Tundra plants and soils may constrain ecosystem response to climate ch
ange. To understand these constraints requires integrated studies of b
iogeochemical cycling in tundra ecosystems. This paper summarizes a 4-
yr irrigation-fertilization experiment designed to evaluate elemental
mobility through small tundra watersheds and examines the ''leaky ecos
ystem-nutrient hypothesis.'' The irrigation treatment increased elemen
tal mobility of Fe and Mn, possibly due to reducing conditions caused
by waterlogging. Ammonium and NO3 concentrations remained above backgr
ound levels for four field seasons following application of a slow-rel
ease NPK fertilizer. Sharp peaks and valleys in NH4 and NO3 concentrat
ions over the four seasons suggest slow fertilizer solubilization due
to low temperatures and mixing of waters of variable composition. Fert
ilizer P showed limited mobility. Soil solutions were saturated with r
espect to gibbsite suggesting that a gibbsite-like mineral may control
Al solubility in these arctic soils. Comparisons of elemental concent
rations among tundra ecosystems indicate that weathering rates are hig
hly individualistic, depending critically on soil parent material. Fur
thermore, tundra ecosystems are leaky with respect to nutrients, which
may have ramifications well beyond the disturbance both with respect
to time and space, and bears attention in assessing the future respons
e of tundra ecosystems.