Search

Made available in DSpace on 2020-12-12T01:33:29Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-11-01 Eucalyptus plantation productivity has increased 4-fold across Brazil as a result of improved genetic selection and intensive silviculture. However, the rate of productivity improvements is slowing and suggests increased production will require planting in areas that have never been previously planted with Eucalyptus with uncertain impacts on local water supplies. This study compares water use and water use efficiency between a drought resistant (E. grandis × E. camaldulensis) and high productivity Eucalyptus clone (E. urophylla) across three sites (dry, mesic and wet) spanning a two-fold precipitation gradient (~600–1350 mm y−1) in Brazil. Each site included a 30% reduced rainfall treatment for each clone to investigate within site response to reduced precipitation. We found that the drought resistant clone used as much or more water than the high productivity clone at all three sites. The drought resistant clone transpired almost 20% percent more water than the high productivity clone at the dry site (3.7 vs. 3.0 m3 y−1) while water use between clones did not differ at the mesic (4.6 vs 4.0 m3 y−1) and wet site (4.0 vs 4.5 m3 y−1). The high productivity clone transpired 54, 46 and 37% of annual precipitation at the dry, mesic and wet sites respectively while the drought resistant clone consumed 67, 47 and 31% of annual precipitation from the dry to the wet site. Although the rain reduction treatments did not impact annual water use for either clone at the dry and mesic sites, there were some differences at a subannual timescale. Rain reduction treatments at the dry site lowered transpiration in the driest, hottest three months for the high productivity clone by 25–32%. Rainfall reduction led to lower water use by the drought resistant clone for only one month during the dry season (18%). The high productivity clone transpired less water (24–26%) in the rainfall reduction treatment at the mesic site during three months and the drought resistant clone transpired 18 and 20% less water in the rainfall reduction treatment during two months. Rainfall reduction reduced annual transpiration by 59% for the high productivity clone at the wet site. Average water use efficiency of the drought resistant clone was almost 70% lower than the high productivity clone at the dry and mesic sites (~5.2 vs. 1.5 kg m−3 y−1) and was due to the combination of low growth and high water use. Higher than expected growth by the drought resistant clone at the wet site resulted in similar water use efficiency between clones (6.3 vs. 5.8 kg m−3 y−1). We discuss potential implications of our findings and highlight several lines of evidence that suggest the drought resistant clone in this study allocates a greater proportion of gross primary productivity below ground allowing it to access deeper reserves of soil water during dry periods. Our results suggest that decisions to plant drought resistant clones in drought prone areas should be carefully considered and also highlight the need for more research on water use and drought tolerance strategies for Eucalyptus clones bred for more arid areas in Brazil. USDA Forest Service Rocky Mountain Research Station, 240 West Prospect Road, Fort Collins Forestry Science and Research Institute (IPEF) Department of Forest Sciences University of Lavras (UFLA), Lavras Department of Forest Science São Paulo State University – UNESP Department of Forest Research and Development Klabin SA, Telêmaco Borba Department of Forest Research and Development International Paper, Mogi Guaçu Department of Forest Science São Paulo State University – UNESP

