Physiological acclimation to decreased water temperature and the relative importance of water viscosity in determining the feeding performance of larvae of a serpulid polychaete

Abstract

Ambient temperature exerts both physiological and mechanical effects on the rates of functional processes of small aquatic ectotherms. Physiological effects of temperature result from its influence on the rates of chemical reactions. Mechanical effects of temperature result from the inverse relationship between the temperature of water and its dynamic viscosity. We measured the relative importance of these components of temperature on the feeding performance of polychaete larvae. Cohorts of larvae were reared for 24 h at 20°C and 10°C in treatments where the physiological and mechanical effects of these temperatures were separated. The feeding performance of these larvae was subsequently measured in treatments where these components of temperature were similarly partitioned. Cold-reared larvae displayed complete acclimation of feeding performance to the physiological effects of decreased temperature: thus, increased viscosity was responsible for 100% of the difference in feeding performance between 20°C and 10°C. The physiological ability of small aquatic ectotherms to acclimate functional processes to temperature variation may be greater than previously thought, and these results have implications for understanding the responses of aquatic ectotherms’ to global temperature change.

Received April 27, 2005; accepted in principle July 26, 2005; accepted for publication August 30, 2005; published online September 6, 2005
 Communicating editor: K.J. Flynn

INTRODUCTION

Ambient temperature profoundly influences rates of physiological processes of aquatic ectotherms (Hochachka and Somero, 2002). Accordingly, higher order functions, such as rates of feeding, swimming, and growth, are also temperature dependent. Although physiological processes of aquatic ectotherms generally operate optimally within narrow temperature ranges, an extensive literature shows that most ectotherms can acclimate to changes in ambient temperature (Huey and Berrington, 1996). Acclimation—the adjustment of physiology to changing environmental conditions (Lincoln et al., 1996)—enhances functional operation over a range of temperatures (e.g. Denhel, 1955; Ament, 1979; Clarke, 1983, 1991) and is frequently assumed to convey fitness benefits. Acclimation is, however, generally considered to be only partial, so that physiological (and functional) processes are not completely de-coupled from temperature (Kingsolver and Huey, 1998).

Ambient temperature also has mechanical effects on aquatic ectotherms by altering the dynamic viscosity of water (from hereon referred to as ‘viscosity’) in which they operate. Viscosity—the inter-laminar stickiness of a fluid—changes inversely with temperature (Vogel, 1994): for example, a decrease of temperature from 20°C to 10°C increases viscosity from 0.0109 Pa s to 0.0139 Pa s. This relationship has important consequences for small aquatic organisms whose feeding and swimming structures (typically cilia, setae, and flagella) operate in a low Reynolds number (Re ≤ 1) hydrodynamic environment, where viscous drag forces predominate, and boundary layers are thick (Emlet and Strathmann, 1985). Temperature-induced changes in viscosity therefore have a marked influence on the drag that operates against the feeding and swimming structures of small aquatic ectotherms.

Previous studies have quantified the relative physiological and physical effects of ecologically relevant decrease in water temperature on feeding and swimming rates of small aquatic ectotherms. In these experiments, water viscosity has been manipulated independently of temperature by the addition of polymers. Thus, the mechanical effects of increased viscosity at low water temperatures have been mimicked independently of the physiological effects of temperature. These studies show that increases in viscosity accounted for around half of the temperature-induced declines in swimming and feeding rates of marine invertebrate larvae (Podolsky and Emlet, 1993; Podolsky, 1994; Bolton and Havenhand, 1997, 1998), and the swimming velocities of small fish (Fuiman and Batty, 1997) and sperm (Kupriyanova & Havenhand, unpublished data).

Most of these studies have incorporated short acclimation periods to new temperatures—a technique which may not, however, allow organisms sufficient time to acclimate their physiological processes to the full extent possible (Selong et al., 2001; Tanaka, 2002). Consequently, the importance of physiological effects of decreases in water temperature to rates of functional processes may have been overestimated. Longer-term acclimation to decreased temperature within the context of its physiological and mechanical effects has only been examined in one study. Podolsky (Podolsky, 1994) reared cohorts of larvae of the echinoid Dendraster excentricus at warm and cold temperatures (∼20°C and 11°C, respectively) and found a degree of acclimation in feeding performance among cold reared larvae. However, the extent to which this acclimation was physiologically or mechanically based remained unknown, because larvae were not reared in conditions where these components of temperature were experimentally separated.

