| dc.description.abstract | This thesis focuses on the estuarine, coastal and marine mortality of Atlantic salmon. The overall objective is to better understand the factors affecting the survival of Atlantic salmon during their early marine phase. To meet this objective, we first review trends in marine survival, and examine theoretical and empirical evidence to identify; (a) potentially important mortality sources, and (b) the timing of high mortality. It is clear that widespread shifts in the marine survival of salmon have occurred, however the timing, magnitude and effect of survival shifts is variable. Likely mortality mechanisms are similarly variable. This review highlights predation during the early marine phase as important for North American salmon; thus granting focus to the remaining research chapters.
We next estimate the survival of postsmolts in selected estuaries and coastal habitats using acoustic telemetry. Simultaneously, we incorporate methods to address the major limitations to estimating survival using acoustic telemetry, including the use of mark-recapture modelling to address the effect of receiver detection performance, and a novel cluster-analysis modelling approach that attempts to quantify the complicating effects of predation. Our findings suggest that the early marine survival of Atlantic salmon, similar to marine mortality as a whole, is highly variable. Cumulative survival through the river, inner estuary, outer estuary and bay habitats ranged from 39·4% to 73·5% in Nova Scotia’s Southern Upland, whereas survival past the outer estuaries of inner Bay of Fundy rivers ranged from 24.3-54.0%. Survival rates followed two patterns: (1) constant rates of survival independent of habitat or (2) low survival most frequently associated with inner estuary habitats. We also examine the potential mortality mechanisms related to predation by examining patterns in the estuarine mortality of acoustically tagged salmon juveniles, using insights from mortality covariates, and the relationship between migratory behaviour and survival. Avian predation appears to be the dominant mortality vector in some estuaries of Nova Scotia’s Southern Upland, with the sudden disappearance of most (75–100%) smolts and post-smolts; which we interpret as evidence of avian predation along with evidence of size-selective survival. Alternatively, predatory striped bass appear to be a major source of mortality for some inner Bay of Fundy salmon populations, with evidence of a minimum of 7.3-27.3% of all tagged smolts being consumed by striped bass, based on migratory movement patterns.
The survival estimates reported in this thesis permit the division of the marine phase into two periods; an early period encompassing estuarine and coastal habitats ( < 1 month), and the remaining time at sea. By comparing estimated survival during these two periods, it appears that estuarine survival cannot be solely responsible for observations of reduced marine survival since approx. 1990.The highest marine mortality must occur outside of estuaries and early coastal habitats. However, efforts directed at reducing estuarine and coastal mortality may be valuable for conservation planning, and may help reduce the risk of extirpation and, in some cases, may lead to viable populations. | en_US |
