Off-channel ponds in the upper reaches of the Coldwater River, British Columbia, were major rearing areas for juvenile coho salmon (Oncorhynchus kisutch). Chinook salmon (Oncorhynchus tshawytscha), steelhead trout (Salmo gairdneri), and Dolly Varden char (Salvelinus malma) were generally scarce in the ponds, although they were numerous in the main river. Coho salmon were predominant at "natural" river sites while steelhead trout was the main species at sites with "rip-rap" bank stabilization. Catches of juvenile coho were much lower in the main river than in the ponds where they were the main species, and were more variable in the river. Population density and biomass estimates of juvenile coho in the ponds ranged from 0.100 fish∙m−2 and 1.00 g∙m−2 to 1.00 fish∙m−2 and 5.15 g∙m−2, compared with density estimates of 0.08–0.23 fish∙m−2 in the river. The coho population in the ponds consisted of 0+ and 1+ age-groups in similar proportions, while in the main river the 0+ age-group was much more abundant. The growth rate of coho in the ponds was faster than in the main river, with pond fish reaching mean lengths of 62–79 mm at the end of the first growing season, compared with 53 mm in the main river. Smolt outmigration from the main study pond occurred in late spring with peak outmigration in May and June coinciding with peak river discharge and increasing water temperatures in the main river and pond.

The winter distribution and abundance of juvenile salmonids was investigated in various main channel and off-channel habitats in the Coldwater and Nicola rivers in the southern interior region of British Columbia. Catches were generally low in all main channel habitats, with coho salmon and steelhead trout being most abundant and chinook salmon and Dolly Varden char being present in smaller numbers. Coho salmon and steelhead trout catches were generally highest in pools with abundant instream and riparian cover. Steelhead trout was the main species in riprap bank protected areas, although catches were generally low. Highest overall catches were recorded in side channels and off-channel ponds, where water temperatures were usually several degrees higher than in the main river. Coho salmon was the main species in the two Coldwater off-channel ponds with overwintering populations of approximately 4000 and 1000 in 1- and 0.1-ha ponds, respectively (0.4–1.5 fish/m2): overwinter survival of coho salmon in the ponds was estimated to be 87 and 54%, respectively. High densities of coho salmon were also recorded in side channels on the Nicola River (1.5–1.8 fish/m2), together with smaller numbers of chinook salmon and steelhead trout (0.2–0.3 fish/m2). Growth in ponds and side channels appeared to be faster than in main channel habitats. We conclude that juvenile salmonids in the rivers investigated showed considerable habitat segregation during the winter. As in coastal rivers, juvenile coho salmon made extensive use of off-channel ponds, while rainbow trout and chinook salmon were generally most abundant in riprap and deep pools containing log debris, respectively.

Attempts to relate the impacts of an experimental weed‐cutting programme, involving the partial and total clearance of aquatic and bankside vegetation, to the abundance of fish stocks in a small lowland river were made difficult by large inherent fluctuations in population levels within the fish community. However, mapping studies showed fish distribution in uncut and partially cut sections to be strongly associated with weed cover, particularly that provided by overhanging backside vegetation. It is suggested that a partial‐cut strategy, in which marginal weed beds and bankside vegetation are left undisturbed, is more beneficial to fisheries than traditional programmes of weed clearance. Increased fish movements during the operation of a weed‐cutting launch suggest that this form of control may result in more disturbance to the fish community than does manual weed control. Weed‐cutting and other weed management practices are discussed in relation to fisheries ecology and conservation.

Fish populations in the River Perry, a small lowland channelized river in England, were affected by long-term river maintenance and improvement works. Habitat diversity at two 'old channelized' sites was low compared with a downstream 'partially channelized' site and an unmodified site, wh,ere natural river features such as the pool-riffle pattern were more apparent. The fish community was low in diversity at all sites, with two running-water cyprinids, dace Leuciscus leuciscus (L.) and chub Leuciscus cephalm (L.), being the dominant fish species. Relative species composition varied between sites, with dace predominating at the 'old channelized' sites and chub increasingly dominant at the 'partially channelized' and unmodified sites. Population density and biomass estimates of dace were similar a t all sites, but chub were up three times more abundant at the 'partially channelized' and unmodified sites than at the 'old channelized' sites. Populations of dace and chub at the 'old channelized' sites contained a higher proportion of small fish than populations at the downstream sites, which exhibited a more balanced population structure. The growth rates of both dace and chub varied between sites. Low recaptures of marked dace suggested a highly mobile movement pattern while higher recapture rates of chub, particularly at the 'partially channelized' and unmodified sites, seemed t o indicate a more sedentary activity pattern. The conclusion of the study is that long-term river maintenance and management works may delay the morphological and biological recovery of lowland channelized rivers.

Fish stocks at a site in the upstream reaches ofthe River Soar, a lowland river in Leicestershire, showed a marked alteration following river channel works associated with a land drainage improvement scheme. The density and standing crop of all 'large-sized' fish species present showed reductions of 15-100% between a 6-month pre-drainage period and a 3-month postdrainage period. The density and standing crop of the total population were reduced from means of 0.160 fish m"^ and 39.0 g m"^ during the pre-drainage period to means of 0.048 fish m'^ and 9.6 g m~^ during the post-drainage period, reductions of 70% and 76%, respectively. It is suggested that modifications to physical features of fish habitat, particularly the elimination of instream cover, were largely responsible for the decreased fish abundance.