SSSAJ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 2 June 2005
Published in Soil Sci Soc Am J 69:1185-1195 (2005)
DOI: 10.2136/sssaj2002.0343
© 2005 Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schilling, E. B.
Right arrow Articles by Lockaby, B. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Schilling, E. B.
Right arrow Articles by Lockaby, B. G.
Agricola
Right arrow Articles by Schilling, E. B.
Right arrow Articles by Lockaby, B. G.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Nutrient Cycling
Right arrow Wetland Soils

Wetland Soils

Microsite Influences on Productivity and Nutrient Circulation Within Two Southeastern Floodplain Forests

Erik B. Schilling* and B. Graeme Lockaby

School of Forestry and Wildlife Sciences, Auburn Univ., Auburn, AL 36849-5418

* Corresponding author (schileb{at}auburn.edu)


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
At broad spatial scales, floodplain forests of the southeastern USA are often perceived as either nutrient poor or nutrient rich. Perceived differences in nutrient availabilities between oligotrophic blackwater and eutrophic redwater floodplain forests suggest rates of aboveground net primary production (ANPP) are lower on the former. However, for these floodplain types, microsite variation may influence relationships between nutrient circulation and ANPP to a greater extent than landscape position and geomorphology of their associated watershed. Therefore, our objectives were to compare ANPP, nutrient circulation through litterfall and decomposition, and the microbial biomass among microsites that differed in terms of soils and hydrology within two riverine floodplain forests. The two floodplains used in this study were located along the Satilla River, a Coastal Plain blackwater system, and the Altamaha River, a Piedmont redwater system, both of which are located in southeastern Georgia. Microsite influences on ANPP were not significantly different during 1999 and 2000 for the Satilla (SAT) and Altamaha (ALT) floodplains. Litterfall production and patterns of P and Ca circulation on the SAT floodplain were significantly lower for the driest microsite type during both years. Microbial biomass N and P were also lowest for the driest microsite, especially in 2000. Overall, microsites on the SAT displayed greater variation in P and Ca cycling than did those of the ALT. Litterfall production, nutrient circulation in litterfall and decomposition, and the microbial biomass were similar among microsites of the ALT floodplain during both years. These findings suggest that spatial variability in nutrient cycling may be greater for oligotrophic floodplain forests.

Abbreviations: ANPP, aboveground net primary production • ALT, Altamaha floodplain • dbh, diameter at breast height • INT, intermediate flooding • SAT, Satilla floodplain


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
IN THE BROADEST OF TERMS, riverine forests may be perceived as oligotrophic (i.e., nutrient poor) or eutrophic (i.e., nutrient rich) based on differences in landscape position, mineralogy, and hydrology of the watershed and it's associated river system (Sharitz and Mitsch, 1993). Watersheds of redwater river systems result in high nutrient loads derived from weathering and leaching of soils and parent materials of the Piedmont Physiographic Region (Kellison et al., 1998). The floodplain forests surrounding these eutrophic rivers have been suggested to be highly productive given high suspended-sediment and nutrient loads associated with their river waters (Wharton et al., 1982). In comparison, low-gradient blackwater rivers arise in the Coastal Plain Physiographic Region where rainfall represents most of the water input to the watersheds (Kellison et al., 1998). Their associated floodplain forests may be less productive given lower concentrations of suspended sediments and inorganics in river waters (Wharton et al., 1982).

While it is evident that differences in nutrition exist within riverine forests of the southeastern USA (Lockaby and Conner, 1999), it remains unclear if variation in soils and hydrology within a floodplain forest influences relationships between nutrient circulation and ANPP to a greater extent than landscape position of the watershed. Therefore, insight into the variation between these two riverine forest types may be gained through scrutiny of microsite ANPP and nutrient circulation. Within floodplain forests there exists an assortment of microtopographies that are defined by differences in hydroperiod and soil type, which often correspond to distinct plant communities (Wharton, 1978). Slight differences in microtopography may be sufficient to induce major changes in soil abiotic factors and hydrology and, consequently, plant species composition, productivity, and nutrient cycling processes (Kellison et al., 1998; Hodges, 1998). Usually ANPP and biogeochemical data reported for particular riverine forests represent composites of the microsite mosaic.

Since production among Southeastern riverine forests varies from 2 to 20 Mg ha–1 yr–1 (Conner, 1994) and our understanding of the processes controlling production rates remains incomplete, investigation of the linkage between biogeochemical processes and ANPP among microsites of these two floodplain forest types is warranted. To examine these relationships, we selected forests along the Satilla and Altamaha Rivers. Having its origins in the Georgia Coastal Plain, the blackwater Satilla River possesses low nutrient and sediment loads. Conversely, the high-gradient Altamaha River originates in the Piedmont producing higher nutrient and sediment loads. Therefore, the two floodplains represent interesting biogeochemical contrasts. Our objectives were to: (i) compare ANPP within each floodplain among microsites that differed in terms of hydrology and soils; and (ii) examine the degree to which nutrient circulation in litterfall, decomposition, and microbial communities among microsites of each floodplain might relate to rates of ANPP. It was hypothesized that microsite variation on both floodplains would result in the intermediately flooded microsites having the highest ANPP, largest amounts of nutrient circulating in litterfall, fastest rate of decomposition, and highest levels of nutrients stored in the microbial biomass.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Satilla and Altamaha Rivers
The Satilla River watershed lies entirely within the Coastal Plain, thus this blackwater river possesses low sediment and nutrient loads (Table 1). Satilla waters possess high dissolved organic C and very low pH (Beck et al., 1974; Cai and Wang, 1998). The Altamaha River watershed arises from the Piedmont and, therefore, has both higher sediment loads and pH (Beck et al., 1974; Cai and Wang, 1998). Altamaha waters carry significant amounts of nutrients, but possess lower organic C, in comparison with those levels of the Satilla River (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Watershed area, discharge, pH, suspended sediment, and nutrient loads for the Satilla and Altamaha Rivers, Georgia.{dagger}

