Environmental trade-offs of livestock intensification in the Northwest Highlands of Vietnam
Emmanuel Mwema
A * , Thi Thu Hang Dao B , An Notenbaert A and Mary Atieno B
A
B
Abstract
Seventy per cent of households in Vietnam’s Northwest Highlands rely on livestock for food security, income, and resilience. However, livestock production drives deforestation, biodiversity loss, and greenhouse-gas (GHG) emissions. Development partners have started promoting improved feeding, genetics, and animal health interventions to boost productivity, but the environmental impacts of this intensification remain uncertain.
This study aimed to evaluate the production and environmental benefits of prevailing livestock production practices and ongoing animal husbandry interventions promoted in Vietnam’s Northwest Highlands.
The study was conducted in Mai Son district, where the CGIAR Sustainable Animal Productivity initiative addresses challenges in livestock value chains. The Rural Household Multi-indicator Survey (RHoMIS) and the Gendered Feed Assessment Survey (G-FEAST) datasets were utilized to develop the following five distinct farm types: A, B1, B2, C1, and C2. This classification was based on several criteria, including market accessibility, production systems, feeding regimes, and herd structure and size. Biophysical characteristics of the farms were run through the Comprehensive Livestock Environmental Assessment for Improved Nutrition, a Secured Environment and Sustainable Development along Livestock Value Chains (CLEANED) tool to quantify land requirements (ha/year; ha/Mg meat), water use (m3/year; m3/kg meat), soil erosion (Mg soil/year; kg soil/kg meat), nitrogen mining (% area mining), and GHG emissions (Mg CO2 eq/year; kg CO2 eq/kg meat) under current and intensified practices. The farm-level footprints were extrapolated to the district level to assess impacts at scale.
Current livestock production practices significantly affect the environment, but integrating improved feeding, genetics, and animal health innovations is expected to go hand in hand with environmental efficiency gains per kilogram of meat. We projected improved resistance to soil erosion in mountainous areas by cultivating improved forages. Additionally, enhancing animal reproductive performance could serve as a climate mitigation strategy in the study sites.
Integrating improved feeding, breeding, and biosecurity interventions improves productivity, environmental efficiency, and climate resilience while enhancing food security and income. However, realizing these benefits requires evidence-based recommendations and collaboration among key stakeholders.
Scaling interventions require investment in capacity building, supportive policies, and market system models. Continuous monitoring and feedback mechanisms are vital to ensure equitable benefits, sustainability, and successful adoption.
Keywords: CLEANED tool, environmental impact assessment, integrated innovations, livestock production, Northwest Highlands, smallholder multispecies systems, sustainable intensification, technology adoption, Vietnam.
Introduction
The Northwest Highlands (NWH) of Vietnam are known for their favorable climatic conditions, soils and rich grassland system ideal for livestock development. Livestock is raised for multifunctional purposes such as draft power, manure, and occasional meat sales, contributing to 22% of the household income (Huyen et al. 2018). Herd size ranges between 2 and 5 for cattle and 5 and 10 for pigs, mostly composed of local breeds partly obtained from the delta provinces (Huyen et al. 2012; Baltenweck et al. 2018). In smallholder pig farms in northern Vietnam, sows produce an average of 18–20 piglets annually, with recent studies reporting pre-weaning mortality rates ranging from 15% to 25%, depending on farm management practices and veterinary support (Ho et al. 2022). Feed availability and quality are critical constraints, with locally available feed sources often resulting in inconsistent growth rates and extended fattening periods. For cattle, productivity remains low, with average annual weight gains of 100–150 kg per animal, which is below potential. Calving intervals are prolonged, often exceeding 18 months, primarily owing to suboptimal nutrition and limited access to artificial insemination or genetically improved bulls (Tran 2020; Huyen et al. 2018). Furthermore, the lack of proper herd health management contributes to disease prevalence, which further hinders growth and reproduction efficiency. Meat production in the NWH is largely dependent on natural grasslands, where 60% of the native grass species have been reported to decline, thus affecting the production potential of the smallholder farmers in the region (Huyen et al. 2018; Blanchard et al. 2019). Feed shortages are exacerbated by overgrazed pastures, restricted access to communal lands, and a lack of technical and financial investment in forage cultivation. Some programs and projects to introduce improved forage and breeds, such as Laisin bulls, have been hindered by low adoption rates because of inadequate resources and support at the farm level (Huyen et al. 2018). Furthermore, although some areas provide extension services focusing on commercialization, their limited scope and scale render them ineffective for the majority of smallholder farms in the region. This underscores the need for policies to enhance feed production, promote genetic improvements, and establish robust market systems for livestock products (Huyen et al. 2010).
