Ammonia excretion in aquatic and terrestrial crabs

Ammonia^†(i.e. the total of NH₃ and NH₄⁺:T_Amm) is highly toxic in most animals. Hydrated NH₄⁺ and K⁺ ions have the same ionic radius(Knepper et al., 1989) and due to their K⁺-like behavior, ammonium ions affect the membrane potential for example, in the giant axon of Loligo pealei(Binstock and Lecar, 1969) and of mammalian neurons (Cooper and Plum,1987). In mammals, elevated ammonia causes major damage in the central nervous system, including changes in blood–brain barrier morphology (Laursen and Diemer,1997). In addition, elevated ammonia levels in mammals have been related to Alzheimer disease (Alzheimer Type II astrocytosis) due to toxic accumulation of glutamine in astrocytes, which leads to cell swelling and cell death (for review see Butterworth,2002). Also, in microglia and astroglioma cell lines, ammonia affects major functional activities, such as phagocytosis and endocytosis. Ammonia also modifies the release of cytokines and increases the activity of lysosomal hydrolases (Atanassov et al., 1994, 1995). Marcaida et al.(1992) speculated that ammonia toxicity is mediated by excessive activation of N-methyl-d-aspartate (NMDA)-type glutamate receptors in the brain. As a consequence, cerebral ATP depletes while intracellular Ca²⁺ increases with subsequent increases in extracellular K⁺ and finally cell death.

In crustaceans, for example in the lobster Homarus americanus(Young-Lai et al., 1991) and the crayfish Pacifastacus leniusculus(Harris et al., 2001),elevated ammonia levels in low-salinity media disrupt ionoregulatory function. Exposure of the green shore crab Carcinus maenas to 1 mmol l^–1 total ammonia leads to increased ion permeability and salt flux across the gill; higher concentrations reduce both variables(Spaargaren, 1990). In fish,branchial gas exchange and oxidative metabolism are disturbed by excess ammonia (Wilkie, 1997).

An effective ammonia detoxification or excretion system is, therefore,essential to maintain cellular functions, and to keep cellular and body fluid ammonia levels within a tolerable range. In most species, including mammals(Cooper and Plum, 1987), fish(Wood et al., 2002) and aquatic crabs (Cameron and Batterton,1978; Weihrauch et al.,1999), the ammonia concentration of the body fluids is typically low (50–400 μmol l^–1; Table 1). Concentrations exceeding 1 mmol l^–1 total ammonia(NH₃+NH₄⁺) are usually toxic to mammalian cells (Hrnjez et al., 1999). In crustaceans, environmental exposure of ammonia is lethal at relatively low doses. For instance, LC₅₀ after 96 h of exposure was determined in the crayfish Orconectes nais at 186 μmol l^–1NH₃ (Hazel et al.,1982), in the Sao Paulo shrimp Penaeus paulensis at 19μmol l^–1 NH₃ and 0.307 mmol l^–1 total ammonia(Ostrensky et al., 1992) and in the redtail prawn Penaeus penicillatus 58 μmol l^–1 NH₃ and 1.39 mmol l^–1 total ammonia (Chen and Lin,1992).

Mammals accrue ammonia both from metabolism and as an influx to the hepatocytes from the gastrointestinal tact. This ammonia is detoxified in the urea cycle, an energy-consuming process, by incorporation into the less-toxic urea. Crustaceans are largely ammonotelic, aquatic species exclusively so, and in water excrete their nitrogenous waste directly to the environment as highly soluble ammonia.

The vast majority of the excreted ammonia originates from the catabolism of proteins and amino acids. There is little evidence to support nitrogenous excretion as amino-nitrogen because consumable amino acids released by crustaceans are more likely a result of passive lost viaamino-acid-permeable structures (Dagg,1976) and via feces(Claybrook, 1983). Some additional ammonia is produced in reactions involving purine and pyrimidine bases (Claybrook, 1983)(Fig. 1). However, ammonia derived via this uricolytic pathway is considered to contribute only a small portion of the total ammonia excreted compared with the predominant production from amino acids (Hartenstein,1970; Schoffeniels and Gilles,1970). Ammonia derives, for the most part, from deamination of glutamine, glutamate, serine and asparagine by the specific enzymes glutaminase, glutamate dehydrogenase, serine dehydrogenase and asparaginase,respectively (King et al.,1985; Krishnamoorthy and Srihari, 1973; Greenaway, 1991, 1999).

