A Comparison of the Respiratory Function of the Haemocyanins of Vertically Migrating and Non-Migrating Pelagic, Deep-Sea Oplophorid Shrimps

The vast pelagic realm of the world ocean is inhabited by a wide variety of animals that are different in lifestyles and physiological adaptations from those that live on the bottom in shallow environments. These animals must be adapted to maintain their vertical position in the water column either by actively swimming or by neutral buoyancy. In addition, a large fraction [off Oahu 47% of total micronekton and 60 % of crustaceans between 400 and 1200 m (Maynard et al. 1975)] of the pelagic animal biomass in the ocean undertakes extensive diurnal vertical migrations. These animals spend the daytime in the cold and often low-O₂ conditions that prevail at depths of several hundred metres and then swim up to. the warmer high-O₂ environment near the surface at night to feed, only to return to depth at dawn (Marshall, 1979). In tropical regions the differences between the day and night habitats can be extreme, with temperature differences of more than 20°C and O₂ ranging from less than 0.67 kPa at depth during the day to more than 17.50 kPa near the surface at night. The adaptations required for this environment are unique because these animals not only go back and forth between these environmental extremes on a daily basis, but at the same time they must remain continuously active to maintain their position in the water column. In addition to the vertical migrators there are also species which remain continuously at depth, sharing their habitat with the migrators during the day, and they too must remain active to maintain their position in the water column.

Although this environment makes unusual physiological demands on its inhabitants, few data are available on the physiology of these animals. In general, the vertically migrating crustaceans are more robust and more active than the non-migrators and have higher metabolic rates than the migrators, even when measured at the same temperatures (Teal and Carey, 1967; Childress and Nygaard, 1974; Childress, 1975; Cowles, 1987). The midwater crustaceans also generally have little anaerobic capacity, but are able to regulate their oxygen consumption down to the lowest to which they are exposed, which can be less than 0.67kPa in some cases (Childress, 1968, 1971,1975; Cowles, 1987). Although these crustaceans have great abilities to remove O₂ from water and transport it to their tissues, there is virtually no information on the properties of that most critical molecule in this process, haemocyanin.

One would expect that such active, aerobic crustaceans living at a range of temperatures and often at low would require very specific adaptations of the O₂-binding properties of haemocyanin. The effect of temperature on the oxygen affinity of crustacean haemocyanin is well known (for reviews see Mangum, 1980, 1983; Table 3 of Bridges, 1986), and differences in the oxygen-binding characteristics of haemocyanins have been found between closely related species which inhabit different thermal environments (Burnett et al. 1988; Sanders et al. 1988; Morris and Bridges, 1989). In crustaceans living in shallow waters, oxygen affinity of haemocyanin is generally lower in species from colder latitudes (Mangum, 1983). However, the few reports of oxygen affinity for haemocyanins from deep sea crustaceans which occupy cool (<6°C), thermally stable (<0.2°C range of interannual variation), low-oxygen (continuously <3.32kPa) environments (Freel, 1978; Arp and Childress, 1985; Sanders and Childress, 1988a; Sanders, 1989) have shown that these species have haemocyanins with relatively high O₂ affinities.

The study of the properties of the haemocyanins of vertically migrating and non-migrating crustaceans provides an opportunity to elucidate some aspects of the temperature adaptations of crustacean haemocyanins by allowing one to compare the properties of haemocyanins from animals which share the same environment during the day but have very different environments at night. Such studies can also give insight into haemocyanin adaptations to regularly fluctuating temperatures. The potential effects of temperature, pH and L-lactate on oxygen binding by haemocyanin are of particular interest in the case of vertical migrators because they encounter a wide temperature range, a wide O₂ range and are almost certainly more active when near the surface. Because of the large temperature and O₂ concentration differences between the day and night depths of vertically migratory shrimps off the Hawaiian Islands, we chose to study this problem off Oahu.

