Population dynamics of the coho salmon and its response to logging in three coastal streams

Abstract

This study examines the ecology and dynamics of coho salmon (Oncorhynchus kisutch) in environments experimentally altered by logging. The objective was to evaluate processes that stabilize or regulate the populations. Two small watersheds in Oregon's Coast Range were logged in 1966, one clear-cut, the other patch-cut. A third adjacent watershed was left uncut as a control. The influence of these treatments on the biology of the coho was assessed. Attention was concentrated on populations of the six year classes 1963 to 1968. The natural variability of streamflow-related conditions influencing both the magnitude and pattern of coho recruitment each year was increased in the logged watersheds. Peak flow during storms increased; intragravel dissolved oxygen levels decreased in the stream draining the clear-cut watershed. These changes, however, were apparently within the range of variation that the coho naturally experience. Increased stream temperatures and mortalities, due to the logging effects, altered the post-recruitment life conditions of the coho in that stream but did not significantly affect the final smolt yield. The nocturnal behavior of recently emerged fry leads to recruitment along the stream length. Fry tend to emerge en masse from the redds at night, and large numbers proceed immediately to disperse downstream. This migration continues for several successive nights, beginning each night soon after dark. Evidence is presented indicating that fry emigration is primarily a dispersal mechanism that distributes fry from redd sites to nursery areas. It is hypothesized that the series of events leading from fry dispersal to be quiescent behavior at night, characteristic of resident fry, is a developmental sequence involving the physiology and maturity of the fry, modified by agonistic activity. Adjustments in coho population size were largely accomplished by fall, resulting in stable and characteristic population levels in each stream. A stable smolt yield was a further result. These adjustments are accomplished through high mortality during the months of the first spring and summer. This mortality is likely density dependent and related to the territorial and agonistic behavior of the fish. Growth, biomass, and net production varied greatly during each year. Seasonal changes in growth rate resulted in seasonal variations in biomass that were in contrast to the stabilized trends of population number. The pattern of net production rate was also largely determined by the seasonal growth pattern, and like biomass, did not show: a tendency to stabilize with time. It averaged 5 g/m² among the three streams for the period June 1 to April 15. The coho populations seem naturally regulated most importantly with respect to number. The patterns of biomass and rate of net production may be understood as an interaction of seasonally variable growth rate with stabilizing population numbers. This study has shown that coho streams normally produce characteristic levels of smolt yield in spite of large natural variations in fry input and conditions for growth. The range of environmental variation for which this result holds may include short-term changes due to logging. However a normal population response to such a severe alteration as occurred on Needle Branch is very likely conditional upon a program that at least includes vigorous stream clearance, the restriction of additional mortality to early summer, when population adjustments are far from complete, and the encouragement of streamside revegetation. A streamside buffer strip of trees is an effective way of protecting aquatic resources.