Influence of local adaptation and interstock hybridization on the cardiovascular performance of largemouth bass Micropterus salmoides

Studies in fishes have documented that substantial genetic variation can exist within a single species (Ryman and Utter, 1987; Philipp et al.,1983), and support the notion that each population has evolved an adaptive suite of genotypic and phenotypic characters for its specific environment (Wright, 1978). These studies also show that differences in performance and fitness occur when these populations are translocated from their point of origin to a novel environment (e.g. Philipp and Whitt,1991; Philipp and Claussen,1995; Philipp et al.,1995; Gharrett et al.,1999). Furthermore, introgression of genetically distinct populations may result in a decrease in overall fitness and performance characteristics (e.g. growth, thermal tolerance, resistance to infectious agents) of the new population (i.e. outbreeding depression) providing further support to the concept of local adaptation(Campton, 1987; Emlen, 1991; Hallerman, 2003; Philipp et al., 2002).

To date, few studies have examined biological differences among fish stocks delineated from the results of molecular genetic analyses, and even fewer have examined the physiological variation among putative stocks or the consequences of mixing them. Theoretically, one would predict that physiological variables would differ among locally adapted and genetically distinct populations, and that mixing individuals from different populations would result in alterations in these variables that could provide a mechanistic explanation for reductions in fitness. Interestingly, there have been few attempts to evaluate population-specific differences in cardiovascular performance, or the cardiovascular consequences of hybridizing fish from different populations. This is surprising since cardiovascular performance is intimately linked to the energetics and physical performance of fish(Thorarensen et al., 1996) and serves as a sensitive and logical technique for assessing intraspecific variation (Gamperl and Farrell,2004). Furthermore, there is a growing interest in intraspecific variation in fish cardiovascular function(Gamperl and Farrell, 2004),and more broadly in the physiological diversity of animals(Spicer and Gaston, 1999).

In this study, we used two genetically distinct stocks of largemouth bass Micropterus salmoides Lacépède that were collected from different geographic regions in the upper midwestern United States to: (1)assess whether cardiovascular function differs in bass from geographic locations with different climatic conditions and (2) quantify the impact that interstock hybridization has on this important physiological characteristic. For both comparisons, we monitored the cardiovascular response of individuals to exhaustive exercise at temperatures of 10 and 20°C.

This experiment took advantage of an extensive population genetic analysis of largemouth bass Micropterus salmoides Lacépède (and other species of fish) throughout the upper midwest USA(Fields et al., 1997). From that molecular genetic study we were able to propose geographic boundaries for 15 putative stocks of largemouth bass in Illinois, Minnesota and Wisconsin. That stock delineation process was based on the distribution of different alleles encoded at variable protein loci, based on starch gel electrophoresis,combined with the distribution of different mitochondrial DNA haplotypes,based on variable restriction endonuclease digestion sites(Fields et al., 1997).

During fall 1997 adult largemouth bass that were fixed for alternative homozygous genotypes at a single locus (MDH-B) were selected from each of two of the proposed stocks: (1) IL: from Central Illinois: Lake Shelbyville in the Kaskaskia River Drainage within the Mississippi River Basin;Genotype=MDH-B²B²; (2) WI: from Southeastern Wisconsin:Big Cedar Lake in the Lake Michigan Drainage within the Great Lakes Basin;Genotype=MDH-B¹B¹.

The WI fish came from a less temperate environment with mean annual temperatures (1971-2000) of 7.5°C and 63.4 days with maximum temperatures<0°C. The IL fish experienced mean annual temperatures (1971-2000) of 11.3°C and 33.9 days with maximum temperatures <0°C. In the spring of 1998, we stocked adult fish (10 males and 12 females, each with the appropriate homozygous genotype) from one or the other stock into each of eight 0.04 ha clay-lined, earthen ponds at the Illinois Natural History Survey Aquatic Research Field Laboratory in Champaign, IL, USA to produce four distinct, genetically tagged experimental stocks: each of the two pure parental (P₁) stocks (IL×IL and WI×WI), as well as both reciprocal F₁ interstock hybrids (WI×IL and IL×WI). In the fall of 1998, production ponds were drained. Fingerlings from each stock were given differential fin clips for external identification and then stocked into a series of ponds in a set of common garden experiments designed to assess their relative survival and growth, as well as their relative fitness(lifetime reproductive success) after they matured. Fish fed on natural invertebrate forage, as well as fathead minnows Pimephales promelasand juvenile bluegill Lepomis macrochirus produced in the ponds. The artificial ponds were substantially smaller than the lentic systems where fish were originally captured; however, the ponds contained substrates and vegetation typical of natural Midwestern systems and were subject to seasonal variation in temperature and photoperiod.