 
Satilla River Study Site
The SAT site is located in Brantley Co., GA (31°13'32'' N lat., 81°52'53'' W long.). The SAT site was partitioned into nine circular plots (10 m radius) according to three wetness site types, or microsites, based on differences in soils (i.e., series, drainage class, depth to redoximorphic features) and vegetation data collected on site before initiation of the study. Satilla soils of the DRY and intermediate flooding (INT) microsites were classified as Kinston series (fine-loamy, siliceous, semiactive, acid, thermic Fluvaquentic Endoaquepts). Kinston series soils are classified as poorly drained. Depth to 10 YR 6/2 redoximorphic features, which are related to the physical and chemical properties of the soil, were recorded in the field at depths of >30 cm for DRY microsite soils and at <15-cm depths for INT microsite soils. The WET microsite soils were classified as Humaqueptic Fluvaquents, which are similar to Kinston series, but possess higher organic matter throughout the profile to a 1-m depth. Within the WET microsite soils, 10 YR 5/8 redoximorphic features were observed at depths of <10 cm.

The forest on the SAT site consisted of sweetbay (Magnolia virginiana L.), black tupelo (Nyssa sylvatica Marshall var. sylvatica), red maple (Acer rubrum L.), laurel oak (Quercus laurifolia Michx.), and sweetgum (Liquidambar styraciflua L.). These species were present in the overstory of all three microsite types. Pondcypress [Taxodium distichum var. nutans (Ait.) Sweet] was uncommon, but occurred within both the INT and WET microsites. Midstory species found across all microsites included titi (Cyrilla racemiflora L.) and buckwheat-tree [Cliftonia monophylla (Lam.) Britton ex Sarg.], whereas the understory and forest floor consisted of Elliott blueberry (Vaccinium elliottii Chapm.) and blue palm [Sabal minor (Jacquin) Pers.].

The percentage of land area within each of the three-microsite categories was estimated from aerial photographs (Paine, 1981). On the SAT site, the INT microsite type covered approximately 64% of the study area followed by the DRY (24%) and WET (12%) microsite types. Across SAT microsites, increment cores taken from laurel oak indicated a stand age for the dominant canopy to be approximately 70 yr. Basal area there averaged 7.7 m2 ha–1. The lacking of a cohort of canopy species in the mid- and understory vegetation suggests an even-aged stand (Nyland, 2002).

Altamaha River Study Site
The ALT site is located within the Griffin Ridge Wildlife Management Area, Georgia Department of Natural Resources, Long Co., GA (31°40'47'' N lat. 81°49'25'' W long.). Nine circular plots (10 m radius) were also assigned according to three wetness site types based on differences in soils and vegetation. Altamaha soils of the DRY microsites were classified as Ocilla series (occasionally flooded, Aquic Arenic Paleudults), which are somewhat poorly drained soils. In the field, 7.5 YR 5/8 redoximorphic features were observed at depths of >30 cm. The INT microsite soils were classified as Osier series (Siliceous, thermic Typic Psammaquents). The poorly drained soils of the INT microsites exhibited redoximorphic features (7.5 YR 5/8) at soil depths >10 cm. Soils of the WET microsites were classified as Bibb series (coarse-loamy, siliceous, active, acid, themic Typic Fluvaquents) with 7.5 YR 5/8 redoximorphic features at soil depths < 10 cm. Bibb series soils of the WET microsites are poorly drained.

Overstory species found within all three ALT floodplain microsites were laurel oak, red maple, and sweetgum. For the INT and WET microsites overcup oak (Quercus lyrata Walt.), ogeechee tupelo (Nyssa ogeche Bartr. ex Marsh.), and water tupelo (Nyssa aquatica L.) were also present. Baldcypress (Taxodium distichum [L.] Rich.) and pondcypress were found only in WET microsites. A woody and herbaceous understory was almost nonexistent on the ALT floodplain; however, on the DRY microsites, blue palm was often present.

Again, using aerial photography, land area within the WET microsite type covered 57% of the ALT study site area, followed by the INT (22%) and DRY (21%) microsite types. Laurel oak increment cores, taken across ALT microsites, indicated the dominant canopy was approximately 75 yr old. Diameter class distributions on the ALT showed a normal distribution, which is typical of even-aged stands (Nyland, 2002). Basal area for ALT averaged 25.1 m2 ha–1.

Hydrology and Flooding
United States Geological Survey discharge data were used to reconstruct flooding regimes for the SAT and ALT floodplain microsites during this study (USGS, 2001). Both the Satilla River (#02228000) and Altamaha River (#02226000) gauging stations were within 0.5 km of the study sites. Stage levels and field observations were matched to discharge levels and used to extrapolate flooding regimes for microsites of both floodplains. Flood events across SAT floodplain microsites were sporadic, reflecting the precipitation-dominated nature of this Coastal Plain blackwater river system (Fig. 1a) . In 1999, floodwaters inundated the WET and INT microsites of the SAT floodplain for 13 and 5 d, respectively. No flooding events were observed for these microsites in 2000. For the DRY microsites, no flooding events occurred during the study period. Flood events on the ALT floodplain were common (Fig. 1b), particularly for WET microsites, which remained flooded for 40 d in 1999 and 37 d in 2000. For 1999 and 2000, the INT microsites were inundated for 21 and 17 d, respectively, while the DRY microsites remained inundated for 10 and 3 d, respectively.