In aquatic crabs, primary urine is formed via ultrafiltration in the antennal gland, which is thought to play the key role in regulation of body water and divalent cations (e.g. Mg²⁺, Ca²⁺; Mantel and Farmer, 1983), but not to contribute significantly to the excretion of nitrogenous waste products(Regnault, 1987). For instance, in the blue crab Callinectes sapidus, <2% of total ammonia is excreted in the urine via the antennal gland system(Cameron and Batterton,1978).

The main site for ammonia excretion by aquatic crabs is the phyllobranchiate gill (Claybrook,1983; Kormanik and Cameron,1981; Regnault,1987), featuring a single-cell-layered epithelium covered by an ion-selective cuticle (Avenet and Lignon,1985; Lignon,1987; Onken and Riestenpatt,2002, Weihrauch et al.,2002). The gills of aquatic crabs are multifunctional organs. In addition to their function in excretion of nitrogenous waste products, they are also responsible for respiratory gas exchange(Burnett and McMahon, 1985),regulation of acid–base balance(Henry and Wheatly, 1992) and osmoregulatory ion transport (Towle,1981; Lucu, 1990, Riestenpatt et al., 1996, Towle and Weihrauch, 2001). Several transporters and enzymes putatively linked and involved in ammonia transport have been shown to be present in the branchial epithelium of crabs,as summarized in Fig. 2 and Table 2.

In solution, both forms of ammonia, non-ionic ammonia (NH₃) and the ammonium ions (NH₄⁺) exist in a pH-dependent equilibrium. As a weak base (pK ≈9.48 at 20°C and NaCl=250 mmol l^–1; Cameron and Heisler,1983) and at a physiological pH of pH 7.8, 98% of total ammonia exists in the ionic form NH₄⁺, whereas only 2% is present as non-ionic NH₃. However, the higher lipid solubility of NH₃ makes it more diffusible through phospholipid bilayers. Kormanik and Cameron (1981)reported that ammonia excretion of seawater adapted blue crabs Callinectes sapidus occurred mainly by diffusion of non-ionic NH₃. An excretion mechanism based predominately on NH₃ diffusion is not likely, however, because membrane permeability of NH₃ is much lower than that of CO₂ (Knepper et al., 1989). Indeed, some plasma membranes of animal epithelia are relatively impermeable to NH₃ as shown for frog oocytes(Burckhardt and Frömter,1992), the renal proximal straight tubules(Garvin et al., 1987) and colonic crypt cells (Singh et al.,1995). Accordingly, other authors have obtained experimental evidence for at least partial excretion of ammonia in its ionic form(NH₄⁺) in Callinectes sapidus(Pressley et al., 1981) and Carcinus maenas (Lucu,1989; Siebers et al.,1995).

Studies on isolated perfused gills of several aquatic crabs showed that ammonia can be excreted actively against a 4–8-fold inwardly directed ammonia gradient across both the anterior and the posterior gills to a similar degree despite their different morphological and physiological characteristics(Copeland and Fitzjarrell,1968; Goodmann and Cavey,1990; Weihrauch et al., 1998, 1999; Towle and Weihrauch, 2001)(Fig. 3). Under physiologically relevant conditions, the potential for active branchial ammonia excretion is significantly greater in the marine Cancer pagurus than in freshwater-acclimated Chinese mitten crabs Eriocheir sinensis,despite the much larger ionic conductance of Cancer pagurus gills(∼250–280 mS cm^–2) compared with that of Eriocheir sinensis gills (∼4 mS cm^–2)(Fig. 4). It is noteworthy that the posterior gills of Carcinus maenas (thought to play the dominant role in osmoregulatory NaCl uptake) and also the anterior gills (thought to be primarily responsible for gas exchange) are equally capable of active ammonia excretion (Weihrauch et al.,1999).

The finding that at least three different crab species with disparate ionic regulatory requirements can actively excrete ammonia raises issues about the metabolic costs involved. The ability to excrete ammonia against a gradient has significant ecological implications with regard to habitats that would,thus, be available to these crabs. Traditionally, ammonia excretion in aquatic animals has been considered to be a passive process driven by diffusion along the partial pressure gradient of NH₃(Pnh₃). Such a model requires environmental concentrations to be kept low, normally by bacterial nitrification of ammonia to nitrite and nitrate. This view is probably justified in pelagic animals colonizing the water column of aquatic habitats because, according to Koroleff(1983), the amounts of NH₄⁺ rarely exceed 5 μmol l^–1 in oxygenated, unpolluted seawater.

