Characterizing Livestock Production in Portuguese Sown Rainfed Grasslands: Applying the Inverse Approach to a Process-Based Model

Spatiotemporal Patterns of Pasture Quality Based on NDVI Time-Series in Mediterranean Montado Ecosystem

The evolution of dryland pasture quality is closely related to the seasonal and inter-annual variability characteristic of the Mediterranean climate. This variability introduces great unpredictability in the dynamic management of animal grazing. The aim of this study is to evaluate the potential of two complementary tools (satellite images, Sentinel-2 and proximal optical sensor, OptRx) for the calculation of the normalized difference vegetation index (NDVI), to monitor in a timely manner indicators of pasture quality (moisture content, crude protein, and neutral detergent fiber). In two consecutive years (2018/2019 and 2019/2020) these tools were evaluated in six fields representative of dryland pastures in the Alentejo region, in Portugal. The results show a significant correlation between pasture quality degradation index (PQDI) and NDVI measured by remote sensing (R2 = 0.82) and measured by proximal optical sensor (R2 = 0.83). These technological tools can potentially make an important contribution to decision making and to the management of livestock production. The complementarity of these two approaches makes it possible to overcome the limitations of satellite images that result (i) from the interference of clouds (which occurs frequently throughout the pasture vegetative cycle) and (ii) from the interference of tree canopy, an important layer of the Montado ecosystem. This work opens perspectives to explore new solutions in the field of Precision Agriculture technologies based on spectral reflectance to respond to the challenges of economic and environmental sustainability of extensive livestock production systems.

Global process-based characterization factors of soil carbon depletion for life cycle impact assessment

Regionalization of land use (LU) impact in life cycle assessment (LCA) has gained relevance in recent years. Most regionalized models are statistical, using highly aggregated spatial units and LU classes (e.g. one unique LU class for cropland). Process-based modelling is a powerful characterization tool but so far has never been applied globally for all LU classes. Here, we propose a new set of spatially detailed characterization factors (CFs) for soil organic carbon (SOC) depletion. We used SOC dynamic curves and attainable SOC stocks from a process-based model for more than 17,000 world regions and 81 LU classes. Those classes include 63 agricultural (depending on 4 types of management/production), and 16 forest sub-classes, and 1 grassland and 1 urban class. We matched the CFs to LU elementary flows used by LCA databases at country-level. Results show that CFs are highly dependent on the LU sub-class and management practices. For example, transformation into cropland in general leads to the highest SOC depletion but SOC gains are possible with specific crops.

Object-Based Classification Approaches for Multitemporal Identification and Monitoring of Pastures in Agroforestry Regions using Multispectral Unmanned Aerial Vehicle Products

Sown Biodiverse Pastures (SBP) are the basis of a high-yield grazing system tailored for Mediterranean ecosystems and widely implemented in Southern Portugal. The application of precision farming methods in SBP requires cost-effective monitoring using remote sensing (RS). The main hurdle for the remote monitoring of SBP is the fact that the bulk of the pastures are installed in open Montado agroforestry systems. Sparsely distributed trees cast shadows that hinder the identification of the underlaying pasture using Unmanned Aerial Vehicles (UAV) imagery. Image acquisition in the Spring is made difficult by the presence of flowers that mislead the classification algorithms. Here, we tested multiple procedures for the geographical, object-based image classification (GEOBIA) of SBP, aiming to reduce the effects of tree shadows and flowers in open Montado systems. We used remotely sensed data acquired between November 2017 and May 2018 in three Portuguese farms. We used three machine learning supervised classification algorithms: Random Forests (RF), Support Vector Machine (SVM) and Artificial Neural Networks (ANN). We classified SBP based on: (1) a single-period image for the maximum Normalized Difference Vegetation Index (NDVI) epoch in each of the three farms, and (2) multi-temporal image stacking. RF, SVM and ANN were trained using some visible (red, green and blue bands) and near-infrared (NIR) reflectance bands, plus NDVI and a Digital Surface Model (DSM). We obtained high overall accuracy and kappa index (higher than 79% and 0.60, respectively). The RF algorithm had the highest overall accuracy (more than 92%) for all farms. Multitemporal image classification increased the accuracy of the algorithms. as it helped to correctly identify as SBP the areas covered by tree shadows and flower patches, which would be misclassified using single image classification. This study thus established the first workflow for SBP monitoring based on remotely sensed data, suggesting an operational approach for SBP identification. The workflow can be applied to other types of pastures in agroforestry regions to reduce the effects of shadows and flowering in classification problems.