Several species of the exclusively pelagic caridean family Oplophoridae, including genera having both migrating and non-migrating species, five at 500–1000 m depth off Oahu during the day. The non-migrators include Acanthephyra curtirostris Wood-Mason and A. acutifrons Bate, which live continuously below 500 m. The vertical migrators include A. smithi Kemp, Systellaspis debilis A. Milne-Edwards and Oplophorus gracilirostris A. Milne-Edwards, which live below 500 m by day, but at night swim up to depths of 50–300 m (Ziemann, 1975). During daytime, these five species share depths where oxygen partial pressures are between 2.66 and 4.00kPa, and temperatures range from 3 to 6°C. At night, the three vertical migrators move to waters with oxygen partial pressures above 13.33kPa and temperatures above 20°C. These vertical migrators have higher metabolic rates at the higher temperatures near the surface and, at the same temperatures, have higher metabolic rates than the non-migrating pelagic crus-taceans (Cowles, 1987). All five of these species are active animals which are denser than sea water and must therefore swim continuously to maintain their position in the water column. They are also able to regulate their oxygen consumption down to the lowest environmental which they encounter (2.66kPa O₂ at 700m depth, Cowles, 1987), but have very limited anaerobic abilities. The functional properties of haemocyanins which might support these respiratory characteristics have not been previously studied.

The effects of temperature and pH on haemocyanin oxygen-binding in five species of vertically migrating and non-migrating oplophorid shrimp were examined, as were the effects of L-lactate in three of these species, to determine the role of the respiratory proteins in enabling these species to support their metabolic oxygen demands in the deep-sea and epipelagic environments they encounter.

Oplophorid shrimps were collected in the summer months of 1986 and 1987 off the leeward side of the Hawaiian island of Oahu (21°15–37′N, 158°15–38′W) from depths of 200–1200 m. Animals were captured in a modified opening-closing Tucker trawl (3.1m square mouth) and were brought to the surface in a thermally insulated cod end (Childress et al. 1978) which kept the samples near the depth temperature. Shrimps were kept alive on board ship in individual 11 containers (5.5°C) until used in experiments.

For pH measurements, haemolymph samples were removed via hypodermic syringe from the ventral sinus of individual shrimps. Without air exposure, the sample was immediately injected into a Radiometer glass capillary electrode (Radiometer America G298A) in a water-jacketed chamber held at the experimental temperature, in conjunction with a reference electrode (Radiometer K171). Precision buffers were used to calibrate the electrode (Radiometer S1500 and S1510). Samples of haemolymph for oxygen equilibrium curves were either dialyzed for immediate use or kept frozen at −80 °C for 12–15 months for later use in experiments.

For each species, the absorbance maximum near 340 nm was measured in a 1:100 dilution of haemolymph in 50 mmol 1⁻¹ Tris buffer (pH 8.9) with EDTA (50 mmol I⁻¹). Haemocyanin concentration was calculated using an extinction coefficient of (Nickerson and Van Holde, 1971). The oxygen-carrying capacity was estimated from the haemocyanin concentration, assuming a subunit mass of 75 000 Da.

The concentrations of the major inorganic solutes of the haemolymph (Na⁺, K⁺, Ca²⁺, Mg²⁺, SO₄²⁻, Cl⁻ and NH₄⁺) as well as trimethylamine were measured with single-column ion chromatography (Sanders and Childress, 1988b), using Wescan cation and anion columns with a conductivity detector. L-Lactate concentrations in fresh, native haemolymph were measured with a Sigma L-lactate test kit (kit no. 826-UV, Sigma Chemical Co.), using the modifications suggested by Graham et al. (1983).