View larger version (36K):
[in this window]
[in a new window]
  		Fig. 1. Satilla River (A) and Altamaha River (B) discharge rates (m3 s–1) from 1990–2000. Flood events occurred within microsites of the Satilla (DRY = dotted line; INT = dashed line; WET = solid line) and Altamaha (DRY = dotted line; INT = dashed line; and WET = solid line) River floodplains when discharge rates of both rivers exceeded each of the aforementioned lines.

 
Soil Physiochemical Properties
From each plot, three soil cores were sampled every other month from April to October 1999. Cores were taken from the upper 0 to 10 cm of mineral soil and analyzed for organic matter, total C, total N, extractable P, and Ca, exchangeable K, and Mg. Total C, N, and organic matter were determined by combustion (Nelson and Sommers, 1996). Extractable P and exchangeable K, Ca, and Mg were determined using inductively coupled argon plasma (ICAP) analysis (Jarrell-Ash 9000, Genesis Laboratory Systems Inc., Grand Junction, CO) after extraction with a double-acid (0.05 M HCl and 0.0125 M H2SO4) reagent (Hue and Evans, 1986). Soil bulk density was measured in April 1999 using a 5-cm diameter double-cylinder push probe. Five cores (0–10 cm depth) were taken from each plot and soils were dried to a constant mass at 105°C and weighed (Blake and Hartge, 1986).

Aboveground Net Primary Production Indices and Litterfall Nutrients
Litterfall was collected monthly from three randomly located 0.25-m2 nylon-screen traps per plot between November 1998 and November 2000. Traps were placed on the forest floor and constructed on a floatable styrofoam base that would rise during flood events. The nylon screen was raised 5 cm above the base to prevent litter from being inundated and the traps were attached with string to nearby trees. Litterfall (i.e., leaf and reproductive components) was oven-dried at 70°C for 48 h, weighed, and ground to pass a 850-µm (20-mesh) sieve. Total C and N were determined using a PerkinElmer Series II CHNS/O Analyzer 2400 (PerkinElmer Corp., Norwalk, CT). Total P was determined using the vanadomolybdate procedure on an HCl extract following dry ashing at 500°C for 4 h (Jackson, 1958). Calcium, Mg, and K samples were dry ashed at 500°C for 4 h, taken up in a double-acid extract (1 M HCl and 1 M HNO3), and determined using ICAP analysis (Hue and Evans, 1986).

For tree species on both floodplains, stem production was determined from annual changes in wood biomass calculated using allometric equations based on diameter at breast height (dbh, approximately 1.3 m) and total height (Clark et al., 1985). For Taxodium spp., annual change in stem biomass was calculated using allometric equations based on dbh alone (Scott et al., 1985). For large buttressed trees, stem diameter was measured 30 cm above the butt swell. Within each plot, dbh and height of all stems ≥10 cm was measured. It was assumed that the contribution of wood production from stems <10 cm dbh and shrubs was a small fraction of ANPP because the growth of saplings and shrubs are typically low and dominated by leaf production (Whittaker et al., 1974), which were included in the litterfall production estimates. Woody debris collected in the litterfall traps was not included in ANPP estimates since all wood production was accounted for by the allometric equations that were based on whole-tree wood biomass measurements. Total ANPP was calculated as the sum of litterfall and stem production.

Leaf Litter Decomposition
Leaf litter from the four major tree species of each floodplain was composited to form weighted average mixtures for inclusion in litterbags. Species included in SAT litterbags were laurel oak, red maple, sweetbay, and black tupelo, while ALT litterbags contained laurel oak, red maple, overcup oak, and sweetgum leaves. The mixed species litterbags (30.5 cm x 45.7 cm with 6-mm and 2-mm openings on the upper and lower sides, respectively) for both floodplains contained approximately 20 g of air-dried leaf litter. Eighty-one litterbags were installed within each floodplain (i.e., nine bags per plot). To examine the influence of litter quality on decomposition, an additional 27 litterbags containing mixed litter from the opposing floodplain were placed within the INT microsites of the other floodplain (i.e., traded litter). One litterbag was collected from each plot per collection period, and collections occurred at 0, 14, 28, 42, 75, 125, 275, 400, and 525 d. Litter was removed from the bags on collection, oven dried at 70°C for 48 h, weighed, and ground to pass a 850-µm (20-mesh) sieve. Analyses for C, N, P, Ca, K, and Mg were the same as those previously mentioned for determining litterfall nutrients. Procedures from Van Soest and Wine (1986) were used to determine the percentage of lignin for litter collected at 525 d. All leaf litter weights were expressed on an ash-free basis.

Microbial Biomass Carbon, Nitrogen, and Phosphorus
Microbial biomass C and N was determined using chloroform fumigation-extraction procedures similar to those outlined by Vance et al. (1987) and Brookes et al. (1985), respectively. In 1999 and 2000, samples were collected in April, May (1999 only), June, August, and October. Microbial biomass P was determined using chloroform fumigation-extraction procedures as outlined by Brookes et al. (1982). In 1999 and 2000, samples were taken in April (2000 only), May (1999 only), June, August, and October. Two soil cores (10-cm depth) were taken from within each plot, at each sample date. Soil cores were sealed in plastic bags, placed on ice, and immediately returned to the lab. Microbial determinations were initiated <24 h after cores were removed from the field.