Detailed global modelling of soil organic carbon in cropland, grassland and forest soils

Assessments of the global carbon (C) cycle typically rely on simplified models which consider large areas as homogeneous in terms of the response of soils to land use or consider very broad land classes. For example, "cropland" is typically modelled as an aggregation of distinct practices and individual crops over large regions. Here, we use the process-based Rothamsted soil Carbon Model (RothC model), which has a history of being successfully applied at a global scale, to calculate attainable SOC stocks and C mineralization rates for each of c. 17,000 regions (combination of soil type and texture, climate type, initial land use and country) in the World, under near-past climate conditions. We considered 28 individual crops and, for each, multiple production practices, plus 16 forest types and 1 grassland class (total of 80 classes). We find that conversion to cropland can result in SOC increases, particularly when the soil remains covered with crop residues (an average gain of 12 t C/ha) or using irrigation (4 t C/ha), which are mutually reinforcing effects. Attainable SOC stocks vary significantly depending on the land use class, particularly for cropland. Common aggregations in global modelling of a single agricultural class would be inaccurate representations of these results. Attainable SOC stocks obtained here were compared to long-term experiment data and are well aligned with the literature. Our results provide a regional and detailed understanding of C sequestration that will also enable better greenhouse gas reporting at national level as alternatives to IPCC tier 2 defaults.

“BalSim”: A Carbon, Nitrogen and Greenhouse Gas Mass Balance Model for Pastures

Animal production systems are increasingly required to co-produce meat products and other ecosystem services. Sown biodiverse pastures (SBP) were developed in Portugal as an improvement over semi-natural pastures (SNP). SBP increase yields and animal intake during grazing, are substantial carbon sinks, and the abundance of legumes in the mixtures provides plants with a biological source of nitrogen. However, the data available and the data demands of most models make integrated modelling of these effects difficult. Here, we developed “BalSim”, a mass balance approach for the estimation of carbon and nitrogen flows and the direct greenhouse gas (GHG) balance of the two production systems. Results show that, on average, the on-farm GHG balance is −2.6 and 0.8 t CO2e/ha.yr for SBP and SNP, respectively. Ignoring the effects of carbon sequestration, and taking into account only non-CO2 emissions, the systems are responsible for 17.0 and 16.3 kg CO2e/kg live weight.yr. The annual analysis showed that non-CO2 emissions were highest in a drought year due to decreased yield and stocking rate. We also showed through scenario analysis that matching the grazing level to the yield is crucial to minimize emissions and ensure reduced feed supplementation while maintaining high soil carbon stocks.

The Effects on Greenhouse Gas Emissions of Ecological Intensification of Meat Production with Rainfed Sown Biodiverse Pastures

Feed production is an important contributor to the environmental impacts caused by livestock production. In Portugal, non-dairy cattle are commonly fed with a mixture of grazing and forages/concentrate feed. Sown biodiverse permanent pastures rich in legumes (SBP) were introduced to provide quality animal feed and offset concentrate consumption. SBP also sequester large amounts of carbon in soils. Here, we used a comparative life cycle assessment approach to test the substitution of concentrate through installation of high-yield SBP. Using field data for the Alentejo region in Portugal, we compare the global warming potential of a baseline scenario where cattle is fed in low-yield, semi-natural pastures supplemented with feeds that vary in the ratio of silage to concentrate, and a second scenario where the feed is substituted with high-yield SBP. Although SBP use more fertilizers and machinery, this replacement avoids the emission of about 3 t CO2eq/ha even after SBP stop sequestering carbon. Using crude fiber to establish the equivalence between scenarios leads to higher avoided impact, owing to the fact that the fiber content of SBP is also higher. SBP can avoid 25% emissions from beef production per kg of live animal weight.

Carbon Footprint of Milk from Pasture-Based Dairy Farms in Azores, Portugal

The dairy sector is a major contributor to greenhouse gas emissions. Pasture-based dairy production is sometimes credited as environmentally friendlier but is less studied than more intensive production systems. Here we characterize and calculate the carbon footprint (CF), using life cycle assessment, of the “Vacas Felizes” pasture-based milk production system, in the Azores archipelago. Impacts were determined for multiple functional units: mass, energy and nutritional content, farm, area and animal. We performed multivariate analysis to assess the contribution of production parameters to the CF. Finally, we performed a literature review to compare these results with other production systems. Results show that emissions from enteric fermentation, concentrated feed production and (organic and mineral) fertilizer application are the three main sources of impact. Milk yield is a key production feature for the determination of emissions. The average CF is 0.83 kg CO2/kg raw milk. At each milk yield level, the farms are approximately homogeneous. Compared with other studies, “Vacas Felizes” milk has a lower CF than 80 (out of 84) published CFs and on average it is approximately 32% lower.