| Collection Mémoires et thèses électroniques | |
| AccueilÀ proposNous joindre |
| Chapitre II. Disparity in the cost of living: Abiotic influences on shell and tissue traits of Dreissena polymorpha (Pallas) and D. bugensis (Andrusov) | ||
|---|---|---|
 |  | |
Table des matières
Les moules du genre Dreissena vivant dans les masses d’eau du fleuve St-Laurent fournissent un modèle pour l’évaluation de l’effet de l’adaptation physiologique différentielle d’un animal largement répandu qui vit dans des environnements hétérogènes. Nous avons émis l’hypothèse qu’il y n’aurait que peu de différences dans la biomasse entre les moules des différentes masses d’eau, mais qu’il pourrait y avoir des coûts physiologiques et bioénergétiques pour maintenir cette uniformité. Contrairement à nos attentes, le profil de changements saisonniers dans la biomasse, le contenu organique et la concentration en glycogène ont tous varié dépendant des conditions locales. Les moules vivant naturellement dans les eaux saumâtres du haut estuaire avaient une masse corporelle plus grande, bien qu’avec de plus faibles réserves en glycogène, que celles de la rivière. Les moules démontrent alors distinctement des profils de réponses différents pour les caractéristiques de la coquille et celles du tissu en relation avec les conditions locales. Une analyse séparée montrait que les traits de la coquille sont aussi sensibles aux facteurs environnementaux, quoique différents de ceux du tissu mou. Contre intuitivement, les différences dans le corps du glycogène correspondaient plus près de la masse de la coquille et le contenu organique du tissu mou correspondait mieux à la finesse de la coquille qu’ à la masse de la coquille. Alors que les réponses du tissu mou de Dreissena polymorpha et D. bugensis étaient similaires, la plus fine, et plus légère coquille de D. bugensis apparaît lui donner un avantage bioénergétique où les deux espèces co-existent.
Dreissenid mussels inhabiting the water masses of St. Lawrence River and estuary providea model forevaluating the effect of physiological adaptability of widespread species distributed across heterogeneous environmental template. Further, because one of the two species Dreissena bugensis is displacing the other D. polymorpha throughout their shared range, this comparison might also detect an underlying physiological explanation. Tissue and shell components were analyzed separately and whereas tissue mass, organic content, and glycogen concentration all differed in relation to location, they were less variable than shell characteristics. In particular, Dreissena from the brackish waters of the upper estuary had heavy bodies but light shells with lower glycogen reserves than those from riverine locations. Mussels from different source populations/water masses also demonstrated distinctly different seasonal and spatial response patterns for shell and tissue characteristics. Though the soft tissue response of Dreissena polymorpha and D. bugensis were similar, the lighter, more fragile shell of D. bugensis appears to give it an advantage where the two species co-occur.
Plastic physiological and life history adaptations that allow the rapid exploitation of novel environments are often associated with widespread species (Kolar and Lodge 2001, Sakai et al. 2001, McMahon 2002). Dreissenid mussels are an example of species that possess a suite of these characteristics: fast growth, early maturity, high fecundity, and the allocation of large amounts of energy to annual reproduction (McMahon 2002). The ecological value of this suite of traits has been seen in the rapid spread of dreissenid mussels from the Great Lakes across the eastern half of North America (Johnson and Carlton 1996). However, these mussels also have a limited physiological ability for facing fluctuations in abiotic conditions meaning that, both demographically and physiologically, their populations can vary dramatically in environments that are unstable and unpredictable (McMahon 2002). Among lotic systems, large rivers are considered among the most abiotically stable and annually predictable due to the combined buffering effect from their large watersheds and channel volumes (Wetzel 2001). Yet these are not homogeneous environments, and certain large river systems include persistent, spatially-discrete water masses that are the result of within-basin differences in tributary and geographic features. In the St. Lawrence River, these regional, physical-chemical differences create a mosaic of local habitats/conditions for Dreissena that range from the cool oligotrophic Great Lakes waters in the upper river, through the warm productive waters of Lac St.-Pierre, to the brackish, turbid and turbulent waters of the fluvial estuary below Quebec City (St. Lawrence Centre 1996; Thorp et al. 2005).
