Quentin Rougemont1, Anne Carrier2, Jeremy Le-luyer3, Anne-Laure Ferchaud1, John M. Farrell4, Daniel Hatin5, Philippe Brodeur6, Louis Bernatchez1
1Département de biologie, Institut de biologie intégrative et des systèmes (IBIS), Université Laval, G1V 0A6, Québec, Canada
2Département de techniques du milieu naturel, Centre d’études collégiales à Chibougamau, Cégep de Saint-Félicien, Chibougamau, G8P 2E9, Canada
3IFREMER, Unité Ressources Marines en Polynésie, Centre Océanologique du Pacifique – Vairao – BP 49 – 98179 Taravao – Tahiti – Polynésie Française
4Department of Environmental and Forest Biology, State University of New York, College of Environmental Science and Forestry, 13210, Syracuse, New York, USA.
5Ministère des Forêts, de la Faune et des Parcs, Direction de la gestion de la faune de l’Estrie-Montréal-Montérégie-Laval, 201, Place Charles-Le Moyne, Longueuil, Québec, J4K 2T5, Canada
6Ministère des Forêts, de la Faune et des Parcs, Direction de la gestion de la faune de la Mauricie et du Centre-du-Québec, 100, rue Laviolette, bureau 207, Trois-Rivières, Québec, G9A 5S9, Canada
Introduction
Over the past decades, an increasing number of fish species have undergone strong decrease in their abundance due to various human activities. Such activities may prevent the free movement of fish, generates pollution and habitat loss, overfishing and many additional problems. To overcome these demographic declines, numerous stocking programs have been implemented to sustain fish populations worldwide. This is the case of the Muskellunge (Esox masquinongy) in the province of Québec, Canada. The species is renowned for his trophy-size specimens which are highly prized by anglers. However, Muskellunge has undergone strong decline in abundance during the first half of the 20th century in the waters of the St. Lawrence River, especially in the greater Montréal region. Consequently, Muskellunge from Ontario and New York State were used for stocking over 1.5 million of individuals from 1950 to 1997. From 1950 to 1965, eggs initially taken from the Chautauqua Lake (New York State, USA) were transferred to the Lachine government hatchery in Québec where fry were reared before being released into the St. Lawrence River, several of its main tributaries and inland lakes. From 1965 to 1986, adults from Lake Joseph were used as source for stocking. Finally, from 1986 to 1997, eggs from Lake Tremblant were used. Muskellunge populations from Joseph and Tremblant Lakes were originally introduced with fish from the Lake Chautauqua source (see details about stocking history in Carrier et al.).
An optimal management of Muskellunge can only be achieved through a detailed understanding of its population structure and of the extant of connectivity between distinct populations. In particular, the existence of genetically differentiated populations with some level of reproductive isolation must be taken into account for sound conservation and management practices. Moreover, genetically distinct groups of fish may evolve to local adaptation in response to particular habitat characteristics (temperature, water chemistry, etc.). Those adaptations allow fish from distinct populations to optimize their reproduction and survival in a given habitat. It is therefore necessary to preserve the natural genetic variation present within a species in order to ensure its potential to evolve and survive in an ever changing environment. Such knowledge will be fundamental to define management units for fishery management, habitat protection and restoration, which is particularly important in highly connected systems such as the St. Lawrence River and its tributaries. Finally, genetics can inform on the extent of hybridization that may have occurred between wild, local fish and fish artificially introduced by stocking.
The genetic structure and diversity of Muskellunge within the St. Lawrence River, its major tributaries and inland lakes of Québec have never been studied. Therefore, the present study has been realised to : 1) evaluate the level of genetic structure in Muskellunge, 2) measure the impact of historical stocking on the genetic structure and diversity, and 3) define evolutionary significant units relevant for population management and to maintain a sustainable resource for angling.
Sampling and genetic characteristics
A total of 662 Muskellunge have been captured in 22 sites for approximately 24 fish per location (Figure 1). Those samples were essentially obtained with the help of professional fishing guides (Mr. Marc Thorpe, Mr. Mike Lazarus, and Mr. Michael Philips), their customers, volunteer sport fishermen, and wildlife technicians. A tiny portion of a pelvic fin was clipped for each fish (1 cm², 100 mg) and preserved in alcohol for further genetic analyses in the laboratory of L. Bernatchez at Laval University in Québec. All fish were released after capture.
