Legacies of agriculture and forest regrowth in the nitrogen of old-field soils

https://doi.org/10.1016/S0378-1127(00)00399-6Get rights and content

Abstract

In the Carolina Piedmont of the USA, agricultural and forest management in the 19th and 20th centuries has greatly altered soil organic nitrogen (N). The objective of this study is to evaluate effects of two centuries of land use on N in upland Piedmont soils that are derived from the region’s most common bedrock, granitic gneiss. Effects of agriculture on total soil N were examined by comparing soils cropped mainly for cotton since about 1800 with soils that remained under hardwood forest without cultivation or fertilization. Effects of forest regrowth on the N of old-field soils were examined in eight permanent plots resampled on seven occasions from 1962 to 1997 at the Calhoun Experimental Forest in South Carolina.

Together, the soil-comparative study and the four-decade field experiment illustrate how soil N in the southern Piedmont has been altered by agricultural management during the 19th and 20th centuries. Not only have agricultural harvests removed considerable N from Piedmont soils, but soil organic matter has been enriched in N by agricultural fertilization, a practice that has now contributed greatly to N cycles of many old-field forests in the region.

In old-field pine stands (Pinus taeda) at the Calhoun Experimental Forest, 40 years of forest growth accumulated 366 kg ha−1 of N (CV=9.3%) in tree biomass and 740 kg ha−1 (CV=9.7%) in forest floor between planting in 1957 and the last sampling in 1997. In the four decades, mineral-soil N was diminished by 823 kg ha−1 (CV=39.5%), a reduction in N accompanied by substantial decreases in mineralizable N as well. On the other hand, N accretion in the whole forest ecosystem averaged 5.9 kg ha−1 per year over this period (significant at a probability of <0.07), an accretion attributed mainly to atmospheric N deposition rather than N2 fixation. Despite the N accretion and legacy of agricultural fertilization, the 40-year-old Calhoun forest has grown into a state of acute N deficiency. Future N research should include support for a network of long-term field studies which investigates N dynamics in forest floor and logging slash, and estimates N-use and N-retention efficiencies of fertilized pine-forest ecosystems.

Introduction

Managing soil nitrogen (N) in the 35-million hectare pine and pine–hardwood forests of the southeastern USA is far from trivial in the 21st century. For example,

  • Acidic, advanced weathering-stage soils (Ultisols) dominate the region. Many of these soils are characterized by generally low native fertility.

  • Most soils in the region are highly disturbed by 150–300 years of agriculture. A legacy of agriculture is prominent in southeastern soils as cultivation, accelerated erosion, fertilizer inputs, and crop harvests have greatly altered soil fertility.

  • Southeastern forests have nearly doubled in standing wood volume between the 1950s and 1990s, as the regional agricultural landscape has changed to one that is largely forested with a mixed set of uses. The nutrient requirements of this rapidly growing forest have made large demands on soil fertility, as have the forest harvests that have accompanied forest regrowth across the region. More wood is currently harvested for industrial products from southeastern forests than from any other wood-producing region on earth, about 250 million m3 per year (Powell et al., 1993; Jaakko Pöyry Group, 1994).

  • The fraction of the southeastern forest that is most intensively managed for high fiber production relies fundamentally on management of soil fertility. A part of the new soil management includes forest N and P fertilization, a practice used in the 1990s on more than a half million hectares of high-productivity southeastern forest.

The forest N cycle, including inputs, returns, and losses, is relatively well quantified throughout the world in studies that span one to several annual cycles (Johnson and Lindberg, 1992). On the other hand, we have hardly any estimates of changes in the forest N cycle and in particular in soil N over time-scales of decades. Although computer simulations are important for understanding changes in N in ecosystems over these time-scales, they are no substitute for direct observations of soil and ecosystem change. Even small errors in annual estimates of N circulation carried over time-scales of decades magnify our uncertainties of whether soil N inputs are outpaced by removals and whether ecosystems are actually gaining or losing N. Ecosystem experiments which quantify soil change over decades of management are therefore invaluable to our perspectives of ecosystem sustainability (Jenkinson, 1991; Leigh and Johnston, 1994; Mitchell et al., 1996).