By contrast, benthic and interstitial animals are often faced with higher ambient ammonia concentrations. High ammonia is especially prevalent in anoxic, deep stagnant water and pore water during periods of high mineralization following collapse of phytoplankton blooms. For example,several investigations of pore water composition in the North Sea showed considerable concentrations of ammonia in a range between 100 and 300 μmol l^–1, but also up to 2500 μmol l^–1 in 4–9 cm sediment depth (Enoksson and Samuelsson, 1987; Lohse et al., 1993).

Like most aquatic crab species, Carcinus maenas, Cancer pagurusand Eriocheir sinenis are benthic-living animals, hiding under stones or burying themselves in the sediment for long periods, for example, during low tide or in the winter season. Under conditions where crabs are situated at sites with low rates of ambient water exchange, plus the fact that the animals produce and excrete metabolic ammonia, the concentration of the ambient ammonia can reach high values.

Considering hemolymph ammonia concentrations of ∼100 μmol l^–1 (Weihrauch et al.,1999; Table 1) of which less than 5 μmol l^–1 exist in the gaseous form NH₃, these crabs may encounter ambient NH₃ and/or NH₄⁺ concentrations exceeding those in their hemolymph. While NH₃ diffuses along its partial pressure gradient across the exposed epithelia, NH₄⁺ follows its electrochemical gradient by either paracellular diffusion or NH₄⁺permeable channels and transporters (see Table 2). An adaptive protection against net ammonia influxes (i.e. an active mechanism for excretion of metabolic ammonia against an inwardly directed gradient,tolerances for high hemolymph ammonia concentrations or efficient detoxification mechanisms) must, therefore, have evolved.

In the blue crab Callinectes sapidus, ammonia excretion rates are correlated with Na⁺ absorption(Pressley et al., 1981). The same result was obtained both for the Chinese crab Eriocheir sinensis(Péqueux and Gilles,1981) and for the shore crab Carcinus maenas(Lucu et al., 1989). Studies employing membrane vesicles from gill epithelia(Towle and Hølleland,1987) and isolated, perfused gills(Lucu et al., 1989) indicated that NH₄⁺ substitutes for K⁺ in activation of the ouabain-sensitive Na⁺/K⁺-ATPase. In gill sections from Callinectes sapidus, this Na⁺/K⁺-ATPase was demonstrated to be located in the basolateral membranes of the branchial epithelial cells (Towle and Kays,1986; Towle et al.,2001). Complete or partial cDNA sequences for the α-subunit of Na⁺/K⁺-ATPase from crab gills have been published in GenBank (see Table 2) thus confirming both its presence in branchial epithelia and its similarity toα-subunits of other species.

Recently, Masui et al.(2002) showed that the branchial Na⁺/K⁺-ATPase from Callinectes danaeis synergistically stimulated by NH₄⁺ and K⁺,increasing its catalytic activity by up to 90%. Masui et al.(2002) came to the conclusion that the two ions bind to different sites of the branchial Na⁺/K⁺-ATPase. This observation was also attributed to the branchial Na⁺/K⁺-ATPase of the freshwater shrimp Macobrachium olfersii by Furriel et al.(2004), who suggested for this species that at high NH₄⁺ concentrations the pump exposes a new binding site for NH₄⁺ which, after binding to NH₄⁺, modulates the activity of the Na⁺/K⁺-ATPase independently of K⁺ ions.

In the marine crab Cancer pagurus, active branchial excretion of ammonia is completely inhibited by ouabain, a specific inhibitor of the Na⁺/K⁺-ATPase(Weihrauch et al., 1999),suggesting this pump is the only driving force for excretion. However, in the gills of Carcinus maenas acclimated to brackish water, both gradient-driven (Lucu et al.,1989) and active ammonia excretion(Weihrauch et al., 1998) are only partially inhibited by ouabain, consistent with a second active mechanism responsible for branchial ammonia extrusion in this species.

The presence of an apically located amiloride-sensitive Na⁺/NH₄⁺ exchanger, transporting NH₄⁺ from the epithelial cell into the ambient medium in exchange for Na⁺, has been suggested for Callinectes sapidus (Pressley et al.,1981) and for Carcinus maenas(Lucu et al., 1989; Siebers et al., 1995). Indeed,branchial mRNA expression of a Na⁺/H⁺-antiporter,putatively transporting also NH₄⁺ ions, was demonstrated in Carcinus maenas (Towle et al.,1997) and in Eriocheir sinensis(Weihrauch and Towle, 2000). However, experiments employing the isolated cuticle from Carcinus maenas have shown that cuticular Na⁺ and NH₄⁺ conductances (G_cut) are inhibited by apically applied amiloride in a dose-dependent manner, with an inhibitor constant K_amiNa+=0.6 μmol l^–1 for sodium ions and K_amiNH4+=20.4 μmol l^–1 for ammonium ions, respectively(Onken and Riestenpatt, 2002; Weihrauch et al., 2002). Differences in K_amiNH4+ and K_amiNa+ are not understood yet. One can speculate that amiloride blocks the passage of cations in the cuticle in a mechanical way like a plug, rather than by blocking a general cation-binding side. Passage of smaller ions (like Na⁺) that carry a larger coat of water molecules is, therefore, possibly easier to block out by lower amiloride concentrations than K⁺ or NH₄⁺ ions, which carry only about half the number of water molecules around their core. However, according to these observations, some of the results obtained by applying amiloride to crab gills (apical) should be interpreted with caution and with special attention to the concentration of this particular inhibitor.

