Control mechanisms of arctic lake ecosystems: a limnocorral experiment

Abstract

Abstract To assess the potential impact of human exploitation on arctic lakes and to determine how these eco systems are regulatedwe initated a limnocorral experiment in Toolik Lake, Alaska, in the summer of 1983. The limnocorrals were 5 m in diameterand from 5–6 m in depth and were open to the sediments. In 1983 four limnocorrals were deployed in an isolated bay of ToolikLake within a cross-classified treatment regime of high and low inorganic nitrogen and phosphorus additions and high and lowfree swimming fish additions. The objective of the nutrient addition was to stimulate phytoplankton growth and determine theextent to which increased plant production was passed through pelagic and benthic food chains. The objective of the fish additionwas to determine the impact of fish predation on large-bodied zooplankton, especially the zooplanktivorous copepod Heterocope, then to study the effect of altered Heterocope densities on small-bodied zooplankton species population dynamics. In 1984 two more limnocorrals were deployed, one a lowfish, 1 nutrient addition treatment and the other a no fish, no nutrient treatment. The fish manipulation was changed toconfining several fish in cages with the cages held in corrals for varying lengths of time.The addition of inorganic nitrogen and phosphorus dramatically increased phytoplankton productivity. This increase in algalbiomass and production greatly altered the light environment and water quality in the nutrient treated limnocorrals. The secchidisk depth in the nutrient treated limnocorrals declined each summer reaching as low as 1 m in 1985. Both oxygen content andpH increased in the nutrient treatment corrals. Corrals not receiving nutrient additions remained near lake concentrationsfor most water quality parameters. While phytoplankton biomass was stimulated in 1983 phytoplankton growth was not sufficientto draw down all the nitrogen and phosphorus added and these nutrients reached high levels in the last half of the summer.In 1984 phosphorus remained above 20 g in the nutrient-treated corrals but ammonia dropped to reference levels by day 25.In 1985 both nutrient concentrations rapidly declined to reference levels.Most pelagic components responded to the nutrient additions. Microbial production was stimulated in the nutrient treated limnocorralsand bacterial population sizes built up to nearly 8–10 times those of the reference corrals. However, microheterotrophs soonincreased in abundance and apparently grazed down bacteria to reference levels. Phytoplankton population density, as estimatedby chlorophyll a determinations, increased dramatically with nutrient addition such that each year the phytoplankton densities were higherthan before. Primary productivity was also stimulated and appeared not to be light limited even when phytoplankton densitiesrose to high levels. In the first two years of the experiment zooplankton densities were little altered by the increased phytoplanktondensities. However, by 1985 daphnid densities were quite a bit higher in the high nutrient addition limnocorrals.The benthic community and sediment response was much less affected by nutrient addition. Overall sediment respiration increasedin the nutrient treated corrals but underlying sediments seemed little affected. Decomposition of Carex litter was likewise little affected by nutrient addition. Benthic invertebrates were also little impacted by the nutrientaddition and increased sedimentation of phytoplankton. However, the response of benthic invertebrates is difficult to assessfully in the current experiment because chironomids, a prominent component of the benthic community, failed to recruit intothe limnocorrals and the corrals physically shifted during ice-out in the spring of 1984 disturbing the sediment in severalcorrals.The fish additions in 1983 of free swimming grayling essentially eliminated large bodied zooplankton, especially Heterocope septentrionalis, from all four limnocorrals. In subsequent summers Heterocope were not so dramatically preyed upon but generally were found in higher densities in the low or no fish treatments. However,either when Heterocope were eliminated in 1983 or were in rough inverse proportion to fish density, altered Heterocope abundance had no obvious affect on small-bodied zooplankton abundance. The fish treatment apparently influenced the zooplanktonresponse to high nutrient addition in 1985. In the high nutrient limnocorrals daphnid populations became very abundant, butin the high fish treatment the daphnid responding was the small-bodied D. longiremis while in the low fish treatment the daphnid responding was the large-bodied D. middendorffiana.Thus we have considerable evidence for bottom up control of phytoplankton density and production. This increased productionultimately, but not for two years, stimulated zooplankton density increases. Increased nutrients had little effect on thebenthos or sediments. Fish manipulations influenced large-bodied zooplankton but had little effect on small-bodied zooplankton.Because grayling are predominantly plankton feeders in lakes, no fish effect on benthic invertebrates was expected.Limnocorrals thus seem good systems to study nutrient-phytoplankton interactions. They are not as suitable for benthic invertebratestudies and fish manipulations may be difficult. Most other limnocorral studies were of brief duration; however, in the presentstudy the limnocorrals seemed to perform well over a three year period.