Statistical Analyses
All data were analyzed using SAS (SAS Institute, 1991). Statistical analyses were conducted separately for each floodplain using a completely randomized design. For both floodplains, three replicates from each wetness category, or microsite type, were used as treatments with the replicates being assigned at random. Statistical analyses for soil variables, ANPP indices, litterfall nutrient content, and the microbial biomass consisted of one-way analyses of variance. The General Linear Model procedure (GLM) was used to test the effect of microsite type on all parameters with mean separation procedures accomplished using Tukey's Honestly Significantly Different. Paired t tests within each floodplain were also used to test for annual variation in ANPP indices for each microsite category.

Nonlinear regression analyses were used to calculate leaf litter decay coefficients (k) for each microsite (e–kt) (Olsen, 1963). Statistical analyses for leaf litter decomposition consisted of analysis of variance comparisons among microsites using the GLM procedure on the percentage of mass and nutrients remaining, litter nutrient ratios at 525 d, and decay coefficients with mean separations done using Tukey's HSD. Mean differences between ALT and SAT floodplain traded litter for percent mass, nutrients remaining, nutrient ratios, and decay coefficients were accomplished using paired t tests. All mean comparisons were reported as significant at the 0.05 probability level.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Microsite Influences on Soil Physiochemical Properties
Soil bulk density on the SAT floodplain was lower (P = 0.0002) for WET microsites compared with the INT and DRY microsites (Table 2). This pattern reflects the significantly greater (P = 0.05) soil organic matter within WET microsite soils of the SAT floodplain. Significantly greater soil C (P = 0.05), N (P = 0.05), extractable P (P = 0.0001), exchangeable Ca (P = 0.0001), and Mg (P = 0.0001) were observed within WET microsites compared with DRY microsites (Table 2). Soil exchangeable K within WET microsites exceeded (P = 0.0001) both DRY and INT microsites.


View this table:
[in this window]
[in a new window]
  		Table 2. Soil bulk density, organic matter, and nutrient concentrations for microsites of the Satilla and Altamaha River floodplains, Georgia to a depth of 10 cm. Standard errors of the means in parentheses.

 
On the ALT floodplain, soil bulk density differed significantly (P = 0.0007) among the three microsite types (Table 2). Soil organic matter was significantly greater (P = 0.05) within INT microsites compared with that of DRY microsites. Total soil C was greatest (P = 0.01) for INT microsites of the ALT floodplain, while total soil N did not differ significantly among microsites (Table 2). Greater extractable P (P = 0.0005) and exchangeable K (P = 0.0001) were observed for INT microsites of the ALT floodplain. Exchangeable Mg also differed significantly (P = 0.0001) among all three ALT microsites (Table 2).

Microsite Influences on ANPP Indices
On the SAT floodplain, lower litterfall production on DRY microsites in 1999 (4.4 Mg ha–1 yr–1, P = 0.03) and 2000 (4.7 Mg ha–1 yr–1, P = 0.04) was observed. The INT and WET microsites did not differ in rates of litterfall production during either year (Table 3).


View this table:
[in this window]
[in a new window]
  		Table 3. Litterfall and stem production, total aboveground net primary production (ANPP), nutrient return in litterfall, and litterfall N/P ratios for microsites of the Satilla and Altamaha River floodplains, Georgia. Standard errors of the means are in parentheses.

 
Microsite did not influence stem production on the SAT in 1999 and 2000 and ranged from 1.9 to 6.1 Mg ha–1 yr–1 (Table 3). Total ANPP ranged from 6.8 to 11.7 Mg ha–1 yr–1 and did not differ among SAT microsites during both years.

Across both years, microsite type did not influence litterfall production on the ALT floodplain, which ranged from 4.9 to 6.4 Mg ha–1 yr–1 (Table 3). Stem production (2.4–8.0 Mg ha–1 yr–1) and total ANPP (7.9–14.4 Mg ha–1 yr–1) also did not differ either year among ALT floodplain microsites (Table 3).

Annual Variation in Microsite ANPP Indices
Differences in annual litterfall production rates among SAT floodplain microsites for 1999 and 2000 were not significant. By contrast, stem production rates declined significantly in 2000 for both the INT (P = 0.02, mean decline of 3.2 Mg ha–1 yr–1) and WET (P = 0.04, mean decline of 4.2 Mg ha–1 yr–1) microsites. The decline in stem production on DRY microsites in 2000 was not significant. Significantly lower total ANPP within INT (P = 0.03, mean decrease of 3.1 Mg ha–1 yr–1) and DRY (P = 0.05, mean decrease of 2.1 Mg ha–1 yr–1) microsites in 2000 was observed.

For ALT floodplain microsites, annual litterfall production rates for 1999 and 2000 were similar for the DRY and WET microsites, but declined significantly for INT microsites in 2000 (P = 0.05, mean decrease of 1.5 Mg ha–1 yr–1). Stem production declined significantly for all ALT microsites (DRY, P = 0.03; INT, P = 0.02; WET, P = 0.04) in 2000. This resulted in significantly lower total ANPP for the DRY (P = 0.02), INT (P = 0.02), and WET (P = 0.03) microsites of the ALT floodplain during that year.