Chemical characteristics which distinguish St. Lawrence water masses and are potentially physiologically stressful for Dreissena include low concentration of ions, calcium in particular (Kilgour et al. 1994, Fong et al. 1995, Dietz et al. 1996, Wilcox and Dietz 1998), associated with the tributary waters flowing south from the Canadian Shield geologic formation (Thorp et al. 2005). Calcium is essential for production of a strong shell necessary for limiting predation and mechanical damage (Boles and Licius 1997, Thorp et al. 1998). However, the energy invested in the production of shell must be balanced against other bioenergetic requirements of soft tissue metabolism, growth, and reproduction. In an environment where calcium is not abundant and thus difficult to obtain and retain, a trade-off between the allocation to shell and soft tissue requirements may ensue with the shell even acting as a calcium reservoir (Russell-Hunter 1981, McMahon 1983). In addition the molluscan shell matrix is not made entirely of inorganic material (Palmer 1992), and the proportion of organic/proteinaceous materials is often variable, ranging between 2–8 % by weight (Wainwright et al. 1976, Beadman et al. 2003), and can vary in an ecologically significant manner with the environment (Brodersen and Madsen 2003). Although the organic material only represents a small proportion of the shell, its energetic cost is disproportionately greater than other shell components (Palmer 1983, 1992).
Under conditions of temperature stress and reduced food Dreissena will allocate energy to reproduction over somatic maintenance and growth (Chase and Bailey 1999b, Stoeckmann and Garton 2001). This suggests that when populations are subjected to persistent stress, local differences in physiological characteristics may develop. In addition to salinity, another significant but site-specific limiting factor for dreissenids is the osmotic challenge presented by fluctuating salinity in the tidally brackish portion of the St. Lawrencefluvial estuary. Dreissenids are passive regulators of K+, Ca++, and Mg++ions (Dietz et al. 1996), and brackish conditions have been shown to cause mortality and physiological malfunction in all stages of thelife cycle (Mackie and Kilgour 1992, Wright et al. 1996, Wilcox and Dietz 1998). These negative effects include decreasing rates of fertilization, increasing larval mortality, and decreasing growth and mass of adults (Mackie and Kilgour 1992, Kilgour et al. 1994, Fong et al. 1995, Wilcox and Dietz 1998). Adult acclimation to slow increases in salinity has, however, been demonstrated in laboratory settings but only under very specific ionic conditions (Kilgour et al. 1994, Fong et al. 1995, Dietz et al. 1996, Wilcox and Dietz 1998).
This study compares spatial and temporal variation in the shell and tissue total mass, shell and tissue organic carbon mass, shell strength, and whole glycogen content of mussels living in four different portions of the river affected by contrasting water masses: the oligotrophic Great Lakes, the low calcium and ionic strength of the Ottawa River, the more productive southern tributaries, and the tidally brackish estuary. We propose that the heterogeneous local environments of the St. Lawrence River will affect the phenotypic characteristics of dreissenid mussels differentially. Specifically the geographic differentiation in phenotype of should show individuals in poorest condition found in the estuary and the best condition upriver, implying a limited ability to adapt phenotypically to the environment of the estuary. Alternatively a lack of differentiation would imply either broad physiological flexibility or potentially the differential use of stored energy reserves such as glycogen to maintain condition across the environmental mosaic.
