Citation:

WICKHAM, J. D., T. G. WADE, AND K. RIITERS. LAND-COVER CHANGE AND ITS IMPACT ON NUTRIENT EXPORT VARIANCE. Presented at International Association for Landscape Ecology, World Congress, Wageningen, NETHERLANDS, July 08 - 12, 2007.

Impact/Purpose:

The objective of this task is to produce land-cover and related products that are needed to meet Annual Performance Goals (APG) under GPRA Goals Clean Air, Clean Water, and Healthy and Safe Communities, and to meet the critical needs of EPA Regional Offices.

Description:

Conversion of natural or semi-natural vegetation to anthropogenic use is widely cited as one of the principal threats to ecosystems worldwide. One consequence of these landcover conversions is increased input of nutrients into surface waters, which promotes eutrophication, noxious algal blooms, and other problems. Despite the pace of land-cover conversion, there is little information on its affect on nutrient export. The few modeling studies that have addressed the question have had difficulty separating the effect of landcover change from other factors such as inter-annual variation in precipitation. Part of the

reason that modeling studies have had difficulty in distinguishing the effect of land-cover change from other factors is that they have been focused on annual averages (e.g., Vuorenmaa et al. 2002). We hypothesize that the principal effect of land-cover change on nutrient export is increased inter-annual variability. A watershed's annual export of nutrients (e.g., kg/ha/yr) fluctuates more widely as natural vegetation is replaced with agriculture and urban use. A consequence of increased inter-annual variability is greater difficulty in meeting

specified nutrient export goals year after year. For example, Fisher et al. (1998) reported 10 years of total nitrogen (TN) and total phosphorus (TP) for the upper Choptank watershed (Maryland USA). The upper Choptank watershed is approximately 50% agriculture and 50%

forest, and land-cover composition has changed little. There was approximately a four-fold range in TN export (3.21-11.5 kg/ha/yr) and approximately a three-fold range in TP (0.19-0.65 kg/ha/yr). Four of the ten years of TN export and two of the ten years of TP export

exceeded management thresholds for the region (Linker et al. 1996). We compiled a large dataset of TN and TP to quantify the impact of land-cover composition and change on nutrient export variance. The dataset includes about 120 sites and about 1200 observations. Most sites contain estimates of TN and TP for several years. The sites span the conterminous United States and also parts of southern Canada. Data sources include (Reckhow et al. 1980, Alexander et al. 1996, Panuska and Lillie 1995,

McFarland and Hauck 2001, and Groffman et al. 2004). Land-cover composition for the site (watershed) had to be homogeneous for inclusion in the dataset. The threshold for

homogeneity was set to 80%. The four main land-cover classes were: forest, urban, agriculture, and range. Range included shrublands and grasslands and were mainly found in

the arid and semi-arid portions of the United States.

The data were fit to three-parameter lognormal distribution by land-cover class and nutrient component. The lognormal models were then used in a Monte Carlo analysis to estimate the variance in nutrient export for mixed land-use watersheds using available landcover

data for the conterminous United States (Vogelmann et al. 2001, Homer et al. 2004). Land-cover change data were also used in a Monte Carlo analysis to quantify the

relationship between land-cover change and change in the variance of nutrient export. Preliminary results indicate that, on average, a 10% loss of forest or range results in a significant increase in TN and TP variance. However, the results were not constant across all watersheds. Much smaller losses produced significant increases in TN and TP variance when the watershed was dominated forest or range. This was consistent with our hypothesis because watersheds dominated by natural vegetation exhibit little inter-annual variability in nutrient export, and adding even small proportions of anthropogenic land cover (urban,

agriculture results in significant increases in TN and TP variance. Conversely, much larger losses of forest and range (i.e., > 10%) were required to significantly increase TN and TP variance in watersheds dominated urban or agriculture. Watersheds dominated by urban and agriculture have characteristically wide TN and TP variance; relatively small losses of the forest or range (i.e., < 10%) that remain do not significantly increase variance.

Record Details: