Devendra M. Amatya

EFFECTS OF CONTROLLED DRAINAGE ON THE HYDROLOGY OF DRAINED PINE PLANTATIONS IN THE NORTH CAROLINA COASTAL PLAIN

Abstract This paper presents results of a 5 year study to characterize the hydrology (rainfall, interception, evapotranspiration (ET), soil water storage, drainage rate, lateral seepage, and water table fluctuations) of three identical drained, pine-forested watersheds in Carteret County, North Carolina. During the 2 year calibration period (1988–1989), all three watersheds were operated in conventional drainage mode with the weirs in the outlet ditch approximately 1.0 m below the soil surface. About 17% of the total rainfall was intercepted and subsequently evaporated and 53% was removed by transpiration and evaporation from the soil during this period. Drainage removed about 28% and the remaining 3% was lost by lateral seepage. During the 3 year controlled drainage treatment period (1990–1993), drainage in Watershed 2, managed for tree growth, was reduced to 21% of gross rainfall as compared with 30.5% for Watershed 1 under free drainage. Watershed 3, managed to minimize offsite impacts, yielded 26% of gross rainfall as drainage. Interception loss accounted for about 14.5% of the gross rainfall. ET amounts computed as the residual in a water balance, were 50%, 60%, and 55% of total rainfall for Watersheds 1, 2, and 3, respectively. The effects of controlled drainage on water table depths, drainage and ET were demonstrated for seasonal and year-to-year variation in rainfall. The controlled drainage treatments affected both drainage volumes and daily peak outflow rates. The treatment in Watershed 3 was more effective in reducing peak outflow rates.

Evaluation of a watershed scale forest hydrologic model

Abstract A watershed scale hydrologic model (DRAINWAT) for drained forested lands was developed by coupling DRAINLOB, a field scale forestry version of DRAINMOD and the ditch and channel routing model section of FLD and STRM. The simulation model was tested with 5 years (1988–1992) of data collected on a 340 ha watershed located near Beaufort in eastern North Carolina. Testing of the model included comparison of observed and simulated daily, monthly, and annual outflows and hourly event hydrographs by three different evapotranspiration (ET) methods. Two of which (Teskey form and GS HR form) are based on the Penman-Monteith method and the third one on the Thornthwaite method. The average absolute deviation in observed and predicted daily outflows for a 5 year period was 0.94 mm day −1 , when the Penman-Monteith methods were used to predict ET. The average absolute deviation in cumulative outflow when ET was predicted by the Thornthwaite method was, respectively, 23% and 50% higher compared with the values obtained with both forms of the Penman-Monteith method. Based on coefficient of determination (R 2 ), coefficient of efficiency ( E ), and root mean square error (RMSE), Teskey and GS HR forms of the Penman-Monteith method performed better than the Thornthwaite method in predicting both daily and monthly outflows. However, the average daily deviations by all three methods were not significantly different at 5% level. Prediction errors in simulating monthly outflows were reduced compared with daily outflows. The predicted mean annual outflow volumes when the GS_HR and Thornthwaite methods were used for ET were in closest agreement with observed data. Statistics showed that errors resulting from use of the Thornthwaite method, with correction factors, were usually within acceptable limits given the large input data required by the Penman-Monteith ET methods. Model prediction of event hydrographs was satisfactory based on different statistical and graphical comparisons. Deviations in predicted and observed results are attributed to errors in both. Errors in the measured outflows occurred for some larger events due to weir submergence. Errors in the simulations resulted from errors in rainfall inputs, and from uncertainties in drainable porosity, hydraulic conductivity and estimates of ET due to a number of factors including approximations of leaf area index (LAI) and stomatal conductance parameters. The model performance as a whole was satisfactory given the complexity of the model, limitations of input data for the watershed, measurement errors in outflow and rainfall, and the fact that the model was not calibrated.

