J. M. Kelly

Aluminum toxicity in forests exposed to acidic deposition: The ALBIOS results

The ALBIOS project was conducted to examine the influence of acidic deposition on aluminum transport and toxicity in forested ecosystems of eastern North America and northern Europe. Patterns of aluminum chemistry were evaluated in 14 representative watersheds exposed to different levels of sulfur deposition. Controlled studies with solution and soil culture methods were used to test interspecific differences in aluminum sensitivity for one indicator species (honeylocust - Gleditsia triacanthos L. ) and six commercial tree species (red spruce - Picea rubens Sarg., red oak - Quercus rubra L., sugar maple - Acer saccharum Marsh., American beech - Fagus grandifolia Ehrh., European beech - Fagus sylvatica, and loblolly pine - Pinus Taeda L. ). Overall, red spruce was the tree species whose growth was most sensitive to soluble aluminum, with significant biomass reductions occurring at Al concentrations of approximately 200–250 umol/L. Analyses of soil solutions from the field sites indicated that the conditions for aluminum toxicity for some species exist at some of the study areas. At these watersheds, aluminum toxicity could act as a contributing stress factor affecting forest growth.

Sulfur Cycling in Five Forest Ecosystems

The cycling and retention of sulfur were studied in five forest ecosystems: a chestnut oak and yellow poplar stand on Walker Branch Watershed, Tennessee; a mixed oak stand on Camp Branch Watershed, Tennessee; and a red alder and Douglas-fir stand at the Thompson site, Washington. Calculations from foliage sulfur turnover indicate that about one-half of total sulfur input was dry in the Tennessee sites, whereas only one-tenth was dry in the Washington sites. Atmospheric sulfur inputs exceeded forest sulfur requirements in all cases, but three sites (chestnut oak, mixed oak, and red alder) showed a net ecosystem retention of atmospherically deposited sulfur. Net ecosystem sulfur retention was consistent with laboratory-determined sulfate adsorption isotherms within a given location (Walker Branch, Thompson site) but not between locations because of differing deposition histories and consequent differing degrees of soil sulfate saturation. No consistent relationships between soil sulfate adsorption capacity and other soil properties (pH, base saturation, iron, and aluminum oxides) were found.

Soil nutrient leaching in response to simulated acid rain treatment

Soil and soil solution nutrient concentrations were evaluated over a 30-mo period to determine the impact of simulated acidic precipitation (70:30 equivalent basis H2SO4: HNO3) at pH values of 5.7, 4.5, 4.0, and 3.5 on forest. microcosms. Soil nutrient analysis indicated significantly lower concentrations of exchangeable Ca and Mg in the top 3.5 cm of the mineral soil after 30 mo of pH 3.5 treatment. Leachate collected from the pH 4.5, 4.0, and 3.5 treatments at the 25 cm depth (below the Å.: horizon) exhibited significant increases in Cl, NH4, PO4, K, and SO4 concentrations compared to the pH 5.7 treatment. At the 50 cm depth (mid-profile) all leachate element concentrations except NH4 increased significantly in response to treatment. At the 100 cm depth (profile bottom), no significant effects of treatment on leachate chemistry were observed. The elevated base cation concentration values found in the 50 cm soil solution samples support at least partially the described reduction in Ca and Mg in the surface soil horizon. The 100 cm concentration data indicate that cations mobilized out of the Å.: and upper B horizon in response to treatment were immobilized before reaching the bottom of the soil profile. Evaluation of nutrient flux out of the microcosm at the 100 cm depth did not indicate any statistically significant response to the treatment. Nitrate rather than SO4 was found to be the dominant anion leaving the microcosm by an average factor of ∼7 to 1.

Elemental patterns in roots and foliage of mature spruce across a gradient of soil aluminium

Tissue concentrations of Al in red and Norway spruce trees were compared across 5 sites in North America and Europe as part of an investigation of Al biogeochemistry in forested ecosystems (ALBIOS). Fine roots and foliage were sampled and analyzed for Al, Ca, Mg, and P, and the chemistry of soil and soil solutions was characterized at each plot by horizon. Sites exhibited a wide range in soil Al saturation and in concentrations of Al and sulfate in lysimeter solutions. Aluminium concentrations in roots were two orders of magnitude higher than those in foliage. Fine roots (<1.0 mm) from B horizons had the highest Al concentrations and appeared to be the best phytoindicators of plant-available Al. Aluminium concentrations in fine roots from B horizons were highly correlated with soil solution monomeric Al, and with Al in 0.01 M SrC2. soil extracts. Stronger soil Al extractants were generally poor predictors of concentrations of Al in plant tissue. Sites with higher levels of plant-available Al supported spruce trees with correspondingly lower foliar levels of Ca and Mg. As such, these field sites provided circumstantial evidence that Al may be interfering with Ca and Mg uptake and transport. No evidence was found of Al interference with P uptake or transport at these sites.

Ozone, acidic precipitation, and soil Mg impacts on soil and loblolly pine seedling nutrient status after three growing seasons

Recent studies have suggested that the growth of loblolly pine (Pinus taeda L.) has declined in the southeastern United States, possibly due to acidic deposition and air pollutants, especially under conditions of low nutrient availability. Consequently, the potential for individual and synergistic impacts of O3, acidic precipitation, and soil Mg status on the nutrient status of loblolly pine seedlings and soil was investigated over a 3 yr study period. Thirty-six open top chambers equipped with a rainfall exclusion/addition system were utilized to administer three levels of O3 (subambient, ambient, or twice ambient) and two acidic precipitation treatments (pH 3.8 or 5.2) to seedlings growing in 24-L plastic pots containing soil having either 35 or 15 mg kg−1 of exchangeable Mg. Each chamber contained 36 pots, and each treatment combination was replicated six times for a total of 1296 individual pots. After three seasons, throughfall and foliar nutrition data indicated that foliar leaching was not accelerated by increasing the acidity of precipitation from pH 5.2 to 3.8 and that increasing O3 did not act to exacerbate foliar leaching. Further, foliar nutrient concentrations were not significantly affected by precipitation pH or O3 treatments. Soil and soil solution data also indicate no accelerated soil leaching associated with chronic acidic precipitation. Differences in soil Mg treatments were reflected in soil solution and seedling Mg contents, but the 15 mg kg−1 soil Mg treatment was not sufficiently low enough to induce Mg deficiency in the seedlings.