Further studies on the branchial ammonia excretion mechanism in Carcinus maenas employing the K⁺ channel blocker Cs⁺ (10 mmol l^–1) revealed that basolateral (but not apical) K⁺ channels play a role in the excretory process(Weihrauch et al., 1998). In addition, experiments inhibiting the branchial V-Type H⁺-ATPase by adding bafilomycin A₁ resulted in a reduction of active ammonia transport by 66%, identifying the H⁺-ATPase as the second active component in the excretory mechanism of the shore crab(Weihrauch et al., 2002). While in Eriocheir sinensis a V-Type H⁺-ATPase has been localized to the apical membrane of the gill epithelium(Onken and Putzenlechner,1995), in Carcinus maenas this pump was found predominantly in the cytoplasm, probably associated with vesicles(Weihrauch et al., 2001a). This latter finding led to the suggestion(Weihrauch et al., 2002) of a vesicular ammonia-trapping mechanism, in which cellular NH₃diffuses into acidified vesicles to be transformed into its membrane-impermeable ionic form, NH₄⁺. For a directed excretion, these NH₄⁺-loaded vesicles would then be transported to the apical membrane for exocytotic release. Such an excretion mechanism was supported by data showing total inhibition of active ammonia excretion by blockers of the microtubule network, including colchicine,thiabendazole and taxol (Weihrauch et al., 2002). The resulting hypothetical model of the ammonia excretion in Carcinus maenas is described in detail in Fig. 5.

For crabs (such as the partially limnic Chinese crab Eriocheir sinensis) that utilize a proton gradient across the apical membrane of the epithelial cell to accomplish NaCl uptake from highly diluted media, it is likely that NH₃ diffuses across the apical membrane along its partial pressure gradient, as shown in freshwater rainbow trout Oncorhynchus mykiss (Wilson et al., 1994).

Recently, Weihrauch and others (D. Weihrauch, unpublished data) have sequenced a full-length cDNA coding for a Rhesus-like protein from Carcinus maenas gills (GenBank Accession number: AF364404), named RhCM (Rhesus-related protein from Carcinus maenas). In mammals Rhesus-related proteins, such as RhGK, have been shown to mediate ammonia(NH₃/NH₄⁺), but not K⁺ or amino acid transport when functionally expressed in yeast mutants lacking endogenous ammonia transporters (triple ΔMep mutant; Marini et al., 2000). However,for this novel Rhesus-like ammonia transporter the detailed transport characteristics (such as mode of transport or kinetics) have not yet been defined. A comparison of the deduced secondary structure of the amino acid sequence of RhCM, and the human ammonia transporter RhGK, showed that 10 out of 12 predicted transmembrane domains are positioned at identical sites of the sequence (Fig. 6). The localization and role of RhCM in branchial ammonia excretion need to be investigated in detail in further studies. One can speculate that the putative ammonia transport of crabs is not localized in the apical membrane of the gill epithelium, because here ammonia (NH₃/NH₄⁺)permeable structures would be of disadvantage allowing ammonia influxes when the animals are exposed to high external concentrations. The human Rhesus-like ammonia transporter RhGK (identical to RhCG), has been described to be localized in the distal tubule and the collecting duct of the kidney in co-localization with a V-type H⁺-ATPase(Eladari et a., 2002). RhCG expressed in Xenopus oocytes facilitates a highly specific NH₃ diffusion via a complex electrogenic NH₄⁺ transport(Bakouh et al., 2004). In addition, Eladari et al.(2002) suggested a secondary active mode of ammonia transport in the distal tubule by acid trapping. According to this assumption, RhGK would promote the transmembrane passage of NH₃. A similar mechanism would be plausible in the gills of the shore crab Carcinus maenas, however, RhCM would be co-localized with the H⁺-ATPase within the membranes of intracellular vesicles to support the proposed vesicular acid-trapping mechanisms. Also, a basolateral localization cannot be excluded, where the putative ammonia transporter might serve as an overflow valve, transporting ammonia back in to the hemolymph,when crabs are exposed to high external ammonia concentrations. Under this condition, intracellular ammonia concentrations might rise to toxic levels due to a passive influx from the apical side, while the Na⁺/K⁺-ATPase is actively pumping NH₄⁺ from the hemolymph space into the cytoplasm. Ammonia directed back into the hemolymph, via RhCM, could probably be buffered, at least for a short term, by incorporation into proteins, for instance glutamine or hemocyanins.