Made available in DSpace on 2020-12-12T01:32:48Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-11-01 Clonal plantations of Eucalyptus are among the most productive forests in the world, with intensification of silviculture and genetic breeding doubling the wood mean annual increments over the past four decades. The TECHS Project demonstrated that even with intensive silviculture, wood production varies by more than two-fold across environmental gradients, and growth of highly selected clones differs by more than two-fold within a site. Wood production accounts for less than half of the photosynthesis of a forest, and we tested two hypotheses about the relation between wood production and the entire carbon balance for five genotypes across four of the TECHS sites, varying in temperature and water availability. We hypothesized that the influence of the environment on carbon fluxes and partitioning related to gross primary production would be consistent across genotypes. We also hypothesized that carbon flux and partitioning would be more sensitive to water stress than temperature. Annual average temperatures ranged from 18 to 27 °C, and annual rainfall ranged from about 600 to 1500 mm yr−1. Water stress was further tested by reduction in rainfall within sites using troughs to capture about 30% of incoming rain. The geographic gradient led to a six-fold range in wood net primary production during the two years of measurement (from age 1.5 to 3.5 years, the period of maximum current annual increment). Gross primary production (GPP) differed only by two-fold, highlighting very large differences among sites in partitioning: wood net primary production (NPP) accounted for 44% of GPP on sites with higher GPP, and only 34% of GPP on lower GPP sites. The average differences for wood NPP among clones was also large, with about half of the differences among clones relating to differences in GPP, and half to differences in the partitioning to wood NPP. The clones showed similar partitioning patterns across sites, supporting our first hypothesis. Differences across sites and clones in partitioning of GPP to wood NPP related inversely to belowground allocation. Belowground partitioning of carbon increased with increasing temperature and increasing water stress. Our second hypothesis was rejected, as patterns across sites related somewhat more strongly to temperature than to water stress. Overall, this ecophysiological investigation in the TECHS Project underscored the importance of understanding how carbon budgets differ across sites (even with intensive silviculture), and why clones can largely differ in wood production. Department of Forest Sciences Federal University of Lavras (UFLA) Department of Forest Soils and Environmental Sciences São Paulo State University (UNESP) Forestry Science and Research Institute (IPEF) School of Forestry Northern Arizona University Natural Resource Ecology Laboratory Colorado State University USDA Forest Service Rocky Mountain Research Station Klabin Telêmaco Borba International Paper Suzano Department of Forest Soils and Environmental Sciences São Paulo State University (UNESP)

Made available in DSpace on 2020-12-12T01:33:53Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-11-01 International Paper Cross-site patterns for multiple sites tend to be more broadly applicable and more useful for constructing and constraining models. We examined cross-site patterns of Eucalyptus plantation response to water supply (including irrigation and 1/3 precipitation removal), mean annual temperature (MAT), vapor pressure deficit during the daytime (VPD), and nutrient addition in eight sites from the Brazil Eucalyptus Productivity Project (BEPP). Mean annual increment (MAI) for all treatments and sites varied from 12.7 to 37.3 Mg ha−1 yr−1 across a 1400 km latitudinal gradient where annual precipitation varied from 940 to 1430 mm yr−1. MAI varied with water added across all sites, with a near linear increase of 1.42 Mg ha−1 yr−1 per 100 mm yr−1 between 640 and 1800 mm yr−1, a MAI plateau of 30.8 Mg ha−1 yr−1 at ~2000 mm yr−1, then decreasing to MAI of 27.2 Mg ha−1 yr−1 at 3060 mm yr−1. Including MAT with the water response showed that MAI decreased at 1.23 Mg ha−1 yr−1 per °C. For the natural precipitation treatments, MAI decreased at 2.52 Mg ha−1 yr−1 per °C, comparable to that for the much broader TECHS Project (Binkley et al., 2020). MAI was not related to temperature for the irrigation treatments, suggesting that some or all of the response of MAI to temperature is related to water supply. The difference between the biomass of the irrigation treatments and that of natural precipitation treatment slowed its increase or stopped increasing at age 30–60 months for five of the eight sites, but increased through the study duration for the other three sites. The difference in biomass between the fertilization and no fertilization treatments increased throughout the study for two sites. Three- and six-month growth was generally unrelated prior 3–12 month weather. Age-related decline was observed for all of the treatments at four of the eight sites. These cross-site comparisons affirm that water supply is the key resource determining levels of plantation productivity in Brazil and that individual site studies are inadequate for understanding many key responses. Department of Ecosystem Science & Sustainability Colorado State University USDA Forest Service Rocky Mountain Research Station Department of Forest Science São Paulo State University – UNESP Forestry Science and Research Institute (IPEF) School of Forestry Northern Arizona University Department of Forest Science São Paulo State University – UNESP

