Effect of Temperature on Filtration in the Blue Mussel, Mytilus edulis—Our Present Understanding

Abstract

The effect of temperature on the filtration rate of filter-feeding bivalves is controversial. Some studies show a clear increase in filtration with an increase in temperature, in agreement with a decrease in kinematic viscosity with increasing temperature and a simultaneous increase in beat frequency of the water-pumping lateral cilia. However, other studies show little or no effect. Here, we present our understanding of the effect of temperature on the filtration rate of the blue mussel, Mytilus edulis, primarily based on our own previously published experimental studies. We point out the possible pitfalls associated with findings of little or no such effects, most likely caused by erroneous use of the method for measuring the filtration rate and failure to separate the factors that may affect it, e.g., the temperature-tolerance interval and the algal-concentration interval within which the mussel is optimally filtering at a specified temperature.

1. Introduction

There are many publications dealing with the effect of temperature on the filtration rate of the blue mussel, Mytilus edulis, and other ciliary filter-feeding bivalves, reviewed by Ali [1], Schulte [2] and Winter [3]. The older literature has been reviewed by Kittner & Riisgård [4], who found that the subject was controversial due to often inconsistent results, likely caused by different experimental conditions and methodological shortcomings.

When dealing with the filtration rate of mussels, it is essential to differentiate between the effect of temperature on the filtration rate measured on optimally filtering mussels and the effect of both high and low temperatures on the valve-opening state, which may significantly affect the filtration (or pumping) rate [4,5,6]. Therefore, it is important to identify the temperature-tolerance interval for optimal filtering. When mussels are exposed to a change in temperature outside the tolerance interval, they may after some time adjust and become tolerant to the new temperature, which is called acclimatization [5]. One of the most important environmental parameters affecting the filtration rate of mussels is the phytoplankton (algal) concentration. Under optimal conditions, M. edulis exploits its filtration capacity, but under sub-optimal conditions, including low algal concentrations, the filtration rate is reduced by a reduction in or closure of the valve gape [7]. The valve closing/opening response has been studied in-depth. The critical phytoplankton biomass below which the mussel closes its valves is about 0.9 chl a l⁻¹ [6,8]. During the winter starvation period in temperate Danish waters, M. edulis reduces its valve gape and thereby its filtration rate [9], which reduces its oxygen uptake (respiration) [10] and thus its body-weight loss [11]. The algal (phytoplankton) biomass in the sea is typically between 1 and 5 µg chl a l⁻¹ [12], but because mussels often live in dense beds their grazing impact on phytoplankton may become pronounced, and likewise, starvation may occur during winter.

The filtration rate (= pumping rate) is sometimes referred to as the clearance rate because it may be measured by the so-called clearance method [13]. The “clearance method” is a reliable and precise method for measuring the filtration rate at known temperatures when correctly used (Figure 1). Here, the clearance rate (Cl) of a group of mussels (n) is estimated from the exponential decline in algal concentration in a well-mixed, strongly aerated aquarium as Cl = aV/n where a = slope of regression line in a semi-ln plot of algal concentration versus time, and V = volume of seawater. A prerequisite for using the clearance method is immediate mixing of once-filtered water in all the water, which is ensured by strong air-mixing. However, other methods, as reviewed by Riisgård [13], may be used to measure the filtration rate of mussels; in particular, the “flow-through chamber method” has frequently been used. In the present account, we present our understanding of the effect of temperature on the filtration rate of the blue mussel, Mytilus edulis, primarily based on our own previously published experimental studies of the filtration rate and beat frequency of water-pumping lateral cilia. We use the following definitions: clearance rate (Cl) = volume of water cleared of suspended particles per unit of time. If algal cells are retained with 100% efficiency, then Cl = F (= pumping rate).

Figure 1. Measurement of the filtration rate in Mytilus edulis using the clearance method. Exponential decrease in algal cell concentration as a function of time due to fully open and actively filtering mussels (length 50.54 ± 5.5 mm, n = 5) during a clearance experiment at 20.3 °C in an aquarium with well-mixed seawater. Regression lines and estimated filtration rates (l h⁻¹ ind.⁻¹) are shown. Arrows indicate addition of algal cells. From [4].