To examine the short-term effects of temperature changes during vertical migration on haemolymph pH in vivo, specimens of Acanthephyra smithi (4–6g animals) and Oplophorus gracilirostris (3–5 g animals) were placed, five per container, in 41 jars of aerated sea water at 5, 10, 15, 20 or 25°C. After 4h, the haemolymph pH of each animal was measured. When a large enough haemolymph sample could be obtained, the concentration of L-lactate in the haemolymph of these individuals was also measured. These animals had been maintained for 24–48 h at 5.5°C prior to their use, to allow them to recover from the stress or capture in a trawl. Because these animals encounter all the experimental temperatures during the course of their migrations each day and we wished to simulate the effect of vertical migration, we did not consider acclimation at each temperature necessary or desirable for these studies.

To determine the effects of pH and temperature on oxygen binding by haemocyanin, haemolymph samples were used in a thin-layer spectrophotometric system (Childress et al. 1984; Sanders et al. 1988). A small sample of haemolymph (5–20 μl) was sandwiched between two layers of 0.006 mm Teflon membrane, and the absorbance at 347 nm of gas-equilibrated samples was recorded with a Tracor Northern diode array spectrophotometer at successively higher oxygen partial pressures. Gas partial pressures (O₂ and N₂) were controlled with a Union Carbide mass flow controller. Oxygen concentrations within the gas-tight sample chamber were monitored with a Systech Instruments zirconium cell oxygen analyzer (Systech model ZR891). The calibration of the oxygen analyzer was verified with 99.99% oxygen gas, air and 99.998% nitrogen gas. A second sample of haemolymph from the same sample was maintained in a gas-tight, water-jacketed tonometer at the same temperature and gas concentrations as the oxygen equilibrium curve sample; the pH of this haemolymph sample was measured near the 50 % saturation point.

Samples used to determine the effects of temperature and pH on oxygen binding by haemocyanin were dialyzed in a buffered, physiological saline prepared from the inorganic ion concentrations measured in the haemolymph of each species. The use of buffered, dialyzed samples removed the possibility of variation between haemolymph samples caused by varying levels of modulators of haemocyanin function and provided control of pH during the determination of O₂ equilibria. A separate dialysis was carried out for each O₂ equilibrium determination. Possible effects of Tris buffer on the haemocyanin properties were evaluated by comparing the log P₅₀-pH relationship of frozen, dialyzed, Tris-buffered A. smithi haemolymph at 5.0°C with the same relationship determined on frozen, dialyzed A. smithi haemolymph whose pH was controlled by varying the in the equilibration gases (method of Sick and Gersonde, 1969, carried out with help of S. Morris).

The salines were prepared from the formulae given (see Table 3) and then titrated with NaOH to the desired pH. Although ammonia and/or trimethylamine were present at low concentrations in the haemolymphs of some oplophorid species, these ions were not used in dialysis media. Since these ions are volatile, their concentrations cannot be maintained in a haemolymph sample used in oxygen-binding experiments, unless the gas mixtures contain ammonia and trimethylamine gases. Although there are few data that show an effect of ammonia or trimethylamine on oxygen binding by haemocyanin (Sanders, 1989), the absence of these ions in dialyzed samples may induce functional differences in haemocyanins. Because of the low concentrations found in these species, these effects are likely to be small in the species studied here.

Samples were dialyzed for 15–18 h in three changes of saline buffered with 0.05 mol I⁻¹ Tris at a ratio of 1000 parts saline to one part sample. Each pH required a separate dialysis with the saline adjusted to the desired pH with NaOH. Since the salines were acidified with a fixed amount of HC1 (see Table 3) and then titrated to the desired pH with NaOH, [Cl⁻] was the same at all pH values. The effects of L-lactate on haemocyanin oxygen-binding were determined by adding 10 μl of 15 or 150 mmol 1⁻¹ L-lactate in saline to dialyzed haemolymph samples and completing spectral analyses, as described above to generate oxygen equilibrium curves. The lactate concentrations in these dialyzed samples were measured after each experiment using a Boehringer L-lactate test kit. Haemolymph samples used in L-lactate experiments had been stored frozen at −80 °C for 12–15 months prior to use.