The CGIAR initiative on Sustainable Animal Productivity for improved Livelihoods, Nutrition and Gender inclusion (SAPLING) brought together a range of partners to address these challenges through a pipeline of co-delivered and demand-driven innovations in the areas of improved feeds and forages, animal breeding, herd health, and competitive, inclusive markets. SAPLING was mainly implemented in Mai Son district, Son La province, and built on the Li-chan project conducted under the CGIAR Research Program on Livestock.
Son La province is the most extensive mountainous province in northern Vietnam, with a total area of 1.4 million hectares and a population of 1.3 million people (General Statistics Office of Vietnam 2019). Eighty three per cent of the population in Son La province are ethnic minorities who suffer from the highest levels of poverty, malnutrition, and gender inequality (Hammond et al. 2021). Communities in the Mai Son district are reliant on agriculture, particularly livestock production, which is dominated by small-scale farmers.
SAPLING targeted two value chains (pigs and beef) considered as the primary source of livelihoods, food nutrition and security, and where small-scale producers can benefit from market growth (Ives et al. 2013; Baltenweck et al. 2018; Hung et al. 2019). Implementing biosecurity measures, artificial insemination (AI), and investment in better feeds and forages are expected to increase cattle and pig productivity significantly (Atieno et al. 2021). However, these benefits could be realized with an environmental cost as livestock production practices, particularly those focused on improving productivity, play a significant role in environmental degradation (Foley et al. 2011). Intensifying livestock production may lead to increased nutrient loads and water use through intensive fodder production, and degradation of grassland as a result of overgrazing and higher stocking densities (Enahoro et al. 2019; Adesogan et al. 2020). Even though greenhouse-gas (GHG) emissions from the livestock sector in Vietnam are lower than those from other sectors, intensification could accelerate GHG emission levels through expansion of land for forage cultivation, manure management challenges and increased synthetic input flows into the systems. There is a risk that increasing demand for meat products and intensification efforts could make livestock production unsustainable for the environment (Smit and Heederik 2017). Therefore, efforts to maximize production must balance with environmental sustainability aspirations before scaling innovations. Given that there have been limited studies that comprehensively assess the environmental footprints of bundling technologies at scale in smallholder multi-species systems, this paper presents a first attempt to assess the potential trade-offs and synergies of bundling technologies within this context to influence policy and decision-making while contributing to the much-needed transformation of the livestock sector in the NWH. This study aims to assess the production and environmental benefits arising from the adoption of animal husbandry interventions (improved feeding, genetics and animal health) in typical beef–pork systems in the NWH of Vietnam. Specific objectives include the following:
Evaluating the baseline environmental situation: land and water use, soil erosion and nitrogen mining, and GHG emissions from cattle and pig production systems.
Projecting the likely changes from adopting an integrated animal husbandry intervention package (improved feeding, genetics, and animal health) promoted by development partners in the NWH.
Assessing the trade-offs and synergies of integrating these technologies at the district scale.
Methods
Study area
Mai Son district in Son La province was selected as the study site to demonstrate challenges and needs related to livestock production in the NWH (Fig. 1). The climate is continental tropical monsoon and is influenced by topography. The cold and dry winters last from October to March, whereas the remaining months are hot, humid and rainy (Hung and Quyen 2017). The average temperature is 21.5°C and the region receives an average of 1400 mm of rainfall annually, with 118 rainy days per year (Kong and Lee 2014). Approximately 80% of the annual rainfall in this region occurs during the monsoon season, which spans from June to September. The population comprises 12 distinct ethnic groups. The Thai ethnic group constitutes 53.6%, followed by the Kinh ethnic group (16.3%), H’mong (16.1%) and other ethnic groups (14%) (General Statistics Office of Vietnam 2019).
The district has a wide range of farming systems, from grazing and extensive systems in the mountainous areas to intensive farms with solid integration of crops and livestock in the valleys. These different farming systems exist in a variety of socio-economic and ecological conditions (Douxchamps et al. 2021). Livestock species include buffaloes, beef cattle, pigs, goats, and poultry, with a combined holding of 0.36 tropical livestock units per capita (Mulisa et al. 2017). Feed resources include natural grazing areas, planted forages, and crop residues (maize, cassava, sugarcane tops, rice straw, banana leaves) (Atieno et al. 2021).