2. Results and Interpretations

Here, we present our understanding of the effect of temperature on mussels, mainly based on our own experimental studies.

In mussels, water-pumping lateral cilia in the gills create the feeding current [14], and because the viscosity of seawater is inversely related to temperature, the increase in filtration rate with temperature in mussels cannot only be accounted for by increased biological activity. The early study by Jørgensen et al. [5] tried to explain in detail whether alteration in filtration rate with temperature in mussels is under chiefly biological/physiological or physical/mechanical control. They showed that the filtration rate of Mytilus edulis increased with temperature (Figure 2) and decreased linearly with the temperature-determined increase in viscosity of seawater (Figure 3). The effects of changes in temperature on filtration rates in mussels acclimatized to seasonally varying temperatures at the collecting site showed that the temperature ranges used in the experiments were those tolerated by the mussels, indicating that they were optimally filtering. Figure 2 shows that in cold-adapted mussels in February the upper temperature tolerated was 16 °C, compared with 22 °C in June. In the lower range, cold-adapted mussels tolerated temperatures approaching zero, whereas the 12 °C-adapted mussels began to reduce or close their valve gape at 4 to 5 °C. Kittner & Riisgård [4] studied the effect of temperature on the filtration rate of M. edulis and likewise found a linear relationship in the temperature-tolerance interval and acclimation ranges (Figure 4). For example, in the lower part of the temperature interval, cold-adapted (11 °C) mussels tolerated temperatures as low as at least 4.1 °C without closing their valves, hence maintaining the maximum filtration rate. In Figure 4B, it appears that the filtration rate (F, ml min⁻¹) versus temperature (T, °C) for a 50.5 mm shell length M. edulis was found to be F = 3.27T + 38.1 (r² = 0.953), which implies F = 54.5 and 103.5 ml min⁻¹ at 5 and 20 °C, respectively, thus indicating a 1.9 times increase in the filtration rate.

Figure 2. Mytilus edulis. Relationship between temperature and pumping rate (=filtration rate) in groups of seasonally acclimated (fully open) mussels of different sizes at different times of year. (○) February, 25 mussels, 31 ± 0.7 (SD) mm in length, acclimated to 6 °C; (x) May, 20 mussels, 39 ± 2.0 mm, acclimatized to 7.5 °C; (●) June, 30 mussels, 29 ± 2.6 mm, acclimatized to 12 °C. It is seen that the temperature-tolerance range within which mussels were fully open varied with the season: cold-adapted mussels tolerated lower temperatures whereas warm-adapted tolerated higher temperatures. From [5].

Figure 3. Mytilus edulis. Relationship between pumping rate (=filtration rate) and temperature-dependent kinematic viscosity of seawater in groups of mussels shown in Figure 2, using the same symbols.

Figure 4. Mytilus edulis. Individual filtration rate of mussels (50 mm shell length, n = 5) as a function of stepwise increasing and decreasing temperature as indicated by inserted numbers in brackets starting with (1). (A) Mussels seasonally acclimated to 18 °C. (B) The same group of mussels after 1 week of acclimation to 11 °C. The regression lines for filtration rate as a function of temperature in the temperature-tolerance interval and their equations are shown. It is seen that there is a linear and reversible effect of temperature in the temperature-tolerance interval and increased tolerance (fully open valves) to temperatures below 8 °C after acclimation to 11 °C. From [4].

Later, Riisgård & Larsen [14] found that the effect of temperature on the beat frequency of the water-pumping lateral cilia in M. edulis gill preparations is controlled by the viscosity of seawater, as seen in Figure 5. Recognizing that the filtration rate of mussels and viscosity depend on temperature may have motivated the development of pumping models as prepared by Jørgensen et al. [5], Riisgård & Larsen [15], and Zhang et al. [16]. Common to these studies, the pump head is equated to the sum of frictional losses in the flow system through a mussel, accounting also for kinetic energy contributions at inhalant and exhalant openings. As viscosity decreases with increasing temperature, friction decreases, and the pumping rate increases as it also does due to increasing cilia beat frequency. This is true within the temperature-tolerance interval, which, however, seems to have been exceeded in the study by Zhang et al. ([16] Figure 4 therein], who found a decrease in the filtration rate at 26 and 28 °C, and at 2 °C, very likely due to reductions in the valve gape.

