Selective foraging by moose on hardwoods and avoidance of conifers alters community composition and structure, which in turn can affect nutrient cycles and productivity. The effect of moose browsing on the nutrient cycles of boreal forests was studied using three 40—yr—old exclosures on Isle Royale, Michigan. Two alternative mechanisms by which moose affect ecosystems were tested: (1) moose depress both the quantity and quality of litter return to the soil, and hence N mineralization and net primary productivity, by browsing on hardwoods and avoiding conifers; (2) moose stimulate N mineralization, and hence net primary productivity, by opening the canopy and by dropping fecal pellets. Soil nutrient availability and microbial activity, including exchangeable cations, total carbon and nitrogen, nitrogen mineralization rates, and microbial respiration rates, were uniformly higher in exclosures than outside. These differences were more significant where browsing intensity was high and less often significant where browsing intensity was low. N mineralization in browsed plots declined with increasing moose consumption rates. Net primary production in exclosures and browsed plots was strongly correlated with N mineralization. N mineralization in turn was positively correlated with litter N return and negatively correlated with litter cellulose content. These differences in litter quantity and quality were caused by an increased abundance of unbrowsed spruce outside the exclosures. Moose pellets alone mineralized less N but more C than soil alone, but pellets combined with soil stimulated N and C mineralization more than the sum of the two separately. However, this did not appear to be sufficient to offset the depression in nitrogen and carbon mineralization in soil resulting from the increased abundance of unbrowsed spruce. We conclude that, in the long term, high rates of moose browsing depress N mineralization and net primary production through the indirect effects on recruitment into the tree stratum, and subsequent depression of litter N return and litter quality. These results suggest that the effects of herbivores on ecosystems may be amplified by positive feedbacks between plant litter and soil nutrient availability.

Large mammalian herbivores can influence the dynamics and structure of ecosystems by selectively removing tissues of specific plant species. The plant community composition can be altered as animals feed on some species but not others, changing the biomass, production, and nutrient cycling of an entire ecosystem. We used four paired moose (Alces alces) exclosures and browsed plots (built between 1948 and 1950) on Isle Royale, Michigan, to examine the influence of moose on aboveground biomass, production, and annual litterfall of boreal vegetation in 1987. Tree biomass was significantly greater (X = 230 vs. 150 Mg/ha, df = 3, P < .05), shrub biomass was significantly less (X = 1.9 vs. 3.1 Mg/ha, P < .05), and herb biomass was significantly less (X = 0.2 vs. 0.8 Mg/ha, P < .05) in exclosures than in browsed plots. Tree production was greater in exclosures than in browsed plots (X = 7.9 vs. 5.0 Mg.ha—1.yr, P = .05), but there was no difference in the production per unit biomass between exclosures and browsed plots. Shrub production in exclosures was similar to that of browsed plots (X = 3.5 vs. 2.3 Mg.ha—1.yr—1, P < .05), despite total vegetation biomass differences between paired plots. There was significantly greater herb litter produced in the browsed plots than in the exclosures (X = 0.7 vs. 0.1 Mg.ha—1.yr—1, P < .05). Moose browsing prevented saplings of preferred species from growing into the tree canopy, resulting in a forest with fewer canopy trees and a well—developed understory of shrubs and herbs. In addition, browsing may have altered the eventual balance of white spruce (Picea glauca) was balsam fir (Abies balsamea), causing an increase in the former and a decrease in the latter. Thus, browsing by moose influences in long—term structure and dynamics of the boreal forest ecosystem, which has important implications for forest ecosystem management, especially where the population dynamics of moose are regulated.

No abstract

In June 1992 we collected 53 blood plasma samples from nestling red‐winged blackbirds (Agelaius phoeniceus), house sparrows (Passer domesticus), and brown‐headed cowbirds (Molothrus ater) at five study sites in northwestern Minnesota to evaluate the feasibility of using plasma and brain cholinesterase (ChE) activity and reactivation as a means of assessing exposure of nontarget organisms to the operational use of ChE‐ inhibiting insecticides in an agricultural setting. Three sites were adjacent to sugar beet fields that were likely to be treated with chlorpyrifos as Lorsban® (an organophosphorus [OP] insecticide) for control of sugar beet root maggot (Tetanops myopaeformis), and two sites were distant from fields likely to be treated (reference sites). Application of chlorpyrifos in fields surrounding study plots was monitored through contact with landowners and direct observations. Cholinesterase activity levels (total cholinesterase [ChE], acetylcholinesterase [AChE], and butyrylcholinesterase [BChE]) in nestling plasma were measured and tested for reactivation (ChE and AChE) in the presence of 2–PAM, an indication of exposure to an organophosphorus insecticide. In addition, 11 nestlings were euthanatized and in these samples we measured brain ChE activity and reactivation, and we analyzed gastrointestinal tracts and carcass washes for chlorpyrifos residues. Total ChE and BChE activity were lowest in similar‐aged nestlings at sites adjacent to treated beet fields (ChE, t = −2.51, d.f. = 21, p = 0.033; BChE, t = −2.56, d.f. = 21, p = 0.043). Nestlings from sites that were near fields where chlorpyrifos was applied were more likely to exhibit plasma AChE reactivation than nestlings from reference sites where OP or carbamate insecticide application was improbable (x2 = 3.805, d.f. = 1, p ∼ 0.05). The magnitude of plasma ChE reactivation was highest within 1 to 3 d of insecticide application, although significant reactivation was measured up to 11 d after application of chlorpyrifos. Plasma AChE reactivation in the presence of pyridine‐2–aldoxime methochloride (2–PAM) appeared to be a more sensitive indicator of exposure to chlorpyrifos than plasma ChE or BChE activity levels.