Divergent long-term trends and interannual variation in ecosystem resource use efficiencies of a southern boreal old black spruce forest 1999-2017

Long-term trends in ecosystem resource use efficiencies (RUEs) and their controlling factors are key pieces of information for understanding how an ecosystem responds to climate change. We used continuous eddy covariance and microclimate data over the period 1999-2017 from a 120-year-old black spruce stand in central Saskatchewan, Canada, to assess interannual variability, long-term trends, and key controlling factors of gross ecosystem production (GEP) and the RUEs of carbon (CUE = net primary production [NPP]/GEP), light (LUE = GEP/absorbed photosynthetic radiation [APAR]), and water (WUE = GEP/evapotranspiration [E]). At this site, annual GEP has shown an increasing trend over the 19 years (p < 0.01), which may be attributed to rising atmospheric CO₂ concentration. Interannual variability in GEP, aside from its increasing trend, was most strongly related to spring temperatures. Associated with the significant increase in annual GEP were relatively small changes in NPP, APAR, and E, so that annual CUE showed a decreasing trend and annual LUE and WUE showed increasing trends over the 19 years. The long-term trends in the RUEs were related to the increasing CO₂ concentration. Further analysis of detrended RUEs showed that their interannual variation was impacted most strongly by air temperature. Two-factor linear models combining CO₂ concentration and air temperature performed well (R² ~0.60) in simulating annual RUEs. LUE and WUE were positively correlated both annually and seasonally, while LUE and CUE were mostly negatively correlated. Our results showed divergent long-term trends among CUE, LUE, and WUE and highlighted the need to account for the combined effects of climatic controls and the 'CO₂ fertilization effect' on long-term variations in RUEs. Since most RUE-based models rely primarily on one resource limitation, the observed patterns of relative change among the three RUEs may have important implications for RUE-based modeling of C fluxes.

Boreal tree hydrodynamics: asynchronous, diverging, yet complementary

Water stress has been identified as a key mechanism of the contemporary increase in tree mortality rates in northwestern North America. However, a detailed analysis of boreal tree hydrodynamics and their interspecific differences is still lacking. Here we examine the hydraulic behaviour of co-occurring larch (Larix laricina) and black spruce (Picea mariana), two characteristic boreal tree species, near the southern limit of the boreal ecozone in central Canada. Sap flux density (Js), concurrently recorded stem radius fluctuations and meteorological conditions are used to quantify tree hydraulic functioning and to scrutinize tree water-use strategies. Our analysis revealed asynchrony in the diel hydrodynamics of the two species with the initial rise in Js occurring 2 h earlier in larch than in black spruce. Interspecific differences in larch and black spruce crown architecture explained the observed asynchrony in their hydraulic functioning. Furthermore, the two species exhibited diverging stomatal regulation strategies with larch and black spruce employing relatively isohydric and anisohydric behaviour, respectively. Such asynchronous and diverging tree-level hydrodynamics provide new insights into the ecosystem-level complementarity in tree form and function, with implications for understanding boreal forests' water and carbon dynamics and their resilience to environmental stress.

Estimating Forest Carbon Fluxes Using Machine Learning Techniques Based on Eddy Covariance Measurements

Approximating the complex nonlinear relationships that dominate the exchange of carbon dioxide fluxes between the biosphere and atmosphere is fundamentally important for addressing the issue of climate change. The progress of machine learning techniques has offered a number of useful tools for the scientific community aiming to gain new insights into the temporal and spatial variation of different carbon fluxes in terrestrial ecosystems. In this study, adaptive neuro-fuzzy inference system (ANFIS) and generalized regression neural network (GRNN) models were developed to predict the daily carbon fluxes in three boreal forest ecosystems based on eddy covariance (EC) measurements. Moreover, a comparison was made between the modeled values derived from these models and those of traditional artificial neural network (ANN) and support vector machine (SVM) models. These models were also compared with multiple linear regression (MLR). Several statistical indicators, including coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE), bias error (Bias) and root mean square error (RMSE) were utilized to evaluate the performance of the applied models. The results showed that the developed machine learning models were able to account for the most variance in the carbon fluxes at both daily and hourly time scales in the three stands and they consistently and substantially outperformed the MLR model for both daily and hourly carbon flux estimates. It was demonstrated that the ANFIS and ANN models provided similar estimates in the testing period with an approximate value of R2 = 0.93, NSE = 0.91, Bias = 0.11 g C m−2 day−1 and RMSE = 1.04 g C m−2 day−1 for daily gross primary productivity, 0.94, 0.82, 0.24 g C m−2 day−1 and 0.72 g C m−2 day−1 for daily ecosystem respiration, and 0.79, 0.75, 0.14 g C m−2 day−1 and 0.89 g C m−2 day−1 for daily net ecosystem exchange, and slightly outperformed the GRNN and SVM models. In practical terms, however, the newly developed models (ANFIS and GRNN) are more robust and flexible, and have less parameters needed for selection and optimization in comparison with traditional ANN and SVM models. Consequently, they can be used as valuable tools to estimate forest carbon fluxes and fill the missing carbon flux data during the long-term EC measurements.

