- Email: [email protected]
Sediment accretion rates and radial growth in natural levee and backswamp riparian forests in southwestern Alabama, USA
Forest Ecology and Management 358 (2015) 272–280
Contents lists available at ScienceDirect
Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco
Sediment accretion rates and radial growth in natural levee and backswamp riparian forests in southwestern Alabama, USA Kathryn R. Kidd ⇑, Carolyn A. Copenheaver, W. Michael Aust Department of Forest Resources and Environmental Conservation, Virginia Tech, Blacksburg, VA 24061, USA
a r t i c l e
i n f o
Article history: Received 20 July 2015 Received in revised form 15 September 2015 Accepted 17 September 2015 Available online 26 September 2015 Keywords: Forested wetlands Sediment accretion Hydrology Dendrochronology Green ash Water tupelo
a b s t r a c t Riparian forested wetlands improve downstream water quality by trapping suspended sediment from adjacent waterways. Waters that transport sediment and nutrients into adjacent wetlands also create wet and dry hydrologic periods and thus, have the potential to impact site productivity. In this study, we used a dendrogeomorphic technique with green ash (Fraxinus pennsylvanica Marsh.) to estimate sediment accretion for two time periods (1881–2012 and 1987–2012) along a natural levee (35 m from river) and backswamp (75 m from river) and identified the influence of hydrology and climate on radial growth in green ash and water tupelo (Nyssa aquatica L.) along the Tensaw River in southwestern Alabama. We detected significantly higher sediment accretion rates for the 1987–2012 time period along the natural levee (p = 0.00; 1.6 cm yr1) and backswamp (p = 0.03; 1.2 cm yr1) than for the 1881 to 2012 period (0.4 and 0.5 cm yr1). Using previously measured (2010) soil bulk density for the site, estimated mass of sediment trapped per unit area ranged from 55–135 Mg ha1 yr1 for the 1987–2012 period and 17–61 Mg ha1 yr1 for the 1881–2012 period. We identified positive, significant correlations between green ash radial growth and the number of days the backswamp was flooded (1.4 m stage), days the Tensaw River was at bankfull (2.1 m), and average daily river stage during the overall growing season (April to August) and for the month of April. Green ash radial growth also illustrated a positive, significant response to April precipitation totals demonstrating the overall role of moisture availability just prior to the onset of xylem formation. Green ash trees along the natural levee and backswamp were more responsive to hydrology and climate than water tupelo trees located further in the backswamp, illustrating the potential resistance of water tupelo to perturbations. This study illustrates the important role forested wetlands play in improving water quality through quantification of sediment accretion rates and the potential impact that introduced disturbances (i.e., forest harvest-related disturbances) can have on ecosystem services. Ó 2015 Elsevier B.V. All rights reserved.
1. Introduction Bottomland hardwoods, a well-known type of riparian forested wetland, provide many ecosystem services to society. Most notably, riparian forested wetlands improve water quality by trapping and storing sediment from adjacent waterways (Boto and Patrick, 1978; Walbridge, 1993). Sediment generated upstream is transported downstream by rivers and eventually released into the ocean unless it is captured and deposited during transport (Walling, 2006). The amount of sediment captured in a riparian forested wetland can change through time if sediment loads in a
⇑ Corresponding author at: Department of Forest Resources and Environmental Conservation (0324), 228 Cheatham Hall, Virginia Tech, Blacksburg, VA 24061, USA. E-mail addresses: [email protected] (K.R. Kidd), [email protected] (C.A. Copenheaver), [email protected] (W.M. Aust). http://dx.doi.org/10.1016/j.foreco.2015.09.025 0378-1127/Ó 2015 Elsevier B.V. All rights reserved.
waterway are altered due to watershed disturbances (e.g., channelization, land use, urbanization) or if trapping efficiency changes due to reasons that increase or decrease surface roughness (e.g., altered vegetation dynamics, microtopography, woody debris dynamics) (Hupp and Morris, 1990; Meade et al., 1990; Hupp and Bazemore, 1993; Kleiss, 1996; Heimann and Roell, 2000; Li and Yang, 2009; Kroes and Hupp, 2010; Ensign et al., 2014). Bottomland hardwood systems are also valued by society as highly productive sites for timber production, wildlife habitat, carbon storage, nutrient cycling, and protection from floodwaters (Walbridge, 1993). Riparian forested wetlands are characterized by hydrologic cycles of wet and dry periods created through variations in onsite flooding, precipitation, and evapotranspiration rates (Broadfoot and Willston, 1973; Wharton et al., 1982; Hupp, 2000; Day et al., 2007). Disturbances which alter the hydrologic regime in a riparian forested wetland have the potential to impact
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
functionality of forested wetlands and thus, the ability to provide ecosystem services (Conner and Day, 1976; Mitsch and Rust, 1984; Hunter et al., 2008; Kroes and Hupp, 2010). Hydroperiod length, flooding occurrence, and streamflow characteristics influence sediment accumulation in forested wetlands (Kleiss, 1996; Hupp, 2000; Hupp et al., 2008). Riparian forested wetlands reduce the velocity and peaks of floodwaters, which transport suspended sediment into adjacent riparian forested wetlands (Kleiss, 1996). As the velocity of the floodwaters is reduced, sediment falls out of suspension and is deposited, thus creating the microtopography associated with bottomlands (Hodges, 1997). Vertical structures such as vegetation, microtopographical features, and woody debris increase surface roughness, contributing to the decreased velocity of floodwaters, and increasing sediment deposition rates (Hupp, 2000; Li and Yang, 2009; Ensign et al., 2014). Distribution and growth rate of individual tree species varies among microtopographic landforms (e.g., levees vs. backswamps) in bottomland forests due to differences in flood tolerance (Broadfoot and Willston, 1973; Wharton et al., 1982; Hodges, 1997). As a result, forested wetlands along meandering coastal plain rivers in the southeastern United States are characterized by a gradient of less flood-tolerant bottomland hardwood species along elevated natural levee and transitional areas, and by more flood-tolerant species at lower elevations (Wharton et al., 1982; Hodges, 1997; Hupp, 2000; Conner et al., 2014). Changes in hydrologic regime alter spatial and temporal sediment deposition and inflow patterns and, as a consequence, can alter tree growth and productivity in forested wetlands (Conner and Day, 1976; Mitsch and Rust, 1984; Conner et al., 2014). Aboveground productivity in forested palustrine wetlands is dependent on several factors including: silvical characteristics for individual species, hydrologic regime, nutrient availability, edaphic conditions, and previous disturbance history (Broadfoot and Willston, 1973; Conner and Day, 1976; Reily and Johnson, 1982; Wharton et al., 1982). Floodwaters transport suspended sediment and additional nutrients into adjacent forested wetlands and thus, are one of the main hydrologic drivers influencing productivity in forested wetlands (Conner and Day, 1976; Hunter et al., 2008). Sediment deposited from floodwaters provides an allochthonous, external source of nutrients, to forested wetlands (Johnston et al., 1984). These additional nutrients enhance aboveground productivity, allowing individual species to achieve maximum growth (Keeland and Shartiz, 1995). Furthermore, these sediments alter microtopography and species distribution over time by filling abandoned river channels, creating and adding to the occurrence of natural levees (Hodges, 1997). Dendrochronological techniques can be used to estimate sediment accretion rates (Shroder, 1980; Hupp and Morris, 1990) and to assess the impacts of altered hydrology through examination of the variation in annual tree-ring growth patterns (Fritts, 1976). A dendrogeomorphic approach that incorporates total vertical sediment deposition and total age of an immediately adjacent tree can be used to estimate sediment accretion rates following one site visit (Hupp and Morris, 1990; USACE, 1993). For instance, this dendrogeomorphic technique has been used to: characterize spatial and temporal sediment deposition patterns (Hupp and Morris, 1990); quantify the influence physical (e.g., elevation, microtopography, distance from waterway or channel, flood duration) and ecological (e.g., vegetation density, stand structure, species composition) factors have on sediment accretion rates (Hupp and Morris, 1990; Kleiss, 1996; Heimann and Roell, 2000; Phillips, 2001; Hupp et al., 2008); and determine the impact of upstream disturbances (e.g., channelization, dam construction, hydrologic alterations, land use) on short- and long-term sediment deposition and subsidence rates (Hupp and Bazemore, 1993;
273
Heimann and Roell, 2000; Kroes and Hupp, 2010). Analysis of the variation in tree ring widths among calendar years has proven useful in identifying environmental influences (e.g., flooding, streamflow, precipitation, temperature) on annual tree growth (Fritts, 1976). Growth rates and aboveground productivity in response to hydrologic regimes have been demonstrated for a variety of tree species through utilization of tree-ring analyses (Reily and Johnson, 1982; Mitsch and Rust, 1984; Anderson and Mitsch, 2008; Predick et al., 2009; Gee, 2012; Keim and Amos, 2012). Riparian forested wetlands located adjacent to rivers in the coastal plains of the southeastern United States are at an increased risk to changes in hydrologic regime and sediment deposition patterns due to their downstream location relative to anthropogenic influences and proximity to oceanic disturbances (Doyle et al., 2007; Stanturf et al., 2007; Conner et al., 2014). Quantification of the amount of sediment being trapped in riparian forested wetlands and impacts of hydrologic fluctuations in response to disturbances through time can aid predictions of future impacts of disturbances on forest productivity, regeneration, carbon storage, bottomland restoration, and water quality (Conner and Day, 1976; Hodges, 1997; Hunter et al., 2008). Therefore, the objectives of this study were to: (1) use the dendrogeomorphic method to compare estimated sediment accretion rates and mass per unit area of sediment trapped between two time periods (1881–2012 and 1987–2012) along a natural levee and backswamp in a riparian forest adjacent to the Tensaw River in southwestern Alabama, and as a result of the occurrence of flooding in this freshwater tidal system, (2) identify the impacts of hydrology and climate on radial growth in green ash (Fraxinus pennsylvanica Marsh.) and water tupelo (Nyssa aquatica L.), two species of differing flood-tolerance that are commonly found in bottomland hardwood systems.
2. Materials and methods 2.1. Study area This study was conducted along the western bank of the Tensaw River within the Mobile-Tensaw River Delta at approximately 1 m asl (Fig. 1). The Mobile-Tensaw Delta is formed by the Mobile, Tensaw, and Middle Rivers and is the second largest delta system in the United States (Smith, 1988). This delta contains approximately 43,000 ha of wetlands, with 75% forested (Smith, 1988). The Mobile-Tensaw Delta is located within the Mobile River Basin, situated below the confluence of the Alabama and Tombigbee River Basins. The Alabama and Tombigbee River Basins combined contain approximately 11.6 million ha (Smith, 1988). Drainage from this area contributes to the estimated 4.3 million Mg yr1 of suspended sediment that is transported downstream into the Mobile Bay and then potentially into the Gulf of Mexico (Fig. 1; Ryan and Goodell, 1972). The study site consisted of a natural levee, a ridge of sediment along the riverbank, and a water tupelo-baldcypress (Taxodium distichum (L.) Rich.) backswamp (30°570 N, 87°530 W) located in Baldwin County, Alabama. This site is approximately 4.5 km southwest of the community of Stockton and 50 km north of the Mobile Bay (Fig. 1). Very poorly drained soils of the Levy (fine, mixed, superactive, acid, thermic Typic Hydraquents) series characterize the site (Aust et al., 2012). Due to its proximity to the river, the natural levee receives greater sand deposition and thus, provides microsite conditions favorable to vegetative species that are less flood-tolerant and survive and grow better in better drained soils. The overstory and midstory on the natural levee are characterized by green ash, overcup oak (Quercus lyrata Walt.), and water oak (Q. nigra L.) with fewer occurrences of swamp tupelo (Nyssa sylvatica var. bicolor (Walt.) Sarg.), black willow (Salix nigra
274
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Fig. 1. Study site (star) located in southwestern Alabama within the Mobile River Basin (shaded) which outflows into the Mobile Bay and subsequently the Gulf of Mexico. Inset (right) illustrates site is located on the west bank of the Tensaw River in the Mobile-Tensaw River Delta and the Barry Steam Plant gage is approximately 11 km northwest of the site.