Figure 5. Mytilus edulis. Beat frequency of water-pumping lateral cilia in mussel-gill preparations stimulated with serotonin (10⁻⁵ M) versus temperature (T) or temperature equivalents (Te); (◊) exposed to different temperatures of seawater; (□) exposed to different dextran-manipulated viscosities of seawater (22 °C, 20 psu); (○) exposed to different PVP-manipulated viscosities of seawater (22 °C, 20 psu). Mean (± 2 SD, n = 8) indicated for PVP and dextran. It is seen that the effect of temperature on lateral cilia activity is controlled by the viscosity of the ambient seawater. From [16].

Water-pumping cilia are not only important in bivalves; beating cilia are also important in other aquatic organisms. Thus, viscosity controls or strongly affects bio-mechanical activity such as the swimming of ciliates and micro- and meso-scale aquatic organisms using cilia [17,18]. In M. edulis, Jørgensen et al. [5] and Kittner & Riisgård [3] found a linear relationship between filtration rate and temperature in the temperature-tolerance interval (Figure 2 and Figure 4), as also found in the soft clam Mya arenaria by Riisgård & Seerup [19], and in Arctic clams Hiatella arctica and Mya sp. by Petersen et al. [20].

3. Discussion and Recommendations

In this section, we discuss the importance of the valve-opening degree and make recommendations to avoid pitfalls in future studies. As shown in the preceding sections, the valve-opening degree of a mussel ultimately controls its filtration rate. To eliminate this variable from the effect of temperature on the water-pumping lateral cilia it is necessary to ensure that the mussel is wide open and exploiting its filtration capacity when the temperature is being varied. When a mussel experiences very low algal concentrations, this eventually leads to a reduced valve gape, or complete closure, along with the cessation of filtration activity, which lasts until the algal concentration is elevated above the lower threshold level.

The opening–closing phenomenon and response times in M. edulis in the presence and absence of algal cells were studied by Riisgård et al. [6] in controlled laboratory experiments. When algal cells were added to unfed mussels they soon after opened their valves, with a strong increase in the filtration rate. Later, when the algal concentration decreased to under a certain level, these mussels reduced their valve-opening degree and eventually stopped filtering. However, new algal additions rapidly restored the maximum filtration rate ([6], Figure 9 therein).

Another study by Riisgård et al. (2006) [8] of valve-gape responses of M. edulis to the presence or absence of algal cells in the ambient water, revealed that the critical algal concentration (C_min) below which the mussel closes its valves is about 700 Rhodomonas salina cells mL⁻¹, equivalent to 0.9 µg chl a l⁻¹. However, high algal concentrations may also result in a reduced filtration rate due to saturation of the digestive system [21]. Therefore, controlled studies on the effect of temperature on the filtration rate should be made on optimally filtering mussels exposed to ambient algal concentrations within this interval.

Using the clearance method implies that algal cells should be added repeatedly to ensure that the mussels, after some time, open their valves and filter at the maximum rate, which should be ensured by the observation of a concentration decrease as parallel linear regression lines in a semi-ln plot after every new algal addition (Figure 1) (Marroni et al. [22], Figure 1 therein). Using only one initial algal cell addition and measurement of only the start and end concentrations after a certain (arbitrary) time without noting the valve-opening degree will inevitably result in questionable data. This seems to have been the case in the study by Pestana et al. [23], who found that the filtration rate was “largely unaffected by temperature” for the golden mussel Limnoperna fortune, while Zhang et al. [16] found a skewed unimodal pattern of increase below and decrease above about 15 °C for this species. Further, (Gopalakrishnan & Kashian [24], Figure 1 therein) found that the filtration rate of the quagga mussel Dreissena rostriformis burgensis increased linearly with temperature from 2 to 25 °C followed by a “decline” at 30 °C, which was outside the mussel’s temperature-tolerance interval.