Mussels were collected in June andOctober of 2002 at sites in each of the four principal physical-chemical water masses St. Lawrence River (Figure 2.1). The GL site (44o59’ 19’’ and 74o49’ 45’’)represented the oligotrophic Great Lakes water mass, the OR site (46o16’ 35’’ and 72o41’ 40”) was located in the low ionic strength Ottawa River plume, the ST site (46o12’ 02’’ and 72o40’ 48”) received more nutrient-rich discharge from southern tributaries including the Richelieu River, and the FE site (46o59’ 47’’ and 69o48’ 21”) represented the highly turbid and tidally brackish conditions of the fluvial estuary. At each site approximately 50 adult mussels (<10 mm) were collected from hard substrata at least 2 m deep. Quagga mussels (Dreissena bugensis) became progressively rarer downstream (Casper, personal observation) and thus analysis of their response was limited to a single site (GL) where they occurred in sufficient numbers. The four sites contrast in terms of ionic content, turbidity, primary productivity and salinity (Table 2.1). We chose to compare changes between June and October because late summer is a physiologically stressful time of the year for dreissenid mussels, when the decline in food availability is compounded by the higher costs of maintenance metabolism related to higher temperatures (Stoeckmann and Garton 1992, 2001). Water temperature showed variation between at three sites (the OR data logger was lost during deployment) between July and September. The ST site was generally warmer due to 4 peaks; in contrast, the temperature at the FE was consistently cooler and the least variable while the GL site was intermediate (Figure 2.2)
Allometry, Mass, and Glycogen
Length, height, and width of all shells were measured to the nearest 0.01 mm with digital calipers.Measurement error for shell length, calculated by repeated measurements 10 times on 6 individuals ranging from 14 to 26 mm in length, was consistently small with a mean standard deviation of 0.075 mm and mean standard error of 0.048 mm.Shell strength was determined by mounting a single valve (dried 48 hours at 60oC) between two flat platens on an Instron load-testing machine and applying force was applied until the shell cracked and deformed at which time the force was recorded. Right and left valves collected during the October sampling were analyzed separately, but no difference between halves was found and thus only the average of the two is reported. The soft body tissue was dissected from shell for random sub-sample at least 25 individuals and the dry weight (DW) was measured after drying the sample for 48 hours at 60oC. Percent loss was used to calculate separate shell and soft tissue estimates of mg organic carbon by multiplying the percent loss after ignition at 550oC for 4 hrs by the total mass of each component. Glycogen content (mg per 22-mm individual) was estimated for whole animals using a modified acid hydrolysis-phenol method (sensu Naimo et al. 1998, Patterson et al. 1999, Okumura et al. 2002) on a separate sub-sample of 10 individuals that were then frozen until analysis.
Statistical Analysis
Analysis of covariance (ANCOVA) using shell length as the covariate and site and date as main effects plus the interactions was chosen for the statistical analysis. Adjusted or least square (LS) means for each site were calculated from untransformed data after they were checked for normality and homoscedasticity by visually examining for outliers and randomness in plots of residuals in a preliminary ANCOVA. The average (LS mean) shell length for the D. polymorpha from all site-date combinations was 22 mm, i.e. values presented are for this standardized size unless otherwise noted. Because D. bugensis was only found in sufficient numbers at the GL site, only June versus October ANOVA of shell length and month at this site were conducted for this species. None of the raw data met the assumptions of parametric statistics until a Box-Cox transformation was applied (Sokal and Rohlf 1995). LS means were tested for significant differences post-hoc, using the Tukey-Kramer method (Zar 1984) unless otherwise noted.
Despite site-specific differences in the length of D. polymorpha collected (Table 2.2), there was still significant variation in shell DW due to both site and month with shell organic content influenced only by site (Table 2.3). In June,OR and FE shells were lightest, those from the GL site intermediate and the heaviest were from the ST site (Figure 2.3). The same pattern was seen in October, except for a decline in shell DW at the GL site to a level equivalent to both the OR and FE sites (Figure 2.3). The force needed to crush a shell paralleled that of October shell DW, being greatest formusselsfrom GL and ST sites (Figure 2.3).In fact, of the species-site combinations analyzed, only D. bugensis from the GL site showed a positive change in shell organic content between June and October (Table 2.3). When the total shell mass of D. bugensis at the GL site was compared between June and October (shell length and month as factors), a significant relationship was present for size (F1,57= 1332.89, p < 0.01) but not month (F1,57= 3.19, p = 0.08). Statistical analysis shows this was due to a significant relationship with shell length (F1,57= 4.89, p = 0.03) not month (F1,57= 2.71, p = 0.11).