Marsh.), American elm (Ulmus americana L.), and hornbeam (Carpinus caroliniana Walt.) (Aust and Lea, 1991). The backswamp is lower in elevation than the natural levee and is primarily composed of naturally regenerated flood-tolerant water tupelo and baldcypress, but also contains a small component of Carolina ash (Fraxinus caroliniana P. Mill.), pumpkin ash (F. profunda (Bush.) Bush.), black willow, red maple (Acer rubrum L.), and water elm (Planera aquatica J.F. Gmel.) (McKee et al., 2012). Buttonbush (Cephalanthus occidentalis L.) and dwarf palmetto (Sabal minor Jacq. Pers.) are common shrubs in the backswamp and natural levee understory. Previous calibration between onsite stage recorders and the Barry Steam Plant gage (USGS #02470630, located 11 km northwest of site: Fig. 1) indicates backswamp flooding begins when a 1.4 m stage is reached with bankfull conditions at a 2.1 m stage (Warren, 2001). Hydrology in tidal freshwater forested wetlands in the southeastern U.S. is influenced by local tides, river discharge, local rainfall, evapotranspiration, surface runoff, and subsurface drainage (Day et al., 2007). The backswamp typically floods seasonally, for prolonged periods, from December to April when the average daily river stage is commonly above 1.4 m (Fig. 2). Water from this flood season period, tends to pond and persist in
the backswamp into summer months when evapotranspiration reduces standing water levels. To illustrate, during the summer months in the backswamp, the water table can range from 25 cm above the surface to 50 cm below the surface (Aust et al., 2006). The Tensaw River experiences semidiurnal tides that can cause the river to fluctuate 20–50 cm near the study site (Aust et al., 2006). Climate in the Mobile-Tensaw Delta region is classified as subtropical. On average, 1750 mm of precipitation is recorded annually with the greatest precipitation recorded during July (NOAA, 2011). Temperatures generally remain mild throughout the year. The average daily temperature is 19.2 °C with the highest temperatures observed during June, July, and August. Only a few weeks of overnight freezing temperatures are observed each year providing an extended growing season for vegetation in this region. 2.2. Field methods During the summer of 2012, total sediment accretion was measured immediately adjacent to 20 green ash trees of seed origin. Ten co-dominant trees were sampled along two 660 m transects; one transect ran parallel to the river along the natural levee and
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
275
Fig. 2. Average monthly river stage (m) at the Barry Steam Generating Electricity Plant in the Mobile-Tensaw Delta and average number of days at the 1.4 m stage per month. Previous calibrations indicate the backswamp becomes flooded when the Barry Steam Plant stage reaches 1.4 m and bankfull conditions are satisfied at a stage of 2.1 m.
one along the edge of the backswamp. Ash trees were selected with a ‘‘telephone pole” like appearance along the base (i.e., lacking basal flare) (Hupp and Morris, 1990). Using a dendrogeomorphic technique that was previously validated by Hupp and Morris (1990), the primary roots were located on two sides of each ash tree. These roots initially formed as rootlets at the point of germination and in mature trees, provide support. Adventitious roots, formed in response to episodic sediment deposition disturbances, were observed, but were smaller in diameter, shorter in length, buried at shallower depths, and occurred in common zones or ‘‘layers” (Shroder, 1980; Hupp and Morris, 1990; Walls et al., 2005). Once the primary roots were located, the vertical distance between the roots and current ground line was measured to the nearest 0.25 cm. Depths were measured 0.5–1.0 m away from the tree base to avoid any direct interference of the base on deposition totals. To determine the year in which each tree was established, a tree core was extracted below ground level using an increment borer. To determine the influence of hydrology and climate on radial growth tree cores were extracted from 20 green ash and 15 water tupelo at diameter at breast height (DBH, 1.4 m above ground level). Green ash cores were taken from the same ash trees sampled using the dendrogeomorphic technique to estimate sedimentation rates along the natural levee and backswamp. Sampled water tupelo trees were located in the backswamp at distances of 160–330 m from river. A minimum of two cores were taken from each tree to account for within-tree variation and to aid in identification of false rings, which are particularly common to water tupelo (Phipps, 1985). 2.3. Laboratory methods Tree cores were air dried and glued to wooden mounts. Progressively finer sand paper was used to surface cores until cellular structures became visible in the cross-sectional view under magnification. Two tree-ring chronologies were developed: (1) green ash using DBH and ground level cores and (2) water tupelo using DBH cores. Cores from each chronology were visually cross-dated using the list method, in which narrow growth rings common among samples were identified and used as signature years to ensure proper alignment of dating (Yamaguchi, 1991). Annual tree ring-widths were measured to the nearest 0.01 mm using the LinTabTM 5 RINNTECHÒ measuring system and TSAP-WinTM software
(v. 4.69). Dated tree-ring width measurements were verified to ensure quality of visual cross-dating using COFECHA software (Holmes, 1983). Dating errors detected by COFECHA caused by misdated, missing, or false rings were corrected. False rings commonly occur in water tupelo, which can make it difficult to accurately date without proper visual cross-dating and statistical verification in COFECHA (Phipps, 1985). Dated series were de-trended using the computer program ARSTAN to remove negative age-related growth trends (Cook, 1985). Through interactive de-trending, a flexible cubic smoothing spline was fit to each series. Cubic smoothing spline lengths ranged from 25 to 50 years, in which 50% of the variance was left in the chronology. This de-trending method allowed for removal of long-term growth trends and variation associated with changes in stand dynamics common to closed-canopy forests. Ring-width index values, actual ring-width value divided by the predicted value, were computed for each series in ARSTAN. Ring-width index values were averaged together to develop a master chronology for green ash and water tupelo. The arstan chronology was selected for further analyses as it is designed to strengthen the common signal in tree-ring records from closed-canopy forests. Mean series length, series intercorrelation, and mean sensitivity statistics were computed for each chronology. The two vertical sediment deposition values measured at each sampled tree were averaged together. Average total sediment deposition was divided by the tree age to estimate sediment accretion rates for the time period in which the tree was living. Sediment accretion values were averaged for trees established (1) prior to 1986 to represent the 1881–2012 time period and (2) after 1986 to represent the 1987–2012 time period. Average sediment accretion rates were determined for the natural levee and backswamp separately for the two time periods. 2.4. Existing datasets Daily stage data for the Barry Steam Plant gage (USGS #02470630, located 11 km northwest of site: Fig. 1) were downloaded from the U.S. Army Corps of Engineers in the Mobile District for the time period 1983–2012 (USACE, 2015). This gage was established in 1951 and until 1986, had been maintained by the U.S. Army Corps of Engineers. This gage is currently maintained by the U.S. Geological Survey. Using this data, the number of days above the 1.4 m and 2.1 m stage and average daily stage were
276
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Table 1 Mean annual sediment accretion rates (cm yr1) and sediment mass trapped per unit area (Mg ha1 yr1) compared between 1881 to 2012 and 1987 to 2012 time periods along the natural levee and backswamp. Different letters within a topographic position indicates values were significantly different at a = 0.05. Standard error presented in parentheses. Topographic position
A
Time period
N
Mean sediment accretion rate (cm yr1) b
MeanA sediment mass trapped per area (Mg ha1 yr1)
Natural levee
1881–2012 1987–2012
6 4
0.4 (0.0) 1.6 (0.2)a
25 (2)b 107 (12)a
Backswamp
1881–2012 1987–2012
5 5
0.5 (0.1)b 1.3 (0.2)a
35 (7)b 82 (13)a
Bulk density (0.66 Mg m3) value used to estimate mass (Mg) was the average obtained at 0–35 cm depths by McKee et al. (2012).
determined for each calendar month and for each growing season (April to August) for each given year. Monthly values for the prior calendar year and growing season were also determined. Stage data were missing for 1992 because the stage recorder was broken and thus, this year was omitted from our analyses. Monthly total precipitation and average daily temperature data were available for the Alabama Gulf Division (08) for years 1895 to 2012 from the National Climatic Data Center (NOAA, 2014).
precipitation, and average daily temperature were analyzed against annual ring width index for green ash and water tupelo using pairwise comparisons. Hydrologic (1983–2012) and climatic (1895–2012) data for all growing season months (April to August) individually and combined, and values for previous years were included in the analyses. Months or years with missing data were excluded from the analyses.
2.5. Data analysis
3. Results and discussion
Dendrogeomorphic sediment accretion rates were derived for the natural levee and backswamp and were compared between the two time periods. Comparisons were made in Wilcoxon–Mann–Whitney tests using the NPAR1WAY procedure at a significance-level of a = 0.05 in SAS 9.3 (SAS, 2012). Exact methods were specified for each nonparametric procedure. To further characterize the magnitude of sediment trapped, annual sediment accretion rates were converted to a mass per unit area (Mg ha1 yr1). Mass was calculated using the average bulk density (0.66 Mg m3) measured at depths of 0–35 cm on this site in 2010 (McKee et al., 2012). Soil bulk density values have been consistently low compared to other wetlands having mineral soils (Mitsch and Gosselink, 2000). In 1986, average bulk density was 0.60 Mg m3 at a distance of 130 m from the river (Aust, 1989). Lower on-site bulk densities are likely a result of the deposition of non-compacted sediment and the shrink–swell nature of the Levy series (Aust, 1989; McKee et al., 2012). To quantify the impacts of hydrology and climate on annual tree growth, Pearson correlation coefficients (r) were evaluated between annual ring width index for green ash and water tupelo and hydrologic and climatic variables. Specifically, the number of days above the 1.4 m and 2.1 m stage, average daily stage, total
3.1. Sediment accretion from 1881 to 2012 vs. 1987 to 2012 along the natural levee and backswamp Sediment accretion rates were significantly higher for the 1987–2012 time period than the 1881–2012 time period along the natural levee (p = 0.00; 124% higher ranging from 1.3 to 2.0 cm yr1) and backswamp (p = 0.03; 72% higher ranging from 0.8 to 1.9 cm yr1) (Table 1). Annual accretion rates for the 1881–2012 time period ranged from 0.3 to 0.5 cm yr1 on the natural levee and 0.3 to 0.9 cm yr1 along the backswamp. Sediment accretion rates were higher on the natural levee than along the backswamp, but were not significantly different for the 1987–2012 (p = 0.29) or the 1881–2012 (p = 0.25) time period. Similar trends were identified when sediment trapped annually was converted to mass per unit area (Table 1). Rates of annual mass of sediment trapped were significantly higher during the 1987–2012 than the 1881–2012 time period both on the natural levee (18–32 Mg ha1 yr1; p = 0.00) and along the backswamp (17–61 Mg ha1 yr1; p = 0.03) (Table 1). For the 1987 to 2012 time period, rates ranged from 85 to 135 Mg ha1 yr1 on the natural levee and 55 to 124 Mg ha1 yr1 along the backswamp.
Fig. 3. Sediment accretion rate (cm yr1) versus establishment year of adjacent green ash trees used in dendrogeomorphic technique prior to and following an adjacent harvest disturbance (dashed line) in 1986. Each data point represents the rate at each sampled location.