Farm types
The study targeted three main farm types categorized on the basis of elevation and accessibility, defined according to the proximity to roads and markets, in the two selected communes (Chieng Chung and Chieng Luong) of Mai Son district (Douxchamps et al. 2021; Hammond et al. 2021). These types included the following: (1) Type A, which are intensive farming systems located in the lowlands with good market access and higher capacity for innovation; (2) Type B, which are mixed crop–livestock systems in mid-altitudes with moderate market access, primarily occupied by Thai ethnic group; and (3) Type C, which are extensive farming systems in highlands with low market access and fragile environments, mainly inhabited by the H’mong ethnic group. Livestock is the second main income source to households across the three farm types, second to cropping activities (Atieno et al. 2021).
In addition to the RHoMIS survey, a gendered feed assessment tool (G-FEAST) was employed as a diagnostic tool to evaluate local feed resources and identify challenges faced by different household types (Lukuyu et al. 2019a, 2019b). Conducted in six villages, including Khoa and Xam Ta in Chieng Chung commune and Mon 1, Mon 2, Oi, and Buom Khoang in Chieng Luong commune, the G-FEAST survey gathered gender-disaggregated data through seven focus group discussions and 110 individual interviews (Atieno et al. 2021). Using the insights on feed availability, feeding practices, constraints in livestock production, and opportunities for interventions, the G-FEAST data, particularly on the feed basket and livestock, were used to further refine the main farm types.
Two of the three main farm types were further divided on the basis of herd size and feeding practices, resulting in five subtypes of A, B1, B2, C1, and C2. The differentiation between B1 and B2 resulted from variations in feed baskets between the villages of Khoa and Oi, whereas the subdivision of Type C into C1 and C2 reflected differences in management systems and herd composition. Five representative farms, one for each of the subtypes, were selected for the environmental assessments (Table 1).
| Description | Type A | Type B1 | Type B2 | Type C1 | Type C2 | |
|---|---|---|---|---|---|---|
| Village | Mon | Khoa | Oi | Buom Khoang | Xam Ta | |
| Farm size (ha) | 2.25 | 0.6 | 2.3 | 1.6 | 6.04 | |
| Management system | Confined | Confined | Confined | Confined and tethering | Grazing/Semi-grazing | |
| Herd composition | Adult cattle – male: 2 Pigs – growers: 7 | Steers/heifers: 3 Pigs – growers: 1 | Cows: 1 Adult cattle – male: 1 Pigs – growers: 10 Pigs – lactating/pregnant sows: 1 | Cows: 2 Calves: 2 Pigs – growers: 2 | Cows: 5 Steers/heifers: 3 Calves: 2 Adult cattle – male: 2 Pigs – growers: 6 Pigs – lactating/pregnant sows: 1 | |
| Breed type | Local breed, crossbreed | Local breed, crossbreed | Local breed, crossbreed | Local breed | Local breed | |
| Feed basket for cattle | Maize cracked grains: 10% Sugarcane tops: 25% Rice straw: 15% Naturally occurring pasture-green fodder: 20% Elephant grass: 20% Grazing: 10% | Rice straw: 20% Naturally occurring pasture-green fodder: 10% Elephant grass: 7% Banana trunk: 3% Grazing: 60% | Sugarcane tops: 43% Rice straw: 7% Naturally occurring pasture-green fodder: 25% Elephant grass: 10% Grazing: 15% | Sugarcane tops: 15% Maize stover: 15% Naturally occurring pasture-green fodder: 10% Elephant grass: 10% Grazing: 50% | Maize cracked grains: 10% Maize stover: 10% Cassava-crop residue: 10% Grazing: 70% | |
| Feed basket for pigs | Banana trunk: 60% Maize – cracked grains: 40% | Banana trunk: 60% Maize – cracked grains: 40% | Banana trunk: 60% Maize – cracked grains: 40% | Banana trunk: 60% Maize – cracked grains: 40% | Banana trunk: 60% Maize – cracked grains: 40% |
The CLEANED approach
This research followed the concepts and guidelines of the Comprehensive Livestock Environment Assessment for improved Nutrition, a secure Environment and Sustainable Development (CLEANED) framework as described by Notenbaert et al. (2014). It is an indicator framework for ex-ante environmental impact assessment, operationalized in Microsoft Excel software. The concept identifies feed production and livestock management as key stages in the livestock value chain where a lot of environmental impacts are felt.