We investigated the relative importance of physiological and mechanical effects of temperature on the feeding performance of larvae of a marine polychaete. Trochophore larvae of Galeolaria caespitosa were reared for 24 h at two temperatures (20°C and 10°C) in which the physiological and mechanical components of temperature were partitioned. Galeolaria caespitosa is distributed intertidally around southern Australia and was chosen as a model for this experiment, because it experiences large seasonal temperature fluctuations.

METHOD

Physiological and mechanical components of water temperature were separated between 20°C and 10°C in two suites of experimental treatments: (i) acclimation treatments and; (ii) subsequent feeding performance treatments on larvae from each acclimation treatment (Fig. 1). In the first suite of three treatments, larvae were acclimated for 24 h to conditions in which physiological and mechanical effects of water temperature were partitioned (acclimation treatments denoted with the letters A, B and C were A = 20°C; B = 20°C with viscosity artificially increased to that of water at 10°C and; C = 10°C). Thus, larvae were acclimated to two temperatures (20°C in treatments A and B; and 10°C in treatment C) and to two viscosities (0.0109 Pa s in treatment A and 0.0139 Pa s in treatments B and C). Treatment B also acted as a control for potential toxic effects of the polymer (Ficoll) used to manipulate water viscosity (Bolton and Havenhand, 1998). The feeding performance of larvae from each acclimation treatment was then determined in feeding treatments where physiological and mechanical effects of temperature were similarly separated (feeding treatments denoted with the numerals 1, 2 and 3 were 1 = 20 °C; 2 = 20°C with viscosity artificially increased to that of water at 10°C, and; 3 = 10°C). Because the viscosity of water cannot be decreased independently of temperature without drastically altering its chemical properties, a balanced experimental design in which viscosity was both increased and decreased independently of temperature was not possible. The combined acclimation and feeding treatments to which larvae were exposed are denoted by their respective acclimation treatment (A, B and C) and subsequent feeding treatment (1, 2 and 3): thus, larvae acclimated at 20°C (acclimation treatment A) for which feeding performance was subsequently measured at 20°C (feeding treatment 1) were denoted A1 (Fig. 1).

Replicate groups of Galeolaria caespitosa larvae (n = 8) were cultured in vitro (Bolton and Havenhand, 1998) from adults collected on consecutive days from pier pilings at Brighton, South Australia. Embryos were incubated at 15°C (i.e. mid-way between the experimental temperatures). Hatching of trochophore larvae occurred after ∼18 h. After 24 h, sub-samples of the larvae were transferred to acclimation treatments for 24 h (concentration = 5 larvae/mL). To ensure that larvae from each acclimation treatment were the same size prior to initiation of feeding experiments, sub-samples of larvae (n = 10) from each treatment were measured (i.e. lengths and widths of the trochosphere, and lengths of prototrochal cilia) with the aid of a microscope (×200 magnification) and compared by one-way analysis of variance (ANOVA) (α = 5%).

Samples of larvae from each acclimation treatment were then transferred to 6 mL vials (5 larvae/mL) containing 5 mL of the respective feeding treatment solutions. To control the temperature of the treatments, vials were suspending in thermostatically regulated water baths (± 0.1°C). Feeding experiments were initiated by addition of polystyrene spheres (3 µm diameter, Duke Scientific) to each treatment (final concentration = 5 spheres/µL). Larvae ingested spheres for 20 min at which time they were killed with formalin. The number of spheres in the guts of sub-samples of larvae (n = 20) from each combination of acclimation and feeding treatment were counted (Bolton and Havenhand, 1998) and compared by one-way ANOVA and subsequent Tukey’s tests (α = 5%). Assumptions of ANOVA were tested prior to analysis (Zar, 1996).