The high degree of conservation with ammonia transporters found in fungi,bacteria and archaebacteria (∼20%), as well as the striking homology to mammalian ammonia transporters (>40%), led to the suggestion that proteins of the Rh-family play a universal role in ammonia transport(Fig. 7).

The mechanisms and processes by which air-breathing crustaceans excrete nitrogenous waste into the terrestrial habitat have been subject to considerable scrutiny (Greenaway, 1988, 1991; Wolcott, 1991, 1992; O'Donnell and Wright, 1995). Terrestrial crabs appear to tolerate considerably greater hemolymph ammonia loads than do aquatic species (Table 1). Conversely, Wolcott(1992) points out that diluting ammonia to non-toxic levels in the urine might require an unsustainable water loss in a land crab.

The probability that branchial NH₄⁺ excretion is linked to sodium transport, both through apical ion exchangers and the basal membrane Na⁺/K⁺-ATPase, is of special importance to air-breathing crabs. In all land crabs examined to date, the gills have become adapted for reabsorption of salt from primary urine directed through the branchial chamber (Wolcott and Wolcott, 1985, 1991; Morris, 2001), allowing diffusive NH₃ loss and NH₄⁺ extrusion in exchange for required ions from the urine. Ocypodid crabs seem exceptional in utilizing the antennal gland for increasing urinary ammonia, although the gills are required to complete the excretory process(DeVries and Wolcott, 1993). Generally, while the gills are initially bathed with a fluid isosmotic with the hemolymph, osmotic concentration may decline by as much as 90%(Wolcott and Wolcott, 1985; Varley and Greenaway, 1994; Greenaway, 1999; Morris et al., 2000; Taylor and Greenaway, 2002; Morris and Ahern, 2003). Even very euryhaline aquatic species do not experience the same range of osmo-concentration as occurs in the extra-branchial fluid of some land crabs.

In terrestrial arthropods as a whole, the primary nitrogenous excretory products are generally purines, whereas in land crabs various mechanisms are employed to permit the continued excretion of ammonia. The single known exception is the terrestrial anomuran Birgus latro, which is purinotelic excreting urate (Greenaway and Morris, 1989) and guanine(Greenaway, 2003). The reasons for the general persistence of ammonotely may be found by examining a continuum of extant species in the transition from aquatic to land crab.

Exposure to air of the aquatic crabs Cancer pagurus and Cancer productus caused hemolymph ammonia to increase by 25 μmol h^–1 (Regnault,1992) and 26 μmol h^–1(deFur and McMahon, 1984),respectively. However, in Cancer pagurus this rate of accumulation was between 15 and 30% of the rate expected by Regnault(1992) on the basis of basal aquatic rates, leading to the suggestion of nitrogen storage in tissues, to avoid toxic hemolymph ammonia loading. Many crustaceans store nitrogen (N) as solid urate (for review see Greenaway,1999) but this seems to be formed primarily as a result of diet rather than being of any large significance in NH₃/NH₄⁺ detoxification(Linton and Greenaway, 1997a)– except possibly under desiccating conditions in land crabs (below). In any case, urate formation was ruled out as a significant contribution in Cancer pagurus (Regnault,1992). Ammonia excretion during air exposure of Cancer pagurus was only 4% of the normal aquatic rate (170–190 μmol kg^–1 h^–1) but when re-immersed they exhibited a very large (50-fold within 5 min) but transient increase to 8860μmol kg^–1 h^–1 (derived from Regnault 1994). Thus, nitrogen storage as NH₄⁺, or as some readily oxidized form, is apparently a normal response to transient air exposure, as is the subsequent pulsatile clearance of ammonia on re-immersion. The possibility that urate can be so rapidly mobilized to NH₄⁺ seems unlikely and expensive. However, the vesicular sequestration of NH₄⁺(see above) has hitherto not been considered in this role.