Twice during fall 2000 (i.e. when ambient water temperatures corresponded to 10 and 20°C), some individuals from each stock were seined from the ponds and held in raceways for 1 week prior to experimentation. Although water temperatures were relatively stable for a period of approximately 2 weeks prior to experimentation, they did vary up to 3°C on a diel basis. During the experiments, water temperatures were controlled so that they varied by no more than 1°C. All experiments were conducted between 10:00 h and 16:00 h. Fish were exposed to the natural photoperiods of Champaign, Illinois during residency in pond and raceway environments, as well as during experimentation. As a result, fish were field-acclimated to conditions in central Illinois. Experiments were conducted between 18 September and 4 December, 2000.

Detailed descriptions of the surgical procedure are provided elsewhere(Schreer et al., 2001; Cooke et al., 2003b). Briefly,each fish was anaesthetized prior to surgery with 60 p.p.m. clove oil(emulsified with ethanol, 9:1 ethanol: clove oil) for approximately 8 min, at which point the fish had lost equilibrium and was non-responsive. Once anaesthetized, each fish was placed on its side on a wetted sponge. The anaesthetized state was maintained during surgery by irrigating the gills with water containing a maintenance concentration of anesthetic (30 p.p.m. clove oil). Connective tissue surrounding the aorta was carefully removed, and a flexible silicone cuff-type Doppler flow probe (sizes from 0.9 to 1.2 mm)subminiature 20 MHz piezoelectric transducer: Iowa Doppler Products, Iowa City, IA, USA) was placed around the aorta. The cuff was secured around the vessel using a single suture, and the lead wire leading to the probe was then sutured to the body of the fish at several locations.

Following surgery, individual fish were placed immediately into a 70 liter tank (50 cm×50 cm) and monitored closely until they had regained equilibrium. Fish were allowed to recover from surgery and to acclimate to the tank for at least 18 h (Cooke et al.,2003b). A darkened area covering approximately 30% of the tank provided cover and ensured that the fish were not disturbed by general laboratory activity. The experimental tanks were continuously supplied with pond water at either 10 or 20°C. To elicit exercise, fish were chased around the tank by hand (Cooke et al.,2003b) until they reached exhaustion; a standard technique in fish physiology (Kieffer, 2000). At this point the fish would no longer swim and would lose equilibrium. Cardiac variables were recorded continuously for at least 1 h prior to the exercise period (the resting period), during the exercise period, and for at least 6 h post-exercise (the recovery period).

Following experimentation, fish were killed with an overdose of anesthetic(180 p.p.m. clove oil), and a postmortem calibration was conducted to convert Doppler shift (in V) to actual blood flow (ml min^-1; Cooke et al., 2003b). The bulbous arteriosus was catheterized with tubing (PE 120) and a constant infusion pump (Harvard Apparatus, South Natick, MA, USA) was used to perfuse anticoagulated blood (2 g sodium oxalate + 0.4 g sodium chloride + 10 ml distilled H₂O l^-1 pig's blood) through the aorta and partial gill arches to maintain output pressure. This procedure permitted the calibration of the probes over a range of flow rates encompassing the rates recorded during the trials. Reference flow rates were analyzed with linear least-squares regression (mean r²=0.971). The ventricles were patted dry and weighed to the nearest 0.001 g and were mass corrected for the size of the fish to represent relative ventricular mass (RVM). To examine the temperature dependence of stock-specific resting cardiac variables, we calculated Q₁₀ rates(Schmidt-Nielsen, 1997).