Results are reported as mean and standard deviation unless otherwise noted. Linear regression analysis was used to fit the data in the figures. Values of n₅₀ were taken from the equations describing the Hill plots between 25 % and 75 % saturation. Analysis of covariance (ANCOVA) was used to test the significance of differences in elevation of regression lines having slopes which were not significantly different. Oxygen equilibrium curves at particular environmental temperatures were constructed using interpolated (for appropriate values of pH) values of P₅₀ and n₅₀. Some custom software provided by S. Morris was used for these calculations.

The concentration of haemocyanin, estimated oxygen-carrying capacity of the haemocyanin, mean haemolymph pH and L-lactate concentrations in samples of haemolymph drawn from animals maintained alive on board ship at 5.5°C for between 24 and 48 h prior to sampling are shown in Table 1. In all cases the animals were swimming in an apparently unexcited manner prior to sampling and the sampling was executed quickly. Higher haemocyanin concentrations were found in the haemolymphs of vertical migrators than in non-migrators (P=0.083, Mann-Whitney U-test). Lactate concentrations varied considerably among the species, but for all except A. smithi showed little variation within species.

The concentrations of the major inorganic ions in the haemolymph of each of the five species are reported in Table 2. In addition to the usually detected ions, NH4⁺ and trimethylamine were detected in all species. Trimethylamine did not exceed the trace level (about 0.1 mmol 1⁻¹) in any species. NH₄⁺ was present at trace levels in two species of vertical migrators (A. smithi and O. gracilirostris) and at higher concentrations in the other three species. The measurements in Table 2 were used to prepare dialysis media having the ion concentrations reported in Table 3 but without NH₄⁺ or trimethylamine.

Results from experiments to determine in vivo haemolymph pH in Acanthephyra smithi and Oplophorus gracilirostris at different experimental temperatures are shown in Fig. 1. The coefficients of the regressions of haemolymph pH as a function of temperature were similar to the slope of the neutral pH of water (Reeves, 1977) and within the range reported by other authors for crustaceans (McMahon and Burggren, 1981; Morris et al. 1985, 1988; Morris and Bridges, 1989).

Concentrations of lactate were low and varied little in O. gracilirostris haemolymph (range 1.01–1.70 mmoll⁻¹, mean 1.29±0.26mmoll⁻¹, N=6, at least one data point from each temperature), showing no significant variation with temperature in these experiments. This pattern is in agreement with the data from animals kept a longer time at 5.5°C (Table 1). In contrast, the haemolymph L-lactate concentrations in A. smithi were relatively high and variable (range 1.39–9.34 mmol 1⁻¹, mean 4.25±2.56mmoll⁻¹, N=17, at least one data point from each temperature). These data are comparable to those presented for individuals of this species kept at 5.5°C for a longer period (Table 1) and showed no significant relationship between temperature and L-lactate. Although the shrimps were swimming in an unexcited fashion before sampling and haemolymph samples were taken quickly from each animal (within seconds of removal from the experimental chamber), handling stress might have played some role in producing the high levels of L-lactate in A. smithi. However, given the similar manner in which the other four species behaved and were treated and the lower, less-variable L-lactate concentrations found in these species (Table 1), this seems unlikely to be the explanation. These data, along with those in Table 1, demonstrate that these species produce different amounts of L-lactate under similar conditions.

