Integrating forest biomass and distance from channel to develop an indicator of riparian condition

Abstract

Living and detrital biomass in a riparian buffer zones ameliorate diffuse-source pollution originating from adjacent landscapes, with higher nutrient removal potential, particularly with respect to nitrogen and sediments, associated with more biomass (i.e., older forests). Pollution removal is mediated by sediment trapping and uptake of nitrogen by plants and by denitrification by microbial communities in root zones using organic matter and root exudates as energy sources. However, as a consequence of various land management activities, the amount of biomass of most forest riparian zones is much lower than its potential, which is mature forest. This makes restoration of riparian buffer zones to mature forest an ideal way to both improve water quality and increase carbon sequestration. This study measures the amount of biomass aboveground and in soils for a variety of common, age-related condition-types associated with riparian buffer zones along low order (headwater) streams in agricultural landscapes of Coastal Plain North Carolina. The data are used to determine pollution removal potential in relation to stand age and distance from channel and a basis for developing an indicator of riparian condition. Mean biomass for common riparian cover types were Mature Forest (>50 y old): 483 Mg/ha; Young Forest (25–50 y old): 257 Mg/ha; Regenerating Forest (5–25 y old): 205 Mg/ha; recently Harvested Forest (0–5 y old): 165 Mg/ha; Perennial Herb: 67 Mg/ha; Shrub/Sapling: 63 Mg/ha; and Annual Rowcrop: 36 Mg/ha. Trees contained >96% of all aboveground biomass and >58% of total biomass present in the forested conditions. Biomass in recently harvested forests were >97% detrital, mostly due to the large amount of slash left after harvesting and the mostly intact soil organic matter. Most (>80%) of the biomass in non-forest cover types was stored in the detrital pool. By partitioning the riparian zone into inverse distance-weighted subzones based on distance from stream channel, we developed an indicator for determining riparian condition based on both biomass and proximity of biomass to channel.

Introduction

The condition of vegetation adjacent to streams has been shown to exert a major effect on water quality in downstream rivers, lakes, and estuaries (Boyer et al., 1994, Garten and Ashwood, 2003, Pinckney et al., 1998, Stanley, 1988, Stanley, 1993, Stow et al., 2001). The condition of headwater riparian ecosystems is particularly important because headwaters constitute the majority of the stream length in watersheds (Brinson, 1993, Gomi et al., 2008, Rheinhardt et al., 2007b). Because most headwater reaches in agricultural landscapes lack sufficient buffers, the headwater portions of most agricultural stream networks provide a direct conduit for the introduction of most diffuse-source pollution into stream networks (Allan, 2004, Freeman et al., 2007, Naiman and Decamps, 1997), particularly nitrogen and sediment. Unfortunately, stream maps are usually too coarse to include the most-headwater reaches of stream networks (Baker et al., 2006). As a result, studies relating buffer condition to water quality have overlooked the most important portion of the stream network, and as a consequence, have likely underestimated the relationship between buffer condition and water quality at a landscape level. Further, many resource agencies have neglected to manage the buffer zone of headwater reaches, which are the places most in need of protection for maintaining or improving water quality.

Forest vegetation in near-stream buffer zones is important because it can assimilate and detain diffuse-source nutrients in surface and surficial groundwater that would otherwise be transported downstream (Dosskey et al., 2010, Jacobs and Gilliam, 1985, Peterjohn and Correll, 1984). Tree roots not only supply exudates that can be used as a carbon (energy) source for microbial denitrification, but they also supply a heterogeneous microenvironment for the denitrifying microbes (Groffman et al., 1996). In contrast, roots of herbaceous vegetation are much less suitable for providing exudates and a heterogeneous microenvironment for denitrifying bacteria, which means that forested riparian zones are much more conducive to denitrification than herbaceous-dominated ones. Detritus, in the form of leaves, snags, and large down wood, also stores nutrients and provides the labile organic carbon needed for denitrification (Tesoriero et al., 2004). Thus, organic matter in a riparian ecosystem, in both living and detrital forms, represents its capacity for assimilating nitrogen (and other nutrients) and for denitrification.