RESULTS AND DISCUSSION

Feeding performance of larvae was lower at 10°C than at 20°C (feeding treatments 1 and 3, respectively) for larvae reared in each of the acclimation treatments (the number of spheres ingested was 55, 57 and 34% lower at 10°C that at 20°C for acclimation treatments A, B and C respectively; ANOVA df = 8, F = 15.6, P = <0.0001) (Fig. 1; Table I). Approximately half of these differences were attributable to increased viscosity for larvae acclimated in treatments A and B (52 and 57% for treatments A and B, respectively). The remaining differences in feeding performance (48 and 43% for acclimation treatments A and B, respectively) were attributable to the physiological effects of decreased temperature. In contrast to these findings, the feeding performance of larvae acclimated at 10°C (treatment C) was not influenced by the physiological effects of temperature: no significant difference was apparent between treatments C2 (where viscosity was artificially increased) and C3 (10°C water) (Fig. 1; Table I). Therefore, 100% of the difference in feeding performance between 20°C and 10°C of larvae acclimated at 10°C was attributable to the mechanical effects of increased viscosity.

Feeding performance of larvae acclimated at 20°C, and 20°C with artificially increased viscosity (treatments A and B respectively), was the same in feeding treatment 2: thus, no acclimation to increased viscosity was apparent. Indeed, as feeding performance of larvae acclimated in treatments A and B was the same in all feeding treatments (i.e. A1 versus B1; A2 versus B2, and; A3 versus B3) (Fig. 1; Table I), we can further conclude that incubation in the polymer used to manipulate water viscosity did not adversely influence the feeding performance of larvae. Sizes of larvae from each acclimation treatment were the same after 24 h (Table II): trochosphere length (df = 9, F = 0.10, P = 0.90), prototrochal width (df = 9, F = 0.30, P = 0.74), and prototrochal cilia length (df = 9, F = 0.03, P = 0.96). Thus, feeding performance experiments were not confounded by differences in the sizes of larvae among acclimation treatments.

Contrary to the widely held notion that physiological acclimation of functional processes is only partial, our findings indicate that larvae of Galeolaria caespitosa can completely acclimate to the physiological effects of decreased temperature on feeding performance. When feeding at 10°C, feeding performance of larvae acclimated at 10°C was greater than that of larvae acclimated at 20°C. Because there was no difference between the feeding performance of cold acclimated larvae at 10°C and that of larvae acclimated at 20°C with artificially increased viscosity, physiological effects of temperature were not responsible for the decrease in feeding performance between these temperatures. Consequently, water viscosity must have exerted the temperature-induced effect on feeding performance of cold-acclimated larvae. In contrast, feeding performance of larvae acclimated to the higher temperature—both 20°C and 20°C with artificially increased viscosity—was decreased by physiological and mechanical effects of temperature in proportions that are in accordance with findings of previous studies that have incorporated relatively short acclimation periods (Podolsky and Emlet, 1993; Podolsky, 1994; Bolton and Havenhand, 1997, 1998; Fuiman and Batty, 1997).

Our findings suggest that the physiological ability of some small aquatic ectotherms to acclimate key functional processes to environmental temperature variation may be greater than previously thought. Why does this matter? The obvious answer is that our current understanding of aquatic ectotherms’ responses to temperature change may be inaccurate—especially if that understanding is based on experiments with relatively short acclimation times. Of equal, if not greater, importance is that an organism’s capacity to acclimate to temperature variation is likely to be of central importance in determining an organism’s ability to withstand climate change. For example, Stillman (Stillman, 2003) showed that porcelain crabs from cooler climates are capable of greater acclimation to temperature changes than crabs from warmer climates. Stillman speculates that these differences among populations may underlie changes in the distribution of species of porcelain crabs (and many other marine invertebrates) along the West Coast of the United States that correlates with global temperature rises. Predictions of the demise or distributional changes of organisms in response to climatic variation are based on their ability to withstand temperature change. Investigations into the functional responses of small aquatic organisms to temperature are clearly not simple: failure to consider physiological and mechanical effects of temperature, as well as an organism’s propensity for physiological acclimation within this context, are likely to lead to erroneous conclusions about the extent of physiologically based changes in functional performance.

ACKNOWLEDGEMENTS

This research was supported by Establishment Funding provided to T. F. Bolton by Flinders University—Adelaide, Australia.

REFERENCES

Author notes

¹Lincoln Marine Science Center, Flinders University, PO Box 2023, Port Lincoln, SA 5606, Adelaide, Australia and ²Tjärnö Marine Biological Laboratory, Department of Marine Ecology, Gothenburg University, 452 96 Strömstad, Sweden