In contrast to shell characteristics, there was a consistent statistically significant interaction between month and site for all the tissue characteristics we measured. However the interaction was not consistent among sites or the three tissue characteristics. Both whole mussel DW and organic carbon content decreased at the GL and FE sites, while there was either an increase or no change at the OR and ST sites (Table 2.4, and Figure 2.4). The opposite was true for whole mussel glycogen content which increased at the GL and FE sites, while declining at both the OR and ST sites (Figure 2.4). Surprisingly, initial soft-tissue DW of mussels at the FE site was 50% greater than for other sites in June, but fell to similar levels in October (Figure 2.4). Initial tissue DW from both Lac Saint-Pierre sites (OR and ST) was equivalent, but the mussels at the ST site became significantly heavier by the October sampling while those at the OR changed little during the summer (Figure 2.4). Additionally, length-adjusted soft-tissue DW at the GL was similar to that of the two Lac Saint-Pierre sites D. polymorpha, but, in contrast to those sites, both species at the GL exhibited seasonal declines (not significant in the case of D. bugensis) that were more consistent with that seen for mussels collected from the FE site. The tissue organic carbon content of D. polymorpha was significantly different but highly variable (Tables 2.4 and 2.5). In fact, only mussels from the OR and ST site had increasing tissue organic carbon content between months (Table 2.4). There was also month and site interaction for whole animal glycogen content for D. polymorpha (Table 2.4). Glycogen levels increased over the summer at the GL site and declined at both the OR and ST sites. Although not significantly different from GL and OR sites, glycogen levels were lowest at the FE site and did not change seasonally (Figure 2.4). When co-occurring D. polymorpha and D. bugensis are contrasted (GL and GLQ), the D. bugensis values were initially higher but unlike GL D. polymorpha, their glycogen did not vary with either shell length(F1,22= 1.46, p = 0.24) or month (F1,22= 0.03, p = 0.85,Figure 2.4).When the dry tissue mass of D. bugensis were compared between June and October at the GL site (shell length and month as factors), a significant relationship was present for size (F1,54= 515.33, p < 0.01) but not month (F1,54= 2.92, p = 0.09). Similarly when organic content of D. bugensis tissue was compared between June and October (shell length and month as factors), a significant relationship was present for size (F1,54= 4.36, p = 0.04) but not month (F1,54= 4.33, p = 0.04).
One of our original goals was to determine if there were spatial patterns in mussel performance that correspond to the abiotic template presented by the St. Lawrence River. Field studies from the lower Mississippi, Rhine, and Ohio Rivers show that most of the increases in shell and soft tissue mass occur on the rising limb of the annual temperature hydrograph, peaking between 10 and 20oC (Sellers 1997, Jantz & Neumann 1998, Thorp et al. 2002). The higher growth rates at the spring maxima can be attributed to a window of elevated levels of energy rich food (i.e., the spring diatom bloom) combined with reduced costs of respiration and maintenance metabolism at moderate ambient temperatures. After this period of high efficiency, growth rate typically declines as temperatures approach either summer or winter extremes. Organisms living in sub-optimal environments often have to pay a bioenergetic price for resisting or tolerating stress, typically adjusting one aspect of their physiology to compensate for demands placed on another (Garland and Carter 1994, Spicer and Gaston 1999, Schluter 2000). Analysis of shell and soft tissue separately allowed us to see an apparent lack of direct positive or negative relationship between the shell and soft tissue compartments for D. polymorpha at two out of four sites.