This extensive sampling made it possible to cover the sections of the St. Lawrence River, from the Thousand Islands region to Lake Saint-Pierre, its major tributaries and some inland lakes of Québec. The major sources of stocking were also sampled : 1) Chautauqua (New York State) and Pigeon (Kawartha Lakes system, Ontario) Lakes, 2) Joseph Lake, and 3) Tremblant Lake. Muskellunge was introduced in these last two lakes and were used as sources for stocking a few years later. Finally, Lake Traverse located in the Mauricie region of Québec has never been stocked to our knowledge and was included in our study.
DNA from individual fish was extracted from the preserved biopsies in the laboratory. This DNA was then characterized using a new sequencing technology allowing to read each DNA variation over a large part of the fish genome. It was then possible to identify for each fish, over 16 000 genetic variants. Such variants were compared among individuals and among sampling sites, which allowed quantifying the genetic diversity of the species, its population structure and document the impact of past stocking events on the genetic makeup of wild populations.
Population genetic structure
The genetic results revealed a moderate level of genetic diversity compared with other freshwater fish species that have been studied using similar methods. The effective size of populations, estimated from genetic data, is the number of broods that reproduce effectively, thus transmitting their genetic background to their offspring. In general, the total number of fish in a population can be 10 to 100 times higher than the number of effective individuals. Effective population sizes were generally quite low among Muskellunge populations, especially for isolated lakes. In the St. Lawrence River, the estimate of the effective size was 669 for all sites grouped together. This value is considered moderate compared to other freshwater species, but reflects the unique characteristics of the Muskellunge life-cycle (high longevity, highest position in the food chain, solitary and territorial behaviours) and its typically low population density. These findings highlight the vulnerability of this species and the importance of applying specific protection measures to ensure its sustainability.
The measures of genetic differentiation and population structure suggest the existence of eight distinct genetic groups in the system under study. The first group includes the Muskellunge used as stocking sources and the sites directly derived from this source, namely Chautauqua, Joseph, Tremblant, Frontière, and Maskinongé Lakes, as well as the Chaudière and Saint-Maurice Rivers. This confirms the common origin of the muskellunge of these water bodies, all derived from the source of Lake Chautauqua (New York State). For the Chaudière and Saint-Maurice Rivers, available knowledge suggest that Muskellunge abundance was initially low in those systems, and that stocking would have established perennial populations. The second group corresponds to the l’Achigan River and the third group to the Yamaska River, which are genetically distinct from the St. Lawrence River. The fourth group consists of all sites within the St. Lawrence River, from the Thousand Islands to Lake Saint-Pierre. The fifth group is Lake des Deux-Montagnes, which is also genetically distinct from the Muskellunge of the whole St. Lawrence River. It is noteworthy that Muskellunge from Lake des Deux-Montagnes show a certain proportion of migration to Lake Saint-Louis. Most of these migratory individuals (83%) were found on the North shore of Lake Saint-Louis, which is fed by water coming from the Ottawa River. The sixth group consists of isolated lakes that have never been stocked, represented in this study by Lake Traverse. This body of water has a unique genetic makeup that needs to be preserved. The seventh group corresponds to Pigeon Lake (Kawartha Lakes system in Ontario), used for stocking to a lesser extent than other water bodies, and the eighth group is Lake Champlain.
Although the St. Lawrence River formed a single population, the genetic differentiation between individuals increased with distance between them. This pattern is a consequence of the geographically reduced dispersal of individuals across the entire St. Lawrence River. In addition, the extant of genetic variation observed in the St. Lawrence River proves to be continuous, that is, there are no real, highly differentiated genetic groups. This suggests that dispersal can occur freely from upstream to downstream, although it is obviously reduced upstream by the presence of the two major obstacles on the St. Lawrence River (Beauharnois and Moses-Saunders dams).
Stocking effects
Fine scale analysis of genetic mixing patterns allowed us to estimate the effect of stocking on the genetic structure of populations (Figure 2). This analysis revealed that stocking had very little effect on the genetic integrity of wild populations in the St. Lawrence River. Indeed, we found very little evidence of genetic mixing of Chatauqua, Joseph or Tremblant Lakes strains used as source populations. Conversely, there is evidence of pronounced genetic mixing in some lakes and tributaries of the St. Lawrence, despite the fact that they have in most cases received smaller quantities of stocked fish than the St. Lawrence River. This is the case for the Saint-Maurice and Chaudière Rivers, as well as for Maskinongé Lake, where there was a mixture of local (represented in black in Figure 2) and introduced (represented in green in Figure 2) genetic makeups. The main hypothesis likely to explain this pattern is that stocking has had variable effects depending on the initial size of the populations being stocked. In general, it is expected that stocking done with individuals from different genetic groups, in this case individuals from distant lakes (differences in climate and habitat types), is potentially ineffective due to lack of adaptation of stocked individuals to local conditions. It is therefore possible that the individuals stocked in the St. Lawrence River had a low reproductive success and/or that hybrids resulting from reproduction were poorly adapted to local conditions, ultimately showing a low survival rate. Thus, non-native Muskellunge may have been displaced in the St. Lawrence, which potentially had a larger population size than isolated lakes or tributaries.