A similar `storage–excretion' is seen in diverse air-breathing crabs(e.g. Potamonautes warreni, Morris and van Aardt, 1998; Austrothelphusa transversa,Linton and Greenaway, 1995; Discoplax hirtipes, Dela-Cruz and Morris, 1997; Cardisoma carnifex,Wood et al., 1986). [Note: Cardisoma hirtipes has been revised to Discoplax hirtipesand Holthuisana to Austrothelphusa(Davie, 2002).] Discoplax hirtipes excretes 99% of its waste as ammonia, but when it is breathing air the rate of nitrogen loss in the urinary flow is only 0.2 μmol kg^–1 h^–1 and NH₃ is volatilized at a very slow rate (0.4 μmol kg^–1 h^–1). On re-immersion, the ammonia excretion rate is transiently elevated to 1100μmol kg^–1 h^–1 (compared with the normal 300 μmol kg^–1 h^–1). Potamonautes warreni also does not excrete while in air but, on return to water,excretes ammonia across the gills at 4900 μmol kg^–1h^–1 (compared with the normal rate in water of 70 μmol kg^–1 h^–1). Artificially irrigating the gills of air-breathing Potamonautes warreni sustained ammonia excretion(Morris and van Aardt, 1998). Gill irrigation appears to be a ubiquitous activity following excursions into the terrestrial environment (Dela-Cruz and Morris, 1997). The requirement to re-immerse, albeit briefly, to accomplish ammonia excretion via the ancestral branchial mechanisms may ultimately limit the duration of air-breathing in these amphibious species. However, Austrothelphusa transversa can spend many months without access to water (Greenaway and MacMillan, 1978) and can forgo ammonia excretion during that time(Linton and Greenaway, 1995). Linton and Greenaway (1995)suggested that the near-cessation of nitrogen excretion in A. transversa implied reduced nitrogen catabolism and temporary nitrogen storage. The speed and brevity of the excretion pulse in P. warrenishowed that wastes stored during terrestrial forays are rapidly excreted on return to water. However, it seems unlikely that this store is accumulated as NH₄⁺/NH₃ within the gill epithelium because hemolymph levels remain low and pH (and therefore Pnh₃) remains unchanged(Adamczewska et al., 1997)although the required enzymes may be present(Linton and Greenaway, 1998). Further study is required to determine the storage product but an accessible intermediate is implicated (such as glutamine rather than, for example,urate). Apical H⁺/NH₄⁺ exchange and V-ATPase-driven cell alkalization have been suggested as likely mechanisms of transbranchial ammonia transport (Linton and Greenaway, 1995), but experimental evidence is required to confirm this. Again, the involvement of a Rhesus-related ammonia transporter needs to be evaluated.

Gecarcinid land crabs recycle their urine over the branchial surfaces,producing a dilute fluid `P' (Wolcott and Wolcott, 1985; for review Morris, 2002). Discoplax hirtipes, a crab that immerses from time to time, can reduce the NaCl concentration of the urine by 90%(Dela-Cruz and Morris, 1997)but this ion pumping does not allow ammonia excretion while in air. For example, while the NH<sub>4</sub><sup>+</sup> content of `P' of Discoplax hirtipes is significantly elevated (5 mmol l^–1) compared with the hemolymph (Table 1),the rates of urinary and `P' flow slow down to almost zero when the animals are in air (Dela-Cruz and Morris,1997). Thus, ammonia excretion becomes severely limited by urine flow and consequent `P' production rates.

In Gecarcoidea natalis, a gercarcinid land crab that does not routinely immerse or have access to pools of water, the primary urine contains 0.36 mmol NH₄⁺ l^–1, which is less than in the blood (Table 1). However, reprocessed `P' contains up to 10.8 mmol l<sup>–1</sup>(Greenaway and Nakamura,1991), and this is sufficient to excrete up to 68% of the total nitrogenous output because `P' production was ∼450 μl kg^–1 h^–1. This flow rate is much greater than the 3 μl kg^–1 h^–1 in Discoplax hirtipes, which is unable to sustain ammonia excretion in air(Dela-Cruz and Morris, 1997). The rate of `P' production in Gecarcinus lateralis was >900 μl kg<sup>–1</sup> h<sup>–1</sup>, which facilitated an excretion rate of 20 μmol kg<sup>–1</sup> h<sup>–1</sup>(Wolcott, 1991), compared with the 25 μmol kg<sup>–1</sup> h<sup>–1</sup> rate in Gecarcoidea natalis (Greenaway and Nakamura, 1991). These authors(Greenaway and Nakamura, 1991; Wolcott, 1991) concluded that acid trapping in the `P' was not involved and, thus, the outward gradient across the gill epithelia is not favorable for gaseous NH₃diffusion. In addition, Wolcott(1991) measured the urine pH of Gecarcinus lateralis and Cardisoma guanhumi, and in both crabs found it to be greater than that of the hemolymph. However, branchial Na⁺/H⁺ exchange would assist in NH₃ diffusion through reciprocal pH changes of intracellular and extra-corporeal fluids. The gills of Gecarcoidea natalis are highly active in Na⁺transport and NH₄⁺ might easily substitute for K⁺ in the basal Na⁺/K⁺-ATPase(Morris, 2001; Morris and Ahern, 2003) but the necessary active transport of NH₄⁺ across the apical membrane into the `P' remains unresolved. A net exchange of Na⁺ for NH₄⁺ (e.g. Pressley et al., 1981) would facilitate salt reclamation and nitrogen excretion, but would be hampered as the external Na⁺ declined. The possibility of exocytotic mechanisms and/or involvement of a Rhesus-related ammonia transporter needs to be evaluated in the excretory processes of these species.

