Discussion
Forest recovery following timber harvesting and wildfire often shifts mixed-species subalpine forests to pine-dominant stands (Collins et al., 2011), and in our study, this transition from old-growth to second-growth forest decreased hillslope DOM export by 89% (Tables 1 and 2, Fig 1). The amount of C contained within the O horizon determines the amount of DOM available for microbial processing and subsurface transport (Lee, Park, & Matzner, 2018; Nave, Vance, Swanston, & Curtis, 2010). As seen elsewhere we found higher O-horizon C stock, DOC concentrations and annual export in the old-growth compared to the second-growth hillslopes (Chatterjee, Vance, Pendall, & Stahl, 2008; Chatterjee, Vance, & Tinker, 2009). In spite of less C contained in mineral soils of old-growth forests, the higher C exported from them underscores the importance of subsurface transport of C from soil O horizons. Our findings suggest that land cover shifts following forest disturbance (McDowell et al., 2020) and subsequent harvesting is likely to decrease the amount of C exported to headwater streams.
The chemical character and biological reactivity of DOM inputs from organic horizons also responded to the land cover change (Supp. Info: T1, Fig 4). The transition from old- to second-growth forest induced a change from recalcitrant DOM inputs to more reactive non-humic and labile compounds such as proteins, small polyphenolics, and carbohydrates released by pine litter (Beggs & Summer, 2011; Berg & McClaugherty, 2014; Reckhow, Singer, Malcom, 1990; Yavitt & Fahey, 1984). The higher respiratory quotient (RQ, ΔCO2/ΔO2) and dissolved oxygen consumption rate of input DOM in the second-growth forest reflect this change in chemical composition and bioavailability (Fig 4) and suggest that labile proteins or carbohydrates are more readily utilized than less reactive components common to old-growth DOM (Berggren & Del Giorgio, 2012). As opposed to the thin pine needle litter of the second-growth pine stand, the deep, stratified, O horizon of the old-growth forests generate less reactive DOM (Table 1).
The forest cover changes altered the reactivity of subsurface DOM exports. Pine litter produces biologically reactive DOM (Fig 4) that is subject to microbial transformation as it infiltrates into soils and moves along subsurface flowpaths (Fig 2 and 3). The greater prevalence of microbially processed DOM we found in second-growth hillslope exports agreed with numerous other studies of DOM composition and export after post-harvest and subsequent forest change (Cawley et al., 2014; Lee & Lajtha, 2016; Williams et al., 2010; Yamashita, Kloeppel, Knoepp, Zausen, & Jaffe, 2011). In contrast, the DOM exported from old-growth forest was less biologically reactive (Fig 3b, Fig 4a), consistent with fluorescence indicators of DOM recalcitrance and complexity (Fig 3a and 3b) and respiratory quotients of litter leachate (Supp. Info: T1, Fig. 4).
We found that DOM reactivity changed in distinct ways in the two forest types as it moved along subsurface flowpaths (Fig 4). The thick O horizons in the old-growth forest released more C, though its recalcitrant, aromatic composition limited the degree of subsurface DOM processing (Fig 4). In contrast, the more reactive DOM of the second-growth forest changed substantially downslope. Both our FRI modelling and earlier work (Cordova et al., 2018; Yano, Lajtha, Sollins, & Caldwell, 2005) demonstrate the conversion of labile DOM from litter into microbial biomass and release of increasingly labile compounds downslope (Fig 2). The highly reactive nature of second-growth DOM inputs may also stimulate turnover of C stored in mineral soil (Evans, Pierson, & Lajtha, 2020; Kuzyakov, 2010) and agrees with studies that indicate that microbially processed C may contribute to vertical C movement from surface litter into the mineral soil profile (Cordova et al., 2018; Cotrufo, Wallenstein, Boot, Denef, & Paul, 2013).
The composition of DOM exported from the old-growth hillslope shifts over the course of the snowmelt season (Fig 3b and c), with a decline in relatively labile byproducts of microbial processing and an increase in recalcitrant compounds. This pattern suggests the presence of distinct DOM sources in this forest type. The DOM transported early in the snowmelt season may be comprised of labile metabolic byproducts that accumulated during overwinter microbial transformation of recalcitrant organic compounds from old-growth litter (Schmidt et al., 2007). After this initial flush, export DOM may originate from unprocessed compounds leached directly from the recalcitrant O horizon (Fig 3b). In contrast, the DOM exported from the second-growth forests was consistently labile (Fig 3c). This uniform compositional pool may result from the absence of a recalcitrant source of litter inputs and the continual turnover of C between labile litter and microbial biomass.