277
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Differences between sediment accretion rates and the mass of sediment trapped per unit area for the two time periods (1881– 2012 and 1987–2012) were most likely a consequence of effects from a nearby timber harvest disturbance that occurred in 1986 (Fig. 3). No harvesting disturbances occurred directly on the natural levee or in the backswamp areas sampled in this study. However, further upstream (60 m from most upstream and 720 m from most downstream portions of the site) an old pull boat run, that was first established in 1915 during a previous harvest, was used for barge access to transport and unload a rubber tired skidder into the backswamp during the 1986 harvest treatments (Mader, 1990; Aust et al., 2006). Old pull boat runs were historically used as a method to pull logs via cable across soft ground from backswamps rather than having to wait for high water to float them to the river for transport (Mancil, 1980; Mader, 1990; Doyle et al., 2007). Repeated dragging of logs along the same run created ditches approximately 2–3 m deep that were filled with a combination of mud and water (Mancil, 1980). Upstream barge entry into this old pull boat channel likely served as a sediment source and altered hydrology at downstream locations; therefore, providing one explanation of higher sediment accretion rates during the 1987–2012 time period. Previous studies have reported increased sediment accretion rates in riparian forests for time periods in which changes in land use or upstream channelization have occurred (Hupp and Morris, 1990; Hupp and Bazemore, 1993; Kleiss, 1996; Heimann and Roell, 2000; Kroes and Hupp, 2010). Upstream anthropogenic disturbances (e.g., agriculture, logging, mining, urbanization) have been shown to increase sediment yields and thus, sediment loads in disturbed watersheds (Meade et al., 1990). Re-use of this old pull boat run re-defined the channel into an adjacent backswamp. This likely altered site hydrology by increasing the flow of water into the site, resulting in at least initial increases in sediment deposition rates. This harvest disturbance also likely increased levels of light reaching the forest floor, which could have increased the density of vegetation in areas immediately adjacent to these channels (Aust et al., 1997). Vertical structure (i.e., vegetation) has been shown to increase surface roughness, which slows floodwaters and enhances deposition of suspended sediment (Hupp, 2000; Li and Yang, 2009; Ensign et al., 2014). Earlier research conducted in this harvested backswamp and surrounding areas illustrated an initial increase in herbaceous and woody vegetation following the 1986 harvest-related disturbances (Aust et al., 1997). During the initial seven years after the harvest-related disturbances, increased sediment accretion rates were directly related to ground flora biomass present (Aust et al., 1997, 2006). Similar initial increases in sediment accretion rates with increased herbaceous vegetation were reported following harvesting disturbances in the blackwater riparian forests adjacent to the Edisto River in South Carolina (Perison et al., 1997). Although we were unable to document changes in vegetation density through time in this particular study, areas immediately adjacent to our sampled locations
demonstrated the initial effects of alter vegetation dynamics that resulted in increased sediment accretion rates (Aust et al., 1997, 2006). Erosion also likely contributed to lower annual sediment accretion rates and mass trapped per unit area for the 1881–2012 time period, particularly, on the natural levee. To illustrate, mean total vertical sediment accretion for 1881–2012 (35 cm; ranged from 17 to 52 cm) was slightly lower than for the 1987–2012 time period (36 cm; ranged from 32 to 47 cm) on the natural levee. This was not the case along the backswamp; where, as anticipated, mean total deposition was greater for the 1881–2012 period (39 cm; ranged from 20 to 55 cm) than the 1987–2012 time period (30 cm; ranged from 20 to 47 cm). Natural levees are created through re-current sediment deposition adjacent to meandering black rivers in the coastal plains (Hupp, 2000). However, these levees are also in direct contact with river waters, which have the potential to erode exposed portions of the natural levee through time. Differences between gross and net sediment accretion rates were previously illustrated in wetlands which received inflow from the adjacent Olentangy River in Ohio, demonstrating the potential impact of erosion on sediment accretion estimates (Mitsch et al., 2014). Other studies that have used multiple estimation methods to determine long-term sediment accretion rates have found estimates from methods that require only one site visit (e.g., dendrogeomorphic, 137Cesium) to be more conservative than methods involving multiple, repeated measurements (Kleiss, 1996; Heimann and Roell, 2000; Kroes and Hupp, 2010). Primary reasons cited were associated with the length of time that the estimate covered. This is due to the increased likelihood of influence of erosion or compaction through time, which is only accounted for in methods which involve relatively frequent, multiple repeated measures (e.g., feldspar clay marker horizons, sediment pins, elevation surveys) (USACE, 1993). 3.2. Impacts of hydrology and climate on radial growth Tree ring chronologies depicted growth responses of green ash (1881–2011) along the natural levee and backswamp and for water tupelo (1890–2011) in the backswamp to hydrologic (days above 1.4 m and 2.1 m stage) and climatic (total precipitation and average daily temperature) variables (Table 2). Green ash ring width index had positive, significant relationships with growing season (April to August) days above the 1.4 m (r = 0.38; p = 0.04) and 2.1 m (r = 0.49; p = 0.00) stage as well as the average daily stage for the overall growing season (r = 0.41; p = 0.03) (Table 3). During the month of April, days above the 1.4 m (r = 0.43; p = 0.02) and 2.1 m (r = 0.53; p = 0.00) stage and average daily stage (r = 0.51; p = 0.00) were significantly correlated with green ash ring width index. In contrast to green ash, water tupelo ring width index was not significantly correlated with overall growing season days above the 1.4 m (r = 0.10; p = 0.61) or 2.1 m (r = 0.20; p = 0.31) stage, or growing season average daily stage (r = 0.16; p = 0.44).
Table 2 Tree ring chronology characteristics for Fraxinus pennsylvanica and Nyssa aquatica growth chronologies developed for stands adjacent to the Tensaw River in southwestern Alabama. Species
a
Distance from river (m)
Chronology span
Series lengths Range (years)
Mean (years)
Number of trees
Number of radii
Total growth rings
Series intercorrelation
Mean sensitivity
Fraxinus pennsylvanicaa
35–75
1881–2011
18–131
55
20
58
2032
0.466
0.424
Nyssa aquatica
160–330
1890–2011
54–122
80
15
30
2391
0.581
0.492
Chronology included belowground cores used in estimation of sediment accretion rates.
278
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Table 3 Pearson correlation coefficients (r) between annual ring width index and hydrologic and climatic variables for Fraxinus pennsylvanica along the natural levee and backswamp. Bolded values indicates significant correlation at a = 0.05. Bolded with asterisk (⁄) indicates p < 0.01. Variable/month Days above 1.4 m stage (backswamp flooded) Growing seasona April May June July August
a
Fraxinus pennsylvanica 0.39 0.43 0.23 0.31 0.16 0.10
Days above 2.1 m stage (bankfull) Growing seasona April May June July August
0.49⁄ 0.53⁄ 0.29 0.35 0.17 0.00
Average daily stage Growing seasona April May June July August
0.41 0.51⁄ 0.13 0.30 0.16 0.02
Total precipitation Growing seasona April May June July August
0.15 0.24⁄ 0.07 0.01 0.18 0.20
Average daily temperature Growing seasona April May June July August
0.16 0.11 0.11 0.11 0.07 0.05
April to August.
Further, no significant relationships were detected between water tupelo ring width index and days above the 1.4 m (r ranged from 0.10 to 0.19) or 2.1 m (r ranged from 0.11 to 0.25) stages or average daily stage (r ranged from 0.09 to 0.13) for any growing season month. Total precipitation and average daily temperature for the overall growing season were not significantly correlated with green ash or water tupelo ring width index. Green ash ring width index did not appear to be significantly influenced by average daily temperature for any month. However, similarly to the relationship with hydrologic variables, green ash was significantly correlated with total precipitation during the months of April (r = 0.24; p = 0.01) and August (r = 0.20; p = 0.04) (Table 3). No significant correlations were identified between water tupelo ring width index and total precipitation or average daily temperature. Lagged values (prior growing season) for days above 1.4 m and 2.1 m stage, average daily stage, total precipitation, and average daily temperature were not significantly correlated with the green ash or water tupelo ring width index. Average annual ring width for green ash in our study (1.5 mm) was similar to rates (1.6 mm yr1) along the Missouri River in North Dakota (Reily and Johnson, 1982), but lower than rates (6.3 mm yr1) along the Kankakee River in Illinois (Mitsch and Rust, 1984). Differences in green ash growth rates among studies are likely attributed to variation in site elevation, microtopography, soil texture, and flooding regime (Anderson and Mitsch, 2008). Stress induced by periodic flooding and sedimentation
disturbances in our study may have initially reduced radial growth of the green ash (Yanosky, 1982; Walls et al., 2005). Although green ash is classified as moderately flood-tolerant, it responds to periodic flooding events through formation of adventitious roots. Thus, belowground allocation of resources in response to flooding and sedimentation disturbances could have a negative effect on aboveground radial growth. Along the Olentangy River in Ohio, floodplain tree species were found to have a number of flood days at which growth was optimal; however, above this number of days, reduced growth became a tradeoff due to increased stress from the additional disturbances (Anderson and Mitsch, 2008). Green ash radial growth appeared to be most influenced by moisture-related hydrologic and climatic variables. Specifically, positive, significant correlations were identified between green ash ring width index and days above the 1.4 m and 2.1 m stage (growing season and April) and average daily stage (growing season and April). Our results agree with previously reported positive, significant correlations between green ash radial growth and mean monthly stage for April, May, and June along the White River in Arkansas (Gee, 2012). Positive relationships were also identified between 5-year green ash average annual basal area increments and average daily streamflow and days of floodplain inundation along the Kankakee River (Mitsch and Rust, 1984). Along the Wisconsin River, green ash radial growth was greater in an active floodplain than a floodwater excluded floodplain (Predick et al., 2009). Relationships between hydrologic variables and green ash ring width index were likely positive in our study due to potential moisture limitations during the growing season along natural levee and backswamp edge. Radial growth rates for water tupelo in forested wetlands have been shown to be greater under both permanently flooded (Dicke and Toliver, 1990; Keeland and Shartiz, 1995) and periodically flooded scenarios (Keeland and Shartiz, 1995; Kroes et al., 2007) than when only flooded seasonally (Dicke and Toliver, 1990). Average annual radial increase for mature water tupelo in our study (2.9 mm) was greater than annual radial growth reported for water tupelo in a continuously flooded (0.7 mm) mature (63 year-old) baldcypress-water tupelo stand in the Atchafalaya River Basin in Louisiana (Dicke and Toliver, 1990). Although in the backswamps along the Tensaw River flooding is characterized as seasonal, residual water from this season typically ponds in the lower backswamp and persists until evaporated during summer months. This coupled with a relatively high water table and influences of Squirrel Bayou (200 m to the west of the site) likely contributed to continuously hydric conditions in the backswamp increasing annual growth rates in our water tupelo. Extended hydroperiods created by these conditions have been shown to provide additional nutrients and dissolved oxygen required to achieve maximum growth in water tupelo (Keeland and Shartiz, 1995; Kroes et al., 2007). Lastly, tidal influences at our site may have resulted in relatively high growth rates in water tupelo. Annual DBH growth rates were found to be higher in tidal freshwater forested wetland (3.9 mm) than in a fluvial freshwater forested wetland (1.7 mm) along the Pocomoke River in Maryland (Kroes et al., 2007). Unexpectedly, no significant relationships were detected between hydrologic variables and water tupelo ring width index. Other studies have illustrated that water tupelo is highly dependent on flooding to transport additional nutrients and dissolved oxygen, particularly in backswamps, which may become stagnant by ponded water (Dicke and Toliver, 1990; Conner and Day, 1992; Keeland and Shartiz, 1995). Lack of significant relationships could have been influenced by defoliation by the endemic forest tent caterpillar (Malacosma disstria) previously observed on site from approximately 1954 to 1979, prior to implementation of aerial treatments (Morris, 1975; Mader, 1990). In southern
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Louisiana, where annual defoliation which has occurred since at least the 1960s, radial growth was shown to cease following defoliation until a new flush of leaves is developed (Conner and Day, 1992). During more severe years, radial growth may stop and re-start up to three times, with the last leaves persisting until the first frost, as late as early November (Dicke and Toliver, 1990; Conner and Day, 1992). In southern Alabama and Louisiana, reductions of 34% or more in radial growth rates were reported as a consequence of these annual defoliations (Morris, 1975). Defoliation by the forest tent caterpillar may have impacted growth of water tupelo in our study; therefore, making relationships between growth response and hydrologic variables vague and difficult to detect. This was the case in a study in the Atchafalaya Basin in Louisiana which failed to detect radial growth response in water tupelo following a thinning treatment due to reduced growth caused by forest tent caterpillar defoliation (Kennedy, 1983). Another reason for the lack of relationships could be related to changes in physiology as trees mature in age and increase in diameter. These changes in physiology may make mature trees more resistant to short-term alterations in environmental conditions than younger trees (Voelker, 2011; Liñán et al., 2012). 4. Conclusions The improvement of downstream water quality and aboveground biomass production are two major ecosystem services provided by riparian forested wetlands (Walbridge, 1993). This study emphasizes the important role that riparian forested wetlands play in improving water quality through trapping sediment from adjacent waterways and illustrates the potential impacts of forest harvest-related disturbances on sediment accretion rates. Further, this study identifies positive, significant relationships between green ash radial growth and hydrologic (days above 1.4 m and 2.1 m stage and average daily stage) and climatic (total precipitation) variables during the early growing season. In our study, water tupelo radial growth was less responsive to hydrologic and climatic fluctuations demonstrating potential resistance of water tupelo to perturbations. Results from this study suggest sediment accretion rates and impacts of hydrologic and climatic fluctuations on aboveground productivity should be considered when managing riparian forested wetlands. Lastly, we effectively used dendrochronological techniques to estimate sediment accretion rates and to determine impacts of hydrology and climate on tree radial growth. These techniques minimized site visits while still providing invaluable pre- and post-disturbance data. Acknowledgements This study was funded by the United States Department of Agriculture, National Needs Fellowship Grant 2010-38420-21851. Previous sediment accretion data collections were supported by the National Council for Air and Stream Improvement, Inc. We would also like to thank the Alabama Department of Conservation and Natural Resources and the Forever Wild Land Trust for granting permission to continue this long-term study. Additional field assistance was provided by John Peterson and Greg Ward. References Anderson, C.J., Mitsch, W.J., 2008. Tree basal growth response to flooding in a bottomland hardwood forest in central Ohio. J. Am. Water Resour. Assoc. 44, 1512–1520. Aust, W.M., 1989. Abiotic functional changes of a water tupelo-baldcypress wetland following disturbance by harvesting. PhD Dissertation, North Carolina State University. Aust, W.M., Lea, R., 1991. Soil temperature and organic matter in a disturbed forested wetland. Soil Sci. Soc. Am. J. 55, 1741–1746.