The tool involves two processes, including collecting and inputting the baseline data and then generating the reports for different scenarios of how livestock production systems might change (Mukiri et al. 2019). The tool thus uses feed (feed items, their production practices, nutritive values and intakes) as the core module, and together with livestock management practices and a set of standard parameters, quantifies environmental footprints of livestock production along several environmental dimensions, including land requirement (ha/year; ha/Mg meat), water use (m3/year; m3/kg meat), soil erosion (Mg soil/year; kg soil/kg meat), nitrogen mining (% area mining), and GHG emissions (Mg CO2 eq/year; kg CO2 eq/kg meat). Land requirements are calculated on the basis of the amount of feed required to sustain the animals, which is compared to the biomass yield of the individual feed items forming the animal diet. A threshold of daily dry-matter (DM) intake for the cattle and pigs was set at 4% of their bodyweight (Onyimonyi and Okeke 2007). Nitrogen mining or leaching is determined by nitrogen balance on farm and considers the crop input flows into the production system calculated using the nutrient monitoring methodologies as described by De Jager et al. (1998). Soil erosion is computed using the revised universal soil loss equation (RUSLE), a widely used equation for estimating soil loss by water (Benavidez et al. 2018). Water use is estimated using the evapotranspiration method, which equates the amount of water lost by forage under normal conditions to its water needs. GHG estimations are calculated using Tier 2 methods of the Intergovernmental Panel on Climate Change (IPCC) (2019) Guidelines, a refinement of the 2006 Guidelines for the National Greenhouse Gas Inventories. The method uses country-specific and detailed approaches to estimate GHG emissions by using disaggregated activity data and emission factors tailored to specific livestock types, production systems, and local conditions. The guidelines account for methane (CH4) emissions from enteric fermentation and manure, nitrous oxide (N2O) from manure and soils, and off-farm fertilizer production. These gases were converted to carbon dioxide equivalents for a global warming potential of 100 years, with 28 kg CO2 per kg CH4 for CH4 and 265 kg CO2 per kg N2O for nitrous oxide (Intergovernmental Panel on Climate Change (IPCC) 2014).
Out-scaling approach
We adopted a consultative approach that incorporated inputs from a diverse range of experts to gather and validate district livestock data and distribution of farm types we assessed. The experts were selected from various levels of governance to ensure diverse representation and expertise. They included (1) representatives from the National Institute for Animal Science (NIAS) for national-level expertise, (2) agricultural officers from the Department of Crop Production, Animal Husbandry, Veterinary Services, and Aquatic Products in Son La province, (3) representatives from the Agricultural Service Center, along with agricultural officers from Mai Son District, and (4) veterinary and agricultural extension officers from various communes in Mai Son district.
To quantify the impacts at scale, we first estimated the total number of cattle and pig farmers in the 22 communes in Mai Son. This was undertaken by multiplying the total number of households in each commune, as provided by the Department of Crop Production in Mai Son district, by the percentage number of households owning cattle and pigs and the average number of animals per household. We then organized experts into groups to first determine the percentage distribution of the main farm types (A, B, and C) across the 22 communes in Mai Son district (Supplementary Table S1). Afterwards, we asked the local experts at the commune level to accurately allocate subtype proportions (B1, B2, and C1, C2) from the main Types B and C on the basis of their deep understanding of the communes and village-specific practices (Table S2).
We then multiplied the number of enterprises per subtype with baseline and intervention footprints to obtain relative impacts at scale. The farm-level footprints were extrapolated to Mai Son district impacts, assuming 100% adoption of interventions and that agricultural practices are consistent across all cattle–pig systems of the same type. This approach has been widely applied in multiscale assessments (Notenbaert et al. 2020; Waha 2020).
Intervention scenarios
The intervention scenario involved the integration of three SAPLING innovation packages, namely, improved feeds and forages, genetics, and animal health. These were implemented between the 2022 and 2023 to reduce mortality, improve animal genetic performance through AI and improve feed resource availability and quality respectively. Below is a description of each innovation package.