The mollusc shell is an important physiological component whose total mass (primarily but not exclusively inorganic) is accumulated over time. An organic-rich shell is more expensive to fabricate (Palmer 1981, 1983), and studies have demonstrated that both shell mineral content and strength are plastic characteristics subject to subtle changes in water chemistry and ecological environment (Rosenberg et al. 2001, Beadman et al. 2003). Experiments have shown that freshwater gastropods grown under low calcium conditions have lower crushing resistance, implying that shell strength is a function of environmental calcium levels (Brodersen and Madsen 2003). Based on this result and the low calcium content at the OR site, we hypothesized this would similarly be true for Dreissena. The crushing force required to deform a mussel from the Ottawa River plume was about 30 % less than for the Great Lakes or southern tributary mussels, about 15% less than needed for estuarine animals, and roughly the same as needed to break the lighter D. bugensis shells. Broderson and Madsen (2003) also noted that in addition to reduced shell calcium content, shell organic content was lower. They attributed this to compensation for the increased energetic cost of dealing with calcium deposition under conditions of low availability. We did not find a similar result. Instead, organic content of shells from the OR site was intermediate, possibly because the low calciumOttawa River plume is located in a productive section of the river (i.e. Lac Saint-Pierre provides increased productivity, potentially compensating for lower ambient calciumavailability in the OR water mass, Basu et al. 2000). We found spatial variation in the condition of the shells (the organic composition and strength), but little difference in shell dry mass environments we hypothesized would be stressful (the brackish OR and FE waters) and those that should be more benign. However, because there is not an obvious environmental limitation such as low calcium corresponding to the lower shell mass, we cannot conclude whether this pattern is due to differential environmental limitations or some sort of sampling artifact.
Unlike studies in other large rivers (Sellers 1997, Jantz and Neumann 1998, Thorp et al. 2002) we did not find a uniform summer spawning-related decline in biomass simultaneous across all of the sites we compared. Instead there was a complex geographic pattern, seen both graphically and in the significant interaction terms in our statistical analysis, which did not easily correspond to water mass. We had initially hypothesized that the fluvial estuary would be the most difficult environment, but that for those individuals capable of tolerating these conditions, the cost of the higher turbidity, salinity, and current found in this freshwater-marine transition zone (Alexander et al. 1994, Dietz et al. 1996, Summers et al. 1996, Wilcox and Dietz 1998) would be offset by access to some of the richest particulate and dissolved food resources in the river (Vincent et al. 1996). Our results support this hypothesis in that estuarine mussels finish the summer have the second heaviest tissue DW and organ carbon mass even though they also have the largest decline in tissue mass, lowest glycogen content, and lightest shells. Based on this we can speculate that the energy associated with the higher estuarine primary productivity is either not being assimilated or is going to a process other than accumulation of shell, tissue, or glycogen. Yet at the other end of the riverine gradient there was not as simple positive connection among measures of condition we analyzed and the local environment. The condition of mussels from the Great Lakes water mass, representing a local environment where Dreissena has been highly successful, not uniformly better; despite increasing higher glycogen levels and greater shell strength, D. polymorpha from the GL site both started and finished the summer with lower total and organic carbon tissue mass.
These results are congruent with the view that under difficult environmental conditions Dreissena allocates energy to production of soft tissue, gametes first followed by somatic, whenever it is available as opposed to relying on stored energy or reduced metabolic expenditure ( Chase and Bailey 1999 , Stoeckmann and Garton 2001,McMahon 2002). This lack of a consistent one-to-one correspondence suggests that while there was a seasonal cost to both the Great Lakes water mass and fluvial estuary (i.e. declining mass), there were also secondary and differing site-specific variables that affected tissue characteristics in different ways. We saw no evidence of a trade-off between shell and soft tissue. Instead we showed that in this river, regardless of local site characteristics, the soft tissue compartment alone drives change in whole animal carbon. Another possible unexplored explanation for apparent lack of connection could be either the use of an alternative bioenergetic currency such as protein or high-energy lipids for storage. Studies using Dreissena (Walz 1978) and Mytilus (Hawkins and Bayne 1991) have postulated that lipid concentrations may have a disproportionally large role in growth despite their generally lower tissue concentrations relative to glycogen. Other research indicates that faster-growing mytilid mussels have lower maintenance metabolisms because protein turnover is reduced in relation to energy assimilation (Hawkins and Bayne 1991, Hawkins and Day 1996). A third possibility stems from our use of larger, presumably older mussels that grow less than smaller, younger individuals. A laboratory study with larger Dreissena found the majority of energy goes to reproduction (Stoeckmann and Garton 1997) that might have obscured our focus on changes in mass. Any of these three possibilities could potentially explain why no decline in organic carbon content or glycogen was seen in the mussels at the southern tributary plume where summer growth in soft-tissue occurred. We attempted to verify this by comparing mussels from adjacent water masses in Lac St-Pierre (Ottawa River and southern tributary) that have much more equivalent rates of primary productivity (Centre Saint-Laurent 1996, Frenette et al. 2003). Despite the contrasts in ionic composition due to watershed and tributary influences, these two water masses are separated by 5 km or less and are very similar in substrate, depth, productivity, and hydrology (Frenette et al. 2003, Thorp et al. 2005). In this way we limited the confounding influences in order to focus on the differences to hypo-ionic conditions more narrowly. Size-adjusted dry weight of southern tributary animals increased although organic content was unchanged whereas the organic content of the Ottawa River mussels increased yet their dry weight was unchanged. While the Ottawa River animals started with lower glycogen stores, the relative change in glycogen at both locations was similar. We conclude that energy stored as glycogen did not go to the accumulation of biomass, but to another aspect of bioenergetic demands such as calciummetabolism, respiration, or reproduction.