Cranford et al. [25] claimed that temperature has “not been identified as an important control on feeding behavior” under natural conditions for mussels, and the authors referred to a figure ([26], Figure 4.6 therein), based on unpublished data) that showed the clearance rates of M. edulis measured in the field over a wide range of water temperatures using the flow-through chamber method that “revealed little dependence of feeding on temperature”. However, the same shortcomings as mentioned below when using this method are likely. Cranford et al. [25] criticized the study by Kittner & Riisgård [4] for being limited to “an artificial condition that does not exist in nature” where the mussels are not “stimulated by a controlled artificial diet”. This criticism is unwarranted. To separate the various factors that may affect the filtration rate, it is necessary to identify (1) the temperature-tolerance interval within which the valves remain fully open, and further, (2) the algal-concentration interval within which the valves are fully open. The separate effects of temperature and viscosity on the water-pumping lateral cilia can only be precisely measured in fully open and optimally filtering mussels. Finally, the use of “artificial diet” (cultivated algal cells) is also justified. Lüskow & Riisgård [26] found in situ filtration rate measurements of M. edulis using ambient natural phytoplankton were comparable to filtration rates obtained in laboratory studies using several methods and controlled diets of cultivated algal cells.

In a recent paper by Rosa et al. [27], the authors conclude that their experimental data indicate that changes in water temperature and the associated changes in water viscosity have no effect on the filtration rate of mussels. This statement is controversial, as it is contrary to other findings [15,17]. To perform precise and reliable laboratory filtration rate measurements on mussels with the ‘flow-through chamber method’ used by Rosa et al. [27], the design of the chamber must ensure that only inflow water reaches the bivalve’s inhalant aperture and that exit flow is fully mixed. These prerequisites can be checked by a plot of clearance rate versus increasing through-flow to reach a plateau, which is the true clearance rate [13,28,29]. This requirement was probably not fulfilled by Rosa et al. [27], who measured the clearance rate of 55.9 mm shell length M. edulis in chambers of undescribed size and shape (possibly allowing the bypass of some water and/or recirculation of once-filtered water). They used a constant through-flow of 150 ml min⁻¹, which resulted in strongly varying clearance rates ranging between 28 to 56 ml min⁻¹ and “no effect of temperature” on the clearance rate. However, the formula for clearance rate used by Rosa et al. [27] appears to be wrong because the particle concentration (C_out) was not measured in the out-flowing water from the chamber but instead measured in water samples taken in the mussel’s exhalant aperture. The particle-capture efficiency of the gills can be estimated as CE = 1 − C_out/C_in, where C_in and C_out are the concentrations in the inhalant and exhalant water, respectively, but the clearance rate using the flow-through method is given by CR = F(1 − C_out/C_in), where F = flow rate through the chamber and C_in and C_out are the particle concentrations in the inflowing and outflowing water into and out from the chamber, respectively. But the formula CR = F × CE, which was apparently used by Rosa et al. [27], is wrong, implying that the clearance rate increases linearly with increasing through-flow without reaching a plateau. The maximum filtration rate (F, l h⁻¹) of M. edulis may be estimated from the shell length (L, mm) as [30]: F = 0.00135L^2.088, which for a 55.9 mm mussel gives 6 l h⁻¹ = 100 ml min⁻¹, or about 2 to 4 times higher than reported by Rosa et al. [27]. Using the correct formula and a chamber of the proper shape and size and checking for optimal flow-through rate to ensure a clearance plateau would probably have resulted in a close correlation between temperature and filtration rate as those found in M. edulis by Jørgensen et al. [5] and Kittner & Riisgård [4].

4. Conclusions

Earlier studies on the effect of temperature on the filtration rate of mussels and other filter-feeding bivalves were typically inconsistent and without information on whether their valves were fully open or partially closed. Here, we point out the importance of taking notice of the valve-opening degree for being able to separate the effect of temperature from other important factors that may affect the filtration rate, namely the temperature-tolerance interval and the algal-concentration interval within which these mussels are fully open and optimally filtering. Further, a serious pitfall associated with observing little or no effect of temperature on the filtration rate may be caused by erroneous use of the flow-through chamber method for measuring the filtration rate. It should also be mentioned that mussels seasonally acclimatized to varying temperatures may increase their temperature-tolerance interval (i.e., keep their valves fully open) after a certain period of acclimation to a lower or higher temperature. Finally, the effect of temperature on the water-pumping lateral cilia, which ultimately determines the filtration rate, can only be evaluated on fully open mussels, and here it has been experimentally proven that water viscosity is the controlling factor.