Management implications
Our results suggest that from a genetic point of view, the entire St. Lawrence River, from the Thousand Islands region to Lake Saint-Pierre, can be considered as a single population within which genetic differentiation of individuals increase slightly with distance. Thus, a single management unit would be sufficient on the St. Lawrence River to ensure the maintenance of genetic diversity in this system. Of course, individuals who are isolated by impassable obstacles should be managed locally. This is particularly the case of Lake St-François, enclosed by dams upstream (Moses-Saunders) and downstream (Beauharnois). The second management unit includes Lake des Deux-Montagnes, which is genetically different from the St. Lawrence River population. The third group consists of the tributaries of the St. Lawrence River, each representing a distinct unit with some nuance depending on the abundance of Muskellunge prior to stocking. Thus, l’Achigan and Yamaska Rivers showed little evidence of hybridization with stocked fish while the Chaudière and Saint-Maurice Rivers have a more pronounced genetic mixing profile with the stocking sources. The fourth group consists of lakes stocked directly from Chautauqua Lake (Joseph, Tremblant, and Frontière Lakes) which all share a strong genetic similarity with Chautauqua Lake. The fifth group includes lakes into which the Muskellunge was initially present (Maskinongé and Champlain Lakes) where apparently only modest mixing occurred. Finally, Traverse Lake is one of the few, if not the only unstocked natural population in Québec with a unique genetic makeup.
In conclusion, in systems previously unoccupied by Muskellunge or with a very low density of individuals, stocking has made it possible to sustain local populations in the long term and therefore, have helped to enhance recreational fishing activities. Although stocking has temporarily contributed to the species recruitment and to the Muskellunge fishery in the Montréal region of the St. Lawrence system (see the article of Carrier et al. in the present issue), it does not appear to have been successful in the long term, possibly because of the poor adaptation of the stocked individuals to the particular local conditions of a large river such as the St. Lawrence. However, they may have contributed to sustain the fishery on the short term (see Carrier et al. in the present issue). During your next fishing trip, for example on the St. Lawrence River or on Lake des Deux-Montagnes, you will be able to assert that you most likely caught native Muskellunge of local origin. Based on the results of this study, we recommend avoiding future stocking without detailed knowledge of stock abundance, diversity and genetic structure, and of the level of exchange between them. Actions aiming habitat protection and restoration should rather be prioritized in order to optimize the success of natural reproduction.
Acknowledgements
We express our gratitude to muskies anglers who collected most of the samples, especially to Marc Thorpe, Mike Lazarus and Michael Phillips. We thank Christopher Legard (New York State Department of Environmental Conservation) for collecting and sharing Chautauqua Lake samples. We thank Samuel Cartier for collecting Lake Champlain fish. Thanks to Chris Wilson (Aquatic Research and Monitoring Section, Ontario Ministry of Natural Resources and Forestry) for sharing Pigeon Lake DNA samples and to Christopher Wilson (Fish Culture Section, Ontario Ministry of Natural Resources and Forestry) for sharing hatcheries and stocking history between Ontario and Québec. Thanks to Shawn Good (Vermont Fish & Wildlife Department) and Jeffrey J. Loukmas (New York State Department of Environmental Conservation) for sharing historical management and stocking history of Lake Champlain. We also thank Nicolas Auclair, Florent Archambault, Rémi Bacon, Christian Beaudoin, Anabel Carrier, Chantal Côté, Julie Deschesnes, François Girard, Guillaume Lemieux, Louise Nadon, Yves Paradis, Geneviève Richard, and Éliane Valiquette for logistic, field and laboratory assistance. Funding was provided by the ministère des Forêts, de la Faune et des Parcs du Québec, Canadian Research Chair in Genomics and Conservation of Aquatic Resources, Fondation héritage faune (Fédération québécoise des chasseurs et pêcheurs), Ressources Aquatiques Québec, and Muskies Inc.