shrimp, a series of controlled laboratory experiments were conducted in which shrimp were provided a choice between sand and peat substrates. In addition, food was introduced on the substrates to determine if its presence could influence substrate selection by the sand shrimp.

Furthermore, sand shrimp have been shown to ingest sand particles during the feeding process, thereby indicating that they do not display any discriminatory capacities in selecting between food and substrate particles (Wilcox and Jeffries, 1974). A similar profile may be applicable in those instances where shrimp feed on peat substrate. This gives cause for concern, as it has already been demonstrated that inclusion of peat particles during the feeding process of other invertebrate species (Olsson and Näslund, 1985; Strychar and MacDonald, 1999) may interfere with the ingestion of nutrients which could, in turn, contribute to a decrease in growth and development. In order to determine if peat particles in the substrate are ingested by the shrimp, histological observations were conducted on the digestive system.

Materials and methods

Laboratory conditioning

Offshore samples of adult sand shrimp (30–50 mm total length) were collected from an open-water rainbow smelt fishery in the Miramichi River, New Brunswick, Canada during the winter of 1998. Bag nets (24 mm stretched mesh) were set in 2.5–6 m deep waters and emptied weekly. Healthy specimens were transported to the Université de Moncton in a cooler filled with brackish water (15 ppt, 5°C). Upon arrival, shrimp were acclimated in aquaria containing water at 12°C and 15 ppt, made from distilled water mixed with artificial salts (Instant Ocean Inc.).

Shrimp were subjected to a 12 h light/dark regime beginning at 7.00 hours for 4 days prior to the beginning of the experiments. They were fed a commercial 400-μm shrimp diet (Zeigler Bros, Inc.) once a day at 19.00 hours. Excess food and feces were removed by siphon each morning. The shrimp used for either behavioural or histological studies were non-ovigerous and had an average total length and standard deviation of 38 ± 3 mm (n=200).

Experimental setup

Experimental aquaria were prepared to determine if sand shrimp displayed a preference between sand and peat substrates. Each aquarium (60 × 31 × 30 cm high) was divided across the middle by a 4 × 31 cm strip of plexiglass resulting in two identical areas of 930 cm². A 5 cm layer of peat and another of sand was deposited on opposite sides, each one flush with the divider. Loose peat was obtained from a peat deposit in Mill Creek, a tributary of the Richibucto River in New Brunswick. The peat had accumulated over the years through the introduction of peat fibers from drainage water from a peat extraction site (Ouellette et al., 1997). Peat fibers, typically dark-brown to black in colour, ranged between 0.01 and 0.05 mm, depending on whether they were whole or fragmented because of the decomposition process. Beige-coloured sand void of any peat fibers and ranging between 0.005 and 0.01 mm was obtained from a local masonry company.

Substrate preference in starving shrimp

The set-up employed consisted of two aquaria. In order to block for any potential directional bias, peat was placed on the left side in one aquarium and on the right in the other. The study was repeated three times, and new specimens were used for each replicate. Feeding stopped prior to the experiment and shrimp were not fed during the duration of the experiment.

Ten shrimp were introduced into each of the two aquaria by way of a perforated plastic cylinder (10 cm wide : 30 cm high) with a closed screened bottom which was positioned in the middle of the aquarium. Starting at 9.00 hours, the animals were left to acclimate to the test aquaria environment for 12 h in the cylinders. The cylinders were subsequently removed and the shrimp were released into the aquaria and allowed to acclimate and explore their environment for another 12 h as of 21.00 hours.

The location of each shrimp was observed for both aquaria at 9.00 hours the following morning and every 12 h thereafter during the first 2 days (i.e. 9.00 and 21.00 hours). A final observation was made 84 h later. A total of five sets of observations were therefore conducted for each aquarium. Upon completion, the shrimp were removed and sediment and water were changed in each aquarium. This procedure was repeated for each of the next two replicates. The duration of each replicate was 4.5 days long (acclimation period and observation time).

Because no differences in preference patterns were detected between aquaria within replicates or between replicates over each of the five observation times (P > 0.10, heterogeneity chi-square), data of the three replicates were pooled for statistical analysis. Deviation from an even distribution of animals between peat and sand was tested separately by chi-square for each observation time. Heterogeneity among observation times was also tested by chi-square.

Substrate preference in feeding shrimp

The study consisted of six aquaria using identical setups to those described above. In three of the aquaria, food in the form of fish flakes (Nutrafin, Inc.) was placed at each end (i.e. on sand and on peat) of the aquarium just prior to the introduction of the animals. Each food patch had an area coverage of 10 cm in diameter. No food was added to the three remaining aquaria, which served as controls.