While this branchial system allows routine excretion of ammonia to air, it also seems to make NH₄⁺ excretion dependent on the urinary flow rate, as well as on the extent of ion re-absorption. For example,in Gecarcoidea natalis, urine and `P' flow can decline to zero under dry season conditions (Morris and Ahern,2003) and so this mode of NH₄⁺ clearance becomes inoperable. In Cardisoma guanhumi the fluid retained within the abdominal flap (∼13.5 mmol l^–1) is contiguous with that in the branchial chamber (∼6.5 mmol l^–1) and further NH₄⁺ excretion may occur via unknown mechanisms(Wolcott, 1991) but, even so,this would be unavailable to land crabs in the dry season. Purine is stored in large amounts in connective tissue cells throughout the bodies of some land crabs (Linton and Greenaway,1997b). In Gecarcoidea natalis this stored purine is normally synthesized de novo, from excess dietary nitrogen(Linton and Greenaway, 1997a). Recent data, including enzyme activities and nitrogen utilization (Linton and Greenaway, 1998, 2000) have lead to the suggestion of a storage–excretion function for the urate accumulated by G. natalis (Greenaway,2003). However, this seems likely to be infrequently called upon(Linton and Greenaway, 1997a)and evidence is required to show that waste amino N is incorporated as well as dietary N.

Other lineages of terrestrial crabs have not been investigated to the same extent as the gecarcinids, but at least one air-breathing ocypodid, Ocypode quadrata, has been shown to sustain ammonia excretion(DeVries and Wolcott, 1993). O. quadrata also recycles the urine to produce `P', which can be as little as 10% of the osmotic strength of the primary urine(Wolcott and Wolcott, 1985). However, the mechanism of NH<sub>4</sub><sup>+</sup> excretion is quite different from that of gecarcinids because the concentration in the primary urine of the ghost crab is extraordinarily high. For example, in O. quadrata (DeVries and Wolcott,1993; DeVries et al.,1994) this reaches 116–212 mmol l<sup>–1</sup>, and in O. ceratopthalma and O. cordimanus (under field conditions)>40 and 27 mmol l<sup>–1</sup>, respectively (S. Morris,unpublished). The primary urine of Ocypode quadrata is unusually acidic (pH 5.36±0.21), providing an `acid-trap' for NH₄⁺. On passage over the gills, the pH is increased (pH 7.01±0.24) and Cl^– (but not Na⁺) is reclaimed, such that the alkalinization of the fluid promotes significant NH₃ volatilization (∼71 μl kg^–1h^–1 in control crabs). While pH 7 is not alkaline, the increase in pH is quite effective. For example, at an ammonia concentration of 116 mmol l^–1 and pH 5.4 for primary urine(DeVries & Wolcott, 1993),if the pH is increased to pH 7 it is possible to estimate Pnh₃ using the pK and solubility for NH₃ provided by Kormanik and Cameron(1981) as used by Varley and Greenaway (1994). In the primary urine Pnh₃=11.6 Pa whereas at pH 7 the Pnh₃=460 Pa, which is a 40-fold increase in potential diffusive gradient. In view of the concomitant increase in the fluid CO₂ concentration and the uptake of Cl^–, the most obvious candidate for the net base excretion is transport by an apical HCO₃^–/Cl^– exchanger(DeVries & Wolcott, 1993). Reclamation of urinary Na⁺ appears to be accomplished within the antennal gland (DeVries et al.,1994). These authors (DeVries et al., 1994) report high activity of Na⁺/K⁺ATPase in the antennal gland of O. quadrata for which NH₄⁺ may substitute for K⁺ in the basal membrane exchange. Furthermore, apical Na⁺/H⁺ antiporters in the antennal gland may sustain both Na⁺ reclamation and acidification of urine to promote NH₄⁺-trapping (Fig. 8).