Improved feeds and forages: involved the selection of the appropriate feed options, ranging from improved forages such as legumes as cover crops and grasses, and utilization of locally available feed resources, to address feed quality and availability, while enhancing productivity gains in the systems. Some feeds and forage interventions included cultivating Mombasa Guinea (Megathyrsus maximus cv. Mombasa), Ubon Stylo (Stylosanthes guianensis var. guianensis cv. Ubon stylo), Mulato II (Urochloa hybrid cv. Mulato II), Green elephant (Cenchrus purpureus), and rice bean (Vigna umbellata), feed fermentation using probiotics, silage making and urea-treated rice straw. The nutritive values for these feed items were sourced from Feedipedia, an open-access online animal feed library database1 and the diet proportions were based on farmer practices a year after the introduction of innovations. Supplementary Figs 1–10 show detailed feed baskets of the baseline and intervention scenarios.
Animal genetic improvement: improved semen supply and training of AI service providers, with a focus on promoting animal breeding and genetics as key enablers of productivity and resource-use efficiency.
Improved animal health: this included measures to improve biosecurity, disease control and prevention, drug management and vaccination rates, by increasing the capacity of animal health workers and local vets, as well as training farmers to improve herd health management practices.
All these interventions were implemented with the assumption that cattle and pig productivity would improve while promoting the environmental sustainability of the value chains. Because most environmental impacts in livestock value chains are associated with feed production and livestock management, feed baskets and livestock data are important inputs for the environmental analysis evaluated in this study. We integrated the packages into the model by adjusting the input parameters of the baseline systems. This adjustment was based on midline survey data collected from the selected case study farms and expert information from NIAS. As a result, we ran a total of five scenarios, one for each farm type. The baseline and after-to-be intervention livestock-related input variables for each production system are in the Table S3.
A relative change from the baseline environmental footprint was calculated, and projected changes at the farm scale were reported on a per-area and livestock product basis. The relative change (expressed in %) was presented in a friendly colour-coded table with three uniform intervals based on the maximum change values. The trade-offs and synergies were visualized by ‘+’ signs to imply a positive environmental change and ‘−’ signs to denote a worsened environmental condition relative to the baseline. The more ‘+’ or ‘−’ signs a scenario has, the greater is the intensity of the environmental change for improved animal health, genetics, and adoption of improved feeds and forages.
Results
Baseline situation
The farms displayed varying levels of productivity (Table 2). The farms on the mountain tops (Type C2) noticed considerable yield owing to the large herd sizes. The farms near midland areas (Type A) were more productive than others. Type B, with upland systems in the valleys, exhibited the lowest production level.
| Farm type | Production | Land requirement | Soil impact | Water impact | GHG emissions | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Meat produced (kg/year) | ha/year | ha/Mg meat | Nitrogen mining (% area) | Erosion (Mg soil/year) | Erosion (kg soil/kg meat) | (m3/year) | (m3/kg meat) | (Mg CO2 eq/year) | (kg CO2 eq/kg meat) | ||
| Type A | 168 | 0.8 | 4.8 | 94.8 | 11.2 | 66.9 | 4162.1 | 24.8 | 3.9 | 23.3 | |
| Type B1 | 96 | 1 | 10.1 | 89.4 | 7 | 72.7 | 3137.6 | 32.7 | 6.9 | 72.4 | |
| Type B2 | 96 | 0.7 | 7.8 | 100 | 45.6 | 475 | 1936.2 | 20.2 | 4.2 | 43.7 | |
| Type C1 | 126.4 | 1.2 | 9.5 | 100 | 45.4 | 359.1 | 2454 | 19.4 | 7.2 | 56.9 | |
| Type C2 | 409.6 | 6.6 | 16.1 | 100 | 87.5 | 213.7 | 15,218.4 | 37.2 | 47.7 | 116.5 | |
The area required for feed production ranged from 0.7 to 6.6 ha. A higher land requirement in Type C2 was linked to high stocking rate and over-reliance on natural pasture as a livestock feeding strategy. Systems that had nutritious and high-yielding forage biomass, for example, Type A, were observed to utilize land more efficiently. Type B2 farm in the valleys with mixed feeding (cultivated forages, crop residues, and limited natural pasture inclusion) had the lowest land requirements.
The mountainous rural areas (Types C1 and C2) and upland areas (Type B2) faced a double soil-health burden from nutrient mining and erosion (Table 2). Because of topography and high precipitation, much soil is lost yearly. Similarly, a substantial amount of soil was also lost in the uplands where sugarcane is continuously grown. Farm input supply, such as nitrogen, is limited in this area because of accessibility constraints. The imbalance of nitrogen flow into the system led to excessive mining.