Five shrimp were introduced into each of the aquaria by way of the cylinder as described above. Shrimp were maintained in the cylinders for 12 h starting at 21.00 hours prior to experimentation. This step allowed for acclimation to the aquarium environment, while preventing the consumption of food. The cylinders were subsequently removed at 9.00 hours the next morning and the animals released. Unlike the first study, observations started immediately after the cylinder was removed and the shrimp were released into the aquaria.

Locations of the five shrimp were recorded in each of the six aquaria over a period of 2 days at different observation times. During the first day, observations were made every 10 min between 9.30 and 11.00 hours, then at 12.00, 14.00, 16.00, 18.00 and 20.00 hours. On the second day, because by this time all the food had been consumed and shrimp movement had become negligible, observations were only conducted at 9.30, 12.00 and 20.00 hours. A total of 18 observations were therefore made for each aquarium for the two replicate experiments. The duration of each replicate was 2 days (acclimation period and observation times).

Preference, defined as whichever substrate had three or more shrimp at each of the 18 observation times for the two replicates, was determined for each aquarium using a binomial test. Because no differences in preference were detected among the 18 observation times (P > 0.10, Bonferroni multiple-comparison test), the total number of significant observation times in which shrimp preferred sand from the three aquaria with food were pooled together, as were those without food. Finally, for each replicate experiment, mean number of observation times in which sand was preferred over peat were compared between fed and unfed aquaria by anova.

Histological study

Eighty shrimp that had previously been starved for 24 h, were dispersed amongst four aquaria, (20 shrimp/aquarium), two containing sand and two with peat. After 24 h acclimation in the experimental aquaria, food in the form of fish flakes was introduced into two of the aquaria, one with sand and one with peat. No further food was provided for the duration of the experiment.

Three starved and three fed shrimp from each of the two types of substrate were sacrificed at intervals of 10, 20, 30, 60, 120, and 360 min (total 72 shrimp) after the introduction of food. Observation times were chosen based on previous experiments conducted on sand shrimp by Wilcox and Jeffries (1974), who identified food in the stomach and the digestive gland of animals within minutes after ingestion and as early as an hour later in the intestine. The sampling regime employed in the current study was used to ensure that if shrimp ingested peat, it would not go undetected.

At each observation time, shrimp were immediately fixed by injecting 15 ml of Davidson's solution (Bell and Lightner, 1988) into the cephalothorax and abdomen, respectively. The specimens were subsequently transferred to 50% ethyl alcohol and stored for periods of 48 h. The cephalothorax was severed from the abdomen and the latter was cut transversely into three equal parts. Six-micrometer sections parallel to the coronal plane of the cephalothorax and transversally to the abdomen were prepared. Sections were stained using a modified Mayer's hematoxylin and phloxine/eosin preparation (Bell and Lightner, 1988) and the digestive tube was examined under the microscope.

Results

Behavioural patterns related to sand and peat substrate

When substrate preference between sand and peat by C. septemspinosa was tested with starved shrimp, it was revealed that shrimp were more often observed on sand than peat (73% of all shrimp observed were on sand, n=60). This was true for all observation times (Fig. 1), which were highly significant for 12, 36, 48, and 84 h (P < 0.001) as well as for 24 h (P < 0.005). Moreover, no significant change in distribution patterns was noted over the five observation times (P > 0.50, heterogeneity chi-square).

When food was introduced on both substrates, similar preference patterns as described above were noted. Data were non-normally distributed but neither anova analysis of proportions of observation time, arc-sin squareroot transformated, nor non-parametric analysis (Kruskal–Wallis) changed the outcome of the results reported. Furthermore, pooling the two replicate experiments also did not change our conclusion. On average, 89% of the starved shrimp preferred sand over peat (Table 1). With the introduction of food, however, 14% fewer shrimp on average were recorded on sand throughout the study (Table 1). Because of the higher densities of shrimp on sand in the early stages of the experiment, the food was depleted more rapidly than over peat. As a result, about 1 h after the beginning of the study, having consumed the food on the sand substrate, shrimp moved toward the other side of the aquarium and continued foraging on the peat where food was still present. This persisted for a period of about 1 1/2 h until all food was exhausted; thereafter, the number of shrimp on sand vs peat assumed their initial distribution pattern, averaging 60% on sand compared with 40% on peat (Fig. 2). Differences in preference between the two treatments (fed shrimp and unfed shrimp) were not significant for either replicate (Table 1).

Presence of peat particles in the digestive system

Peat fibers were identified in the gut contents of the stomach and the gastric mill of fed shrimp removed from the aquarium containing peat substrate (Fig. 3a). While sand particles were also observed in the gut content of shrimp fed over sand, they occupied less area and were not as widespread (Fig. 3b), resulting in a more even food distribution. Shrimp that were starved throughout the study and sampled from either substrate did not display peat or sand particles in the gut.