Study of the more-terrestrial grapsid, Geograpsus grayi, has revealed further modifications of the NH₃/NH₄⁺ excretory system(Greenaway and Nakamura, 1991; Varley and Greenaway, 1994). G. grayi is a highly active carnivorous land crab and also reprocesses the urine to reclaim salts via branchial uptake(Greenaway and Nakamura, 1991)but, unlike Ocypode sp., does not employ ion reclamation within the antennal gland (Varley and Greenaway,1994). Clearly Ocypode and Geograpsus represent separate radiations in to the terrestrial habitat, but with superficially analogous ammonia excretion physiology. G. grayi volatilizes NH₃ from the limited volume of `P' within the branchial chamber and, thereby, increases the effective NH<sub>4</sub><sup>+</sup> capacity of the fluid, which may achieve concentrations in excess of 80 mmol l<sup>–1</sup> compared with <1 mmol l<sup>–1</sup> in the urine (Varley and Greenaway,1994). However, G. grayi manages a rate of ammonia excretion comparable to that of aquatic crabs in water (107–220 μmol kg<sup>–1</sup> h<sup>–1</sup>; Greenaway and Nakamura, 1991; Varley and Greenaway, 1994). Gaseous ammonia contributes ∼78% of this total excretion in a discontinuous process over 3 h to 3 days(Varley and Greenaway, 1994)although urine flow is apparently limited, restricting fluid available for `P'formation. The pH of this fluid (pH 8.07) is higher than that of the hemolymph(pH 7.66–7.59) and at the same time the CO₂ content (36 mmol l^–1) is considerably greater than that of the hemolymph(13.7–17.2 mmol l^–1). Amiloride reduced NH₄⁺ efflux by 83% in this system and reduced unidirectional Na⁺ uptake. Thus, NH₃ volatilization is achieved by raising the fluid pH towards the pK, such that gaseous NH₃ becomes 8% of the total ammonia, creating a diffusive gradient(Pnh₃ ∼3.3 Pa) into the convective air stream. Varley and Greenaway(1994) discussed the difficulties in transporting NH₃/NH₄⁺ outward into this fluid in the absence of `acid-trapping' and proposed a net excretion reaction NH<sub>4</sub><sup>+</sup>+HCO<sub>3</sub><sup>–</sup>→H<sub>2</sub>O+CO<sub>2</sub>↑+NH<sub>3</sub>↑. The volatilization of NH<sub>3</sub> and CO<sub>2</sub>, together with formation of water, all contribute to lowered ionic strength in the extrabranchial fluid. Again, apical NH<sub>4</sub><sup>+</sup> transport,possibly mediated by a putative Rhesus-related ammonia transporter protein,needs to be investigated in further studies. This mechanism in G. grayi effectively increases the functional volume of the `P' offering some escape from the dependency on regular and significant urine flow rate. Thus, there are significant increases in the amount of ammonia excreted per unit volume of `P', while retaining the advantages of ammonia excretion,thereby allowing a more terrestrial habit(Fig. 9). At the same time, the system remains potentially limited by the supply of Na⁺ and Cl^– in a lowered supply of urine.

The most successful terrestrial animals have abandoned NH₃ as an N-excretory vehicle in favor of urea or purines. The anomuran Birgus latro is the only identified purinotelic crab(Greenaway and Morris, 1989)but is sympatric with several species of gecarcinids that retain NH₃ excretion. Excreting purines allows greater flexibility of urinary water flow independently of urine reprocessing and salt reclamation,but significant energetic advantages may accompany ammonotely.

Decapod crustaceans exhibit a wide variety of ammonia excretion mechanisms and consequently provide good models for general investigation of nitrogen excretion. The debate as to whether ammonia is lost viaNH₃ diffusion or by NH₄⁺ transport remains active, but the answer may be both or either depending on circumstance and species. The ability to move ammonia against its gradient is obviously essential. There is a clear continuum of increased terrestriality accompanied by managed and active excretion with lowered water loss. This continuum represents multiple transitions onto land and is underpinned by phylogenetic differences. The increased application of molecular and post-genomic methodologies to the question will reveal, for example, the role of Rhesus-related proteins and vesicular transport systems in the physical extrusion of ammonia.

The authors' research is supported by the National Science Foundation(IBN-9807539 to D.W.T. and D.W.), the National Center for Research Resources(D.W.T.), and the Maine Science and Technology Foundation (D.W.T. and D.W.).