International audience; Leaf mass per area (LMA) and leaf equivalent water thickness (EWT) are key leaf functional traits providing information for many applications including ecosystem functioning modeling and fire risk management. In this paper, we investigate two common conclusions generally made for LMA and EWT estimation based on leaf optical properties in the near-infrared (NIR) and shortwave infrared (SWIR) domains: (1) physically-based approaches estimate EWT accurately and LMA poorly, while (2) statistically-based and machine learning (ML) methods provide accurate estimates of both LMA and EWT. Using six experimental datasets including broadleaf species samples of >150 species collected over tropical, temperate and boreal ecosystems, we compared the performances of a physically-based method (PROSPECT model inversion) and a ML algorithm (support vector machine regression, SVM) to infer EWT and LMA based on leaf reflectance and transmittance. We assessed several merit functions to invert PROSPECT based on iterative optimization and investigated the spectral domain to be used for optimal estimation of LMA and EWT. We also tested several strategies to select the training samples used by the SVM, in order to investigate the generalization ability of the derived regression models. We evidenced that using spectral information from 1700 to 2400 nm leads to strong improvement in the estimation of EWT and LMA when performing a PROSPECT inversion, decreasing the LMA and EWT estimation errors by 55% and 33%, respectively. The comparison of various sampling strategies for the training set used with SVM suggests that regression models show limited generalization ability, particularly when the regression model is applied on data fully independent from the training set. Finally, our results demonstrate that, when using an appropriate spectral domain, the PROSPECT inversion outperforms SVM trained with experimental data for the estimation of EWT and LMA. Thus we recommend that estimation of LMA and EWT based on leaf optical properties should be physically-based using inversion of reflectance and transmittance measurements on the 1700 to 2400 nm spectral range.

Made available in DSpace on 2020-12-12T01:16:20Z (GMT). No. of bitstreams: 0 Previous issue date: 2020-04-15 Colorado State University North Carolina State University Universidade Federal de Lavras Universidade Federal do Rio Grande do Norte U.S. Forest Service Universidade de São Paulo The TECHS project spanned a 3500 km gradient from the Amazon to Uruguay, examining the influence of stresses from temperature and water supply on clonal plantations of Eucalyptus, with and without rain reduction, and across a stocking gradient. The whole-rotation mean annual increment (MAI) showed a humped pattern in relation to temperature, rising from about 18 Mg ha−1 yr−1 of stemwood production when mean annual temperatures were near 16 °C, to 27 Mg ha−1 yr−1 at 20 °C, and then falling to less than 15 Mg ha−1 yr−1 above 24 °C. The age trend in growth showed a steeper initial rise in the warmer tropical sites (reaching a peak current annual increment, CAI, of 27 Mg ha−1 yr−1, at age 2–3 years), but the slower early growth in the cooler subtropical sites had a higher peak (CAI of 32 Mg ha−1 yr−1, at 4 years) and slower decline, giving 15% higher MAI for the cooler region. Whole-rotation MAI declined by about 2.2 Mg ha−1 yr−1 for each 1 °C increase in temperature (in the range between 19.5 and 23.5 °C), and MAI declined by 0.5 Mg ha−1 yr−1 for each 100 mm yr−1 decline in rain. The effect of reducing ambient rain was also a loss of 0.5 Mg ha−1 yr−1 for each 100 mm yr−1 reduction in rain, though the effect was small on low productivity sites (

Made available in DSpace on 2018-12-11T16:41:14Z (GMT). No. of bitstreams: 0 Previous issue date: 2015-08-25 Understanding the consequences of changes in climatic and biological drivers on tree carbon and water fluxes is essential in forestry. Using a metamodeling approach, sensitivity and uncertainty analyses were carried out for a tree-scale model (MAESPA) to isolate the effects of climate, morphological and physiological traits, and intertree competition on the absorption of photosynthetically active radiation (APAR), gross primary production (GPP), transpiration (TR), light use efficiency (LUE), and water use efficiency (WUE) in clonal Eucalyptus plantations. The metamodel predicting daily TR was validated using one year of sap flow measurements and showed close agreement with the measurements (mean percentage error = 11%, n = 2155). Simulations showed that APAR, GPP, and TR were very sensitive to the tree morphology and to a competition index representing its local environment. LUE and WUE were, in addition, very sensitive to the natural variability of the physiological leaf and root parameters. A maximum percentage error of 10% in these parameters leads to 18%, 17%, 16%, 9%, and 18% uncertainty for APAR, GPP, TR, LUE, and WUE, respectively. The uncertainties in TR were highest for the smallest trees. This study highlighted the need to take account of the spatial and temporal variability of tree traits and environmental conditions for simulations at the tree scale. UMR Eco and Sols CIRAD, 2 place Viala SupAgro Montpellier, 2 place Viala Departamento de Ciencias Atmosfericas Universidade de São Paulo ESALQ Universidade de São Paulo Forest Science Department UNESP Department of Forestry and Environmental Resources North Carolina State University Forestry Science and Research Institute-IPEF Forest Science Department UNESP