The Hubbard Brook Ecosystem Study 40 Years of Applied Biogeochemical Research Chris E. Johnson Syracuse University

http://www.lternet.edu

The Hubbard Brook Experimental Forest Established in 1955 by the USDA Forest Service for hydrologic research. Hubbard Brook Ecosystem Study Initiated in 1963 using the small watershed approach to study hydrologic cycle-element interactions in small undisturbed and human-manipulated forest ecosystems.

Characteristics of the Hubbard Brook Experimental Forest Bedrock Quartz Mica Schist and Quartzite Landscape Till-Mantled Glacial Valleys Soils Spodosols (Typic Haplorthods) pHw %BS Oa 3.9 50 Mineral Soil 4.3 12 Vegetation Northern Hardwood Forest; Cutting 1915-17; 80-90% Hardwoods, 10-20% Conifers Climate Humid Continental, Mean Precipitation 1400 mm

Characteristics of Monitored Watersheds Watershed Number Size (ha) Year Started Treatment 1 11.8 1956 Calcium silicate addition 1999 2 15.6 1957 Clear-felled in ‘65-66, no products removed, herbicide application ‘66,67, 68. 3 42.4 1958 None – Hydrologic reference. 4 36.1 1961 Clear-cut by strips in three phases – ‘70,72,74. Timber products removed. 5 21.9 1962 Whole-tree clear-cut in 1983-84. Timber products removed. 6 13.2 1963 None – Biogeochemical reference. 7 76.4 1965 None 8 59.4 1969 9 68.0 1994 101 12.1 1970 Clear-cut as block in 1970. Timber products removed.

Hubbard Brook Ecosystem Study Study Objectives: To assess the structure and function of a northern hardwood forest ecosystem in various stages of development. To quantify hydrologic and element cycle interactions in undisturbed watershed and lake ecosystems. To evaluate the short-term and long-term responses of ecosystems to disturbance (i.e., clear-cutting, changes in land use, air pollutants, insect outbreaks, climate).

Changes in Lead (Pb) Cycling with Decreasing Atmospheric Inputs Background Major sources of Pb: Smelting, Battery Production, Paints, Gasoline (Petrol). Alkyl-Pb compounds used as anti-knock additives in gasoline beginning in 1923. With increasing automobile/truck traffic, gasoline became principal source of atmospheric Pb in USA. 1970: Clean Air Act; General Motors announces intent to comply by installing catalytic converters beginning in 1974. Other auto makers follow. US Pb consumption declines > 90% 1975 – 1985. Natural ecosystem experiment…

Pb concentration in bulk precipitation has declined to ~1% of 1975 values

Precipitation Pb concentration may be stabilizing

Streamwater Pb concentrations have been low throughout the study period

Remarkably… Pb input in precipitation has declined by more than 98% BUT Precipitation input continues to exceed stream output So, what is happening in the soil?

The Pb content of the O horizon has declined by 40% since 1976 (…or has it?) Johnson et al. (1995) - updated

Pb inputs are related to consumption Johnson et al. (1995)

Reconstructed Pb input chronology Johnson et al. (1995)

Modern (1926-1997) Lead Budget for the HBEF All values: kg/ha 1. Atmospheric Deposition – 1926-1997 8.76 2. Pb in Forest Floor - 1997 6.80 3. Estimated Pb in Forest Floor - 1926 1.35 4. Net Accumulation of Pb in Forest Floor (2) – (3) 5.45 5. Estimated Stream Flux – 1926-1997 (0.7 mg/L, 87 cm yr-1 runoff) 0.43 6. Flux to Mineral Soil (1) – (4) – (5) 2.88

Pb fractionation through selective extraction

Most of the Pb in Hubbard Brook soils is in the crystalline form Johnson and Petras (1998)

‘Labile’ Pb is largely bound with organic matter and amorphous oxides Johnson and Petras (1998)

Enrichment of Ti during weathering

The depletion of crystalline Pb can be used to estimate post-glacial weathering release Johnson et al. (in press)

Post-Glacial Pb Budget Weathering release 14.1 kg/ha 20th Century deposition 8.7 kg/ha Sum 22.8 kg/ha ‘Labile’ Pb in soil 22.1 kg/ha This ignores pre-industrial inputs and outputs. 1999-2001: Bulk Precipitation: 250-750 ng/L Streamwater: 100-300 ng/L

Changes in Soil Chemistry 15 Years after Whole-Tree Clear-felling

Nutrient Pools – Biomass vs. Soil (kg/ha)

Nutrient Release after Clear-felling Bormann & Likens: Pattern and Process in a Forested Ecosystem (1979)

Hypotheses (Soil Biogeochemistry) Whole-tree clear-felling results in increased leaching losses of N, Ca, K, Mg. Leaching losses and uptake by regrowing vegetation result in significant decreases in exchangeable Ca (and other nutrient cations).

W5 Whole-Tree Harvest 1983-84

Hubbard Brook Experimental Forest, NH (1997) W6 (Reference) W5 W2

Measurements

Stream Chemistry W5 () W6 () Annual volume-weighted average concentrations Thanks: Gene Likens

No Change in Exchangeable Ca Pools (N = 60 0 No Change in Exchangeable Ca Pools (N = 60 0.5 m2 pits per sampling year) Johnson et al. (1991, 1996) updated

Redistribution of Ca within the Profile Johnson et al. (1991, 1996) updated

Ca Saturation of CEC

1984-99: 15-year Cumulative Fluxes

New Hypotheses & Questions Retention of exchangeable cations after harvesting was related to changes in soil organic matter structure/chemistry (13C-NMR studies, HA/FA/Humin fractionation, titration expts.) Why have elevated Ca losses in stream water persisted so long after harvesting? (It’s not pH!) (How) are stream fluxes related to soil Ca pools? Budgets vs. Geochemistry

“Swimming Rock” Mirror Lake Questions?

Simple, 1st- Order Box Model

Accumulation of Pb in forest floor soils

Use total Pb to estimate post-glacial gain/loss throughout the profile

Conclusions Concentrations of Pb in precipitation have declined by 99% since 1975. Streamwater Pb concentrations – low throughout the study period – remain lower than precipitation. Ecosystem is a sink for Pb. The pool of Pb in the forest floor has declined by ~40% since 1976. Labile Pb in the mineral soil is dominated by organically bound Pb and Pb bound to/in amorphous oxides (consistent with podzol soil processes). Post-glacial weathering release and 20th-century Pb deposition can account for all labile Pb in the soil. Pb fractionation studies may be a useful tool for monitoring the fate of anthropogenic Pb in forest soils.

Conclusions Exchangeable pools of Ca and other bases did not change significantly after harvesting. Large fluxes of Ca in stream water and into regrowing vegetation were balanced largely by release from decomposing residual biomass. We did observe redistribution of exchangeable bases within the profile. Calcium saturation of CEC increased in mineral soils after harvesting. This may be a more sensitive indicator of changing base status than the pool size.