Dreissena bugensis is morphologically and ecologically very similar to D. polymorpha (McMahon 2002). Even though D. bugensis is believed to have arrived in the Great Lakes system later and has been spreading more slowly, it has been displacing D. polymorpha in many locations (Mills et al. 1999, Ricciardi and Whorisky 2004, Wilson et al. 2006) suggesting they have a competitive advantage. From our work it appears that the species-specific differences in soft tissue growth and metabolism noted from the laboratory (Baldwin et al. 2002, Stoeckmann 2003) do not clearly differentiate these two species in situ in the Great Lakes water mass, where the two species have equivalent size-adjusted soft tissue weight, seasonal patterns of change in organic content and glycogen reserves. However the thinner, lighter shells of D. bugensis shouldrequire lower bioenergetic investment in their fabrication, implying that more energy can be put towards higher growth rates (MacIsaac 1994, Baldwin et al. 2002 , Stoeckmann 2003). This idea of a species difference in energy allocation is further supported by observations of a greater loss soft-tissue mass in D. polymorpha compared to co-occurring D. bugensis the summer. Even in the river where the changes are not as well documented as in the Great Lakes, quagga mussels have gone from a being a rarity to comprising 50% or more of co-occurring animals upriver of Montreal within a 6-y period (Ricciardi and Whoriskey 2004). Despite the general increase in the abundance of quagga mussels upriver, repeated sampling during the same period in the fresh-water and brackish portions of the estuary found <5% D. bugensis. We speculate that the rarity of D. bugensis in the fluvial estuary could indicate either the presence of important physiological differences between the ability of the adults of the two species to tolerate estuarine instead of riverine conditions or, alternatively, that due to species differences in dispersal, D. bugensis has simply not yet arrived in the estuary.
We show that dreissenid physiological characteristics varied across the environmental mosaic created by the water masses of the St. Lawrence River. Whereas this situation can promote the development of divergent and heritable population characteristics, a common microevolutionary response seen in other model systems (Schluter 2000), we have only focused on the adult mussels in this study.Although large lowland river systems are often perceived as being well mixed with strong downstream transport, there are a variety of hydrologic mechanisms that can lead to retention or at least greatly decreased advective transport, of planktonic organisms and life stages (Stoeckel et al. 1997, Speirs and Gurney 2001, Schneider et al. 2003, Carr et al. 2004). Thus while it is clear that there are localized differences, it is not clear from our results whether the site-specific differences are due to either physiological plasticity or pre- or post-larval differentiation. A second conclusion is that shell and tissue physiology may be simultaneously responding to different spatially-varying environmental constraints and thus may be uncoupled instead of being accurate predictors of each other. Shell length is the most widely used measure for estimating Dreissena population demographics, biomass and production, yet our data show the relationship between shell and tissue can be both specific to the environment and, independently, the season. The important implication is that using a single formula from an unrelated location to estimate population density-environmental impact relationships at widely varying locations may introduce significant errors (sensu Young et al. 1996). Geographic variation in the shell-tissue relationship can have ecosystem level ramifications because biomass and related physiological condition can also be an additional covariable for explaining ecological impacts. One example of this is the allometric pattern that larger, heavier mussels have a greater filtration rates (Lei et al. 1996, Young et al. 1996, Diggins 2001). Thus allometric or demographic differences among populations may modify the functional role of mussels in large rivers, especially the coupling of the pelagic and benthic compartments through the filtration of suspended particles (Welker and Walz 1998, Ietswaart et al. 1999, Strayer et al. 1999, Thorp and Casper 2002, 2003).