Microsite Influences on Litterfall Nutrient Return
Annual SAT litterfall N returned for DRY microsites was lower (P = 0.04) than WET microsites in 1999. However, similar rates of N returned in litterfall were observed among SAT microsites in 2000 (Table 3). In 1999, litterfall P return for DRY microsites was significantly lower (P = 0.03) than that of INT and WET microsites. The DRY microsites also exhibited lower litterfall P in 2000 (Table 3). Litterfall N/P ratios for the SAT microsites ranged from 8.7 to 10.5 and did not differ either year. Annual litterfall Ca returned for INT and WET microsites significantly exceeded that of DRY microsites for 1999 (P = 0.009) and 2000 (P = 0.003, Table 3). Litterfall K and Mg returns for the SAT floodplain averaged 8.9 and 13.5 kg ha–1 yr–1, respectively for both years. In 2000, litterfall K returned within DRY microsites was lower (P = 0.05) than levels within INT and WET microsites (Table 3). Litterfall Mg return within the DRY microsites was lower (P = 0.04) than WET microsites in 1999, but not in 2000 (Table 3).

On the ALT floodplain, microsite did not significantly influence litterfall N return, which averaged 36.4 kg ha–1 yr–1 during 1999 and 2000 (Table 3). Amounts of P returned in litterfall on the ALT microsites were also similar in 1999. However, in 2000, litterfall P returned on WET microsites was greater (P = 0.02) than that of DRY and INT microsites (Table 3). Litterfall N/P ratios did not differ significantly among ALT microsites in 1999 but, in 2000, litterfall N/P ratios for the WET microsites were lower (P = 0.01) than DRY and INT microsites. Annual litterfall Ca and Mg returns averaged 63.9 and 10.3 kg ha–1 yr–1, respectively and did not differ among ALT microsites during either year (Table 3). In 2000, annual litterfall K returned for WET microsites was greater (P = 0.03) than that of DRY and INT microsites.

Microsite Influences on Decomposition
For the SAT floodplain, the percentage of mass remaining after 525 d on INT microsites was significantly greater (P = 0.0008) compared with that of DRY and WET microsites (Table 4). Decay rates for both the WET and DRY microsites were faster (P = 0.002) than those of INT microsites (Table 4). However, microsite did not influence the percentage of C, N, and Mg remaining on the SAT floodplain at 525 d. In contrast, WET litter possessed significantly greater P (P = 0.006) and Ca (P = 0.009) remaining than DRY microsites. Decomposing WET microsite litter contained significantly more K (P = 0.009) than INT microsites at 525 d (Table 4). While litter C/N, N/P, lignin/N, and lignin/P ratios were similar among SAT microsites, litter C/P ratios were greater (P = 0.04) for DRY compared with INT microsites (Table 4).


View this table:
[in this window]
[in a new window]
  		Table 4. Percentage of mass, C, N, P, Ca, K, and Mg remaining, C/N, N/P, C/P, lignin/N, and lignin/P ratios, and decay coefficients (k) for litter within microsites of the Satilla and Altamaha River floodplains, Georgia after 525 d. Standard errors of the means are in parentheses.

 
In contrast to the patterns observed on the SAT floodplain, no differences in the percentage of mass remaining at 525 d and rates of decay were observed among ALT floodplain microsites (Table 4). Microsite also did not influence ALT floodplain percentages of nutrients remaining at 525 d. Additionally, no differences in litter C/N, N/P, C/P, lignin/N, and lignin/P ratios were observed.

The ALT traded litter possessed lower percentage of mass remaining (P = 0.006) and decomposed at a faster rate (P = 0.01) than the SAT traded litter suggesting lower litter quality on the latter floodplain (Table 4). Comparisons of the percentages of N remaining between ALT and SAT traded litter showed no statistical separation. However, the amount of P remaining for SAT traded litter was significantly greater (P = 0.0008). No differences in percentages of Ca, K, and Mg remaining for the ALT and SAT traded litter were observed at 525 d (Table 4). Litter N/P and C/P ratios were also similar for the two litter types, while litter C/N ratios were significantly lower (P = 0.03) for the ALT traded litter.

Microsite Influences on Microbial Biomass Carbon, Nitrogen, and Phosphorus
In 1999, microbial biomass C within WET microsites was significantly greater (P = 0.002) than the INT and DRY microsites (Table 5). Microbial biomass C did not differ among SAT microsites in 2000. Within the WET microsites, microbial biomass N was greater (P = 0.0001) than both the DRY and INT microsites in 1999. In 2000, microbial biomass N differed significantly (P = 0.0001) among all three SAT microsites (Table 5). Similar levels of microbial biomass P were observed among SAT microsites in 1999, although, WET microsites had significantly greater (P = 0.03) microbial P than DRY microsites in 2000.


View this table:
[in this window]
[in a new window]
  		Table 5. Microbial biomass C, N, and P for microsites of the Satilla and Altamaha River floodplains, Georgia for 1999 and 2000. Standard errors of the means are in parentheses.

 
For the ALT floodplain, significantly greater (P = 0.002) microbial biomass C was observed within INT microsites compared with DRY microsites in 1999 and 2000 (Table 5). Microbial biomass N within INT microsites in 1999 significantly exceeded (P = 0.0001) levels within the DRY and WET microsites (Table 5). No differences in microbial biomass N were observed among ALT microsites in 2000. Microbial biomass P also did not differ among ALT microsites during both years (Table 5).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Microsite Influences on Soil Nutrients
Percentage of soil organic matter and soil C increased moving from drier to wetter microsites on the SAT floodplain. This was also the case for the ALT floodplain where the INT microsites possessed higher soil organic matter and soil C in the mineral soil. However, a considerable organic layer (approximately 10 cm) atop the mineral soil horizons of WET microsites was present, but was not included in soil analyses. The presence of this organic horizon above the mineral soil of the WET microsites of the ALT floodplain reflects the effects of prolonged soil saturation on soil organic matter and C formation (Megonigal and Day, 1988).