Negative impacts of high temperatures on growth of black spruce forests intensify with the anticipated climate warming

An increasing number of studies conclude that water limitations and heat stress may hinder the capacity of black spruce (Picea mariana (Mill.) B.S.P.) trees, a dominant species of Canada's boreal forests, to grow and assimilate atmospheric carbon. However, there is currently no scientific consensus on the future of these forests over the next century in the context of widespread climate warming. The large spatial extent of black spruce forests across the Canadian boreal forest and associated variability in climate, demography, and site conditions pose challenges for projecting future climate change responses. Here we provide an evaluation of the impacts of climate warming and drying, as well as increasing [CO2 ], on the aboveground productivity of black spruce forests across Canada south of 60°N for the period 1971 to 2100. We use a new extensive network of tree-ring data obtained from Canada's National Forest Inventory, spatially explicit simulations of net primary productivity (NPP) and its drivers, and multivariate statistical modeling. We found that soil water availability is a significant driver of black spruce interannual variability in productivity across broad areas of the western to eastern Canadian boreal forest. Interannual variability in productivity was also found to be driven by autotrophic respiration in the warmest regions. In most regions, the impacts of soil water availability and respiration on interannual variability in productivity occurred during the phase of carbohydrate accumulation the year preceding tree-ring formation. Results from projections suggest an increase in the importance of soil water availability and respiration as limiting factors on NPP over the next century due to warming, but this response may vary to the extent that other factors such as carbon dioxide fertilization, and respiration acclimation to high temperature, contribute to dampening these limitations.

Autumn warming reduces the CO2 sink of a black spruce forest in interior Alaska based on a nine-year eddy covariance measurement

Nine years (2003-2011) of carbon dioxide (CO2) flux were measured at a black spruce forest in interior Alaska using the eddy covariance method. Seasonal and interannual variations in the gross primary productivity (GPP) and ecosystem respiration (RE) were associated primarily with air temperature: warmer conditions enhanced GPP and RE. Meanwhile, interannual variation in annual CO2 balance was controlled predominantly by RE, and not GPP. During these 9 years of measurement, the annual CO2 balance shifted from a CO2 sink to a CO2 source, with a 9-year average near zero. The increase in autumn RE was associated with autumn warming and was mostly attributed to a shift in the annual CO2 balance. The increase in autumn air temperature (0.22 °C yr(-1)) during the 9 years of study was 15 times greater than the long-term warming trend between 1905 and 2011 (0.015 °C yr(-1)) due to decadal climate oscillation. This result indicates that most of the shifts in observed CO2 fluxes were associated with decadal climate variability. Because the natural climate varies in a cycle of 10-30 years, a long-term study covering at least one full cycle of decadal climate oscillation is important to quantify the CO2 balance and its interaction with the climate.

Changes in net ecosystem productivity of boreal black spruce stands in response to changes in temperature at diurnal and seasonal time scales

Net ecosystem productivity (NEP) of boreal coniferous forests is believed to rise with climate warming, thereby offsetting some of the rise in atmospheric CO(2) concentration (C(a)) by which warming is caused. However, the response of conifer NEP to warming may vary seasonally, with rises in spring and declines in summer. To gain more insight into this response, we compared changes in CO(2) exchange measured by eddy covariance and simulated by the ecosystem process model ecosys under rising mean annual air temperatures (T(a)) during 2004-2006 at black spruce stands in Saskatchewan, Manitoba and Quebec. Hourly net CO(2) uptake was found to rise with warming at T(a) < 15 degrees C and to decline with warming at T(a) > 20 degrees C. As mean annual T(a) rose from 2004 to 2006, increases in net CO(2) uptake with warming at lower T(a) were greater than declines with warming at higher T(a) so that annual gross primary productivity and hence NEP increased. Increases in net CO(2) uptake measured at lower T(a) were explained in the model by earlier recovery of photosynthetic capacity in spring, and by increases in carboxylation activity, using parameters for the Arrhenius temperature functions of key carboxylation processes derived from independent experiments. Declines in net CO(2) uptake measured at higher T(a) were explained in the model by sharp declines in mid-afternoon canopy stomatal conductance (g(c)) under higher vapor pressure deficits (D). These declines were modeled from a hydraulic constraint to water uptake imposed by low axial conductivity of conifer roots and boles that forced declines in canopy water potential (psi(c)), and hence in g(c) under higher D when equilibrating water uptake with transpiration. In a model sensitivity study, the contrasting responses of net CO(2) uptake to specified rises in T(a) caused annual NEP of black spruce in the model to rise with increases in T(a) of up to 6 degrees C, but to decline with further increases at mid-continental sites with lower precipitation. However, these contrasting responses to warming also indicate that rises in NEP with climate warming would depend on the seasonality (spring versus summer) as well as the magnitude of rises in T(a).