The water usage in the systems varied on the basis of rainfall and feed type. Large herd size and natural pasture feeding drove higher water needs in Type C2. The midland areas also observed a water resource burden, owing to intensified feed cultivation. Water-use efficiency varied significantly among farms, with C1 and C2 being the most and least efficient respectively (Table 2).
The GHG emissions footprint of Type C2 was at least five times higher than that of other systems, because of its low-quality diet and relatively larger herd size, which produced high CH4 concentrations through enteric fermentation. Generally, all systems reported high emissions from CH4 enteric fermentation followed by manure management. The lowest carbon footprint was exhibited in Type A, owing to its relative share of improved feeding and energy-efficient animals. A comparison of GHG emissions per kilogram of meat produced showed that the most GHG-efficient systems were the intensified Type A and the least efficient were the extensive Type C2.
Trade-offs in environmental impacts
The effects of integrated improved feeding, genetics, and animal health (biosecurity) on animal productivity varied across different systems (Table 3). Meat production increased in Type B1 and B2 because of an increased share of improved breeds. Under the market-oriented Type A system, there was a notable increase in productivity. The breeding investment was low in non-market-oriented Type C1 and C2, resulting in the least improvement.
| Farm type | Meat production | Land requirement | Soil impact | Water impact | GHG emissions | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| (kg/year) | (ha/year) | (ha/Mg meat) | Nitrogen mining (% area) | Erosion (Mg soil/year) | Erosion (kg soil/kg meat) | (m3/year) | (m3/kg meat) | (Mg CO2 eq/year) | (kg CO2 eq/kg meat) | ||
| Type A | ++ | + | ++ | + | + | ++ | + | ++ | – | + | |
| Type B1 | +++ | – | + | – | – | + | – | + | + | ++ | |
| Type B2 | +++ | – | + | + | – | + | – | + | – | ++ | |
| Type C1 | + | + | + | – | – | + | + | – | + | ||
| Type C2 | + | + | + | + | + | + | + | + | + | + | |
| Legend (%) | Best case scenario | ||||||||||
| +++ | < −112 |  | |||||||||
| ++ | −57 to −112 | ||||||||||
| + | −0.01 to −56 | ||||||||||
| 0 |  | No effect | |||||||||
| – | 0.01–56 |  | |||||||||
| – | 57–112 | ||||||||||
| – | >112 | ||||||||||
| Worst case scenario | |||||||||||
Integrated innovation packages showed Types C2 and B1 as the most and least environmentally efficient enterprises respectively. In Type C2, at least 50% of cattle and pig diets were improved while confining pigs to reduce disease and heat stress, thus enhancing productivity. Types B1 and B2 required more land and water resources to meet the meat production levels. This was driven by increased demand for feed owing to increased livestock numbers.
Annual soil loss increased in Types B1, B2, and C1 because of topography and limited cover cropping practices, except for Types A and C2. Introducing cover crops in mountainous regions was observed to be an effective method of controlling soil erosion. Soil loss efficiency improved for each kilogram of meat produced in all systems except for Type C1, which increased by 35%.
The system GHG emissions increased except for the total absolute reductions in Types B1 and C2. Type A system reported the highest increases (up to 54%), whereas the least was in Type B2 (6%). Improved animal reproductive performance, achieved through biosecurity, better nutrition, and breeding, has the potential to offset carbon emissions by 5% in Type B1 and 41% in Type C2. Overall, GHG emissions efficiency improved across the systems, with the highest mitigation in Types B1 and B2, driven primarily by high meat production.
Environmental impacts at scale in Mai Son district
Integrated interventions implemented at the landscape level resulted in synergies between production and environmental sustainability (Table 4). Meat production efficiency demonstrated low environmental impacts at scale across the measured indicators. At the landscape level, production increased by 55%, resulting in a decrease in both total absolute and relative environmental footprints. Efficiency gains in land use, erosion control, water conservation, and GHG emissions reduction ranged from 15.53% to 43.64% per unit of meat produced.
