>Proportion of Watershed as Agricultural/Urban

Proportion of Watershed as Agricultural/Urban


Urban development occurs at various intensities, from low density rural residential development to commercial/industrial development. Urban and agricultural areas both disturb various ecological processes and attributes.
Category: 
Landscape Condition

The metric is a combination of the percent area of sub-watershed as agriculture and the weighted percentage area as urban, where highest density urbanization gets the highest weighting.

QT: Indicators

What is it?

Land intensively used by humans for agriculture and urban development has direct impacts on the functions of natural ecosystems. The proportions of subwatersheds in use for urban and agricultural activity were combined to produce a measure that reflects the amount of land governed primarily by natural processes and those governed by anthropogenic processes. Regions that are not classified as urban or agriculture may be in use for rural activities that also impact natural ecosystems (Byron and Goldman 1989). To gage the impact of rural human activities, housing unit density is used as a measure of rural disturbance.

Why is it important?

Agricultural and urban land uses changes sediment characteristics and loading in adjacent rivers, increases nitrogen loading (Ahearn et al., 2005), modifies in-stream and hyporheic hydrology, delivers pesticides and other chemicals, modifies nutrient cycling and other ecological processes, fragments terrestrial and aquatic habitat, and has caused the loss of productive habitat.

Land that is not currently developed for urban, suburban, rural or agricultural uses provides habitat for native plant and animal communities and provides the opportunity for healthy and natural population dynamics in the community and natural processes. With loss of habitat, the services provided by these communities are reduced and, with fragmentation of the remaining habitat, the impact is even greater. Negative impacts on water quality are correlated with landscape disturbance and, though land-cover is only a general indicator of these impacts, it is regularly updated and made public.

One of the most important ways that developed areas impact streams and watershed processes is through the creation of impervious surface (surfaces that are not pervious or less pervious to water than soil). Recent studies elsewhere in California suggest that the amount and the spatial distribution of impervious area influence hydrologic response (Coleman et al., 2005). The amount of impervious surface development per unit watershed area is one important measure. Past studies have suggested that the range in proportion is from any development at all affecting some aspect of aquatic systems (Booth et al., 2002) to >25% development making a watershed non-supporting to aquatic life (Schueler, 2000). Intermediate development (3-5% impervious cover) has intermediate impacts on aquatic organisms (Stein and Zaleski, 2005).

Target or Desired Condition

The desired condition of the landscape, from an ecological health standpoint, is to have no land in use as urban or agriculture. Likewise, any area with no housing units is deemed desirable. These conditions are used as the baseline for the scoring of the regions. This means that an area that has no urban or agricultural landuse and no housing units will have a score of 100. This is not to say that human development activities should be prohibited, but rather to recognize that when development occurs, it invariably brings some level of impact and measuring and tracking that impact is important.

What can influence or stress condition?

Any sort of development (urban or agricultural) and any sort of human activity on the landscape that disturbs the natural state will impact the overall health of streams in the watershed. Also, any efforts to mitigate existing disturbance will serve to improve the health of the watershed. Inefficient land consumption through low-density development will tend to exacerbate effects to streams. But even dense development can have focused effects on adjacent and nearby streams.

Data sources

Landuse data was sourced mainly from DWR. The counties used are Placer (1994), Sierra (2002), Yuba (1995), Butte (2004), Plumas (1997). The CalFire data was used to source the Nevada county landuse data (2002).

The US Census was the source for housing units at the block level and can be accessed as a prepared spatial product from the CalFire FRAP website.

Analyses

DWR conducts periodic land use surveys of California counties that have substantial agricultural activities. These data were combined into one dataset which comprised portions of Placer, Sierra, Yuba, Butte and Plumas counties. Land use data for Nevada County was sourced from CalFire and processed to a comparable format (converted from raster to vector) and combined with the remaining county data. Landuse categories were combined into three categories including agriculture, urban and other. The proportion of each landuse type was calculated for each planning watershed as well as for the subwatersheds. Housing unit counts from the 2000 Decennial Census were calculated for each planning watershed. These data were used to generate housing unit density (units per acre) for each planning watershed and for each subwatershed (see equations below).

Scoring

The landuse proportion was multiplied by 100 to generate a score between 0 and 100. The housing unit density scores for the subwatersheds were normalized by dividing the calculated density by the maximum observed density (at the planning watershed scale), then multiplying by 100. These two scores were combined by averaging the two values (Figure 2).

  • Proportion Ag/Urban Landuse = (Areaurban + Areaagriculture) / Areatotal
  • Housing Unit Density = Unit count / Area
  • Landuse Score = 100 x (1 – Proportion)
  • Housing Score = 100 x (Unit Density / 1.061); 1.061 is the highest observed density (units/acre) at the planning watershed scale.
  • Combined score = (Landuse score + Housing score) / 2

This indicator's data resource has not been uploaded yet.

What did we find out/How are we doing?

The scores for subwatersheds in the foothills and mountains are high due to the low proportion of urban and agricultural land use and the low population density (Table 1 and Figure 1). This status is not likely to change much in the next 10 to 20 years at the subwatershed scale. The scores of subwatersheds in the valley are lower and reflect the high proportion of urban and agricultural land use and the higher population density. These may change in subsequent analyses as cities grow and more areas are farmed on the valley floor. It is anticipated that declines in the indicator will be evident over periods of decades with the lower elevation and flatter watersheds seeing the most rapid declines in the indicator.

Temporal and spatial resolution

Landuse and housing data are spatially continuous for the subwatersheds, meaning that the whole watershed is represented. Temporal resolution for the housing-density data is 10 years as the data are sourced from the US Census. Landuse data are updated on an irregular interval by both the DWR and the CA Dept of Forestry and Fire Protection (CalFire). The nominal time span between landuse datasets is about 6 years.

The relative condition of the subwatersheds must be established over several data collection events. These data can be updated using the decennial census and using the periodic land use assessments provided by both the DWR land use survey and the CalFire Fire and Resource Assessment Program (FRAP) land use survey.

How sure are we about our findings (Things to keep in mind)

Due to the spatial extent and resolution of the data, confidence in the results is high. However, the Census housing unit density data are from 2000 (to be updated in 2010) and will be more accurate with the inclusion of 2010 Census data in subsequent analyses. One caveat to our finding is that the score is an average of two measures that overlap each other. Proportion of the landscape that is agriculture or urban overlaps with housing unit density in urban areas, essentially double-counting this small part of the total landscape. We used both measures because of the extensive rural development in the watershed.