A flowmeter (545C-4 Directional Pulsed Doppler Flowmeter: Bioengineering,The University of Iowa, Iowa City, IA, USA) and a digital strip-chart recorder(LabVIEW, Version 4.0.1, National Instruments Corporation, Austin, TX, USA)were used to record real time data on cardiac output. Heart rate(fh) was determined by counting the number of heart beats(flow trace peaks) over a 60 s period and stroke volume(Vs) was calculated as cardiac output(Q̇)/fh.

To determine recovery times, traces for Q̇, fh and Vs, adjusted to resting (100%), were plotted for each fish and evaluated visually. A fish was considered to be recovered when values returned to resting and became stable (within 10 % of resting values; Schreer et al., 2001; Cooke et al., 2003b). Maximal disturbance was determined as the greatest change in a cardiac variable(either positive, >100% basal or negative, <100% basal) during the recovery period. Data were visually assessed for normality using quantile plots and homogeneity of variance using residual plots (SYSTAT, V8.0, SAS Institute). The premise of all analyses was to test two null hypotheses: (1)that there were no differences in cardiovascular performance among molecularly defined P₁ stocks and (2) there were no differences in performance of F₁ hybrids compared to the P₁ stocks. All tests were conducted using two-way analysis of variance with stock being the main effect and water temperature the secondary effect (JMPIN, V4.01, SAS Institute). We used planned contrasts to examine where specific differences of interest occurred. All values reported are means ± s.e.m. Tests were considered significant at P=0.05.

Although fish used in the experiments were all age-2 and generally of similar size, both length and mass were observed to vary with temperature and by stock (Tables 1, 2). RVM was generally similar among stocks at both temperatures (Tables 1, 2). However, an interaction revealed that at 20°C IL×WI fish had higher RVM than all other stocks.

Cardiac output and heart rate values were significantly lower at 10 than 20°C for all four stocks, whereas stroke volume, differed significantly between 10 and 20°C only for the IL×WI stock(Table 2, Fig. 1). Resting cardiac output for the IL×IL stock was consistently lower than for the other three stocks. The only departure from this pattern was that at 20°C, resting cardiac output rates were similar between both P₁ stocks. Heart rate followed a similar pattern to cardiac output, with the IL×IL fish exhibiting the lowest resting heart rates; however, fewer stock-specific differences were noted (Fig. 1). Resting stroke volume values did not vary significantly among stocks (Table 2, Fig. 1). Q₁₀ values for resting cardiac output and heart rate ranged from 1.33 to 1.65, and were highest for IL×IL fish (Fig. 1). Resting stroke volume Q₁₀ values ranged from 0.89 to 1.01, with values being highest (i.e. close to 1) for both P₁stocks (Fig. 1).

IL×WI and WI×WI fish both exhausted more rapidly at 20°C than at 10°C, whereas there was no temperature-specific difference among other stocks (Tables 1, 2). There was, however, a strong but nonsignificant trend towards the IL×IL fish taking longer to exhaust than the other stocks at 20°C(Table 2). In general, maximal cardiac output and heart rate were significantly higher at 20°C than at 10°C, whereas stroke volume was only significantly different among IL×IL and WI×WI fish across the two temperatures(Table 2, Fig. 2). Maximal cardiac output values were lower for IL×IL fish than for IL×WI fish. For heart rate, the only significant difference detected was between IL×IL and WI×IL at 20°C (Fig. 2). Maximal change in stroke volume did not differ among stocks at either temperature (Table 2, Fig. 2). Scope for cardiac output at 10°C for both WI×IL and WI×WI were higher than at 20°C (Tables 1, 2). No other differences were noted for cardiac scope (Tables 1, 2).

Cardiac recovery times for IL×WI fish varied by temperature for heart rate and cardiac output, but in the opposite direction to what was observed for WI×WI fish (Fig. 3). At 10°C, cardiac output recovered similarly quickly for both P₁stocks, but significantly more slowly, for both hybrids. At 20°C, cardiac output for the IL×IL stock recovered more rapidly than all other stocks. The same temperature-specific patterns of recovery observed for cardiac output were also observed for heart rate (Tables 1, 2, Fig. 3). Stroke volume at 10°C recovered most rapidly for both P₁ stocks, although the WI×WI recovery rate did not differ from the WI×IL stock(Table 2, Fig. 3). At 20°C, the IL×IL stock recovered more rapidly than the WI×IL and WI×WI stock.