Table 2.1. Characteristics of the St. Lawrence River water masses as they relate to the physiological tolerances of adult Dreissena (GL = Great Lakes, OR = Ottawa River, ST = southern tributaries, FE = fluvial estuary, NA = data not available).
|
|
Table 2.2. Mean and standard error of shell lengths (mm) of mussels analyzed for shell and soft tissue characteristics from each site-month from the principal water masses of the St. Lawrence River (Q = D. bugensis, GL = Great Lakes, OR = Ottawa River, ST = southern tributaries, FE = fluvial estuary).
|
|
Table 2.3. ANCOVA model statements, mean squares, and F values for analysis of Box-Cox transformed shell characteristics of D. polymorpha (LS mean = 22 mm, * indicates significance atα0.05, NA = not applicable because crushing force was not determined for mussels collected in October).
|
|
Table 2.4. Assessment of changes between shell and soft tissue physiological components of a 22 mm D. polymorpha and D. bugensis based on changes in the organic carbon content between June and October (Q = D. bugensis, GL = Great Lakes, OR = Ottawa River, ST = southern tributaries, FE = fluvial estuary).
|
|
Table 2.5. ANCOVA model statements, mean squares, and F values for analysis of Box-Cox transformed tissue dry weight (mg), organic carbon content (mg), and whole mussel glycogen content (mg) of a 22-mm D. polymorpha (* = significance atα0.05). Note: The glycogen analysis was done on a separate sample of mussels, thus whereas the model is unchanged, though the error degrees of freedom becomes 72.
|
|
Figure 2.1. Map of the St. Lawrence river-estuary system with the general location of sample sites are indicated. GL for the Great Lakes water mass near Massena, NY, OR for the Ottawa River water mass along the north shore of Lac Saint-Pierre, ST for the southern tributary water mass along the south shore of Lac Saint-Pierre, and FE for the fluvial estuary at the eastern tip of Île d’Orléans, Québec.
Figure 2.2. Continuous water temperatures July 10 - September 4, 2002 for the Great Lakes water mass (GL) near Massena, NY, the southern tributary water mass (ST) along the south shore of Lac Saint-Pierre, and the fluvial estuary (FE) at the eastern tip of Île d’Orléans, Québec.
Figure 2.3. Size corrected (LS means) for shell dry weight and the force (in Newtons) needed to crush one valve of a 22-mm Dreissena polymorpha or D. bugensis from the principal water masses of the St. Lawrence River. Crushing forces was determined for shells sampled in October only.(Letters indicate significant differences between site means, * Indicates significant difference between June and October at a site, Tukey-Kramerα= 0.05, Q = D. bugensis, GL = Great Lakes, OR = Ottawa River, ST = southern tributaries, FE = fluvial estuary).
Figure 2.4. Size corrected (LS means) for tissue dry weight and whole animal glycogen content for a 22-mm Dreissena polymorpha or D. bugensis from the principal water masses of the St. Lawrence River (Letter indicates differences between initial site means, * indicates significant difference between June and October at a site, Tukey-Kramerα= 0.05, Q = D. bugensis, GL = Great Lakes, OR = Ottawa River, ST = southern tributaries, FE = fluvial estuary).
© Andrew F Casper, 2007
| | | |
| Chapitre I. Introduction générale |  | Chapitre III. Life at the edge: Physiological constraints on freshwater mussels (Dreissena polymorpha Pallas and D. bugensis Andrusov) in a fluvial estuary |