High organic matter typically results in higher soil N (Patrick, 1981) while increasing soil saturation generally increases P availability (Mitsch and Gosselink, 2000). These trends were clear for SAT microsites, but were not evident for ALT microsites, suggesting that differences in soil saturation there were not sufficiently divergent to influence N and P availability on that floodplain (Table 2). For both floodplains, higher soil base cation availability was also associated with increased soil saturation (Table 2). Low soil O2 levels may result in soils being highly reduced, thereby increasing nutrient mobilization and availability (Mitsch and Gosselink, 2000).

Soil base cations on the SAT floodplain increased in a stepwise, and significant, fashion moving from drier to wetter microsites. For the ALT microsites, increasing soil wetness generally increased soil base cation availability. However, on the ALT floodplain, exchangeable Ca, K, and Mg were greatest for INT microsites. Patterns of increased base cation availability with increasing soil wetness and flooding on both floodplains follow patterns observed for redwater (Patrick, 1981) and blackwater (Wharton et al., 1982) floodplain forests.

Microsite Influences on ANPP Indices
Litterfall production varied significantly among SAT microsites during both years (Table 3) with DRY microsites having significantly lower litterfall than INT microsites in 1999, and both the INT and WET microsites in 2000. Decreased litterfall production in floodplain forests often results from prolonged periods of soil saturation (Conner and Day, 1992). However, for the DRY microsites of the SAT floodplain it appears that decreased litterfall production may be more reflective of decreased soil moisture. It also appears that the differences in flood duration for the SAT floodplain INT and WET microsites were not great enough to influence production (Fig. 1a).

Among ALT microsites, litterfall production did not differ significantly during either year. Furthermore, litterfall production within the DRY and WET microsites remained similar numerically across both years (Table 3). Decreased litterfall production within INT microsites from 1999 to 2000 corresponded to a decline in the reproductive component of litterfall (1.17 vs. 0.37 Mg ha–1 yr–1, respectively). Annual variation in the reproductive component (i.e., flowers and fruit) of litterfall often increases variability in litterfall estimates (Peterson and Rolfe, 1982).

Stem production did not differ significantly among microsites of either floodplain or year, further suggesting that differences in flooding and soil saturation were not large enough to influence production rates (Table 3). While rates of stem production were similar among microsites of both floodplains, rates of stem production did significantly decline for all microsites in 2000. For Southeastern floodplain forests, fluctuating annual stem production rates among floodplain topographic positions is often observed (Megonigal et al., 1997). The variation in stem production for all microsites may suggest that the two-floodplain forests allocated C between roots and stems differently between years. Additionally, decreased stem production in 2000 may also reflect changes in temperature and precipitation during the growing season that may have created more stressful growing conditions with little change to patterns of C allocation.

Total ANPP was similar among ALT microsites during both years due to similar rates of litterfall and stem production. For the SAT microsites, differential litterfall production was not large enough to significantly influence total ANPP rates given similar rates of stem production. These findings indicate that rates of ANPP for both floodplains were not influenced by microsite type and generally reflect those of Megonigal et al. (1997) who noted that periodic flooding and flowing water did not increase productivity compared with non-flooded, or upland sites.

Microsite Influences on Decomposition
Decomposition rates often differ among floodplain forest microsites due to differences in hydrology (Battle and Golladay, 2001). This was the case for the SAT floodplain where decomposition within DRY and WET microsites had significantly lower percentage of mass remaining and faster decay rates compared with INT microsites. However, for the ALT floodplain, this was not the case since decomposition parameters were similar among microsites there.

Reasons for the lack of differences in mass remaining and decay rates between the DRY and WET microsites of the SAT floodplain and the microsites of the ALT floodplain are unclear. Litter quality plays a strong role in regulating decomposition, particularly if microsites are large enough to alter species composition (Lockaby and Walbridge, 1998). Furthermore, Day (1982) compared decomposition rates among four communities within the Great Dismal Swamp, VA that differed substantially in terms of species and hydrology and noted that differences in litter quality influenced loss rates to a greater degree than the direct effects of flooding. For this study, similar microsite species compositions for both floodplains suggest that the lack of differences in the percentage of mass remaining and rate of decay might be a reflection of similar microsite litter qualities.

Leaf litter quality influences on decomposition become clearer when examining the traded floodplain litter. The lower percentage of mass remaining and faster decay rates for ALT traded litter compared with SAT traded litter in addition to lower C/N, lignin/N, and lignin/P ratios suggests better quality litter on the ALT floodplain (Table 4). This was particularly evident on the SAT floodplain where INT microsites had significantly slower decomposition rates compared to non-traded litter.

Microsite Influences on Nitrogen Circulation
Microsite influences on N circulation within the SAT floodplain showed numerical increases moving from drier to wetter microsites across both years (Table 3). Differences in rates of N returned through litterfall among SAT microsites in 1999 appeared to be primarily driven by differences in biomass production given similar (P = 0.08) litterfall N concentrations for the DRY (9100 mg kg–1) and WET (8100 mg kg–1) microsites. In contrast to patterns of N circulation observed for the SAT microsites, annual litterfall N returned among ALT microsites generally increased as microsites became drier. However, litterfall N concentrations did not differ significantly among ALT microsites averaging 8647 mg kg–1.