The effects of pH and temperature on oxygen binding by haemocyanin in dialyzed haemolymph samples (never frozen) are shown in Fig. 2 for vertical migrators and non-migrators. Moderate Bohr effects (ΔlogP₅₀/ΔpH) were found for vertical migrators at all experimental temperatures (Fig. 2). Analysis of covariance confirmed the significance of the large decreases in oxygen affinities of the haemocyanins of vertical migrators at higher temperatures (Fig. 2, P ⩽0.005 for all three species, comparing values from 5 to 15°C, 15 to 25°C, 5 to 25°C). Oxygen affinities of haemocyanins in dialyzed haemolymph samples from nonmigrators were also moderately affected by changes in pH, but the decreases in oxygen affinities of haemocyanins at higher temperatures were smaller and not highly significant (Fig. 2; A. acutifrons, P₅₀ values from 5 to 25°C and A. curtirostris, P₅₀ values from 5 to 15°C, P⩽0.05, ANCOVA; Acanthephyra acutifrons, comparing P₅₀ values from 5 to 15°C, 15 to 25°C and A. curtirostris from 5 to 25°C, P⩽0.10, ANCOVA). Cooperativity of oxygen binding by haemocyanin in dialyzed haemolymph was not significantly affected by temperature (P⩾0.25, ANCOVA) for vertical migrators or non-migrators (Fig. 2), except for a slight increase in cooperativity with temperature found for Oplophorus gracilirostris at 5°C (P<0.01, ANCOVA).

The effects of temperature on oxygen affinity of haemocyanins from vertical migrators and non-migrators, at three constant pH values, were analyzed by van’t Hoff plots (Fig. 3). Relative to other crustacean species (see Mangum, 1983; Table 3 of Bridges, 1986), the ΔH values (kJ mol⁻¹) at a constant pH (Fig. 3) are moderate for vertical migrators and low for non-migrators.

Comparisons of the oxygen-binding properties of haemocyanin in ‘fresh’ (never frozen) dialyzed haemolymph, and haemolymph which had been dialyzed after being frozen and thawed once (frozen at −80°C for 12 –14 months) are shown in Fig. 4 for Acanthephyra smithi, A. acutifrons and Oplophorus gracilirostris. For A. smithi, ANCOVA showed no significant differences in cooperativity (n₅₀ values) and affinity values) of oxygen binding (P>0.25, 5 and 25°C). Significant decreases in haemocyanin cooperativity (P<0.005, ANCOVA) and increases in haemocyanin oxygen-affinity (P<0.005, ANCOVA) were found for frozen haemolymph from A. acutifrons and O. gracilirostris (5 and 25°C) (Fig. 4).

The effects of Tris buffer and CO₂ on the O₂ affinity of the haemocyanin of A. smithi were evaluated by comparing the log P₅₀-pH relationship generated at 5 °C using Tris-buffered, previously frozen samples with no CO₂ present (Fig. 4) with one generated using previously frozen samples buffered with 0.002 mol 1⁻¹ bicarbonate whose pH was controlled by varying (Fig. 4). Both data sets had six points at pH values ranging from 7.1 to 8.1. ANCOVA showed that they did not differ significantly in slope (P>0.25) or elevation (P=0.73). Thus, neither Tris nor CO₂ have apparent specific effects (independent of pH) on the affinity of this haemocyanin for oxygen.

The effects of L-lactate on haemocyanin oxygen-affinity were measured in frozen, dialyzed haemolymph of Acanthephyra smithi, Oplophorus gracilirostris and A. acutifrons (Fig. 5). The oxygen affinity of the haemocyanin was increased slightly in A. acutifrons (Δlog P₅₀/Δlog[lactate] =−0.08 at pH7.4 and −0.05 at pH 7.9) and O. gracilirostris haemocyanin (ΔlogP5o/Δlog[lactate] = −0.03 at pH7.4 and 7.9, 5°C). These increases in affinity were, however, significant only at high concentrations of lactate for both species (P<0.005, ANCOVA). Lactate had a moderate, but significant (P<0.005, ANCOVA), effect on the oxygen affinity of the haemocyanin of A. smithi at 5 °C, and a smaller, but significant (P<0.01, ANCOVA), effect at 25°C (Δlog P₅₀/ Δlog[lactate] = −0.11 at pH7.4 and −0.17 at pH7.9, 5°C; −0.04 at pH7.4 and 7.9, 25°C). The cooperativity of oxygen binding by haemocyanin was not significantly altered by the presence of L-lactate (P>0.25, ANCOVA) in any of the three oplophorid species examined (Fig. 5).