279
Aust, W.M., Schoenholtz, S.H., Zaebst, T.W., Szabo, B.A., 1997. Recovery status of tupelo-cypress wetland seven years after disturbance: silvicultural implications. For. Ecol. Manage. 90, 161–169. Aust, W.M., Fristoe, T.C., Gellerstedt, P.A., Giese, L.A.B., Miwa, M., 2006. Long-term effects of helicopter and ground-based skidding on site properties and stand growth in a tupelo-cypress wetland. For. Ecol. Manage. 226, 72–79. Aust, W.M., McKee, S.E., Seiler, J.R., Strahm, B.D., 2012. Long-term sediment accretion in bottomland hardwoods following timber harvest disturbances in the Mobile-Tensaw River Delta, Alabama, USA. Wetlands 32, 871–884. Boto, K.G., Patrick, W.H., 1978. Wetland functions and values: the state of our understanding. Am. Water Resour. Assoc., 479–489 Broadfoot, W.M., Willston, H.L., 1973. Flooding effects on southern forests. J. For. 9, 584–587. Conner, W.H., Day Jr., J.W., 1976. Productivity and composition of a baldcypresswater tupelo site and a bottomland hardwood site in a Louisiana swamp. Am. J. Bot. 63, 1354–1364. Conner, W.H., Day Jr., J.W., 1992. Diameter growth of Taxodium distichum (L.) Rich. and Nyssa aquatica L. from 1979 to 1985 in four Louisiana swamp stands. Am. Midl. Nat. 127, 290–299. Conner, W.H., Duberstein, J.A., Day Jr., J.W., Hutchinson, S., 2014. Impacts of changing hydrology and hurricanes on forest structure and growth along a flooding/elevation gradient in a south Louisiana forested wetland from 1986 to 2009. Wetlands 34, 803–814. Cook, E.R., 1985. A time series approach to tree ring standardization. PhD Dissertation, University of Arizona, Tuscon. Day, R.H., Williams, T.M., Swarzenski, C.M., 2007. Hydrology of tidal freshwater forested wetlands of the southeastern United States. In: Conner, W.H., Doyle, T. W., Krauss, K.W. (Eds.), Ecology of Tidal Freshwater Forested Wetlands of the Southeastern United States. Springer, New York, pp. 29–63. Dicke, S.G., Toliver, J.R., 1990. Growth and development of bald-cypress/watertupelo stands under continuous versus seasonal flooding. For. Ecol. Manage. 33– 34, 523–530. Doyle, T.W., O’Neil, C.P., Melder, M.P., From, A.S., Palta, M.M., 2007. Tidal freshwater swamps of the southeastern United States: effects of land use, hurricanes, sealevel rise, and climate change. In: Conner, W.H., Doyle, T.W., Krauss, K.W. (Eds.), Ecology of Tidal Freshwater Forested Wetlands of the Southeastern United States. Springer, New York, pp. 1–28. Ensign, S.H., Hupp, C.R., Noe, G.B., Krauss, K.W., Stagg, C.L., 2014. Sediment accretion in tidal freshwater forests and oligohaline marshes of the Waccamaw and Savannah Rivers, USA. Estuar. Coast. 37, 1107–1119. Fritts, H.C., 1976. Tree Rings and Climate. Academic Press, New York. Gee, H., 2012. The effects of hydrologic modifications on floodplain forest tree recruitment and growth in the Mississippi River Alluvial Valley, USA. PhD Dissertation, Louisiana State University. Heimann, D.C., Roell, M.J., 2000. Sediment loads and accumulation in a small riparian wetland system in northern Missouri. Wetlands 20, 219–231. Hodges, J.D., 1997. Development and ecology of bottomland hardwood sites. For. Ecol. Manage. 90, 117–125. Holmes, R.L., 1983. Computer-assisted quality control in tree-ring dating and measurement. Tree-Ring Bull. 43, 69–78. Hunter, R.G., Faulkner, S.P., Gibson, K.A., 2008. The importance of hydrology in restoration of bottomland hardwood wetland functions. Wetlands 28, 605–615. Hupp, C.R., Morris, E.E., 1990. A dendrogeomorphic approach to measurement of sedimentation in a forested wetland, Black Swamp, Arkansas. Wetlands 10, 107–124. Hupp, C.R., Bazemore, D.E., 1993. Temporal and spatial patterns of wetland sedimentation, West Tennessee. J. Hydrol. 141, 179–196. Hupp, C.R., 2000. Hydrology, geomorphology, and vegetation of coastal plain rivers in the south-eastern USA. Hydrol. Process. 14, 2991–3010. Hupp, C.R., Demas, C.R., Kroes, D.E., Day, R.H., Doyle, T.W., 2008. Recent sedimentation patterns within the central Atchafalaya Basin, Louisiana. Wetlands 28, 125–140. Johnston, C.A., Bubenzer, G.D., Lee, G.B., Madison, F.W., McHenry, J.R., 1984. Nutrient trapping by sediment deposition in a seasonally flooded lakeside wetland. J. Environ. Qual. 13, 283–290. Keeland, B.D., Shartiz, R.R., 1995. Seasonal growth patterns of Nyssa sylvatica var. biflora, Nyssa aquatica, and Taxodium distichum as affected by hydrologic regime. Can. J. For. Res. 25, 1084–1096. Keim, R.F., Amos, J.B., 2012. Dendrochronological analysis of baldcypress (Taxodium distichum) responses to climate and contrasting flood regimes. Can. J. For. Res. 42, 423–436. Kennedy, H.E., 1983. Water tupelo in the Atchafalaya Basin does not benefit from thinning. Res. Note SO-298. New Orleans, LA. U.S. Department of Agriculture, Forest Service, Southern Forest Experiment Station. Kleiss, B.A., 1996. Sediment retention in a bottomland hardwood wetland in eastern Arkansas. Wetlands 16, 321–333. Kroes, D.E., Hupp, C.R., Noe, G.B., 2007. Sediment, nutrient, and vegetation trends along the tidal, forested Pocomoke River, Maryland. In: Conner, W.H., Doyle, T. W., Krauss, K.W. (Eds.), Ecology of Tidal Freshwater Forested Wetlands of the Southeastern United States. Springer, New York, pp. 113–137. Kroes, D.E., Hupp, C.R., 2010. The effect of channelization on floodplain sediment deposition and subsidence along the Pocomoke River, Maryland. J. Am. Water Resour. Assoc. 46, 686–699. Li, H., Yang, S.L., 2009. Trapping effect of tidal marsh vegetation on suspended sediment, Yangtze Delta. J. Coast. Res. 25, 915–924.
280
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Liñán, I.D., Gutiérrez, I.H., Andreu-Hayles, L., Muntán, E., Campelo, F., Helle, G., 2012. Age effects and climate response in trees: a multi-proxy tree-ring test in old-growth life stages. Eur. J. For. Res. 131, 933–944. Mader, S.F., 1990. Recovery of ecosystem functions and plant community structure by a tupelo-cypress wetland following timber harvesting. PhD Dissertation, North Carolina State University. Mancil, E., 1980. Pullboat logging. J. For. Hist. 24, 135–141. McKee, S.E., Aust, W.M., Seiler, J.R., Strahm, B.D., Schilling, E.B., 2012. Long-term site productivity of a tupelo-cypress swamp 24 years after harvesting disturbances. For. Ecol. Manage. 265, 172–180. Meade, R.H., Yuzyk, T.R., Day, T.J., 1990. Movement and storage of sediment in rivers of the United States and Canada. In: Wolman, M.G., Riggs, H.C., (Eds.), Surface Water Hydrology: Boulder Colorado, vol. O-1, Geological Society of America, The Geology of North America. Mitsch, W.J., Rust, W.G., 1984. Tree growth responses to flooding in a bottomland forest in northeastern Illinois. For. Sci. 30, 499–510. Mitsch, W.J., Gosselink, J.G., 2000. Wetlands, third ed. John Wiley, New York. Mitsch, W.J., Nedrich, S.M., Harter, S.K., Anderson, C., Nahlik, A.M., Bernal, B., 2014. Sedimentation in created freshwater riverine wetlands: 15 years of succession and contrasts of methods. Ecol. Eng. 72, 25–34. Morris, R.C., 1975. Tree-eaters in the tupelo swamps. For. People 25, 22–24. National Oceanic and Atmospheric Administration (NOAA), 2011. Summary of Monthly Normals, 1981–2010, National Climatic Data Center. Available online at (accessed on 22 February 2015). NOAA, 2014. Divisional data for Gulf of Alabama (08) for 1895–2012, National Climatic Data Center. Available online at (accessed on 4 February 2015). Perison, D., Phelps, J., Pavel, C., Kellison, R., 1997. The effects of timber harvest in a South Carolina blackwater bottomland. For. Ecol. Manage. 90, 171–185. Phillips, J.D., 2001. Sedimentation in bottomland hardwoods downstream of an east Texas dam. Environ. Geol. 40, 860–868. Phipps, R.L., 1985. Collecting, preparing, crossdating, and measuring tree increment cores. US Geological Survey Water-Resources Investigations Report 85-4148. Lakewood CO. Predick, K.I., Gergel, S.E., Turner, M.G., 2009. Effect of flood regime on tree growth in the floodplain and surrounding uplands of the Wisconsin River. River Res. Appl. 25, 283–296. Reily, P.W., Johnson, W.C., 1982. The effects of altered hydrologic regime on tree growth along the Missouri River in North Dakota. Can. J. Bot. 60, 2410–2423.
Ryan, J.J., Goodell, H.G., 1972. Marine geology and estuarine history of Mobile Bay, Alabama, Part 1. Contemporary sediments. Geol. Soc. Am. Mem. 133, 517–554. SAS, 2012. SAS/STAT 9.3 User’s Guide: The NPAR1WAY Procedure. SAS Institute, Inc., Cary, North Carolina. Shroder Jr., J.F., 1980. Dendrogeomorphology: review and new techniques in treering dating. Prog. Phys. Geol. 4, 161–188. Smith, W.E., 1988. Geomorphology of the mobile delta. Geol. Survey Alabama Bull. 132 (Tuscaloosa, AL). Stanturf, J.A., Goodrick, S.L., Outcalt, K.W., 2007. Disturbance and coastal forests: a strategic approach to forest management in hurricane impact zones. For. Ecol. Manage. 250, 119–135. U.S. Army Corps of Engineers (USACE), 1993. Methods of measuring sedimentation rates in bottomland hardwoods. WRP Tech. Note SD-CP-4.1. Vicksburg, MS. U.S. Army Corps of Engineers Waterways Experiment Station. USACE, 2015. Barry steam plant gage data (U.S. Geological Survey Gage #02470630), water years 1983–2012. Available online at (accessed 02/22/2015). Voelker, S.L., 2011. Age-dependent changes in environmental influences on tree growth and their implications for forest responses to climate change. In: Meinzer, F.C., Lachenbruch, B., Dawson, T.E. (Eds.), Size- and Age-Related Changes in Tree Structure and Function, Tree Physiology, vol. 4. Springer, New York, pp. 455–479. Walbridge, M.R., 1993. Functions and values of forested wetlands in the southern United States. J. For. 91, 15–19. Walling, D.E., 2006. Human impact on land-ocean sediment transfer by the world’s rivers. Geomorphology 79, 192–216. Walls, R.L., Wardrop, D.H., Brooks, R.P., 2005. The impact of experimental sedimentation and flooding on the growth and germination of floodplain trees. Plant Ecol. 176, 203–213. Warren, S.E., 2001. Sedimentation in a tupelo-baldcypress forested wetland 12 years following harvest disturbance. MS Thesis, Virginia Tech. Wharton, C.H., Kitchens, W.M., Pendleton, E.C., Sipe, T.W., 1982. The ecology of bottomland hardwood swamps of the southeast: a community profile. US Dept. of Interior, Fish and Wildlife Service, Biol. Serv. Prog., Washington, D.C. FWS/ OBS-81-37. Yamaguchi, D.K., 1991. A simple method for cross-dating increment cores from living trees. Can. J. For. Res. 21, 414–416. Yanosky, T.M., 1982. Hydrologic inferences from ring widths of flood-damaged trees, Potomac River, Maryland. Environ. Geol. 4, 43–52.