| Extrapolation | Production | Land requirement | Soil impact | Water impact | GHG emissions | ||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Meat produced (kg/year) | Feed area (ha/year) | (ha/Mg meat) | Nitrogen mining (% area) | Erosion (Mg soil/year) | Erosion (kg soil/kg meat) | (m3/year) | (m3/kg meat) | (Mg CO2 eq/year) | (kg CO2 eq/kg meat) | ||
| Total baseline | 4.1 million | 0.047 million | 9.66 | 96.85 | 0.67 million | 237.46 | 131 million | 26.84 | 0.32 million | 62.56 | |
| Total intervention | 6.4 million | 0.035 million | 5.47 | 87.02 | 0.56 million | 184.97 | 120 million | 16.77 | 0.24 million | 35.26 | |
| Relative change (%) | 55% | −25.21% | −43.42% | −10.14% | −15.53% | −22.11% | −8.77% | −37.52% | −25.17% | −43.64% | |
At the district level, improved feeding, animal genetics, and reduced mortality led to a 55% increase in production, while significantly reducing the pressure on environmental resources. The total feed area and water use decreased by 21.25% and 8.77% respectively. Furthermore, soil health improved significantly on a larger scale, with a decrease of 10.14% in nitrogen mining and a 15.53% drop in soil erosion, thus contributing to the overall productivity of the systems. Soil health management practices, such as green manuring, mulching, and constant manure recycling, were ideal for landscape restoration. The district also managed to reduce emissions by 25.27% despite the increase in production. This was achieved through integrated innovations that offer climate mitigation potential for the area.
Discussion
Our findings indicated that current livestock production practices negatively affect the environment. However, integrating cultivated forages, breeding, and biosecurity measures into small-scale, multi-species farming systems in the NWH region can substantially benefit both the environment and productivity. However, this will require a targeted investment in farmer awareness and capacity building to adopt and scale these technologies. Additionally, promoting community-driven AI and a One Health approach, which fosters collaboration among veterinarians, health scientists, and environmental experts to optimize animal and ecosystem health, could accelerate the adoption of these practices across the district. Establishing seed system structures would further support and speed up this adoption.
Although the government offered to increase land allocation for livestock feed production (Government of Vietnam 2013), our analysis generated statistical evidence for a potential land resource use efficiency pathway that can influence current and future decisions on land allocations. If the government’s decision is implemented, these innovations will help the government allocate land more efficiently. By reducing the land pressure on feed production through improved biomass and feed quality, certain areas can be managed more efficiently. For example, land can be converted into conservation zones, promoting biodiversity and carbon offsetting, especially from livestock production activities (Estacio et al. 2021).
In prioritizing the implementation of the National Livestock Development Strategy (NLDS) for 2021–2030 (Government of Vietnam 2022), the government should make data-driven decisions that promote a fair transition. In the cattle-rich corridor of NWH, significant investment is necessary to improve the reproductive performance of cattle and pigs. Growing high-yielding and nutrient-rich forages alongside improving animal reproductive health and performance, can offer multiple environmental and productivity benefits. The forages selected, including Mulato II, Mombasa Guinea, Green Napier, and Ubon Stylo, offer higher-quality biomass than do current farmer feeding practices, mainly relying on low-quality Napier and crop residues. This aligns well with the desired feed interventions for the country (Government of Vietnam 2022). However, seasonal variations can limit the reliability of grasses and legumes as consistent feed sources. Maize silage could therefore serve as a complementary, high-energy feed, offering a more stable alternative to mitigate seasonal deficits in feed quality and quantity (Atieno et al. 2023).
Integrating innovative practices can be a solution to reduce soil erosion, which is a major production constraint in the NWH region. These practices would be most effective in mono-cropped systems situated in valleys and mountainous areas. Improving soil health has been described as a major strategy for enhancing landscape fertility and production (Lal 2016). Although at scale our study shows high production efficiency with low water and land footprints, there is need for careful management of quality and stocks of these resources. Efficient water usage could help alleviate competition for water resources between livestock production and other uses, such as household consumption. This would have a positive impact on overall water sustainability in the region.
Integrating innovative packages into small-scale farming systems is not only beneficial for increasing productivity but also helps reduce GHG emissions (Notenbaert et al. 2020). Increasing carbon accumulation in soil and biomass enhances GHG balance, enabling farming systems to adapt to climate variability and change (Soussana et al. 2010). Improved animal breeding can mitigate climate change by enhancing feed conversion efficiency, which reduces CH4 emissions (Fennessy et al. 2018; Hegarty et al. 2021; Honan et al. 2021).
