Contemporary genetic structure of Atlantic salmon (Salmo salar L.) in the River Moy in Ireland is shown here to be strongly related to landscape features and population demographics, with populations being defined largely by their degree of physical isolation and their size. Samples of juvenile salmon were collected from the 17 major spawning areas on the river Moy and from one spawning area in each of five smaller nearby rivers. No temporal allele frequency differences were observed within locations for 12 microsatellite loci, whereas nearly all spatial samples differed significantly suggesting that each was a separate population. Bayesian clustering and landscape genetic analyses suggest that these populations can be combined hierarchically into five genetically informative larger groupings. Lakes were found to be the single most important determinant of the observed population structure. Spawning area size was also an important factor. The salmon population of the closest nearby river resembled genetically the largest Moy population grouping. In addition we showed that anthropogenic influences on spawning habitats, in this case arterial drainage, can affect relationships between populations. Our results show that Atlantic salmon biodiversity can be largely defined by geography and thus knowledge of landscape features (for example, as characterised within Geographical Information Systems) has the potential, to predict population structure in other rivers without an intensive genetic survey, or at least to help direct sampling. This approach of combining genetics and geography, for sampling and in subsequent statistical analyses, has wider application to the investigation of population structure in other freshwater/anadromous fish species and possibly in marine fish and other organisms.

Identifying differences in quantitative life history traits between cultured and native or non-native wild populations is important in assessing the impact of accidental and deliberate introductions of hatchery-reared fish into the wild. As the ability to exploit the marine environment is the defining life history characteristic of anadromous salmonids, knowledge of variation in smoltification characteristics among populations is crucial in determining how these introductions affect fitness in recipient populations. Data are presented here describing the timing and extent of the autumn migration; the propensity for male parr maturation; the timing of the spring migration; and the size of autumn and spring migrants from Atlantic salmon (Salmo salar) populations from various genetic backgrounds. These experiments were carried out under common garden conditions over a decade in the Srahrevagh River in the west of Ireland. Population specific genetically determined differences in quantitative life history traits associated with smoltification were apparent. These differences may reflect smolt quality and therefore impact on marine survival and ultimately lifetime fitness. Both hatchery domestication and geography (different selective environments) were found to be important factors determining smolt phenotypes, although it was difficult to measure the relative contribution of each. These results indicate that farm, native hatchery, non-native wild salmon (even from a neighbouring catchment) and their hybrids with native wild fish, are likely to produce less well adapted and thus poorer quality smolts than native wild populations and, where wild populations are extant, such stocks should not be used for enhancement purposes.

The implementation of many fisheries management-related activities in fresh water depends on habitat area inventories over extensive geographical scales. While river lengths are readily available, representative widths, necessary for area calculations, are difficult to obtain. As field surveys to collect this information are resource intensive, a predictive model was developed to enable the calculation of river wetted width using GIS-derived values for catchment and river descriptors. A model containing upstream catchment area and the Shreve river drainage network index accounted for 88% of the variation in field measured river wetted width. Comparisons in Irish and Scottish rivers between modelled and measured widths were highly correlated and suggest that the model may be transferable to neighbouring geographic areas. As an example, the model is applied to provide an estimate of the usable fluvial habitat available to Atlantic salmon in Ireland.

The assessment report of the 4th International Panel on Climate Change confirms that global warming is strongly affecting biological systems and that 20–30% of species risk extinction from projected future increases in temperature. It is essential that any measures taken to conserve individual species and their constituent populations against climate-mediated declines are appropriate. The release of captive bred animals to augment wild populations is a widespread management strategy for many species but has proven controversial. Using a regression model based on a 37-year study of wild and sea ranched Atlantic salmon (Salmo salar) spawning together in the wild, we show that the escape of captive bred animals into the wild can substantially depress recruitment and more specifically disrupt the capacity of natural populations to adapt to higher winter water temperatures associated with climate variability. We speculate the mechanisms underlying this seasonal response and suggest that an explanation based on bio-energetic processes with physiological responses synchronized by photoperiod is plausible. Furthermore, we predict, by running the model forward using projected future climate scenarios, that these cultured fish substantially increase the risk of extinction for the studied population within 20 generations. In contrast, we show that positive outcomes to climate change are possible if captive bred animals are prevented from breeding in the wild. Rather than imposing an additional genetic load on wild populations by releasing maladapted captive bred animals, we propose that conservation efforts should focus on optimizing conditions for adaptation to occur by reducing exploitation and protecting critical habitats. Our findings are likely to hold true for most poikilothermic species where captive breeding programmes are used in population management.

Oligotrophic catchments with short spatey streams, upland lakes and peaty soils characterise northwest European Atlantic coastal regions. These catchments are important biodiversity refuges, particularly for sensitive diadromous fish populations but are subject to changes in land use and land management practices associated with afforestation, agriculture and rural development. Quantification of the degree of catchment degradation resulting from such anthropogenic impacts is often limited by a lack of long-term baseline data in what are generally relatively isolated, poorly studied catchments. This research uses a combination of palaeolimnological (radiometrically-dated variations in sedimentary geochemical elements, pollen, diatoms and remains of cladocera), census, and instrumental data, along with hindcast estimates to quantify environmental changes and their aquatic impacts since the late 19th century. The most likely drivers of any change are also identified. Results confirm an aquatic biotic response (phyto- and zooplankton) to soil erosion and nutrient enrichment associated with the onset of commercial conifer afforestation, effects that were subsequently enhanced as a result of increased overgrazing in the catchment and, possibly, climate warming. The implications for the health of aquatic resources in the catchment are discussed