Long-term hydrology and water quality of a drained pine plantation in North Carolina

Long-term data provide a basis for understanding natural variability, reducing uncertainty in model inputs and parameter estimation, and developing new hypotheses. This article evaluates 21 years (1988-2008) of hydrologic data and 17 years (1988-2005) of water quality data from a drained pine plantation in eastern North Carolina. The plantation age was 14 years at the beginning of the investigation (1988) and 34 years at the end (2008). The 21-year average rainfall of 1517 mm was 9% higher than the 50-year (1951-2000) long-term average of 1391 mm observed at the nearest U.S. Weather Bureau station in Morehead City, North Carolina. Annual rainfall varied from 852 mm in the driest year (2001) to 2331 mm in the wettest year (2003) during the study period and was affected by several hurricanes and tropical storms. The runoff coefficient (ROC; drainage outflow expressed as a fraction of rainfall) varied from 0.05 in the driest year to as high as 0.56 in the wettest year (2003), with an average ROC of 0.32. Annual outflow (runoff) on this watershed was primarily subsurface flow to drainage ditches and was strongly correlated with rainfall (R2 = 0.81). Outflows were greater, more continuous, and longer in winter than in other seasons. Outflow in winter was 59% of rainfall on average. March was the only month that never produced zero outflow. The lowest mean outflow occurred in the spring and was significantly different from the other three seasons. Consistent with theory for subsurface drainage, outflow from this poorly drained land is dependent on water table elevation and occurs when the water table is within about 1.1 m of the surface. The water table tended to be close to the surface during the winter and early spring with low ET demands, and during summer with hurricanes and tropical storms producing large outflows, but was drawn down to depths much deeper than the drains during long dry periods in summer and fall. As a result, annual outflow and annual average water table depth were only weakly correlated (R2 = 0.52). There was no relationship (R2 = 0.01) between the annual average water table depth and the annual average evapotranspiration (ET), calculated as the difference between annual rainfall and outflow. The estimated average annual ET of 1005 mm was close to the Penman-Monteith based average annual potential ET (PET) of 1010 mm for a grass reference. Although nitrogen (N) levels in the drainage water were elevated after fertilization of the stand in late 1988, these elevated levels declined substantially by 1995. Average annual concentrations of total N ranged from 0.51 to 2.23 mg L-1 with a long-term average of 1.10 mg L-1. Annual average values for total P ranged from 0.01 to 0.12 mg L-1 with an average of 0.04 mg L-1. The highest average annual concentrations for N and P occurred in 1989 (N) and 1990 (P) following fertilization in spring of 1989. The average annual total N and P loadings were 6.5 ±5.3 kg ha-1 and 0.17 ±0.11 kg ha-1, respectively. Both concentrations and annual loadings were similar to other forested sites in the region. These long-term data should be useful for assessing the effects of land use change and management treatments on the hydrology and water quality of similar lands in the coastal region.

Hydrology of Poorly Drained Coastal Watersheds in Eastern North Carolina

A 10,000 ha lower coastal plain land near Plymouth in eastern North Carolina has been intensively nmonitored since 1996 to measure hydro-meteorological parameters including outflows and quality of water ndrained from fields and subwatersheds with varying land management practices. This study summarized the ndata for a six-year period (1996-2001) for a 2950 ha forested, a 710 ha agricultural subwatershed and a 8140 nha watershed comprised of agricultural, forested, and riparian lands. The period covered a wide range of nweather conditions from a dry year with annual rainfall of 775 mm in 2001 to a wet year with 1512 mm of rain nin 1996 with two hurricanes. While 1998 with 1242 mm of annual rain experienced a wet winter and a nprolonged dry summer-fall, the conditions were opposite in 1999 (1302 mm of rain) with a dry winter-spring nand three hurricanes in the summer and fall. A near normal rainfall (1219 mm of rain) was observed in year n2000. The average annual PET for the site was estimated to be 1000 mm. Variability in annual rainfall was nfound to have greater effect than the land use type on annual outflows drained from these three watersheds. nThe average annual runoff/rainfall ratio for the managed pine forest watershed was the lowest compared to ntwo other watersheds. Both the magnitude and frequency of peak flow rates were highest for the agricultural nwatershed, as expected. Average annual ET, calculated as difference of rainfall and outflow, was 922mm, 714 nmm, and 727 mm for forested, agricultural and mixed land use watersheds, respectively. Annual ET estimated nby the method suggested by Zhang et al. (2001) were in close agreement with the water balance for all six nyears when a plant-available water coefficient value of 3.0 was used for the managed pine forest. However, nfurther tests of this ET model are suggested in other watersheds. These results will be valuable for estimating nnutrient exports from the watershed as well as verifying watershed scale hydrologic and water quality models.

