Nine years of in situ soil warming and topography impact the temperature sensitivity and basal respiration rate of the forest floor in a Canadian boreal forest

Abstract

The forest floor of boreal forest stores large amounts of organic C that may react to a warming climate and increased N deposition. It is therefore crucial to assess the impact of these factors on the temperature sensitivity of this C pool to help predict future soil CO2 emissions from boreal forest soils to the atmosphere. In this study, soil warming (+2-4°C) and canopy N addition (CNA; +0.30-0.35 kg·N·ha-1·yr-1) were replicated along a topographic gradient (upper, back and lower slope) in a boreal forest in Quebec, Canada. After nine years of treatment, the forest floor was collected in each plot, and its organic C composition was characterized through solid-state 13C nuclear magnetic resonance (NMR) spectroscopy. Forest floor samples were incubated at four temperatures (16, 24, 32 and 40°C) and respiration rates (RR) measured to assess the temperature sensitivity of forest floor RR (Q10 = e10k) and basal RR (B). Both soil warming and CNA had no significant effect on forest floor chemistry (e.g., C, N, Ca and Mg content, amount of soil organic matter, pH, chemical functional groups). The NMR analyses did not show evidence of significant changes in the forest floor organic C quality. Nonetheless, a significant effect of soil warming on both the Q10 of RR and B was observed. On average, B was 72% lower and Q10 45% higher in the warmed, versus the control plots. This result implies that forest floor respiration will more strongly react to changes in soil temperature in a future warmer climate. CNA had no significant effect on the measured soil and respiration parameters, and no interaction effects with warming. In contrast, slope position had a significant effect on forest floor organic C quality. Upper slope plots had higher soil alkyl C:O-alkyl C ratios and lower B values than those in the lower slope, across all different treatments. This result likely resulted from a relative decrease in the labile C fraction in the upper slope, characterized by lower moisture levels. Our results point towards higher temperature sensitivity of RR under warmer conditions, accompanied by an overall down-regulation of RR at low temperatures (lower B). Since soil C quantity and quality were unaffected by the nine years of warming, the observed patterns could result from microbial adaptations to warming.

Fig 1. Schematic representation of the field experimental design at the Simoncouche research station, Quebec, Canada.

Squares represent 7.5 ×7.5 m experimental plots, and circles show the location of soil samples outside the experimental plots as controls. N+: canopy N addition (CNA); W+: soil warming; W+N+: combined CNA and soil warming; C: control (no CNA nor soil warming). The four types of experimental plots were replicated three times in three slope positions: upper slope, back slope and lower slope.

Fig 2. Relationship between forest floor mean respiration rate (Forest floor RR, μg C-CO₂·g^-1·C·h^-1) and incubation temperatures (°C).

Relationships are shown for the four experimental treatments (C: control; N+: CNA; W+: soil warming; W+N+: combined soil warming and CNA) with (red lines) or without (black lines) RR values obtained from incubations at 40°C. The rows and the columns show the treatments and replicates for each treatment, respectively. Curves were obtained by fitting a first-order exponential equation (RR = Be^k.T).

Fig 3. Relationship between Q₁₀ and natural log(B).

Both parameters were calculated from the relationship between forest floor respiration rate and incubation temperatures for each treatment (3 replicates × 4 treatments, n = 12) and for samples collected outside of experiment plots (i.e., controls, n = 12). Unit for parameter B is μg C-CO_2·g^-1 C·h^-1. *** indicates P<0.001.

Fig 4. Impact of soil warming and CNA on the temperature sensitivity and the basal rate of forest floor respiration.

(A) Mean (± SD) Q₁₀ and (B) mean (± SD) B values in “warmed” and “unwarmed” experimental plots with no canopy N addition (CNA-) or with canopy N addition (CNA+). Different letters indicate a significant effect of soil warming on Q₁₀ and B values (ANOVA with warming and CNA treatments as independent variables; P < 0.05).

Fig 5. Principal component analysis (PCA).

Projection of eight soil variables (Q₁₀, B, C:N and alkyl:O-alkyl C ratios, and percentages of alkyl, O-alkyl, carboxyl and aromatic compounds), and sample scores for inside and outside experimental plots (n = 24) along the two first axes of a PCA. Red, black and blue symbols show upper, back and lower slope positions, respectively.

Fig 6. Effect of the slope position on forest floor organic C quality.

(A) Forest floor alkyl:O-alkyl C ratio (mean ± SD; n = 24) and (B) B parameter (mean ± SD; n = 24) in the upper, back, and lower slope positions. Values not sharing the same letters are significantly different (ANOVA followed by Tukey’s HSD test; P < 0.05).