There is a paucity of research on population-specific differences in the cardiovascular physiology of fishes (see Gamperl and Farrell, 2004). However, in well-studied species such as rainbow trout Oncorhynchus mykiss, cardiovascular variables and responses vary widely among studies perhaps reflecting population level differences (e.g. Thorarensen et al., 1996; Brodeur et al., 2001) that clearly exist for swimming performance and metabolic rate(Taylor and McPhail, 1985; Lee et al., 2003). Formal analyses of population level differences in cardiovascular performance are limited to two recent accounts for largemouth bass(Cooke et al., 2003b) and rock bass Ambloplites rupestris (Bunt et al., 2004), but neither study was designed primarily to assess intra-specific variation in a controlled manner. For the first time, we present data from a common garden experiment designed to explicitly evaluate the role of local adaptation on basic cardiovascular function using two genetically distinct stocks of largemouth bass and their reciprocal hybrids.

Our data illustrate that when translocated to novel environments, there are substantial differences in the cardiovascular performance of these largemouth bass relative to locally adapted individuals in the recipient population. Furthermore, when these two stocks were mixed, the interstock hybrids performed either similarly to the Wisconsin stock or more poorly than both P₁ stocks. These data suggest that interstock hybrids did not experience heterosis (i.e. hybrid vigor; Emlen, 1991; Thornhill, 1993) and instead exhibited performance indicative of a brake down of co-adapted gene complexes(Templeton, 1986; Hallerman, 2003) even at the F₁ generation. These data emphasize the fact that cardiovascular performance characteristics exist in locally adapted populations and that outbreeding occurs when fish from different populations are mixed. In Illinois, locally adapted bass generally exhibited the lowest resting cardiac activity, the lowest maximal cardiac activity in response to exhaustive exercise, and recovered almost 25-35% faster than other stocks examined. Interestingly, at the lower temperature (10°C), the northern P₁stock from Wisconsin recovered in the same time period as did the more southern P₁ stock (IL×IL). At the warmer temperature(20°C), however, the northern stock took longer to recover than the southern stock. Data presented here provide clearer evidence than has been previously reported for largemouth bass using swimming performance and respirometry (i.e. Cooke et al.,2001) that the translocation of fish and their subsequent hybridization with locally adapted stocks has physiological consequences that may have direct fitness implications, since cardiovascular performance is a determinant of aerobic swimming ability(Keen and Farrell, 1994; Kolok and Farrell, 1994).

When we evaluated resting cardiac activity, cardiac output and heart rate values for IL×IL fish were consistent with previous studies on largemouth bass (Cooke et al.,2003b) and were approximately 30% less than both of the interstock hybrids. The elevated resting cardiac output and heart rate of the WI×WI fish and in particular, the interstock hybrids, could result in greater myocardial energetic costs, but this would be minimal in the overall energetic budget as myocardial metabolism is generally less than 4.5% of total metabolic rate (Farrell and Jones,1992). The maximal values for cardiac output and heart rate were also frequently higher for the interstock hybrids than the P₁stocks, potentially yielding additional energy costs. However, without further analyses of oxygen uptake, transport and use, it is not possible to determine if the bioenergetic consequences of these cardiovascular performance differences are ecologically meaningful. Despite having higher resting rates and maximal rates, the scope for cardiac output and heart rate were similar among all four stocks. The interstock hybrids, therefore, maintained scope,but likely did so with a higher cost because they were operating at higher resting rates. Largemouth bass (Cooke et al., 2003a,b)and other centrarchid fishes (e.g. smallmouth bass Micropterus dolomieu, Schreer et al.,2001; black crappie Pomoxis nigromaculatus, Cooke et al., 2003a) are frequency modulators, responding to performance challenges typically by increasing heart rate and decreasing stroke volume. In this study, the pattern of largemouth bass being frequency modulators was maintained in the translocated stock (WI×WI) as well as the interstock hybrids and was similar to that of the IL×IL fish.