A simplified, generic N budget illustrates how managed southeastern pine forests may gain or lose N over time-scales of decades (Table 1). Nitrogen removals in timber harvest plus hydrologic leaching range from <120 to 400 kg ha−1 in pine forest stands harvested on a 25-year rotation (Table 1). Compensating these removals are inputs of N from atmospheric deposition plus N2 fixation that range from 5 to 10 kg ha−1 per year (Binkley et al., 1989a; Johnson and Lindberg, 1992; Fox and Mikler, 1995; Richter and Markewitz, 1995), inputs that over a 25-year forest-rotation may amount to 150–300 kg ha−1. The N balance hypothetically ranges from a surplus of 180 kg ha−1 to a deficit of 250 kg ha−1, a range that depends substantially on ecosystem productivity, harvest utilization, and measurement errors. Table 1 helps to demonstrate why long-term soil–ecosystem experiments are important to quantifying ecosystem-N dynamics over time-scales of decades (Stone, 1975; Powers and Van Cleve, 1991; Richter and Markewitz, 2000).

In this paper, we estimate changes in soil N in a four-decade soil–ecosystem experiment (at the Calhoun Experimental Forest in South Carolina), and we place these relatively recent changes in soil N in the context of soil N as it was affected by more than a century of prior agricultural use (1800–1950).

In the upland Piedmont of southeastern USA, changes in soil N are particularly interesting because three distinctly different ecosystems (Fig. 1) correspond to major changes in the regional N cycle over time-scales of millennia, centuries, and decades (Table 2):

  • A primary forest ecosystem, mainly of oak and hickory species, that covered the upland Piedmont for thousands of years prior to 1700.

  • An agro-ecosystem, that dominated the southern Piedmont region from about 1800 to the early 20th century, a system managed for cotton, tobacco, corn, wheat, and other agricultural products.

  • A secondary, old-field forest ecosystem, that in the late 20th century grows on >35 million hectares, much of which was established in pine between 1920 and 1960.

Of the paper’s five specific objectives, two evaluate changes in soil N associated with forest clearing and agricultural use of these soils for more than a century, two investigate changes in soil N in aggrading pine stands that grow on former agricultural fields abandoned since the 1950s, and one integrates these changes in soil N over time-scales of centuries and decades. Explicitly stated, the objectives are to estimate:

  • 1.

    soil N under unfertilized and uncultivated deciduous hardwood forests, selected to represent conditions close to those of the primary forest before 1700;

  • 2.

    changes in soil N that are associated with long-term cotton and other agricultural uses from about 1800 to present;

  • 3.

    changes in mineral-soil N in old-fields over four decades of pine-forest development;

  • 4.

    N accretion in the four-decade-old pine ecosystem (biomass plus soil); and

  • 5.

    a N budget of upland Piedmont soils that spans the period from pre-1800 to the present.

Section snippets

Methods

To accomplish the paper’s objectives, we used a soil-comparative study and a long-term field experiment with all plots on or near the Calhoun Experimental Forest, SC, USA (Richter and Markewitz, 2000). The fifth objective combines data from both the comparative and experimental studies with information from the history of cotton cultivation in the Piedmont and across the southeastern USA (Vance, 1929; Gray, 1933; Sheridan, 1979; Mitchell et al., 1996).

Legacies of primary forest and agricultural use in modern Calhoun soils

The hardwood forest soils (Table 3), which we assume have never been cultivated or fertilized, are our closest approximations to pre-agricultural soil conditions in these Piedmont ecosystems. Total N in the upper 30-cm soil of hardwood forests (Table 4) averaged 384 μg g−1 and 1551 kg ha−1 (CV=36.4%). Soils that support hayfields or row cropped corn have higher contents of soil N (Table 4), 2337 (CV=19.5%) and 1765 kg ha−1, respectively, no doubt due to long-term N fertilization. Old-field pine

References (43)

  • C.E. Pehl et al.

    Total biomass and nutrients of 25-year-old loblolly pines (Pinus taeda L.)

    For. Ecol. Mgmt.

    (1984)
  • B.N. Richards

    Nitrogen fixation in the rhizosphere of conifers

    Soil Biol. Biochem.

    (1973)
  • D.H. Van Lear et al.