It is now well established that freezing haemolymph samples prior to their use in oxygen-binding experiments can result in changes in haemocyanin cooperativity (Bridges et al. 1984; Morris et al. 1985; Morris, 1988), but freezing often does not significantly alter affinity. In the present study, haemocyanin oxygen-affinity, as well as cooperativity, was significantly altered by freezing for two of the three species examined, emphasizing the importance of using, whenever possible, fresh haemolymph samples in descriptions of oxygen-binding properties of haemocyanins from species whose haemocyanins have not been previously studied. A comparison of the oxygen affinity of the haemocyanin of fresh Oplophorus gracilirostris haemolymph at 5°C and once-frozen haemolymph at 25 °C, for example, would obscure the actual effects of temperature on affinity shown by direct comparison in fresh haemolymph samples for this species (Figs 2 and 4). As shown by Morris (1988), the effects of freezing vary among species, making examination of the oxygen-binding properties of haemocyanin in fresh haemolymph samples necessary prior to evaluation of the oxygen-binding properties of haemocyanin in frozen haemolymph samples.

It is evident from Figs 2 and 3 that oxygen affinities of the haemocyanins from the vertical migrators Acanthephyra smithi, Systellaspis debilis and Oplophorus gracilirostris are temperature dependent. Values for ΔH at constant values of pH are comparable to those of other crustaceans from environments with variable temperature (Mangum, 1983; Morris et al. 1985; Bridges, 1986; Burnett et al. 1988). When the changes in haemolymph pH with temperature are taken into account, the predicted changes with increasing temperature in oxygen affinities of the haemocyanins of vertical migrators are even greater (Figs 6 and 7).

These temperature-induced changes in the oxygen affinities of the haemo-cyanins are adaptive for the vertical migrators in the context of diurnal changes in their oxygen consumption rates and habitat (Cowles, 1987). At daytime depths off Hawaii, environmental values are in the range 2.66–4.00kPa, and temperatures are 3–6°C. The haemocyanin of Acanthephyra smithi can, however, be 95 % saturated at an internal of 2.66kPa at 5°C, pH7.82 (Fig. 6), allowing it to be effective, we suggest, in taking up oxygen from the lower- water. Conversely, at night-time depths, the vertical migrators are more active in warmer, high- surface waters and the greater demand for oxygen requires a haemocyanin with a lower oxygen affinity. This facilitates off-loading at the tissues while not jeopardizing maximal saturation at the gills due to the increased environmental At 25 °C the haemocyanin of A. smithi will be highly saturated with oxygen at intermediate O₂ levels (Fig. 6). If a temperature effect resulting in increased oxygen affinity at reduced temperatures were not present, the haemocyanin of A. smithi could be only poorly saturated at the low temperatures found at this species’ daytime depths (Fig. 6). Therefore, the substantial temperature effects on the haemocyanins of the vertical migrators result in the same haemocyanins exhibiting very different, yet adaptive, properties under the diverse environmental conditions they encounter at day and night depths.

The situation for non-migrators is somewhat different. Oxygen consumption rates and in vivo pH measurements could not be made at higher temperatures because these animals do not survive in temperatures above 10 –15°C (Cowles, 1987). At 5°C, Acanthephyra acutifrons and A. curtirostris are less active and have lower oxygen consumption rates compared to vertical migrators at the same temperature. This is reflected, for non-migrators, in the presence of haemocyanins with higher oxygen affinities (Figs 2 and 7), which are adaptive for the uptake of oxygen from the low-temperature, low- waters in their environment. At the predicted in vivo pH of 7.8 at 5°C (Fig. 1), the haemocyanin of A. acutifrons can be fully saturated at 2.66 kPa O₂ (Fig. 6). Because they are relatively temperature insensitive, however, these haemocyanins do not have the greatly reduced affinities at higher temperatures (Fig. 7) which would be necessary to support the oxygen demand present in vertical migrators when they move to near-surface waters. Low temperature sensitivity may be a general property of deeper-living crustaceans which do not migrate into the surface waters. Similar low-temperature effects have been found in a physiological pH range for the deep-sea benthic glyphocrangonid shrimp Glyphocrangon vicaria Faxon (Arp and Childress, 1985) and the pelagic lophogastrid mysid Gnathophausia ingens Dohrn (Sanders, 1989), which also five in thermally stable environments.