Contents lists available at ScienceDirect
Forest Ecology and Management journal homepage: www.elsevier.com/locate/foreco
Sediment accretion rates and radial growth in natural levee and backswamp riparian forests in southwestern Alabama, USA Kathryn R. Kidd ⇑, Carolyn A. Copenheaver, W. Michael Aust Department of Forest Resources and Environmental Conservation, Virginia Tech, Blacksburg, VA 24061, USA
a r t i c l e
i n f o
Article history: Received 20 July 2015 Received in revised form 15 September 2015 Accepted 17 September 2015 Available online 26 September 2015 Keywords: Forested wetlands Sediment accretion Hydrology Dendrochronology Green ash Water tupelo
a b s t r a c t Riparian forested wetlands improve downstream water quality by trapping suspended sediment from adjacent waterways. Waters that transport sediment and nutrients into adjacent wetlands also create wet and dry hydrologic periods and thus, have the potential to impact site productivity. In this study, we used a dendrogeomorphic technique with green ash (Fraxinus pennsylvanica Marsh.) to estimate sediment accretion for two time periods (1881–2012 and 1987–2012) along a natural levee (35 m from river) and backswamp (75 m from river) and identified the influence of hydrology and climate on radial growth in green ash and water tupelo (Nyssa aquatica L.) along the Tensaw River in southwestern Alabama. We detected significantly higher sediment accretion rates for the 1987–2012 time period along the natural levee (p = 0.00; 1.6 cm yr1) and backswamp (p = 0.03; 1.2 cm yr1) than for the 1881 to 2012 period (0.4 and 0.5 cm yr1). Using previously measured (2010) soil bulk density for the site, estimated mass of sediment trapped per unit area ranged from 55–135 Mg ha1 yr1 for the 1987–2012 period and 17–61 Mg ha1 yr1 for the 1881–2012 period. We identified positive, significant correlations between green ash radial growth and the number of days the backswamp was flooded (1.4 m stage), days the Tensaw River was at bankfull (2.1 m), and average daily river stage during the overall growing season (April to August) and for the month of April. Green ash radial growth also illustrated a positive, significant response to April precipitation totals demonstrating the overall role of moisture availability just prior to the onset of xylem formation. Green ash trees along the natural levee and backswamp were more responsive to hydrology and climate than water tupelo trees located further in the backswamp, illustrating the potential resistance of water tupelo to perturbations. This study illustrates the important role forested wetlands play in improving water quality through quantification of sediment accretion rates and the potential impact that introduced disturbances (i.e., forest harvest-related disturbances) can have on ecosystem services. Ó 2015 Elsevier B.V. All rights reserved.
1. Introduction Bottomland hardwoods, a well-known type of riparian forested wetland, provide many ecosystem services to society. Most notably, riparian forested wetlands improve water quality by trapping and storing sediment from adjacent waterways (Boto and Patrick, 1978; Walbridge, 1993). Sediment generated upstream is transported downstream by rivers and eventually released into the ocean unless it is captured and deposited during transport (Walling, 2006). The amount of sediment captured in a riparian forested wetland can change through time if sediment loads in a
⇑ Corresponding author at: Department of Forest Resources and Environmental Conservation (0324), 228 Cheatham Hall, Virginia Tech, Blacksburg, VA 24061, USA. E-mail addresses: [email protected] (K.R. Kidd), [email protected] (C.A. Copenheaver), [email protected] (W.M. Aust). http://dx.doi.org/10.1016/j.foreco.2015.09.025 0378-1127/Ó 2015 Elsevier B.V. All rights reserved.
waterway are altered due to watershed disturbances (e.g., channelization, land use, urbanization) or if trapping efficiency changes due to reasons that increase or decrease surface roughness (e.g., altered vegetation dynamics, microtopography, woody debris dynamics) (Hupp and Morris, 1990; Meade et al., 1990; Hupp and Bazemore, 1993; Kleiss, 1996; Heimann and Roell, 2000; Li and Yang, 2009; Kroes and Hupp, 2010; Ensign et al., 2014). Bottomland hardwood systems are also valued by society as highly productive sites for timber production, wildlife habitat, carbon storage, nutrient cycling, and protection from floodwaters (Walbridge, 1993). Riparian forested wetlands are characterized by hydrologic cycles of wet and dry periods created through variations in onsite flooding, precipitation, and evapotranspiration rates (Broadfoot and Willston, 1973; Wharton et al., 1982; Hupp, 2000; Day et al., 2007). Disturbances which alter the hydrologic regime in a riparian forested wetland have the potential to impact
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
functionality of forested wetlands and thus, the ability to provide ecosystem services (Conner and Day, 1976; Mitsch and Rust, 1984; Hunter et al., 2008; Kroes and Hupp, 2010). Hydroperiod length, flooding occurrence, and streamflow characteristics influence sediment accumulation in forested wetlands (Kleiss, 1996; Hupp, 2000; Hupp et al., 2008). Riparian forested wetlands reduce the velocity and peaks of floodwaters, which transport suspended sediment into adjacent riparian forested wetlands (Kleiss, 1996). As the velocity of the floodwaters is reduced, sediment falls out of suspension and is deposited, thus creating the microtopography associated with bottomlands (Hodges, 1997). Vertical structures such as vegetation, microtopographical features, and woody debris increase surface roughness, contributing to the decreased velocity of floodwaters, and increasing sediment deposition rates (Hupp, 2000; Li and Yang, 2009; Ensign et al., 2014). Distribution and growth rate of individual tree species varies among microtopographic landforms (e.g., levees vs. backswamps) in bottomland forests due to differences in flood tolerance (Broadfoot and Willston, 1973; Wharton et al., 1982; Hodges, 1997). As a result, forested wetlands along meandering coastal plain rivers in the southeastern United States are characterized by a gradient of less flood-tolerant bottomland hardwood species along elevated natural levee and transitional areas, and by more flood-tolerant species at lower elevations (Wharton et al., 1982; Hodges, 1997; Hupp, 2000; Conner et al., 2014). Changes in hydrologic regime alter spatial and temporal sediment deposition and inflow patterns and, as a consequence, can alter tree growth and productivity in forested wetlands (Conner and Day, 1976; Mitsch and Rust, 1984; Conner et al., 2014). Aboveground productivity in forested palustrine wetlands is dependent on several factors including: silvical characteristics for individual species, hydrologic regime, nutrient availability, edaphic conditions, and previous disturbance history (Broadfoot and Willston, 1973; Conner and Day, 1976; Reily and Johnson, 1982; Wharton et al., 1982). Floodwaters transport suspended sediment and additional nutrients into adjacent forested wetlands and thus, are one of the main hydrologic drivers influencing productivity in forested wetlands (Conner and Day, 1976; Hunter et al., 2008). Sediment deposited from floodwaters provides an allochthonous, external source of nutrients, to forested wetlands (Johnston et al., 1984). These additional nutrients enhance aboveground productivity, allowing individual species to achieve maximum growth (Keeland and Shartiz, 1995). Furthermore, these sediments alter microtopography and species distribution over time by filling abandoned river channels, creating and adding to the occurrence of natural levees (Hodges, 1997). Dendrochronological techniques can be used to estimate sediment accretion rates (Shroder, 1980; Hupp and Morris, 1990) and to assess the impacts of altered hydrology through examination of the variation in annual tree-ring growth patterns (Fritts, 1976). A dendrogeomorphic approach that incorporates total vertical sediment deposition and total age of an immediately adjacent tree can be used to estimate sediment accretion rates following one site visit (Hupp and Morris, 1990; USACE, 1993). For instance, this dendrogeomorphic technique has been used to: characterize spatial and temporal sediment deposition patterns (Hupp and Morris, 1990); quantify the influence physical (e.g., elevation, microtopography, distance from waterway or channel, flood duration) and ecological (e.g., vegetation density, stand structure, species composition) factors have on sediment accretion rates (Hupp and Morris, 1990; Kleiss, 1996; Heimann and Roell, 2000; Phillips, 2001; Hupp et al., 2008); and determine the impact of upstream disturbances (e.g., channelization, dam construction, hydrologic alterations, land use) on short- and long-term sediment deposition and subsidence rates (Hupp and Bazemore, 1993;
273
Heimann and Roell, 2000; Kroes and Hupp, 2010). Analysis of the variation in tree ring widths among calendar years has proven useful in identifying environmental influences (e.g., flooding, streamflow, precipitation, temperature) on annual tree growth (Fritts, 1976). Growth rates and aboveground productivity in response to hydrologic regimes have been demonstrated for a variety of tree species through utilization of tree-ring analyses (Reily and Johnson, 1982; Mitsch and Rust, 1984; Anderson and Mitsch, 2008; Predick et al., 2009; Gee, 2012; Keim and Amos, 2012). Riparian forested wetlands located adjacent to rivers in the coastal plains of the southeastern United States are at an increased risk to changes in hydrologic regime and sediment deposition patterns due to their downstream location relative to anthropogenic influences and proximity to oceanic disturbances (Doyle et al., 2007; Stanturf et al., 2007; Conner et al., 2014). Quantification of the amount of sediment being trapped in riparian forested wetlands and impacts of hydrologic fluctuations in response to disturbances through time can aid predictions of future impacts of disturbances on forest productivity, regeneration, carbon storage, bottomland restoration, and water quality (Conner and Day, 1976; Hodges, 1997; Hunter et al., 2008). Therefore, the objectives of this study were to: (1) use the dendrogeomorphic method to compare estimated sediment accretion rates and mass per unit area of sediment trapped between two time periods (1881–2012 and 1987–2012) along a natural levee and backswamp in a riparian forest adjacent to the Tensaw River in southwestern Alabama, and as a result of the occurrence of flooding in this freshwater tidal system, (2) identify the impacts of hydrology and climate on radial growth in green ash (Fraxinus pennsylvanica Marsh.) and water tupelo (Nyssa aquatica L.), two species of differing flood-tolerance that are commonly found in bottomland hardwood systems.
