Time scale of land use change and export of N and P from coastal plain basins to the coastal zone

(Funded by NASA MTPE)

participants | abstract | introduction | results | conclusions | publications

Participants

Thomas R. Fisher, Professor, principal investigator
Margaret M. Norton, graduate student (PhD 1999)
Heather L. Berndt, graduate student (MS 1999)
Kuang-Yao Lee, graduate student (PhD 2000)
Emma J. Rochelle-Newall, graduate student (PhD 2000)
Jorge A. Benitez, graduate student (PhD 2001)
Adrienne Sutton, graduate student (PhD 2005)
Jason J. Traband, graduate student (MS 2003)
Anne B. Gustafson, faculty research assistant
Gregory M. Radcliffe, faculty research assistant
Mark T. Lighten, intern (Pembrokeshire College, UK)
Alessandro R. Liboni, intern (Politechnico di Torino, Italy)
Roger Stone, collaborator (US Fish Wildlife Service)
Antony P. Goodyear, intern (Parkside High School, Salisbury MD)

Abstract

Over the last 350 years the Atlantic coastal plain was largely deforested. European settlers cleared the region primarily for agriculture due to the relatively good soils, although urbanization has increasingly claimed more area in the last 100 years. These changes led to greater export of N and P in stream discharge and declining water quality of lakes and estuaries. We documented these land cover and use changes (LCLUC) in the Choptank and Chester basins on the Delmarva Peninsula and modeled the water quality impacts. LCLUC were estimated or measured using crop rotation models and socio-economic statistics (1665-1830), historical maps (1850, 1900), aerial photos (1937), and satellite imagery (1972-1996). These data showed that deforestation of the Delmarva Peninsula occurred prior to 1800 as human populations and demand for tobacco and wheat exports increased. Between 1800 and 1900 there was a stable period of agricultural dominance and low urbanization, but in the 20th century there has been increasing urbanization and aforestation, although agriculture is still the dominant land use. Using the hydrochemical model GWLF we estimated the impacts of the LCLUC. We found small changes in water budgets primarily related to deforestation and aforestation, and sewage from human populations in urbanizing areas increased P export. In contrast, large impacts on N export were related to increasing fertilizer applications on existing agricultural areas after 1950. This increased intensity of an existing land use, rather than LCLUC, was responsible for 10-fold increases in N export over the last 50 years.

Introduction

The first phase of this project investigated the causes of historical increases in N and P inputs to estuaries. There were two goals of this phase:

• to examine historical land cover/land use changes (LCLUC)
• to estimate the water quality consequences of the LCLUC with a hydrochemical model

We used two coastal plain basins in the Chesapeake Bay drainage (see figure 1) as field verification sites to evaluate regional patterns of N and P export from the mid-Atlantic coastal plain

Study Sites

coastal plain basins representative of coast typically <20% of basins are gauged (see figure 2) inputs from most of basin are estimated

Historical LCLUC was obtained from a variety of sources:

• 1665 to 1820: estimated using crop rotation models (see figure 3)
• 1850 to 2000: census data
• 1850: historical maps
• 1900: historical maps
• 1937: aerial photos (NAPP)
• 1972 to 2000: landsat imagery

Crop rotation models employed the following variables: (see figure 3)

• human populations (total, taxable, laborers)
• socio-economic statistics (output per laborer, crop yields)
• crop husbandry (strategy, rotations, sequences)

Four crop rotation models were used for 1665-1830
• tobacco+corn (tobacco dominant, see figure 3)
• tobacco+corn+wheat (tobacco dominant)
• tobacco+corn+wheat (wheat dominant)
• corn+wheat (wheat dominant)

Historical changes in N and P export were estimated for 1850 - 2000:

• The hydrochemical model GWLF (Generalized Watershed Loading Functions) (see figure 4)
• Inputs:

-land cover/land use
-human populations
-wastewater (point) sources
-soil characteristics
-fertilizer application rates
-weather
-local water quality observations