When exposed to a performance challenge such as exhaustive exercise,recovery times provide information on the duration that individual fish are forced to deal with the challenge, precluding them from allocating resources to other functions. In our study, recovery times were consistently about 25-35% longer for both interstock hybrids than for the Illinois stock. Previous work on largemouth bass revealed that cardiac recovery time was remarkably consistent across temperatures when exposed to 150 s of exercise and 30 s of air (∼135 min for Q̇ and fh and ∼110 min for Vs; Cooke et al., 2003b). In the current study, the largemouth bass cardiac recovery times for translocated and interstock hybrids (∼110 min for Q̇and fh and ∼70 min for Vs) were more variable than those of Cooke et al.(2003b). Although these were more rapid recovery times than the previous research (i.e. Cooke et al., 2003b), in the current study fish were not exposed to air, which is an additional stressor. Furthermore, when contrasted with the relatively rapid recovery of the locally adapted Illinois stock, the recovery times of the Wisconsin stock and interstock hybrids are indicative of major physiological differences in how they respond to exercise. During this period of lengthened recovery,interstock hybrids could exhibit reduced digestion(Randall and Daxboeck, 1982)and food intake (Smart, 1981; Barton and Schreck, 1987),negatively affecting acquisition of new energy resources. Furthermore, because cardiovascular performance is strongly correlated with metabolic rate, when the fish are at an elevated stage on their scope, fish may have less scope for activity available to escape predators or deal with other stressors (Priede, 1977, 1985). Although cardiac performance is a factor in determining oxygen consumption and metabolic rate,there are also other factors that we did not measure here that could exacerbate or mitigate the consequences of intra-specific variation in cardiovascular performance.

Evidence suggesting that inbreeding and outbreeding depression may be manifested as alterations in cardiovascular performance in higher organisms comes from selective breeding experiments with mice Mus domesticus(e.g. Mattson, 2001) or observations of high rates of cardiac abnormalities in some inbred organisms such as Florida panther Felis concolor coryi(Roelke et al., 1993). To our knowledge, however, no information on the cardiovascular consequences of inbreeding or outbreeding currently exist for teleost fish(Waldman and McKinnan, 1993). The only research that may be relevant is the assessment of cardiovascular performance in transgenic and domesticated fish. Triploid brown trout Salmo trutta appeared to have reduced factorial metabolic scope relative to diploid fish (Altimiras et al.,2002), while domesticated rainbow trout reared in a hatchery did not exhibit appreciable differences in cardiac morphometrics or performance relative to wild fish (Gamperl and Farrell, 2004). While these studies did evaluate the cardiovascular consequences of selective breeding, they did not, however,evaluate the effects of interstock hybridization. The approach used in our study differs in that it provides a possible proximal mechanism for reductions in fitness that can result from such hybridization. Because there were minimal differences in cardiac morphometrics among the treatment groups we examined,the variation in cardiac performance may represent disruption in humoral/neurological regulation and/or efficiency of enzymatic and other cellular processes, consistent with the breakdown of co-adapted gene complexes(Templeton, 1986).

In addition to providing insight into the fundamental physiological diversity and intraspecific cardiovascular performance of fish, this study also has implications for conservation scientists. Our findings suggest that molecular genetic techniques are able to delineate populations that have sufficiently different physiologies to warrant separate management and conservation activities (e.g. Nielsen,1995). The current study supported previous findings (i.e. fitness, Philipp and Claussen,1995; swimming performance, Cooke et al., 2001) in that the northern stock of largemouth bass (WI×WI) did not perform as well in Illinois as the Illinois stock. For largemouth bass, therefore, local adaptations in cardiovascular performance exist among populations that occur within a relatively small geographic region so translocations should me minimized. Finally, our results illustrate the negative consequences of mixing fish from different locally adapted populations. Future efforts focusing on fish from the F₂ generation will provide additional insight into intra-specific cardiovascular performance and the potential long-term consequences of fish translocation practices.

We thank K. Deters, C. Koschman, E. Grant, T. Kassler, and E. De La Corpuz for expert field assistance. This work was conducted in accordance with the University of Illinois Office of Lab Animal Research. The University of Illinois Research Board provided funding for purchase of equipment through support to D.P.P. and S.J.C. Additional funding for the project was provided by the Illinois Natural History Survey. S.J.C. was supported by a Natural Sciences and Engineering Research Council of Canada Julie Payette Postgraduate Fellowship.