2. Materials and methods 2.1. Study area This study was conducted along the western bank of the Tensaw River within the Mobile-Tensaw River Delta at approximately 1 m asl (Fig. 1). The Mobile-Tensaw Delta is formed by the Mobile, Tensaw, and Middle Rivers and is the second largest delta system in the United States (Smith, 1988). This delta contains approximately 43,000 ha of wetlands, with 75% forested (Smith, 1988). The Mobile-Tensaw Delta is located within the Mobile River Basin, situated below the confluence of the Alabama and Tombigbee River Basins. The Alabama and Tombigbee River Basins combined contain approximately 11.6 million ha (Smith, 1988). Drainage from this area contributes to the estimated 4.3 million Mg yr1 of suspended sediment that is transported downstream into the Mobile Bay and then potentially into the Gulf of Mexico (Fig. 1; Ryan and Goodell, 1972). The study site consisted of a natural levee, a ridge of sediment along the riverbank, and a water tupelo-baldcypress (Taxodium distichum (L.) Rich.) backswamp (30°570 N, 87°530 W) located in Baldwin County, Alabama. This site is approximately 4.5 km southwest of the community of Stockton and 50 km north of the Mobile Bay (Fig. 1). Very poorly drained soils of the Levy (fine, mixed, superactive, acid, thermic Typic Hydraquents) series characterize the site (Aust et al., 2012). Due to its proximity to the river, the natural levee receives greater sand deposition and thus, provides microsite conditions favorable to vegetative species that are less flood-tolerant and survive and grow better in better drained soils. The overstory and midstory on the natural levee are characterized by green ash, overcup oak (Quercus lyrata Walt.), and water oak (Q. nigra L.) with fewer occurrences of swamp tupelo (Nyssa sylvatica var. bicolor (Walt.) Sarg.), black willow (Salix nigra
274
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Fig. 1. Study site (star) located in southwestern Alabama within the Mobile River Basin (shaded) which outflows into the Mobile Bay and subsequently the Gulf of Mexico. Inset (right) illustrates site is located on the west bank of the Tensaw River in the Mobile-Tensaw River Delta and the Barry Steam Plant gage is approximately 11 km northwest of the site.
Marsh.), American elm (Ulmus americana L.), and hornbeam (Carpinus caroliniana Walt.) (Aust and Lea, 1991). The backswamp is lower in elevation than the natural levee and is primarily composed of naturally regenerated flood-tolerant water tupelo and baldcypress, but also contains a small component of Carolina ash (Fraxinus caroliniana P. Mill.), pumpkin ash (F. profunda (Bush.) Bush.), black willow, red maple (Acer rubrum L.), and water elm (Planera aquatica J.F. Gmel.) (McKee et al., 2012). Buttonbush (Cephalanthus occidentalis L.) and dwarf palmetto (Sabal minor Jacq. Pers.) are common shrubs in the backswamp and natural levee understory. Previous calibration between onsite stage recorders and the Barry Steam Plant gage (USGS #02470630, located 11 km northwest of site: Fig. 1) indicates backswamp flooding begins when a 1.4 m stage is reached with bankfull conditions at a 2.1 m stage (Warren, 2001). Hydrology in tidal freshwater forested wetlands in the southeastern U.S. is influenced by local tides, river discharge, local rainfall, evapotranspiration, surface runoff, and subsurface drainage (Day et al., 2007). The backswamp typically floods seasonally, for prolonged periods, from December to April when the average daily river stage is commonly above 1.4 m (Fig. 2). Water from this flood season period, tends to pond and persist in
the backswamp into summer months when evapotranspiration reduces standing water levels. To illustrate, during the summer months in the backswamp, the water table can range from 25 cm above the surface to 50 cm below the surface (Aust et al., 2006). The Tensaw River experiences semidiurnal tides that can cause the river to fluctuate 20–50 cm near the study site (Aust et al., 2006). Climate in the Mobile-Tensaw Delta region is classified as subtropical. On average, 1750 mm of precipitation is recorded annually with the greatest precipitation recorded during July (NOAA, 2011). Temperatures generally remain mild throughout the year. The average daily temperature is 19.2 °C with the highest temperatures observed during June, July, and August. Only a few weeks of overnight freezing temperatures are observed each year providing an extended growing season for vegetation in this region. 2.2. Field methods During the summer of 2012, total sediment accretion was measured immediately adjacent to 20 green ash trees of seed origin. Ten co-dominant trees were sampled along two 660 m transects; one transect ran parallel to the river along the natural levee and
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
275
Fig. 2. Average monthly river stage (m) at the Barry Steam Generating Electricity Plant in the Mobile-Tensaw Delta and average number of days at the 1.4 m stage per month. Previous calibrations indicate the backswamp becomes flooded when the Barry Steam Plant stage reaches 1.4 m and bankfull conditions are satisfied at a stage of 2.1 m.
one along the edge of the backswamp. Ash trees were selected with a ‘‘telephone pole” like appearance along the base (i.e., lacking basal flare) (Hupp and Morris, 1990). Using a dendrogeomorphic technique that was previously validated by Hupp and Morris (1990), the primary roots were located on two sides of each ash tree. These roots initially formed as rootlets at the point of germination and in mature trees, provide support. Adventitious roots, formed in response to episodic sediment deposition disturbances, were observed, but were smaller in diameter, shorter in length, buried at shallower depths, and occurred in common zones or ‘‘layers” (Shroder, 1980; Hupp and Morris, 1990; Walls et al., 2005). Once the primary roots were located, the vertical distance between the roots and current ground line was measured to the nearest 0.25 cm. Depths were measured 0.5–1.0 m away from the tree base to avoid any direct interference of the base on deposition totals. To determine the year in which each tree was established, a tree core was extracted below ground level using an increment borer. To determine the influence of hydrology and climate on radial growth tree cores were extracted from 20 green ash and 15 water tupelo at diameter at breast height (DBH, 1.4 m above ground level). Green ash cores were taken from the same ash trees sampled using the dendrogeomorphic technique to estimate sedimentation rates along the natural levee and backswamp. Sampled water tupelo trees were located in the backswamp at distances of 160–330 m from river. A minimum of two cores were taken from each tree to account for within-tree variation and to aid in identification of false rings, which are particularly common to water tupelo (Phipps, 1985). 2.3. Laboratory methods Tree cores were air dried and glued to wooden mounts. Progressively finer sand paper was used to surface cores until cellular structures became visible in the cross-sectional view under magnification. Two tree-ring chronologies were developed: (1) green ash using DBH and ground level cores and (2) water tupelo using DBH cores. Cores from each chronology were visually cross-dated using the list method, in which narrow growth rings common among samples were identified and used as signature years to ensure proper alignment of dating (Yamaguchi, 1991). Annual tree ring-widths were measured to the nearest 0.01 mm using the LinTabTM 5 RINNTECHÒ measuring system and TSAP-WinTM software
(v. 4.69). Dated tree-ring width measurements were verified to ensure quality of visual cross-dating using COFECHA software (Holmes, 1983). Dating errors detected by COFECHA caused by misdated, missing, or false rings were corrected. False rings commonly occur in water tupelo, which can make it difficult to accurately date without proper visual cross-dating and statistical verification in COFECHA (Phipps, 1985). Dated series were de-trended using the computer program ARSTAN to remove negative age-related growth trends (Cook, 1985). Through interactive de-trending, a flexible cubic smoothing spline was fit to each series. Cubic smoothing spline lengths ranged from 25 to 50 years, in which 50% of the variance was left in the chronology. This de-trending method allowed for removal of long-term growth trends and variation associated with changes in stand dynamics common to closed-canopy forests. Ring-width index values, actual ring-width value divided by the predicted value, were computed for each series in ARSTAN. Ring-width index values were averaged together to develop a master chronology for green ash and water tupelo. The arstan chronology was selected for further analyses as it is designed to strengthen the common signal in tree-ring records from closed-canopy forests. Mean series length, series intercorrelation, and mean sensitivity statistics were computed for each chronology. The two vertical sediment deposition values measured at each sampled tree were averaged together. Average total sediment deposition was divided by the tree age to estimate sediment accretion rates for the time period in which the tree was living. Sediment accretion values were averaged for trees established (1) prior to 1986 to represent the 1881–2012 time period and (2) after 1986 to represent the 1987–2012 time period. Average sediment accretion rates were determined for the natural levee and backswamp separately for the two time periods. 2.4. Existing datasets Daily stage data for the Barry Steam Plant gage (USGS #02470630, located 11 km northwest of site: Fig. 1) were downloaded from the U.S. Army Corps of Engineers in the Mobile District for the time period 1983–2012 (USACE, 2015). This gage was established in 1951 and until 1986, had been maintained by the U.S. Army Corps of Engineers. This gage is currently maintained by the U.S. Geological Survey. Using this data, the number of days above the 1.4 m and 2.1 m stage and average daily stage were
276
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Table 1 Mean annual sediment accretion rates (cm yr1) and sediment mass trapped per unit area (Mg ha1 yr1) compared between 1881 to 2012 and 1987 to 2012 time periods along the natural levee and backswamp. Different letters within a topographic position indicates values were significantly different at a = 0.05. Standard error presented in parentheses. Topographic position
A
Time period
N
Mean sediment accretion rate (cm yr1) b
MeanA sediment mass trapped per area (Mg ha1 yr1)
Natural levee
1881–2012 1987–2012
6 4
0.4 (0.0) 1.6 (0.2)a
25 (2)b 107 (12)a
Backswamp
1881–2012 1987–2012
5 5
0.5 (0.1)b 1.3 (0.2)a
35 (7)b 82 (13)a
Bulk density (0.66 Mg m3) value used to estimate mass (Mg) was the average obtained at 0–35 cm depths by McKee et al. (2012).
determined for each calendar month and for each growing season (April to August) for each given year. Monthly values for the prior calendar year and growing season were also determined. Stage data were missing for 1992 because the stage recorder was broken and thus, this year was omitted from our analyses. Monthly total precipitation and average daily temperature data were available for the Alabama Gulf Division (08) for years 1895 to 2012 from the National Climatic Data Center (NOAA, 2014).
precipitation, and average daily temperature were analyzed against annual ring width index for green ash and water tupelo using pairwise comparisons. Hydrologic (1983–2012) and climatic (1895–2012) data for all growing season months (April to August) individually and combined, and values for previous years were included in the analyses. Months or years with missing data were excluded from the analyses.
2.5. Data analysis
3. Results and discussion
Dendrogeomorphic sediment accretion rates were derived for the natural levee and backswamp and were compared between the two time periods. Comparisons were made in Wilcoxon–Mann–Whitney tests using the NPAR1WAY procedure at a significance-level of a = 0.05 in SAS 9.3 (SAS, 2012). Exact methods were specified for each nonparametric procedure. To further characterize the magnitude of sediment trapped, annual sediment accretion rates were converted to a mass per unit area (Mg ha1 yr1). Mass was calculated using the average bulk density (0.66 Mg m3) measured at depths of 0–35 cm on this site in 2010 (McKee et al., 2012). Soil bulk density values have been consistently low compared to other wetlands having mineral soils (Mitsch and Gosselink, 2000). In 1986, average bulk density was 0.60 Mg m3 at a distance of 130 m from the river (Aust, 1989). Lower on-site bulk densities are likely a result of the deposition of non-compacted sediment and the shrink–swell nature of the Levy series (Aust, 1989; McKee et al., 2012). To quantify the impacts of hydrology and climate on annual tree growth, Pearson correlation coefficients (r) were evaluated between annual ring width index for green ash and water tupelo and hydrologic and climatic variables. Specifically, the number of days above the 1.4 m and 2.1 m stage, average daily stage, total
3.1. Sediment accretion from 1881 to 2012 vs. 1987 to 2012 along the natural levee and backswamp Sediment accretion rates were significantly higher for the 1987–2012 time period than the 1881–2012 time period along the natural levee (p = 0.00; 124% higher ranging from 1.3 to 2.0 cm yr1) and backswamp (p = 0.03; 72% higher ranging from 0.8 to 1.9 cm yr1) (Table 1). Annual accretion rates for the 1881–2012 time period ranged from 0.3 to 0.5 cm yr1 on the natural levee and 0.3 to 0.9 cm yr1 along the backswamp. Sediment accretion rates were higher on the natural levee than along the backswamp, but were not significantly different for the 1987–2012 (p = 0.29) or the 1881–2012 (p = 0.25) time period. Similar trends were identified when sediment trapped annually was converted to mass per unit area (Table 1). Rates of annual mass of sediment trapped were significantly higher during the 1987–2012 than the 1881–2012 time period both on the natural levee (18–32 Mg ha1 yr1; p = 0.00) and along the backswamp (17–61 Mg ha1 yr1; p = 0.03) (Table 1). For the 1987 to 2012 time period, rates ranged from 85 to 135 Mg ha1 yr1 on the natural levee and 55 to 124 Mg ha1 yr1 along the backswamp.
Fig. 3. Sediment accretion rate (cm yr1) versus establishment year of adjacent green ash trees used in dendrogeomorphic technique prior to and following an adjacent harvest disturbance (dashed line) in 1986. Each data point represents the rate at each sampled location.