Within forest ecosystems, leaf litter with C/N ratios above 50 tends to exhibit N immobilization during decomposition (Paul and Clark, 1989). Furthermore, decomposing leaf litter possessing C/N values below this threshold suggests N mineralization. Baker et al. (2001) reported that initial mixed species litter C/N ratios of 52 and 57 were initially N limited, while N was the least limiting with a narrower litter C/N ratio of 46. Therefore, the initial leaf litter mixtures for microsites of the SAT (C/N = 70) and ALT (C/N = 72) floodplain appear to have been N limited. Even though litter C/N ratios for the SAT and ALT floodplain microsites narrowed considerably after 525 d, it appears that N was still slightly limiting decomposition on the INT microsites of both floodplains and to a lesser extent, the DRY microsites of the SAT floodplain (Table 4). Additionally, lignin/N ratios for both SAT and ALT floodplain microsites exceed those reported for the Ogeechee River floodplain (lignin/N = 20), where vegetation production has been theorized to be primarily N limited (Lockaby et al., 1996).

For the ALT floodplain, microbial biomass C and N values for INT microsites were significantly larger than those of the DRY and WET microsites in 1999, and the DRY microsites in 2000 (Table 5). Increased soil C levels within INT microsites may have resulted in the observed increased microbial biomass given that hydrology and litter quality are key factors in the regulation of microbial pools (Groffman et al., 1996). Higher microbial biomass N on INT microsites suggests increased N immobilization there. Differences in microbial N dynamics among SAT microsites for 1999 and 2000, suggest that the microbial communities on WET microsites had access to greater quantities of N, which may be a reflection of the higher total N in soils.

Microsite Influences on Phosphorus Circulation
On both floodplains, the wetter microsites tended to exhibit the highest amounts of litterfall P (Table 3). However, on the SAT floodplain, P levels on the least wet (DRY) microsite were significantly lower than those of the other microsites. In contrast, the only significant difference among ALT microsites was that of the WET microsite during 2000. The general tendency for higher litterfall P to be associated with wetter microsites may reflect the low frequency of flooding on both floodplains during the study period. It is interesting to note that the levels of P returned in litterfall of both floodplains were high compared with those of other blackwater and redwater floodplain forests. For example, SAT litterfall P returned is higher than that reported by Post and de la Cruz (1977) for a blackwater stream in the coastal plain of Mississippi (0.9 kg ha–1 yr–1), and ALT microsite litterfall P return is nearly twice that reported by Clawson et al. (2001) for the redwater Flint River floodplain in central Georgia (4.0–4.4 kg ha–1 yr–1).

Decomposing litter within microsites of both floodplains showed strong trends of P immobilization at 525 d (Table 4). However, P immobilization increased in litter on the SAT floodplain moving from drier to wetter microsites. This trend was not observed for the ALT floodplain since rates of P immobilization were similar for the DRY and WET microsites.

Temporal trends for P remaining suggest some divergence between SAT microsites in terms of immobilization tendencies. After 125 d, INT and WET microsite litter had higher proportions of original P remaining (140 and 135%, respectively) compared with that of DRY microsites (120%). These two wetter microsite types continued to accrue P steadily up to 525 d while litter of the least wet DRY microsite did not. This may be a reflection of the significantly lower soil extractable P found within DRY microsites. While reasons for the litter on that microsite to display differential P dynamics cannot be specified, there may be some tendency for soil P availability to be reduced under drier conditions (Wright et al., 2001).

Microsite influences on microbial biomass P were small for the ALT floodplain. By contrast, P assimilated into the microbial biomass of the SAT floodplain tended to increase from drier to wetter microsites, a trend particularly evident in 2000 (Table 5). As was the case with leaf litter P dynamics, this suggests that microbial populations on the wetter SAT microsites either had access to greater quantities of P or displayed preferential accumulation of a deficient nutrient. Increased P availability within saturated soils may be the result of increased P solubility under flooded conditions (Lockaby and Walbridge, 1998) and also due to decreases in vegetation demand for P under anaerobic soil conditions (Mitsch and Gosselink, 2000).

Microsite Influences on Calcium, Potassium, and Magnesium Circulation
For the SAT floodplain, litterfall Ca return increased moving from drier to wetter microsites. Additionally, SAT floodplain microsites displayed more variation in Ca dynamics in decomposition, with the drier DRY and INT microsites mineralizing Ca, and the wetter WET microsites immobilizing Ca. Increased Ca mineralization in the decomposing litter within the drier DRY microsites may be a response to lower soil Ca availability and decreased Ca inputs from floodwaters resulting in lower Ca circulating in litterfall. Immobilization of Ca in the vegetation woody biomass and/or the microbial biomass may have also limited Ca availability. Furthermore, the comparative levels of soil K and Mg relative to that of Ca (Table 2) may have interfered with Ca uptake by vegetation within the DRY microsites (McLaughlin and Wimmer, 1999). The accrual of Ca within WET microsites on the SAT floodplain reflects the higher soil organic matter present since organic matter is one of the primary controllers of Ca availability in soils (Ulrich and Matzner, 1986).

Litterfall Ca content for the ALT floodplain was similar among microsites, particularly in 2000. Decomposing litter on the ALT floodplain strongly accrued Ca at 525 d. Opportunities for Ca input via floodwaters (Fig. 1b, Table 1) suggest that Ca exchange between floodwaters and ALT floodplain microsites are high. Therefore, chelation of Ca by organic matter may have been a mechanism resulting in Ca immobilization within ALT floodplain litter (Brinson, 1977) since Ca actively binds with organic matter (McLaughlin and Wimmer, 1999). Additionally, immobilization of Ca in leaf litter may have also resulted from microbial demand.

