Land Cover along an Urban-rural Gradient: Implications for Water Quality

Developtlicnt pressures in rural mountainous areas o f the United States hold crucial implications for water quality. Especially important arc changes in the extent and pattern o f various land uses. W e cxamille how position along an urban-rural gradient affects landscape pattcrns in a southern Appalachian watershed, first by testing for the effect o f distance from an urban center on land-cover change probabilities and then simulating the implied development o f a landscape at regular distance intcrvals. By simulating a common hypothetical landscape we control for variable landscape conditions and define how land development might proceed in the hture. Results indicate that position along thc urban-rural gradient has a significant effect on land-cover changes on private lands but not on public lands. Furthermore, position along the gradient has a compounding effcct on laridcover changes through interactions with other variables such as slope. Simulation results indicate that these differences in land-cover changes would give rise to uniquc "landscape signatures" along the urban-rural gradient. By examining a development sequence, we identify patterns o f change that may be most significant for water quality. Two locations along the urban-rural gradient may hold disproportionate influence over water quality in the future: ( 1 ) at the most remote portion o f the landscape and (2 ) at the outer envelope o f urban expansion. These findings demonstrate how landscape simulation approaches can be used to identify where and how land use decisions may have critical influence over environniental quality, thereby focusing both future research and monitoring efforts and watershed protection measures. Water connc~tj and focuscq actlv~tlcs throughout a watershed, definlng rhc curnulat~ve Impact5 o f humdn use o f land and resources Thc use and condlt~on o f land 111 part~cular have a profound lnflucncc on watcr quahty (Allan and Flecker 1997, Hunsakcr and Lev~nc 1995) Intcgrat~ng soc~al and cnv~ronnicntal v~tallty wlth~n properly scaled ecosystems requlres a d~fferent focus for science, ~nvolv~ng both dcfin~ng lncchan~slns that o1gan1i.e human bchav~or and thcn linklng human endcavols and cnvlronrucntal quahty T h ~ s hnkagc bctween people dnd ~ h c urlvlrunmrnt 17 b~d~rcct~onal, Involvlng both the cffccts o f peoplc on ccosystcnis and the effects o f ecosystem qtructurc and cnv~ronmcntnl quality on people's well-bc~ng Because water quahty 1s so ciedrly ~mpacted by pcople's use o f resources, and because human well-be~ng is so clearly hnkcd to the avalldblhty o f clean watcr, watcr prov~dcs perhaps the strongest demand for social-ecological intcrdisciplinary research (Naiman et al. 1995). This article examines factors that organize human behavior in a developing landscape and examines thc potential consequences for watcr quality. The overarching theme is that by undcrstanding pattcrns o f devcloplnent and linking them to thcir implications for watcr quality, planning and regulatory efforts thcn can be focused on areas critical in deterniining cnvironmental quality. For example, knowlcdgc o f this sort can be used to identify where land use is 11iost likely to be intensified within a watershed and, conversely, whcrc it will likely remain stable or change in only trivial ways. In a world o f limited rcsourccs, this kind o f targeting may be much more effective than broad regulations intcnded to protect watcr quality (Wear et al. 1994). Our specific focus is on understanding pattcrns o f land use change along the urban-rural gradient in the Manuscript received 6 February 1557; revised 4 Dcccrr~bcr Southern Appalachian Highlands. Portions o f this re1907; accepted 22 December 1997; final version received 7 gion have experienccd nearly exponential population January 1988. I:or reprints of this Invited Feature, see foottrowth with associated development o f land since the 11otc I . pagc 569. 1950s. Growth ccntcrs have included I-Iendersunville Embgcrl Applications Vol. 8. No. 3 and Boonc, North Carolina, USA. In areas where population growth and dcvclopment arc in initial pliases, it niay be possiblc to anticipate and thercby redirect solnc dcveloptncnt that would prove detrimcntal to environmental and aesthetic quaiitics. Wc test various liypothcscs regarding the way that dcvclopment, Incas ~ ~ r c d as extent and patterns of land-cover changc, is organized along the urban-rural gradient and then examine how these changes could he related to planning for the protcction of water quality in a developing landscape. In particular, we use statistical niodels of landcover change to test wlietlier: (1) land-cover changes are generally dctcrtnincd by topographic and locational features, (2) land-covcr rcgimes are influenccd by position along the urban-rural gradient, and (3) land-covcr changes differ betweeri public and private lands at various positions along thc urban-rural gradient. We usc a simulation experiment to test for: (1) differences in the implied cquilibrium land cover at different positions along the urban-rural gradient and (2) Implied changes in equilibrium land cover as development procecds. Findings from these tests are then used to explore the iniplieations that continued development night have for water quality in the southern Appala-