Hydrology and water quality of two first order forested watersheds in coastal South Carolina

Two first-order forested watersheds (WS 80 and WS 77) on poorly drained pine-hardwood nstands in the South Carolina Coastal Plain have been monitored since mid-1960s to characterize the nhydrology, water quality and vegetation dynamics. This study examines the flow and nutrient ndynamics of these two watersheds using 13 years (1969-76 and 1977-81) of data prior to Hurricane nHugo (1989) and nearly 10 years (1990-1992, 1996-99, and 2003-06) immediately after Hugo. WS n80 remained as a control throughout the study period, whereas WS 77 underwent several treatments nincluding prescribed burning, partial harvest, salvage logging and prescribed fire for red-cockaded nwoodpecker habitat management. Depending upon the antecedent moisture conditions, both the nwatersheds were highly responsive of rainfall events throughout the periods. Accordingly, annual noutflows varied from 5% in 1981 to 59% in 1998 with an average of 22% of the annual precipitation nfor the control (WS 80) and from 9% in 2004 to 44% in 1991, with an average of 27% for the ntreatment watershed (WS 77). The coefficient of variation (COV) on WS 80 was higher (55%) ncompared to 36% for the WS 77. Annual rainfall variation was much lower (COV = 14%) than the variation in stream outflows. Post Hugo average outflow from WS 77 increased relative to WS 80 nuntil 1992. By the regeneration period of 1996 reversal in outflow was noticed with the higher noutflows on WS 80 than on the WS 77. While prescribed burning of WS 77 in a course of five years n(1977-81) did not affect on stream outflows and chemistry, mastication in course of nine months in n2001 followed by another prescribed burning of 84% of WS 77 on May 10, 2003 seemed to have nincreased the outflows on WS 77 both in 2004 (64%) and 2005 (70%). Average nutrient nconcentrations were similar on both watersheds although there was a wide variability in NH4-N on the ntreatment watershed (WS 77) compared to WS 80. pH was slightly lower on the WS 77 (5.4) than on nWS 80 (6.8). Both NO3-N and NH4-N concentrations were very low for both the watersheds, before nand after Hugo, with organic nitrogen as the dominant factor on both watersheds. Phosphate was nalso very low (0.02 mg L-1, on average) on both the watersheds during both the periods. Hurricane nHugo substantially increased the nutrient loads primarily due to increase in outflows. Although data npresented herein may serve as baseline information for assessing impacts of both the developments nand natural disturbance in the region, further studies and analysis with additional data should be nconducted to verify some results such as the reversal of flow pattern after the hurricane Hugo that nmay have changed the dynamics of regenerated vegetation after Hugo possibly affecting stream noutflows via evapotranspiration (ET) on these humid coastal plain watersheds.

HYDROLOGIC AND WATER-QUALITY RESPONSE OF FORESTED AND AGRICULTURAL LANDS DURING THE 1999 EXTREME WEATHER CONDITIONS IN EASTERN NORTH CAROLINA