CO2 efflux from deciduous forest litter and soil in response to simulated acid rain treatment

Using both field and laboratory measurements of CO2 evolution as an index of decomposer activity, forest microcosms were used to evaluate the impact of simulated acidic precipitation on decomposition. The following pH treatments: 5.7, 4.5, 4.0, and 3.5 annual average were applied for a 30 mo period. No statistically significant effect of treatment on decomposition could be found in the field measurements. When the microcosm was partitioned into 01 and 02 litter, mineral soil (A and B horizons), and roots within the mineral soil horizons for laboratory determination of CO2 efflux, only the 02 litter exhibited a statistically significant decrease as a function of treatment. The data collected do not allow a complete evaluation of the potential impact of this decrease. However, efflux of CO2 from the 02 layer was small compared to the other layers, and this may account for the failure to detect a significant response in the field measurements. Although the field data did not exhibit a significant response, there is sufficient question concerning the 02 response to warrant additional investigation, especially since many plants derive a major portion of their nutritional requirements directly from the 02 litter layer.

Influence of elevated ecosystems levels on litter decomposition and mineralization

Litter decomposition was studied at two forested watersheds in east Tennessee which differed primarily in their past history of atmospheric S input. Cross Creek Watershed, located near a large coal-fired power plant, has received greater S inputs than the more remote Camp Branch Watershed. Decomposition was estimated through the measurement of forest floor respiration, litter microflora populations, litter and soil microarthropod populations, and litter nutrient status. Average forest floor respiration rates were very similar, 6.78 g CO2 m−2 day−1 or 2472 g m−2 yr−1 at Camp Branch and 6.86 g CO2 m−2 day−1 or 2505 g M−2 yr−1 at Cross Creek. Fractional loss rates provided estimates of annual decay rates (k) of 0.35 and 0.39 for Camp Branch and Cross Creek, respectively. Litter decomposition was estimated to contribute 23% of the total CO2 output at Camp Branch and 26% at Cross Creek, while root respiration accounts for about 43 to 46%. Bacterial and fungal populations were about equal in size at both watersheds, with bacteria averaging 100 × 106 g−1 of litter and fungi 23 × 106 g−1 of litter. Total numbers of arthropods averaged 34% greater at Camp Branch. Acarina populations averaged 59% higher at Camp Branch, while Collembola numbers were about equal at the two watersheds. Nutrient mobility in the litter and soil was similar at both watersheds. The order of decreasing mobility was K, Mg, Ca, S, N, and P. Litterfall nutrient concentrations were slightly higher for all elements at Cross Creek, resulting in greater litter concentrations of Ca and Mg. Litter concentrations of S and N, however, were significantly greater at Camp Branch, indicating watershed differences in the loss rates and cycling processes of these elements. There were no differences between the loss rates or litter concentrations of P, K, and Na at either site. Overall, decomposition was similar at the two watersheds. Historic S inputs do not appear to have had a major effect on decomposition rate or decomposer organisms with the possible exception of lowered arthropod populations at Cross Creek.

A short-term microcosm evaluation of CO2 evolution from litter and soil as influenced by SO2 AND SO4 additions

Segments of the forest floor (litter and top 5 cm of the mineral soil) collected from two forest soils (one impacted by 25 yr of elevated S input, the other representing background conditions) were placed in 16 × 27 cm plastic boxes, and maintained in a controlled environment while being subjected to various S additions over a 12-wk period to simulate a growing season response time. CO2 evolution, as measured by the soda-lime technique, was used as an index of decomposer organism response to S additions. High-S (0.45 ppm SO2,0.4 g SO4 m−2 wk−1) inputs reduced respiration rates. This rate is equivalent to an annual S input in excess of ∼ 66 kg ha−1 yr−1 Initial S additions at all levels produced a slight stimulation of respiration, but low to moderate inputs produced no lasting effects. However, there does appear to be a threshold level below which the addition of S has a neutral to positive impact on decomposition but beyond which negative impacts on respiration rates will begin to occur. Microcosms taken from the two study sites showed no differences in physical characteristics, respiration rates, or response to simulated growing season S treatments even though one site had been subjected to approximately 25 yr of elevated S input.

Litterfall sulfur and nitrogen inputs as influenced by power plant proximity

The mass of litterfall input, and its weight and concentration of S and total N, were evaluated over a 3 yr period at two locations on the Cumberland Plateau in Tennessee to determine if proximity to a major coal-fired power plant would have a significant impact on elemental input to the forest floor via litterfall. Higher levels of atmospheric S and N input have been reported for the Cross Creek site versus the Camp Branch location. As a general rule, the mass of litterfall and the amount of S and N transferred from standing biomass to the forest floor by litterfall did not differ in a statistically significant (P = 0.05) manner as a function of year, cover type, or location. The limited number of significant responses which were observed were confined for the most part to a single cover type and were attributed to factors other than atmospheric inputs. These and other data collected from the same sites indicate that at least for the elements evaluated, no statistically significant effect on litterfall S and N due to local emissions could be detected.