Although the difference between mature and younger forests in their capacity to remove nutrients is unknown, in general, biomass accumulation in regenerating forests should allow a net uptake of nutrients, while more mature forests would be more active in detaining nutrients through recycling mechanisms, assuming that they are already at steady state (Fisk et al., 2002, Vitousek and Reiners, 1975). Also, the root systems of older trees provide more complex microenvironments for denitrification than younger trees and although nitrogen retention is affected by site-specific conditions of soil, hydrology, and vegetation (Groffman et al., 1996, Ranalli and Macalady, 2010), the capacity to remove and recycle nutrients over sustained periods would be expected to be greater in mature forests relative to younger forests and greater in forested reaches than in reaches dominated by herbaceous vegetation. Thus, streams with natural channels and buffered by forest have the capacity to intercept and remove nutrients and sediments originating from uplands, and help protect downstream water quality (Brinson et al., 2006). This is particularly true in agricultural landscapes where excess nitrogen applied to crops commonly flows into headwater streams from adjacent cropland and where sediment often is eroded from plowed fields. Further, not only does biomass increase as a forest matures (Bradford and Kastendick, 2010), but also older forests tend to have higher structural diversity (resulting in more niches and higher species richness) (Ecke et al., 2002, Ishii et al., 2004), and a P:R ratio closer to one than younger forests and altered non-forests (Odum, 1969). Thus, habitat condition (biological integrity) also increases as forested ecosystems mature, which means that riparian zone biomass can be an indicator of both habitat condition and pollution retention potential.

Considering that problems with water quality are so pervasive nationwide, many regulatory and planning agencies have been developing riparian assessment approaches to characterize watershed condition and identify problems and possible solutions. While the emphasis of indicator development has usually focused on chemical water quality, a more recent focus has been to also develop indicators of habitat quality to characterize biological integrity, sensu Section 101 of the Clean Water Act. The strongest predictors of condition would be those that could characterize the degree to which a channel is hydrologically connected to its floodplain and the condition of the associated riparian buffer, including vegetation type and maturity (Brinson et al., 2006, Mayer et al., 2005), distance from stream channel (King et al., 2005), and spatial arrangement (Baker et al., 2006, Van Sickle and Johnson, 2008).

Numerous stream and wetland assessment procedures are in use today, each designed to meet a particular regulatory or planning objective (Stein et al., 2009) and most of these protocols incorporate biological and physical data to determine condition of stream resources (Somerville and Pruitt, 2004). Riparian assessment procedures could benefit from an approach that incorporates an indicator of the capacity of riparian zone vegetation to ameliorate nonpoint source pollution, based on biomass (a surrogate for age and condition) and distance from stream channel. Therefore, the objectives of this study were to: (1) measure and relate living and detrital biomass to forest age since clearcut and non-forest cover-types, (2) characterize the composition of forests along a successional gradient from recently clearcut to mature forest, and (3) using biomass as a surrogate for nutrient removal and habitat quality potential, develop a forest biomass/inverse-distance indicator to (a) characterize the capacity of a riparian buffer to ameliorate pollution (nutrients and sediment) from upslope areas and (b) characterize habitat quality related to physical structure.

All reaches for biomass were located within or near the Little Contentnea drainage basin, centered on Farmville, North Carolina (Fig. 1). This basin is located in the Inner Coastal Plain Physiographic Province, in the Atlantic Coast Flatwoods Major Land Resource Area (MLRA 153A) (USDA, 2006), and the Southeastern Evergreen Forest Region (Braun, 1950). The drainage basin is approximately 4800 km² with 75–85% of stream length comprising first and second order streams (Rheinhardt et al., 2007a). These low order reaches, also herein referred to as headwater reaches, begin in the flat landscape of interstream divides and in the areas between the divides and the broad bottomlands of higher order streams.

In the small agricultural watersheds of the study area, vegetation in most buffer zones have been altered, even along reaches where floodplains have not been converted to cropland (Brinson et al., 2006). In addition, headwater channels have often been deepened and straightened to accelerate the removal of water from arable cropland. Such channelization increases the slope of the water table, allowing nutrient-laden groundwater (from fertilizers applied to increase crop production) to move rapidly from cropland to streams (Spruill, 2000), thus bypassing the root zone of riparian vegetation where denitrification potential is greatest. Most higher-order streams are either deeply channelized, with levees comprised of spoil material along their channels, or if unchannelized, support wide, forested floodplains too wet to farm.

Urban headwater stream networks have been truncated with impervious surfaces, converted to underground pipes and culverts (Brinson et al., 2006, Hardison et al., 2009), and vary more widely in buffer zones cover-types than rural reaches. Because urban riparian zones are so different from rural ones, this paper will only focus on developing an indicator for rural riparian condition.