277
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Differences between sediment accretion rates and the mass of sediment trapped per unit area for the two time periods (1881– 2012 and 1987–2012) were most likely a consequence of effects from a nearby timber harvest disturbance that occurred in 1986 (Fig. 3). No harvesting disturbances occurred directly on the natural levee or in the backswamp areas sampled in this study. However, further upstream (60 m from most upstream and 720 m from most downstream portions of the site) an old pull boat run, that was first established in 1915 during a previous harvest, was used for barge access to transport and unload a rubber tired skidder into the backswamp during the 1986 harvest treatments (Mader, 1990; Aust et al., 2006). Old pull boat runs were historically used as a method to pull logs via cable across soft ground from backswamps rather than having to wait for high water to float them to the river for transport (Mancil, 1980; Mader, 1990; Doyle et al., 2007). Repeated dragging of logs along the same run created ditches approximately 2–3 m deep that were filled with a combination of mud and water (Mancil, 1980). Upstream barge entry into this old pull boat channel likely served as a sediment source and altered hydrology at downstream locations; therefore, providing one explanation of higher sediment accretion rates during the 1987–2012 time period. Previous studies have reported increased sediment accretion rates in riparian forests for time periods in which changes in land use or upstream channelization have occurred (Hupp and Morris, 1990; Hupp and Bazemore, 1993; Kleiss, 1996; Heimann and Roell, 2000; Kroes and Hupp, 2010). Upstream anthropogenic disturbances (e.g., agriculture, logging, mining, urbanization) have been shown to increase sediment yields and thus, sediment loads in disturbed watersheds (Meade et al., 1990). Re-use of this old pull boat run re-defined the channel into an adjacent backswamp. This likely altered site hydrology by increasing the flow of water into the site, resulting in at least initial increases in sediment deposition rates. This harvest disturbance also likely increased levels of light reaching the forest floor, which could have increased the density of vegetation in areas immediately adjacent to these channels (Aust et al., 1997). Vertical structure (i.e., vegetation) has been shown to increase surface roughness, which slows floodwaters and enhances deposition of suspended sediment (Hupp, 2000; Li and Yang, 2009; Ensign et al., 2014). Earlier research conducted in this harvested backswamp and surrounding areas illustrated an initial increase in herbaceous and woody vegetation following the 1986 harvest-related disturbances (Aust et al., 1997). During the initial seven years after the harvest-related disturbances, increased sediment accretion rates were directly related to ground flora biomass present (Aust et al., 1997, 2006). Similar initial increases in sediment accretion rates with increased herbaceous vegetation were reported following harvesting disturbances in the blackwater riparian forests adjacent to the Edisto River in South Carolina (Perison et al., 1997). Although we were unable to document changes in vegetation density through time in this particular study, areas immediately adjacent to our sampled locations
demonstrated the initial effects of alter vegetation dynamics that resulted in increased sediment accretion rates (Aust et al., 1997, 2006). Erosion also likely contributed to lower annual sediment accretion rates and mass trapped per unit area for the 1881–2012 time period, particularly, on the natural levee. To illustrate, mean total vertical sediment accretion for 1881–2012 (35 cm; ranged from 17 to 52 cm) was slightly lower than for the 1987–2012 time period (36 cm; ranged from 32 to 47 cm) on the natural levee. This was not the case along the backswamp; where, as anticipated, mean total deposition was greater for the 1881–2012 period (39 cm; ranged from 20 to 55 cm) than the 1987–2012 time period (30 cm; ranged from 20 to 47 cm). Natural levees are created through re-current sediment deposition adjacent to meandering black rivers in the coastal plains (Hupp, 2000). However, these levees are also in direct contact with river waters, which have the potential to erode exposed portions of the natural levee through time. Differences between gross and net sediment accretion rates were previously illustrated in wetlands which received inflow from the adjacent Olentangy River in Ohio, demonstrating the potential impact of erosion on sediment accretion estimates (Mitsch et al., 2014). Other studies that have used multiple estimation methods to determine long-term sediment accretion rates have found estimates from methods that require only one site visit (e.g., dendrogeomorphic, 137Cesium) to be more conservative than methods involving multiple, repeated measurements (Kleiss, 1996; Heimann and Roell, 2000; Kroes and Hupp, 2010). Primary reasons cited were associated with the length of time that the estimate covered. This is due to the increased likelihood of influence of erosion or compaction through time, which is only accounted for in methods which involve relatively frequent, multiple repeated measures (e.g., feldspar clay marker horizons, sediment pins, elevation surveys) (USACE, 1993). 3.2. Impacts of hydrology and climate on radial growth Tree ring chronologies depicted growth responses of green ash (1881–2011) along the natural levee and backswamp and for water tupelo (1890–2011) in the backswamp to hydrologic (days above 1.4 m and 2.1 m stage) and climatic (total precipitation and average daily temperature) variables (Table 2). Green ash ring width index had positive, significant relationships with growing season (April to August) days above the 1.4 m (r = 0.38; p = 0.04) and 2.1 m (r = 0.49; p = 0.00) stage as well as the average daily stage for the overall growing season (r = 0.41; p = 0.03) (Table 3). During the month of April, days above the 1.4 m (r = 0.43; p = 0.02) and 2.1 m (r = 0.53; p = 0.00) stage and average daily stage (r = 0.51; p = 0.00) were significantly correlated with green ash ring width index. In contrast to green ash, water tupelo ring width index was not significantly correlated with overall growing season days above the 1.4 m (r = 0.10; p = 0.61) or 2.1 m (r = 0.20; p = 0.31) stage, or growing season average daily stage (r = 0.16; p = 0.44).
Table 2 Tree ring chronology characteristics for Fraxinus pennsylvanica and Nyssa aquatica growth chronologies developed for stands adjacent to the Tensaw River in southwestern Alabama. Species
a
Distance from river (m)
Chronology span
Series lengths Range (years)
Mean (years)
Number of trees
Number of radii
Total growth rings
Series intercorrelation
Mean sensitivity
Fraxinus pennsylvanicaa
35–75
1881–2011
18–131
55
20
58
2032
0.466
0.424
Nyssa aquatica
160–330
1890–2011
54–122
80
15
30
2391
0.581
0.492
Chronology included belowground cores used in estimation of sediment accretion rates.
278
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Table 3 Pearson correlation coefficients (r) between annual ring width index and hydrologic and climatic variables for Fraxinus pennsylvanica along the natural levee and backswamp. Bolded values indicates significant correlation at a = 0.05. Bolded with asterisk (⁄) indicates p < 0.01. Variable/month Days above 1.4 m stage (backswamp flooded) Growing seasona April May June July August
a
Fraxinus pennsylvanica 0.39 0.43 0.23 0.31 0.16 0.10
Days above 2.1 m stage (bankfull) Growing seasona April May June July August
0.49⁄ 0.53⁄ 0.29 0.35 0.17 0.00
Average daily stage Growing seasona April May June July August
0.41 0.51⁄ 0.13 0.30 0.16 0.02
Total precipitation Growing seasona April May June July August
0.15 0.24⁄ 0.07 0.01 0.18 0.20
Average daily temperature Growing seasona April May June July August
0.16 0.11 0.11 0.11 0.07 0.05
April to August.
Further, no significant relationships were detected between water tupelo ring width index and days above the 1.4 m (r ranged from 0.10 to 0.19) or 2.1 m (r ranged from 0.11 to 0.25) stages or average daily stage (r ranged from 0.09 to 0.13) for any growing season month. Total precipitation and average daily temperature for the overall growing season were not significantly correlated with green ash or water tupelo ring width index. Green ash ring width index did not appear to be significantly influenced by average daily temperature for any month. However, similarly to the relationship with hydrologic variables, green ash was significantly correlated with total precipitation during the months of April (r = 0.24; p = 0.01) and August (r = 0.20; p = 0.04) (Table 3). No significant correlations were identified between water tupelo ring width index and total precipitation or average daily temperature. Lagged values (prior growing season) for days above 1.4 m and 2.1 m stage, average daily stage, total precipitation, and average daily temperature were not significantly correlated with the green ash or water tupelo ring width index. Average annual ring width for green ash in our study (1.5 mm) was similar to rates (1.6 mm yr1) along the Missouri River in North Dakota (Reily and Johnson, 1982), but lower than rates (6.3 mm yr1) along the Kankakee River in Illinois (Mitsch and Rust, 1984). Differences in green ash growth rates among studies are likely attributed to variation in site elevation, microtopography, soil texture, and flooding regime (Anderson and Mitsch, 2008). Stress induced by periodic flooding and sedimentation
disturbances in our study may have initially reduced radial growth of the green ash (Yanosky, 1982; Walls et al., 2005). Although green ash is classified as moderately flood-tolerant, it responds to periodic flooding events through formation of adventitious roots. Thus, belowground allocation of resources in response to flooding and sedimentation disturbances could have a negative effect on aboveground radial growth. Along the Olentangy River in Ohio, floodplain tree species were found to have a number of flood days at which growth was optimal; however, above this number of days, reduced growth became a tradeoff due to increased stress from the additional disturbances (Anderson and Mitsch, 2008). Green ash radial growth appeared to be most influenced by moisture-related hydrologic and climatic variables. Specifically, positive, significant correlations were identified between green ash ring width index and days above the 1.4 m and 2.1 m stage (growing season and April) and average daily stage (growing season and April). Our results agree with previously reported positive, significant correlations between green ash radial growth and mean monthly stage for April, May, and June along the White River in Arkansas (Gee, 2012). Positive relationships were also identified between 5-year green ash average annual basal area increments and average daily streamflow and days of floodplain inundation along the Kankakee River (Mitsch and Rust, 1984). Along the Wisconsin River, green ash radial growth was greater in an active floodplain than a floodwater excluded floodplain (Predick et al., 2009). Relationships between hydrologic variables and green ash ring width index were likely positive in our study due to potential moisture limitations during the growing season along natural levee and backswamp edge. Radial growth rates for water tupelo in forested wetlands have been shown to be greater under both permanently flooded (Dicke and Toliver, 1990; Keeland and Shartiz, 1995) and periodically flooded scenarios (Keeland and Shartiz, 1995; Kroes et al., 2007) than when only flooded seasonally (Dicke and Toliver, 1990). Average annual radial increase for mature water tupelo in our study (2.9 mm) was greater than annual radial growth reported for water tupelo in a continuously flooded (0.7 mm) mature (63 year-old) baldcypress-water tupelo stand in the Atchafalaya River Basin in Louisiana (Dicke and Toliver, 1990). Although in the backswamps along the Tensaw River flooding is characterized as seasonal, residual water from this season typically ponds in the lower backswamp and persists until evaporated during summer months. This coupled with a relatively high water table and influences of Squirrel Bayou (200 m to the west of the site) likely contributed to continuously hydric conditions in the backswamp increasing annual growth rates in our water tupelo. Extended hydroperiods created by these conditions have been shown to provide additional nutrients and dissolved oxygen required to achieve maximum growth in water tupelo (Keeland and Shartiz, 1995; Kroes et al., 2007). Lastly, tidal influences at our site may have resulted in relatively high growth rates in water tupelo. Annual DBH growth rates were found to be higher in tidal freshwater forested wetland (3.9 mm) than in a fluvial freshwater forested wetland (1.7 mm) along the Pocomoke River in Maryland (Kroes et al., 2007). Unexpectedly, no significant relationships were detected between hydrologic variables and water tupelo ring width index. Other studies have illustrated that water tupelo is highly dependent on flooding to transport additional nutrients and dissolved oxygen, particularly in backswamps, which may become stagnant by ponded water (Dicke and Toliver, 1990; Conner and Day, 1992; Keeland and Shartiz, 1995). Lack of significant relationships could have been influenced by defoliation by the endemic forest tent caterpillar (Malacosma disstria) previously observed on site from approximately 1954 to 1979, prior to implementation of aerial treatments (Morris, 1975; Mader, 1990). In southern
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Louisiana, where annual defoliation which has occurred since at least the 1960s, radial growth was shown to cease following defoliation until a new flush of leaves is developed (Conner and Day, 1992). During more severe years, radial growth may stop and re-start up to three times, with the last leaves persisting until the first frost, as late as early November (Dicke and Toliver, 1990; Conner and Day, 1992). In southern Alabama and Louisiana, reductions of 34% or more in radial growth rates were reported as a consequence of these annual defoliations (Morris, 1975). Defoliation by the forest tent caterpillar may have impacted growth of water tupelo in our study; therefore, making relationships between growth response and hydrologic variables vague and difficult to detect. This was the case in a study in the Atchafalaya Basin in Louisiana which failed to detect radial growth response in water tupelo following a thinning treatment due to reduced growth caused by forest tent caterpillar defoliation (Kennedy, 1983). Another reason for the lack of relationships could be related to changes in physiology as trees mature in age and increase in diameter. These changes in physiology may make mature trees more resistant to short-term alterations in environmental conditions than younger trees (Voelker, 2011; Liñán et al., 2012). 4. Conclusions The improvement of downstream water quality and aboveground biomass production are two major ecosystem services provided by riparian forested wetlands (Walbridge, 1993). This study emphasizes the important role that riparian forested wetlands play in improving water quality through trapping sediment from adjacent waterways and illustrates the potential impacts of forest harvest-related disturbances on sediment accretion rates. Further, this study identifies positive, significant relationships between green ash radial growth and hydrologic (days above 1.4 m and 2.1 m stage and average daily stage) and climatic (total precipitation) variables during the early growing season. In our study, water tupelo radial growth was less responsive to hydrologic and climatic fluctuations demonstrating potential resistance of water tupelo to perturbations. Results from this study suggest sediment accretion rates and impacts of hydrologic and climatic fluctuations on aboveground productivity should be considered when managing riparian forested wetlands. Lastly, we effectively used dendrochronological techniques to estimate sediment accretion rates and to determine impacts of hydrology and climate on tree radial growth. These techniques minimized site visits while still providing invaluable pre- and post-disturbance data. Acknowledgements This study was funded by the United States Department of Agriculture, National Needs Fellowship Grant 2010-38420-21851. Previous sediment accretion data collections were supported by the National Council for Air and Stream Improvement, Inc. We would also like to thank the Alabama Department of Conservation and Natural Resources and the Forever Wild Land Trust for granting permission to continue this long-term study. Additional field assistance was provided by John Peterson and Greg Ward. References Anderson, C.J., Mitsch, W.J., 2008. Tree basal growth response to flooding in a bottomland hardwood forest in central Ohio. J. Am. Water Resour. Assoc. 44, 1512–1520. Aust, W.M., 1989. Abiotic functional changes of a water tupelo-baldcypress wetland following disturbance by harvesting. PhD Dissertation, North Carolina State University. Aust, W.M., Lea, R., 1991. Soil temperature and organic matter in a disturbed forested wetland. Soil Sci. Soc. Am. J. 55, 1741–1746.