    Comparison of biomass equations for planted vs natural loblolly pine stands of sawtimber size

    For. Ecol. Mgmt.

    (1986)
  • V.C. Baldwin

    Green and dry-weight equations for above-ground components of planted loblolly pine trees in the West Gulf region

    South. J. Appl. Forestry

    (1987)
  • Binkley, D., Driscoll, C.T., Allen, H.L., Schoeneberger, P., McAvoy, D., 1989a. Acid Deposition and Forest Soils....
  • B.T. Bormann et al.

    Rapid N2 fixation in pines, alder, and locust: evidence from the sandbox ecosystem study

    Ecology

    (1993)
  • Buol, S.A., Hole, F.D., McCracken, R.J., 1989. Soil Genesis and Classification. University of Iowa Press, Ames,...
  • Fox, S., Mikler, R.A., 1995. Impact of Air Pollutants on Southern Pine Forests. Ecological Studies, Vol. 118. Springer,...
  • Gnau, C.B., 1992. Modeling the hydrologic cycle during 25 years of forest development. Masters Project. School of the...
  • D. Grant et al.

    Rates of free-living nitrogen fixation in some Piedmont forest types

    For. Sci.

    (1987)
  • Gray, L.C., 1933. History of Agriculture in the Southern United States to 1860. The Carnegie Institution, Washington,...
  • Jaakko Pöyry Group, 1994. Global fiber resources situation: the challenges for the 1990s. Unpublished report. Jaakko...
  • Jenkinson, D.S., 1971. The accumulation of nitrogen in soil left uncultivated. Report of the Rothamsted Experiment...
  • D.S. Jenkinson

    Rothamsted classical experiments: are they still of use?

    Agron. J.

    (1991)
  • Johnson, D.W., Lindberg, S.E., 1992. Atmospheric Deposition and Nutrient Cycling in Forest Ecosystems. Springer, New...
  • J. Jorgensen et al.

    Nutrient changes in decomposing loblolly pine forest floor

    Soil Sci. Soc. Am. J.

    (1980)
  • P.R. Kapeluck et al.

    A technique for estimating below-stump biomass of mature loblolly pine plantations

    Can. J. For. Res.

    (1995)
  • Leigh, R.A., Johnston, A.E. (Eds.), 1994. Long-term Experiments in Agricultural and Ecological Sciences. CAB...
  • D. Markewitz et al.

    Three decades of observed soil acidification at the Calhoun Experimental Forest: has acid rain made a difference?

    Soil Sci. Soc. Am. J.

    (1998)
  • R.J. McCracken et al.

    Undisturbed soils, landscapes, and vegetation in a North Carolina Piedmont virgin forest

    Soil Sci. Soc. Am. J.

    (1989)
  • Mitchell, C.C., Arriaga, F.J., Entry, J.A., Novak, J.L., Goodman, W.R., Reeves, W., Runge, M.W., Traxler, G.J., 1996....
  • Cited by (136)

    • Mapping depth to the argillic horizon on historically farmed soil currently under forests

      2020, Geoderma
      Citation Excerpt :

      The predominant series have an abrupt or clear boundary (<5 cm transition) between the A or E and Bt that is generally smooth (i.e., few irregularities). Soil series include Cecil, Appling, and Madison, which are fine, kaolinitic, thermic Typic Kanhapludults, and Wilkes, which is a loamy, mixed, active, thermic, shallow Typic Hapludalf (Richter et al., 2000). Soil augerings were used to investigate the spatial variations in the depth-to-argillic layer.

    • Potential release of legacy nitrogen from soil surrounding onsite wastewater leaching pools

      2020, Water Research
      Citation Excerpt :

      Our observation is also consistent with previous reports regarding the slow legacy nitrogen leaching estimations. It was reported that the nitrogen loss from water drainage over the life of a forest was very little and dissolved organic nitrogen leached from forest soil was estimated to be < 30 mg-N/m2 (Richtera et al., 2000). Meter at al. found out there was a growing pool of soil organic nitrogen (142 ⨯ 10 12 g-N over 30 years) in agricultural soils of Mississippi river basin that will have a 35 years lag time to deplete even if the addition of organic nitrogen was stopped at the time of study (Meter et al., 2016).

    View all citing articles on Scopus
    View full text