For the SAT floodplain, microsite influences on K circulation were not as strong as those observed for Ca. In 2000, K returned in litterfall for the drier DRY microsites was lower than the INT and WET microsites. On the ALT floodplain, similar patterns of K circulation in litterfall and decomposing litter were evident among microsites. However, significantly larger amounts of K did circulate in litterfall of WET microsites in 2000. It is noteworthy that K circulation in litterfall for both the ALT and SAT microsites was lower than levels reported for other floodplain forests (16–22 kg ha yr–1) (Brinson et al., 1980; Peterson and Rolfe, 1982; Shure and Gottschalk, 1985). Litter from all three SAT microsites mineralized K, with mineralization rates being fastest within INT microsites. Potassium is rapidly lost from decomposing litter due to leaching, and this was observed on both the SAT and ALT floodplain microsites as evidenced by <50% of the original K remaining in the litter after 75 d. Similar patterns of rapid K loss have been observed for other floodplain forests (Day, 1982; Peterson and Rolfe, 1982).

Patterns of Mg circulation through litterfall and litter decomposition were similar for microsites of both floodplains, although significantly less Mg circulated through litterfall on DRY microsites of the SAT floodplain in 1999. For the SAT floodplain microsites, litterfall Mg contents are in the range (14–18 kg ha–1 yr–1) reported for other riverine forests (Brinson et al., 1980; Peterson and Rolfe, 1982; Shure and Gottschalk, 1985). By comparison, however, the ALT floodplain microsites returned lower amounts of Mg through litterfall compared with the aforementioned studies.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
For both floodplains, microsite type did not influence total ANPP during either year. This suggests that the differences in hydrological regimes among microsites of both floodplains were not strong enough to influence production rates. However, on the SAT floodplain, litterfall production was lowest for the driest DRY microsites during both years, suggesting that foliar biomass production there may have been limited by moisture and/or nutrient availability. Although rates of litterfall production remained similar, changing levels of annual stem production may suggest that vegetation of both floodplains shifted resources between above and belowground components equally.

Microsite influences on N circulation in litterfall and decomposition were minor on the SAT, and not evident on the ALT floodplain. Litterfall N/P and leaf litter C/N and lignin/N ratios for ALT floodplain microsites suggest that increased N availability may stimulate ANPP there. These same indices also appear to indicate N availability limits production for the SAT floodplain microsites, albeit to a lesser extent.

Large litter C/P and lignin/P ratios found for the SAT microsites suggests that P availability may limit production. Furthermore, the SAT traded litter immobilized P very strongly, also suggestive of P limitation. For the DRY microsites of the SAT floodplain, low amounts of P circulation in litterfall, P immobilization in leaf litter, and the microbial biomass suggest greater P limitation for this microsite type. By contrast, high rates of P returned in litterfall, as well as narrow litterfall and leaf litter N/P and lignin/P ratios for the ALT microsites suggest that P is not limiting production there.

Differences among SAT microsites in terms of base cation circulation, particularly for Ca, were more pronounced than those of N and P. This was particularly evident for DRY microsites on the SAT floodplain, which displayed lower Ca returned in litterfall and more rapid Ca circulation in leaf litter. High Ca returns in litterfall and litter immobilization of Ca within ALT floodplain microsites appear to be in response to greater soil Ca availability, as well as increased opportunities for geochemical exchange of Ca between the ALT floodplain and floodwaters. For both floodplains, circulation of K in litterfall increased moving from drier to wetter microsites and decomposing litter showed strong patterns of K mineralization. Circulation of Mg in litterfall on the SAT floodplain increased moving from the drier to wetter microsite types and leaf litter decomposition indicated Mg mineralization for the three-microsite types. On the ALT floodplain, circulation of Mg in litterfall and decomposition was similar among microsites.

Overall, the INT microsites of both floodplains did not exhibit higher levels of ANPP or show patterns of increased nutrient circulation in litterfall. In fact, ANPP and patterns of nutrient circulation in litterfall were similar among ALT floodplain microsites during both years. On the SAT floodplain levels of productivity and nutrient circulation were often similar for the INT and WET microsites. Given the similar N circulation patterns observed among microsites of both floodplains, the differences in ANPP and biogeochemistry that exist between these two riverine forest systems appear to be reflective of differences in P and Ca circulation. In broad terms, the chief biogeochemical distinction between microsites of both systems appears to be differences in Ca circulation, and to a slightly lesser extent P. Additionally, differential P and Ca circulation through litterfall and litter decomposition among SAT floodplain microsites suggests that spatial variability in nutrient cycling for this oligotrophic floodplain forest was greater than that observed for the ALT floodplain.

Received for publication October 4, 2002.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (2)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Schilling, E. B.
Right arrow Articles by Lockaby, B. G.
Right arrow Search for Related Content
PubMed
Right arrow Articles by Schilling, E. B.
Right arrow Articles by Lockaby, B. G.
Agricola
Right arrow Articles by Schilling, E. B.
Right arrow Articles by Lockaby, B. G.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Nutrient Cycling
Right arrow Wetland Soils

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Journal of Natural Resources
and Life Sciences Education		 Vadose Zone Journal
Journal of Plant Registrations Journal of
Environmental Quality		 The Plant Genome