Introduction
Dissolved organic matter (DOM) derived from vegetation and soils accounts for up to 90% of the carbon (C) found in surface water (Aitkenhead-Peterson, McDowell, & Neff, 2003) and DOM character regulates microbial C processing in aquatic environments (Benner, 2003; D’Andrilli, Junker, Smith, Scholl, & Foreman, 2019; Lennon & Pfaff, 2005). While encompassing less than 1% of the global land surface, inland waters regulate carbon cycling from landscape to continental scales (Battin et al., 2009; Cavallaro et al., 2018; Cole et al., 2007; Raymond et al., 2013), and headwater streams drain 70% of North American land area (Colvin et al., 2019). Changes to forest land cover that affect terrestrial DOM character will influence aquatic ecosystem metabolism (Cawley, Yamashita, Maie, & Jaffe, 2014; Lajtha & Jones, 2018; Williams, Yamashita, Wilson, Jaffe, & Xenopoulos, 2010), and deciphering these relationships are crucial for predicting C cycling in inland waters.
Vegetation responses to land management can alter watershed hydrology and biogeochemistry for decades to centuries (Chantingy, 2003; Lee & Lajtha, 2016; Rhoades, Hubbard, & Elder, 2017; Troendle & King, 1985; Wilm & Dunford, 1948). In the mountain ranges of the American West, where the short growing season and cold, dry climate limit forest regrowth, the consequences of timber harvesting on snowpack, streamflow, and nutrient export as well as on potential wildfire behavior and susceptibility to insect outbreaks are long-lasting (Stottlemyer & Troendle, 1999; Rhoades et al., 2017; Wilm & Dunford, 1948). Timber harvesting decreases canopy interception and evapotranspiration resulting in more snowpack water storage, and increased soil moisture and streamflow (Troendle & King, 1985) with effects that remain measurable for more than 50 years. Timber harvesting also shifts the species composition of these forests and their development trajectory for a more than a century (Collins, Rhoades, Hubbard, & Battaglia, 2011; Lotan & Perry, 1983).
Harvest of subalpine forests comprised of subalpine fir (Abies lasiocarpa ), Engelmann spruce (Picea engelmannii ) and lodgepole pine (Pinus contorta ) typically regenerate into lodgepole pine-dominated stands (Collins et al., 2011). While the effects of land cover, age, and species change on terrestrial nutrient cycling and organic matter decomposition are well studied (Chantigny, 2003 for a review), their consequences on adjacent aquatic ecosystem C cycling are not. The chemical characteristics of litter in regenerating forests alter hillslope DOM inputs relative to those in old-growth forest (Beggs & Summers, 2011; D’Andrilli et al., 2019), likely regulating the reactivity of DOM exported to streams at watershed scales. The highly polyphenolic and protein-like DOM released from lodgepole pine litter in regenerating stands, for example, is more reactive than that generated by old-growth, mixed-conifer forests (Beggs & Summers, 2011; Yavitt & Fahey, 1984 & 1986). Connecting the effects of land cover type on the reactivity of DOM inputs from the forest floor to the reactivity of hillslope DOM exports will increase understanding of the effects of land cover change on watershed-scale C cycling.
The consequences of land cover and tree species shifts on watershed-scale DOM dynamics benefit from techniques that differentiate DOM reactivity of hillslope inputs and exports. Analysis of fluorescing components of DOM (FDOM) has become a relatively routine, low-cost approach to characterize the relative abundance of biologically reactive organic molecules (Smith et al. 2018). Correlation between fluorescing and non-fluorescing DOM molecules of similar chemical structures and biological reactivities expands the utility of FDOM as an index of overall DOM quality (Stubbins et al., 2014). Optical characterization can be used in conjunction with heterotrophic microbial oxygen consumption assays to identify which DOM components are microbially created or utilized (D’Andrilli et al., 2019).
In this study we compare DOM inputs from litter leachate and the concentration, character and biological reactivity of DOM exports in subsurface flow between adjacent old-growth subalpine and second-growth pine forests. Early work at these sites documented that clear cut harvesting resulted in lasting increases in snow accumulation, subsurface discharge, and nutrient export (Reuss, Stottlemyer, & Troendle, 1997; Starr, 2004; Troendle & Reuss, 1997). The amount of DOC exported from forests is determined by O horizon C content, with old-growth forests exporting more DOC compared to regenerating ones (Cawley et al., 2014; Lajtha & Jones, 2018; Pacific, Jensco, & McGlynn, 2010). However, DOC reactivity is affected by C character and processing within subsurface flow paths (Lehmann & Kleber, 2015). Because old-growth forests generally have deeper O-horizons composed of more recalcitrant DOM (Jandle et al., 2007; Johnson, Johnson, Huntington, & Siccama, 1991), we expect greater flux but lower biological reactivity of DOC from these forests. Conversely, we hypothesize that DOM exported from pine-dominated, second-growth forest to contain proportionally more biologically reactive DOM typical of the water-soluble proteins and polyphenolics found in pine needle litter (Beggs & Summer, 2011). Western conifer forests often require a century to regrow to their pre-harvest stand structure (Burns & Honkala, 1990), so the extent of old-growth forest cover has decreased during more than a century of timber harvesting and most forests are in intermediate stages of recovery (Anderson‐Teixeira et al., 2013; Hurtt et al., 2011; McDowell et al., 2020). This study evaluates how forest change following clear-cut harvesting alters the reactivity of DOM from the landscape to aquatic ecosystems.