279
Aust, W.M., Schoenholtz, S.H., Zaebst, T.W., Szabo, B.A., 1997. Recovery status of tupelo-cypress wetland seven years after disturbance: silvicultural implications. For. Ecol. Manage. 90, 161–169. Aust, W.M., Fristoe, T.C., Gellerstedt, P.A., Giese, L.A.B., Miwa, M., 2006. Long-term effects of helicopter and ground-based skidding on site properties and stand growth in a tupelo-cypress wetland. For. Ecol. Manage. 226, 72–79. Aust, W.M., McKee, S.E., Seiler, J.R., Strahm, B.D., 2012. Long-term sediment accretion in bottomland hardwoods following timber harvest disturbances in the Mobile-Tensaw River Delta, Alabama, USA. Wetlands 32, 871–884. Boto, K.G., Patrick, W.H., 1978. Wetland functions and values: the state of our understanding. Am. Water Resour. Assoc., 479–489 Broadfoot, W.M., Willston, H.L., 1973. Flooding effects on southern forests. J. For. 9, 584–587. Conner, W.H., Day Jr., J.W., 1976. Productivity and composition of a baldcypresswater tupelo site and a bottomland hardwood site in a Louisiana swamp. Am. J. Bot. 63, 1354–1364. Conner, W.H., Day Jr., J.W., 1992. Diameter growth of Taxodium distichum (L.) Rich. and Nyssa aquatica L. from 1979 to 1985 in four Louisiana swamp stands. Am. Midl. Nat. 127, 290–299. Conner, W.H., Duberstein, J.A., Day Jr., J.W., Hutchinson, S., 2014. Impacts of changing hydrology and hurricanes on forest structure and growth along a flooding/elevation gradient in a south Louisiana forested wetland from 1986 to 2009. Wetlands 34, 803–814. Cook, E.R., 1985. A time series approach to tree ring standardization. PhD Dissertation, University of Arizona, Tuscon. Day, R.H., Williams, T.M., Swarzenski, C.M., 2007. Hydrology of tidal freshwater forested wetlands of the southeastern United States. In: Conner, W.H., Doyle, T. W., Krauss, K.W. (Eds.), Ecology of Tidal Freshwater Forested Wetlands of the Southeastern United States. Springer, New York, pp. 29–63. Dicke, S.G., Toliver, J.R., 1990. Growth and development of bald-cypress/watertupelo stands under continuous versus seasonal flooding. For. Ecol. Manage. 33– 34, 523–530. Doyle, T.W., O’Neil, C.P., Melder, M.P., From, A.S., Palta, M.M., 2007. Tidal freshwater swamps of the southeastern United States: effects of land use, hurricanes, sealevel rise, and climate change. In: Conner, W.H., Doyle, T.W., Krauss, K.W. (Eds.), Ecology of Tidal Freshwater Forested Wetlands of the Southeastern United States. Springer, New York, pp. 1–28. Ensign, S.H., Hupp, C.R., Noe, G.B., Krauss, K.W., Stagg, C.L., 2014. Sediment accretion in tidal freshwater forests and oligohaline marshes of the Waccamaw and Savannah Rivers, USA. Estuar. Coast. 37, 1107–1119. Fritts, H.C., 1976. Tree Rings and Climate. Academic Press, New York. Gee, H., 2012. The effects of hydrologic modifications on floodplain forest tree recruitment and growth in the Mississippi River Alluvial Valley, USA. PhD Dissertation, Louisiana State University. Heimann, D.C., Roell, M.J., 2000. Sediment loads and accumulation in a small riparian wetland system in northern Missouri. Wetlands 20, 219–231. Hodges, J.D., 1997. Development and ecology of bottomland hardwood sites. For. Ecol. Manage. 90, 117–125. Holmes, R.L., 1983. Computer-assisted quality control in tree-ring dating and measurement. Tree-Ring Bull. 43, 69–78. Hunter, R.G., Faulkner, S.P., Gibson, K.A., 2008. The importance of hydrology in restoration of bottomland hardwood wetland functions. Wetlands 28, 605–615. Hupp, C.R., Morris, E.E., 1990. A dendrogeomorphic approach to measurement of sedimentation in a forested wetland, Black Swamp, Arkansas. Wetlands 10, 107–124. Hupp, C.R., Bazemore, D.E., 1993. Temporal and spatial patterns of wetland sedimentation, West Tennessee. J. Hydrol. 141, 179–196. Hupp, C.R., 2000. Hydrology, geomorphology, and vegetation of coastal plain rivers in the south-eastern USA. Hydrol. Process. 14, 2991–3010. Hupp, C.R., Demas, C.R., Kroes, D.E., Day, R.H., Doyle, T.W., 2008. Recent sedimentation patterns within the central Atchafalaya Basin, Louisiana. Wetlands 28, 125–140. Johnston, C.A., Bubenzer, G.D., Lee, G.B., Madison, F.W., McHenry, J.R., 1984. Nutrient trapping by sediment deposition in a seasonally flooded lakeside wetland. J. Environ. Qual. 13, 283–290. Keeland, B.D., Shartiz, R.R., 1995. Seasonal growth patterns of Nyssa sylvatica var. biflora, Nyssa aquatica, and Taxodium distichum as affected by hydrologic regime. Can. J. For. Res. 25, 1084–1096. Keim, R.F., Amos, J.B., 2012. Dendrochronological analysis of baldcypress (Taxodium distichum) responses to climate and contrasting flood regimes. Can. J. For. Res. 42, 423–436. Kennedy, H.E., 1983. Water tupelo in the Atchafalaya Basin does not benefit from thinning. Res. Note SO-298. New Orleans, LA. U.S. Department of Agriculture, Forest Service, Southern Forest Experiment Station. Kleiss, B.A., 1996. Sediment retention in a bottomland hardwood wetland in eastern Arkansas. Wetlands 16, 321–333. Kroes, D.E., Hupp, C.R., Noe, G.B., 2007. Sediment, nutrient, and vegetation trends along the tidal, forested Pocomoke River, Maryland. In: Conner, W.H., Doyle, T. W., Krauss, K.W. (Eds.), Ecology of Tidal Freshwater Forested Wetlands of the Southeastern United States. Springer, New York, pp. 113–137. Kroes, D.E., Hupp, C.R., 2010. The effect of channelization on floodplain sediment deposition and subsidence along the Pocomoke River, Maryland. J. Am. Water Resour. Assoc. 46, 686–699. Li, H., Yang, S.L., 2009. Trapping effect of tidal marsh vegetation on suspended sediment, Yangtze Delta. J. Coast. Res. 25, 915–924.
280
K.R. Kidd et al. / Forest Ecology and Management 358 (2015) 272–280
Liñán, I.D., Gutiérrez, I.H., Andreu-Hayles, L., Muntán, E., Campelo, F., Helle, G., 2012. Age effects and climate response in trees: a multi-proxy tree-ring test in old-growth life stages. Eur. J. For. Res. 131, 933–944. Mader, S.F., 1990. Recovery of ecosystem functions and plant community structure by a tupelo-cypress wetland following timber harvesting. PhD Dissertation, North Carolina State University. Mancil, E., 1980. Pullboat logging. J. For. Hist. 24, 135–141. McKee, S.E., Aust, W.M., Seiler, J.R., Strahm, B.D., Schilling, E.B., 2012. Long-term site productivity of a tupelo-cypress swamp 24 years after harvesting disturbances. For. Ecol. Manage. 265, 172–180. Meade, R.H., Yuzyk, T.R., Day, T.J., 1990. Movement and storage of sediment in rivers of the United States and Canada. In: Wolman, M.G., Riggs, H.C., (Eds.), Surface Water Hydrology: Boulder Colorado, vol. O-1, Geological Society of America, The Geology of North America. Mitsch, W.J., Rust, W.G., 1984. Tree growth responses to flooding in a bottomland forest in northeastern Illinois. For. Sci. 30, 499–510. Mitsch, W.J., Gosselink, J.G., 2000. Wetlands, third ed. John Wiley, New York. Mitsch, W.J., Nedrich, S.M., Harter, S.K., Anderson, C., Nahlik, A.M., Bernal, B., 2014. Sedimentation in created freshwater riverine wetlands: 15 years of succession and contrasts of methods. Ecol. Eng. 72, 25–34. Morris, R.C., 1975. Tree-eaters in the tupelo swamps. For. People 25, 22–24. National Oceanic and Atmospheric Administration (NOAA), 2011. Summary of Monthly Normals, 1981–2010, National Climatic Data Center. Available online at
Ryan, J.J., Goodell, H.G., 1972. Marine geology and estuarine history of Mobile Bay, Alabama, Part 1. Contemporary sediments. Geol. Soc. Am. Mem. 133, 517–554. SAS, 2012. SAS/STAT 9.3 User’s Guide: The NPAR1WAY Procedure. SAS Institute, Inc., Cary, North Carolina. Shroder Jr., J.F., 1980. Dendrogeomorphology: review and new techniques in treering dating. Prog. Phys. Geol. 4, 161–188. Smith, W.E., 1988. Geomorphology of the mobile delta. Geol. Survey Alabama Bull. 132 (Tuscaloosa, AL). Stanturf, J.A., Goodrick, S.L., Outcalt, K.W., 2007. Disturbance and coastal forests: a strategic approach to forest management in hurricane impact zones. For. Ecol. Manage. 250, 119–135. U.S. Army Corps of Engineers (USACE), 1993. Methods of measuring sedimentation rates in bottomland hardwoods. WRP Tech. Note SD-CP-4.1. Vicksburg, MS. U.S. Army Corps of Engineers Waterways Experiment Station. USACE, 2015. Barry steam plant gage data (U.S. Geological Survey Gage #02470630), water years 1983–2012. Available online at