Low levels of L-lactate were found in haemolymph of four of these oplophorids, while high concentrations were measured in Acanthephyra smithi. The absence of higher L-lactate concentrations at higher temperatures in A. smithi and O.gracilirostris suggests that L-lactate is not produced in response to the higher metabolic demands at higher temperatures.

Among the three species for which the specific effects of L-lactate on haemocyanin oxygen-binding were investigated using frozen material, only in A. smithi was a moderate effect of lactate (Δlog P₅₀/Δlog[lactate] = − 0.17 at 5 °C, pH7.9) present. Although the effect of freezing on lactate sensitivity is unknown, this value is at the lower end of the range of lactate coefficients measured under physiological conditions of temperature and pH for other active crustacean species (−0.63 for Palaemon elegans, Bridges et al. 1984; −0.50 for Hyas araneus, Morris and Bridges, 1989; −0.25 for Cancer magister, Graham et al. 1983; −0.096 for Carcinus maenas, Truchot, 1980; see Bridges and Morris, 1986, for a review). Oxygen affinity of the haemocyanin of A. smithi was significantly increased by L-lactate, particularly at 5°C, and this may be important in allowing sufficient oxygen uptake when the animal descends to its daytime depths. In contrast, a significant effect of L-lactate on the oxygen affinity of the haemocyanin was present only at very high concentrations in O. gracilirostris and A. acutifrons (Fig. 5), suggesting that L-lactate is not an important modulator of haemocyanin function in these species which apparently produce little L-lactate. These data also suggest that there is considerable variation in the importance of lactate as a metabolite and as a modulator of haemocyanin O₂-affinity in midwater crustaceans.

Data are available on the oxygen-binding properties of the haemocyanins of crustaceans from a wide variety of habitats characterized by different physical and chemical conditions (Jokumsen et al. 1981; Jokumsen and Weber, 1982; Mangum, 1982, 1983; Arp and Childress, 1985; Morris et al. 1985; Bridges, 1986; Burnett et al. 1988; Sanders et al. 1988; Morris and Bridges, 1989; Sanders, 1989). In those cases where the biology of the subject species has been sufficiently studied, the authors have generally been able to put forward quite plausible explanations of the adaptive value of the haemocyanin properties found. More general trends in haemocyanin properties have also been enunciated, based upon data from shallow-living, non-pelagic species. In particular, it has been suggested that there is an inverse relationship between affinity and environmental temperature, so that higher O₂ affinities are found in species from warmer habitats and lower affinities in species from colder ones (Redmond, 1968; Mangum, 1982; Mauro and Mangum, 1982). It has also been suggested that there is an inverse relationship between the temperature sensitivity and pH sensitivity of haemocyanin O₂-binding, so that animals with large Bohr coefficients have small ΔH values and vice versa, minimizing the indirect effect of temperature on O₂ affinity via haemolymph pH changes (Burnett et al. 1988). The species upon which these generalizations are based are benthic animals which live at shallow depths. Although they are often exposed to low O₂ concentrations or variable temperatures, they are able to survive under unfavourable conditions by becoming quiescent and/or relying on anaerobic metabolism for a short time, options that are usually not available to midwater crustaceans.