Results

What socio-economic data are available for 1665-1830?
• human populations (total, taxable, number of laborers) (see figure 5)
• food consumption per person (kg person-1 y-1)
• crop yields per laborer (kg ha-1 laborer-1 y-1)
• farm management practices (crop rotations, years in crop)
• crop exports (kg y-1)

Smoothing
• transitions between models smoothed (±10 y)
• smoothing of model output by regression (see figure 6)

Initial deforestation of Choptank region (1665-1850) (see figure 7)
• good agreement between model output and census data
• deforestation virtually complete by1800
• agricultural maximum in 1900
• slow decline in agriculture during the 20th century

Historical maps (1850, 1900) (see figure 8)
- base maps created for first complete bathymetric survey of Chesapeake Bay
- recorded land use composited for Choptank basin
- Aerial photos (NAPP 1937)
-Satellite imagery (1976, 1988, 1996)

Summary of maps, aerial photos, and satellite imagery (see figure 9)
• no significant changes in agriculture and forest during 1850-2000
• slight decline in agriculture from 1900 to 2000
• significant urbanization during entire period (see figure 10)

Synthesis of LCLUC (see figure 11)
• deforestation primarily occured between 1665 and 1800 due to increasing human
populations and demand for export of tobacco and grains (primarily wheat)
• little change in agriculture and forest land use, 1800 to 1900 (agriculture dominant)
• aforestation 1900-present due to abandonment of marginal agricultural lands
• urbanization due to increasing human populations

The hydrochemical model GWLF:

• land use has strong impacts on stream chemistry (see figure 12)

• fertilizer applications strongly increase groundwater nitrate (see figure 13) independent of land use change

• modeled effect of fertilizer on groundwater NO3 (see figure 14)

Application of GWLF to the Choptank basin:
• only a small portion of the Choptank basin is gauged
• extrapolation of N yields over ungauged areas (see figure 15)
• N export higher outside gauged area due to higher populations and more agriculture
• P export varied with soil drainage characteristics, land use, and human populations

• little evidence for large changes in climate over 200 y (see figure 16)
• used average weather data for 1980-1990 to emphasize effects of LCLUC

historical changes in export (see figure 17)
• water yields increased in summer in 20th century due to increased water retention
as aforestation proceded
• P export increased in 20th century due to increasing P content of soils and sewage
from increasing human populations
• N export increased dramatically after 1950 due to application of commercial fertil-
izers on agricultural areas, independent of land use change (intensification of agri-
culture)

Conclusions

1. Settlement of the region was virtually complete by 1800 due to economic pressure to produce food for local consumption (corn) and to export products for commerce (tobacco, wheat).

2. The maximum extent of agriculture was observed in 1900 due to the rapid expansion of the wheat exports at the end of the 19th century.

3. Agriculture has slowly declined in the 20th century, replaced by afforestation and urbanization.

4. Urbanization has occurred slowly in this rural area, and small towns of <20,000 now occupy ~10% of the basin area.

5. The major impacts on watershed export of water, N, and P occurred at varying times due to a combination of LCLUC and other human activities.

6. For water yields, aforestation in the 20th century increased summer stream flows due to increased water retention.

7. P export occurs episodically during storm events, and enhanced P export may have occurred during 1750-1800 when 60% of the basin was deforested for wheat production. Combined with poor land management and soil exhaustion, major P export may have occured at this time in the 18th century, but we have insufficient information to estimate this process. However, in the 20th century, higher export rates were related to increases in surface soil P and sewage inputs from urbanized areas.

8. N export increased rapidly after the application of commercial fertilizers following WWII. Fertilizer use has resulted in large increases in nitrate in ground waters and 10 x transport of N in streams.

Publications

• Fisher, T. R., K.-Y. Lee, H. Berndt, J. A. Benitez, and M. M. Norton. 1998. Hydrology and chemistry of the Choptank River basin in the Chesapeake Bay drainage. Water Air Soil Poll. 105: 387-397