This study evaluated hydrologic and water-quality data collected on a coastal-plain research watershed during a series of hurricanes and tropical storms that hit coastal North Carolina in 1999, including hurricanes Dennis, Floyd, and Irene. During September and October 1999, the research watershed received approximately 555 mm of rainfall associated with hurricanes. This was the wettest such period in a 49-year historical weather record (1951-1999). Prior to the hurricanes, the watershed experienced a dry late winter, spring, and summer (565 cm for Feb.-Aug.). This was the third driest such period in the 49-year record. Maximum daily flow rates measured across the research watershed were greater during hurricane Floyd than for any other time in a four-year (1996-1999) study of the watershed. Daily flows observed for an agricultural subwatershed were generally greater than for a forested subwatershed throughout the study, and during the hurricanes of 1999. Daily nutrient loads measured across the research watershed were greater during hurricane Floyd than for any other time in the study. In general, the two-month period of hurricanes produced total nitrogen and total phosphorus loads nearly equal to average loads for an entire year. Total annual nitrogen export from an agricultural subwatershed was 18 kg/ha in 1999, with 11 kg/ha (61%) lost during September and October. Total annual nitrogen export from a forested subwatershed was 15 kg/ha in 1999, with 10 kg/ha (67%) lost during September and October. The nitrogen export observed in the forested subwatershed was high compared to other forested areas, likely due to the highly permeable organic soils in the watershed. Total annual phosphorus export from an agricultural subwatershed was 0.9 kg/ha in 1999, with 0.7 kg/ha (78%) lost during the hurricanes/tropical storms. Total annual phosphorus load from a forested subwatershed was 0.1 kg/ha in 1999, with 74% of the load exported during the months of September and October. Hurricanes and floods occur with some regularity in North Carolina, but the effects are infrequently documented. This study provides information that will contribute to greater understanding of how watersheds respond to these events.

Leaf Area Index (LAI) of Loblolly Pine and Emergent Vegetation Following a Harvest

Forests provide goods and services to society and, often, refugia for plants and animals; forest managers utilize silviculture to provide ecosystem services and to create habitat. On the Coastal Plain of North Carolina, forest management objectives typically include wood fiber production but may also include the maintenance of environmental quality and, sometimes, species diversity. Silvicultural prescriptions alter stand structure and development trajectories by influencing the competitive interactions among plant species for site resources. Early site intervention may include nutrient additions and/or vegetation control; in coastal loblolly pine (Pinus taeda L.) stands, herbaceous and arborescent species can dominate the site leaf area index (LAI) for many years after a harvest (followed by planting). LAI is an important structural and functional component of a forest stand. Many eco‐hydrologic and water quality models do not accurately account for LAI as the process driver to evapotranspiration (ET), and thus they ignore the ecophysiological effects of LAI on site water balance and nutrient loading. We measured LAI of emergent vegetation following a harvest, mechanical site preparation, and then pine planting for a drained loblolly pine plantation in coastal North Carolina. For six years monthly, growing season estimates of LAI were obtained using a LI‐COR LAI 2000 Plant Canopy Analyzer (PCA) for control (D1), thinned (D3), and harvested (D2) watersheds. In this article, we present results from the D2 treatment. In D2, we “harvested” all emergent vegetation in 18 randomly placed 1 m2 clip plots for three growing seasons where we estimated LAI using species‐pooled estimates of specific leaf area and total leaf dry mass (i.e., LAICLIP); PCA measurements were recorded prior to clipping (LAIPCA). We also simulated loblolly pine seedling growth and development using the biogeochemical process model SECRETS‐3PG to examine site differentiation in LAI. Four years post‐harvest maximum LAICLIP exceeded 8 m2 m‐2 (projected area basis). LAIPCA underestimated LAICLIP; LAICLIP = 1.436 × LAIPCA (r2 = 0.53; p < 0.0001; n = 195). Corrected LAIPCA estimates exceeded simulated pine LAI (LAISIM) for ~4.5 years post‐planting. Emergent vegetation dominated the site for nearly five years and likely exerted a strong influence over site water balance and nutrient use during early stand development.