The data presented here, while they appear to indicate highly adaptive properties for the haemocyanins studied, do not readily fit into the generalizations derived from studies on shallow-living benthic species. These pelagic crustaceans experience very different conditions from those experienced by shallow-living benthic species, and this has apparently resulted in selection for different combinations of adaptive properties. The oplophorids considered here are active animals which must swim continuously to maintain their position in the water column and live continuously at low values of when at depth. As a result they must regulate their O₂ consumption down to environmental values since they are dependent on aerobiosis, having, like most other midwater crustaceans (Childress, 1971, 1975), little anaerobic capacity (Cowles, 1987). Thus, the adaptive value of the high O₂ affinities of the haemocyanins of these species is clear. Belman and Childress (1976) have, in fact, demonstrated that another active, pelagic deep-sea crustacean, the mysid Gnathophausia ingens, requires a high-affinity oxygen-binding protein to survive at the. low values in its environment. The adaptive value of the haemocyanin properties described here is underscored by the finding of similar properties of high affinity and low temperature sensitivity (under depth conditions) in this distantly related crustacean from the same midwater habitat (Freel, 1978; Sanders, 1989). In this context, and given the aerobic nature of these midwater crustaceans, the higher oxygen affinities of the haemocyanins of these species under the conditions prevailing at depth appear to be specifically adapted to the low values found in the deep-sea oxygen minimum layer environment. The higher affinities found for the non-migrators indicate that adaptation to an aerobic lifestyle at low O₂ has taken precedence over the general trend of lower affinity at lower temperature found in studies of shallower species.

The trend of an inverse relationship between the pH sensitivity and the temperature sensitivity of haemocyanin O₂-affinity is also not found in these midwater shrimps, all of which have moderate Bohr coefficients but vary considerably in temperature sensitivity. This may be the result of the Bohr shift having an important functional role in these continuously active animals so that it cannot readily be reduced in the migrators. The moderate size of the Bohr effect in all these species may represent an adaptation to limit indirect temperature effects which would be more important if the animals had larger Bohr effects.

This comparison of the data presented in the literature leads one to the conclusion that generalizations about the adaptive value of haemocyanin properties in particular environments are now beginning to become apparent. Generalizations which span the full range of environments, however, require considerably more knowledge about the habitats, habits, metabolic poise and respiratory adaptations of crustaceans from as wide a range of habitats as possible. We would suggest that in addition to temperature range and magnitude it is essential to consider the activity level, the reliance on anaerobic metabolism, the environmental O₂ partial pressures and the pattern of environmental fluctuation of the crustaceans studied to evaluate the adaptive nature of the properties of their haemocyanins. In this effort the study of closely related species from different habitats or with different habits is likely to be especially rewarding. It is also essential to study species from as wide a range of habits and habitats as can be found.

The marked differences in oxygen-binding properties of the haemocyanins of vertically migrating and non-migrating oplophorids reflect the lifestyles and habitats of each. Both groups have haemocyanins with moderately high oxygen affinity and moderate Bohr shifts at low temperature, enabling both the binding of oxygen from low- waters, and the off-loading of oxygen to the tissues via a reduction in haemolymph pH. The non-migrators, which are limited by other physiological constraints to the low temperatures at depth, appear to maintain stable respiratory function within the narrow limits of their temperature range by having haemocyanins whose oxygen affinity is almost independent of temperature. In contrast, vertical migrators, which encounter very different environments and have much higher metabolic rates at night, achieve suitable, but very different, haemocyanin properties in both of these habitats by means of large temperature effects on the oxygen affinity of their haemocyanins.

This research was supported by NSF grant OCE 8500237 to JJC and a NSF Predoctoral Fellowship and McNaughton Scholarship for Oceanography to NKS.

Thanks to A. Alldredge, A. Kuris and B. Robison for critical review of this manuscript. Special thanks to the captain and crew of the R/V New Horizon for assistance in animal collection. We thank Dr S. Morris for help in making O₂ equilibria measurements in nondialyzed haemolymph, use of statistical software and critical review of this manuscript.