Peat particles were observed in the digestive tract of shrimp in the stomach (Fig. 4a) and also in the gastric mill along with the rest of the gut content where they are crushed into minute bodies (Fig. 4b). Because of the crushing effect of the gastric mill, the stomach contents are broken down into very fine particles. Peat fibers lose their characteristically fibrous structure and become indistinguishable from other debris. Minute black particles representing fine remnants of peat were also observed in the digestive gland and intestine; these particles were not observed in the digestive system of shrimp fed on sand.

Discussion

Influence of peat particles on the sand shrimp

Crangon septemspinosa is an ubiquitous, epibenthic species that is widespread in estuarine and marine environments. While it has been described as occurring on sandy bottoms (Sanders, 1958; Wilcox and Jeffries, 1974; Haefner, 1979; Corey, 1981), it has also been observed on other types of soft substrates.

Efforts to gauge the effects of peat particles on the distribution of Crangon in laboratory experiments confirmed that shrimp display a marked preference for sand vs peat. These experiments also indicated, however, that shrimp will move onto a peat substrate to forage if food is present only there. This observation suggests that, although there is a marked preference for sand, peat is not completely repellent to them. Such observations are consistent with field studies (Ouellette, 2002) which show that shrimp occur in much higher densities over sand with no peat, or with sand mixed with low to moderate peat concentrations, than on substrates containing high levels of peat. Preference for sand can be largely explained by two phenomena, best categorized as physical and/or physiological responses to the substrate.

An important physical consideration is movement, which includes burrowing and displacement actions. Sand shrimp, like other epibenthic decapod crustaceans, burrow into the substrate for shelter. The nature and size of sediment particles can affect locomotive and burrowing ability (Trueman and Ansell, 1969; Ruello, 1973; Moller and Jones, 1975; Aziz and Greenwood, 1982; Pinn and Ansell, 1993). Less suitable substrates can decrease their burrowing efficiency, thereby creating an inferior cover. This in turn could lead to less protection against non-burrowing predators and poorer coverage in preparation for ambush of passing prey (Minello et al., 1987).

The burrowing behaviour of sand shrimp was observed on both sand and peat during experimentation, and is similar to the mechanism described by Trueman and Ansell (1969) for crustaceans, and more specifically for brown shrimp Crangon crangon as recorded by Pinn and Ansell (1993). In sand, shrimp use their pereopods (walking legs) to brace themselves on the sediment, scraping and loosening the sediment under their body. During this time, their pleopods (swimming appendages) and uropods create a current of water that displaces sand particles from under the animals and out of the burrow. Shrimp use their appendages to force themselves downwards and the sand particles quickly redeposit on top of the animals, exposing only their rostrum, eyes and antennae. Finally, the antennae move in a parallel motion to the substrate, sweeping sand particles toward their body, thereby covering themselves completely.

The burrowing mechanism of shrimp is less effective in peat than in sand substrate. When they burrow, peat fibers tend to drift away because of their buoyancy and do not redeposit themselves directly on the shrimp, leaving them partly exposed. Furthermore, their pale pigmentation, which normally blends in well with a sandy substrate acting as a desirable camouflage, contrasts with the dark colour of peat. In nature, therefore, both these situations would reduce their coverage and more readily expose them to predation.

The second phenomenon, which might explain the shrimp's preference for sand vs peat, can be classed as physiological and is primarily related to potential interference with respiratory and feeding activities, as shown to be the case for other species (Olsson and Näslund, 1985; Olsson and Persson, 1986; Strychar and MacDonald, 1999). Histological observations confirmed the presence of peat in the shrimp's digestive system. These findings, therefore, lend credence to the initial suggestion that shrimp may be incapable of selecting food without also ingesting peat fibers. The consequence of this is that peat particles reduce available space for food and undergo breakdown in the gastric mill. Such activity requires energy during the crushing action by the gastric mill. Olsson and Näslund (1985) and Strychar (1997) demonstrated that peat fibers are a low quality food source for invertebrates, and this may very well be applicable to sand shrimp. If that is the case, there would be a loss of energy with no assimilative benefit to the animal. Interestingly, present findings also revealed that when shrimp were starved, they did not resort to ingesting peat fibers; therefore peat is not considered to be a food source by sand shrimp.

In an ecological context, the implications for shrimp having to forage for food on peat substrate, thereby subjecting themselves to a less favorable situation, may merit further exploration. Future studies that assess the energy required by shrimp to process the peat fibers ingested with food on substrates displaying varying concentrations of peat, may provide valuable information on the limitations imposed by this substance on the animal's growth and development.