Agricultural sources as major supplies of atmospheric phosphorus to Lake Kinneret

Abstract

Phosphorus (P) deposition through the atmosphere to fresh water bodies is a major supply pathway and, in some cases, can accelerate eutrophication processes. However, the quantification of atmospheric P fluxes to lakes is often not well established and the major P sources are generally poorly recognized. Here we present measurements of the atmospheric P deposition flux of two consecutive years in two sampling stations around Lake Kinneret, a large fresh water body in Northern Israel and quantified the dust daily P load emitted from 14 potential source sites upwind to the lake that include agricultural soils, pasture lands, livestock feeding zone sites and uncultivated lands. In addition, we analyzed phosphate oxygen isotopes (δ18OP) in the deposited dust and in surface material from potential P sources around the lake such as cultivated and uncultivated soils, cow manure and compost. The mean daily dust P deposition to the lake ranged from 150 to 770 μg P m−2 day−1, which accounts for ca. 5%–35% of the P concentrations in the lake water during summer. Dust deposition showed strong daily and monthly variations that were not associated with wind speed alone, indicating that local anthropogenic disturbances have a major control over the daily dust P flux. Simultaneous measurements of atmospheric P in different locations upwind to Lake Kinneret shows that the bulk of the P originated from sources in close proximity to the lake, with negligible P transported from distal sources. Quantification of the atmospheric P load emitted from potential sources upwind to the lake indicate that the highest daily loads of P are emitted from livestock feeding zone sites, probably because of a continuous accumulation of P in these soils. δ18OP values (ranging from 19.1‰ to 24.8‰) in deposited dust samples were significantly different than the values reported for natural soils, coal combustion and airborne tree pollen but in the in the same range as the values found for surface samples taken from agricultural lands fertilized with phosphate fertilizer, compost or cow manure or from uncultivated soils nearby, probably contaminated by agricultural P from nearby sites. Overall, our results show that deposition of dust originating from agricultural sources in close proximity to the lake supply substantial quantities of P to Lake Kinneret. Based on our results we suggest that when considering atmospheric P inputs to fresh water bodies more attention should be paid to land use in sites which are located upwind to the lake basin.