Effects of Land Use on Soil Properties and Hydrology of Drained Coastal Plain Watersheds

Some of the worlds most productive cropland requires artificial or improved drainage for efficient agricultural production. Soil hydraulic properties, such as hydraulic conductivity and drainable porosity, are conventionally used in design of drainage systems. While it is recognized that these soil properties vary over a relatively wide range within a given soil series, it is generally assumed they can be approximated based on soil type, independent of crop or land use. Effects of land use on hydrology of drained soils in the North Carolina lower coastal plain were investigated by comparing hydrologic measurements on drained agricultural cropland, drained forest land (Loblolly pine), and an undrained forested wetland. Higher ET on the drained pine forest site resulted in reduced drainage outflow and deeper water tables compared to the agricultural site. Measurements for the wetland site showed water tables near the surface but annual outflows similar to the drained forest site. Field effective hydraulic conductivity in the top 70 cm of the drained forest site was more than two orders of magnitude greater than that of corresponding layers of the soil on the agricultural site. Drainable porosity, based on measured soil water characteristics, was also much higher for the forested sites. Long term (50-year) DRAINMOD simulations predicted average annual drainage outflow of 51.4 cm for the agricultural field as compared to 37.6 cm for the forested site. The difference resulted primarily from greater ET predicted for the forested site. Because of the high hydraulic conductivity of the surface layers and large surface depressional storage, predicted surface runoff from the forested site was nil, compared to an average annual runoff of 13 cm for the drained cropland site. Results of long-term simulations were used to analyze these effects for the widely variable seasonal and annual weather conditions of eastern North Carolina.

Advances in forest hydrology: challenges and opportunities

Forests are an integral component of the landscape, and maintaining their functional integrity is fundamental for the sustainability of ecosystems and societies alike. Tools, innovations, and practices, analogous to those developed to improve agricultural production and quantify environmental impacts, are needed to ensure the sustainability of these forested landscapes as well as the ecosystem goods and services they produce. This article introduces ten technical articles on critical ecohydrologic processes, protection and restoration, and the effects of management practices on the hydrology and water quality of forests and forested wetlands, using both monitoring and modeling approaches. Prepared by experts in forest science, forest and agricultural hydrology, and water management, the studies reported in this special collection are concentrated in the Atlantic Coastal plain and focus on forests with shallow water tables. Experimental studies describe the effects of riparian vegetation harvest, human disturbance, and future climatic change on groundwater, the significance of emergent vegetation after harvest, and long-term hydrologic water balance of a managed pine forest. Modeling studies use the SWAT model to predict streamflow dynamics of a less disturbed, coastal forested watershed, and DRAINMOD to determine the impacts of minor silvicultural drainage on wetland hydrology and to improve wetland restoration. Finally, a study describes potential uncertainties associated with infrequent water sampling of nutrient loads from drained forested watersheds. This introductory article summarizes these studies of shallow water table forests and relates them to the broader field of forest hydrology, including its challenges and opportunities, while identifying pressing issues of land use and climate change. The results from these studies should help guide management and restoration of forest wetland ecosystems and direct future forest hydrologic research, including research in large prior converted agricultural landscapes.

Testing of DRAINMOD for Forested Watersheds with Non-Pattern Drainage

Models like DRAINMOD and its forestry version, DRAINLOB, have been specifically developed as na field scale model for evaluating hydrologic effects of crops (trees), soil, and water management practices for nlands with pattern drainage (i.e. with parallel ditches) on relatively flat, high water table soils. These models nconduct a water balance between the ditches to predict water table depths, drainage rates, surface runoff, ET, nand soil water storage. However, a vast landscape under silvicultural management in the coastal plains along nthe southeast and Gulf Coast region consists of lands with non-pattern drainage systems. Reliable models are nneeded to determine the processes and water quality impacts of management practices on these lands also. nIn this study data from two naturally drained forested watersheds without the pattern drainage in Florida and nSouth Carolina coasts were used to test the ability of DRAINMOD to predict water table depths and drainage noutflow rates. A large ditch spacing and a shallow drain depth were assumed to simulate outflows from both of nthese flat depressional wet sites. With minimal field calibration, the model’s predictions of daily outflows from nSantee watershed in South Carolina were satisfactory. The predicted outflows for Bradford watershed on nFlorida flatwoods with a higher level of field calibration were found to be better, as expected, for wet, dry, and nnormal years. The results suggest that reliable estimates of surface storage and PET inputs in DRAINMOD nwith a surface flow routing component may further enhance the flow predictions on these watersheds.