Our objective is to use a full-sib half-sib experiment to determine the genetic component of variation in heat tolerance and the heritability of heat tolerance in the zebra mussel. We will meet the objective above by testing the following hypothesis. Heat tolerance has a genetic component and a portion of the phenotypic variability is due to additive genetic variance and is therefore, heritable.

Zebra mussels, Dreissena polymorpha , have established populations in the Laurentian Great Lakes and in the Mississippi River drainage over the past 14 years (Ludyanskiy et al. 1993). During that time they have had considerable impact on unionid populations and on human endeavors (Holland 2000). Recently, it has been proposed that the continuing spread of these populations can be controlled using lethal temperatures (Harington et al. 1997, Mathews and McMahon 1999). Previous research by one of us has indicated that zebra mussels may be able to adapt to high temperature environments (Elderkin 2001). If this is the case, then high temperatures will be ineffective as a method of control.

In order for selection to act on a phenotypic character such as heat tolerance, there must be variation for the proposed character within and among populations (Falconer 1981). Variation in phenotypic characters can be divided into three components: variation due to maternal effects (VM), variation due to environmental effects (VE), and variation due to genetic components (VG). Selection acts on a sub-factor of the genetic variation of a phenotypic character called additive variance (VA). A measure of the amount of additive genetic variation contributing to a phenotypic character is called heritability (h2).

Falconer (1981) has proposed that using a full-sib half-sib experimental design is most appropriate for determining heritability of a character. Using this approach, a researcher mates one male with several females to create half-sib families among females, and all families are kept in a "common garden", where all individuals are kept under the same conditions. This design is useful because it eliminates most of the variation due to maternal effects (VM) and due to environmental factors (VE). We propose to spawn zebra mussels in the laboratory and combine males and females into half-sib families to determine if heat tolerance is a heritable character, and therefore has a genetic component. Previously, we completed a preliminary study of heritability of temperature tolerance on three full-sib families of zebra mussels. Our analysis revealed that the estimated heritability was 0.20 (±0.26 SE). We strongly suspect that the variation in these data reflect the small sample size and design of the experiment, and do not reflect the true heritability estimate of heat tolerance for zebra mussels. Power analysis confirmed that the power of this analysis was low (b = 0.12), and that the experiment should be repeated using a larger number of families. Using information published in Klein (1973, 1974) we determined that we could greatly increase the power of our estimate by using a full-sib half-sib design with 12 half-sib families producing 50 offspring from each female.

Lethal high temperature has been proposed as a control for zebra mussels. This method has been used in controlled environments, such as intake pipes. In addition, researchers have proposed that warm environments may naturally control the spread of these mussels (Strayer 1991). Our research has indicated that zebra mussels may have a capacity to rapidly adapt to these high temperature environments, and therefore, decrease the efficacy of high temperature control techniques. Further development of such control strategies requires understanding of both the physiological tolerances of zebra mussels, and their ability to adapt to the environmental changes that form the basis of these strategies.

In addition to our preliminary experiments indicating heritability of heat tolerance, we have performed experiments showing that zebra mussels may be adapting to high temperatures (Elderkin 2001). Zebra mussels show an allele frequency shift along the latitudinal gradient formed by the Mississippi River. This shift is seen in allele frequencies at the leucine aminopeptidease (Lap) locus for individuals from high-temperature (southern) populations (Elderkin et al. In Press). We also experimentally determined that mussels sampled from Baton Rouge, LA were more tolerant of high temperatures than mussels sampled from two northern locations. Both of these studies support the hypothesis that zebra mussels are under selection for heat tolerance. Therefore, it is important to determine the amount of additive genetic variance zebra mussels have relative to their heat tolerance. Until we know whether such tolerance is heritable, we cannot be confident that use of high temperatures will be a feasible strategy for control of zebra mussels.

To meet our objective, we will collect zebra mussels in spawning condition and maintain them in the laboratory at 9°C until the experiment starts. We have used this technique previously to maintain zebra mussels in spawning condition for several months. We will spawn > 100 individuals and combine gametes from one male with every three females to form half-sib families. In order to generate the maximum number of spawning individuals, mussels will be warmed to room temperature over several days. After this temperature acclimation, groups of 20 mussels will be placed into a 10^-3 solution of serotonin (5-hydroxytriptamine), which has been shown to induce spawning (Fong et al. 1995, Ram et al. 1996). After gametes are combined, zygotes will be kept in 1-L containers at 22°C for approximately 24 hrs. During this time, veligers are extremely sensitive and disturbance will be kept to a minimum. After veligers reach the D or hinge stage, we will expose family groups to a lethal temperature of 32°C. We will measure the number of individuals in each family surviving every 6 hours for 3 days and use survival analysis (SAS, Sas Institute, Cary NC) to determine the time-to-death (TTD) for each individual. Survival analysis is also helpful as it calculates the within and among group variation which is used to determine heritability.