Smallholders play a vital role in the economic development of developing countries such as Vietnam (Tuan 2012; Fan and Rue 2020). However, for their efforts to yield results, it is essential to involve multiple stakeholders. To promote a transformative agenda for livestock, it is essential to establish a level playing field and a policy environment that enables it (Pingali 2015). In Vietnam, sectoral policies such as the Law on Animal Husbandry and the National Agricultural Restructuring Plan explicitly promote the development of local feed resources, encourage public–private partnerships, and support value-chain integration (Government of Vietnam 2018, 2021). One practical strategy is to develop commune-level aggregation hubs co-managed by farmer groups and logistics providers that could help bulk livestock and feed products weekly or monthly and link them to a mobile-based platform for real-time price discovery and transport coordination (Baumüller 2018; Abraham et al. 2022). This aligns with Vietnam’s efforts under Decision 150/QĐ-TTg, which emphasizes building market linkages and logistics systems in rural areas (Government of Vietnam 2022). To enhance the cattle and pig value chains, a public–private consortium could train and certify AI technicians and operate regional semen depots with cold-chain storage (Engidawork 2018). Mobile veterinary clinics in outreach vehicles could deliver on-farm preventive care and diagnostics under blended public–private cost-recovery schemes (Karimuribo et al. 2016; Omondi et al. 2021). Additionally, community seed fairs at aggregation hubs backed by village-based inspection networks could distribute certified forage seeds, addressing gaps in feed access and quality as identified in the National Action Plan 2021–2030 on Climate Change Adaptation in Agriculture (Government of Vietnam 2020; Croft et al. 2021).
Conclusions
Integrating feed interventions, breeding, and biosecurity measures in smallholders’ multi-species systems improves productivity and enhances environmental efficiency. Increased farm productivity leads to higher farmer incomes, lower poverty rates, and improved food security and nutrition, aligning with Vietnam’s development priorities. However, adopting these innovative practices must be tailored to the local context, because there is no universal solution for all settings. In systems where resources are limited, it is advisable to have a proactive approach to manage the available stocks efficiently.
Mountainous areas, particularly those with steep slopes and prone to soil degradation, are recommended to diversify and include cover cropping and terracing with dense forages. Although these interventions offer numerous benefits, their implementation on a larger scale must account for farmers’ limited technical expertise and associated enablers for adoption. Government and development actors’ priorities should be centred around low-cost capacity-building approaches such as group-based training, peer-to-peer extension models, demonstration sites and use of digital extension services to enhance technical support and innovations uptake. Forage breeders should prioritize their efforts on identifying and selecting high-quality forages that are well-suited to cold and dry conditions during the winter season to reduce variability. Additionally, farmers would need to diversify their livestock and understand available breeding technologies such as AI. The evidence presented here can assist the local government and development actors in determining climate mitigation and adaptation strategies for livestock systems in the NWH region. However, establishing continuous monitoring and feedback mechanisms is crucial to ensure the sustainability of these systems.
Data availability
The data that support this study will be shared upon reasonable request to the corresponding author.
Declaration of funding
This work was funded by the CGIAR Trust Fund through the CGIAR Initiative on Sustainable Animal Productivity for Livelihoods, Nutrition and Gender inclusion (SAPLING). The funders had no role in the writing of the paper or in the decision to publish the results.
Acknowledgements
This work was part of the CGIAR Initiative on Sustainable Animal Productivity for Livelihoods, Nutrition and Gender Inclusion (SAPLING). CGIAR research is supported by contributions to the CGIAR Trust Fund. CGIAR is a global research partnership for a food-secure future dedicated to transforming food, land, and water systems in a climate crisis. We express our gratitude to the local authorities of the Sub-Department of Animal Health, Animal Husbandry and Aquaculture (Sub-DAH), National Institute of Animal Science in Vietnam (NIAS) and Northern Mountainous Agriculture and Forestry Science Institute, Vietnam (NOMAFSI), for verifying the input data used in the analyses presented in this paper.
References
Abraham M, Verteramo Chiu L, Joshi E, Ilahi MA, Pingali P (2022) Aggregation models and small farm commercialization – a scoping review of the global literature. Food Policy 110, 102299.
| Crossref | Google Scholar |
Adesogan AT, Havelaar AH, McKune SL, Eilittä M, Dahl GE (2020) Animal source foods: sustainability problem or malnutrition and sustainability solution? Perspective matters. Global Food Security 25, 100325.
| Crossref | Google Scholar |
Atieno M, Mai T, Douxchamps S, Peters M, Duncan A (2021) Rapid survey of livestock feed resource availability and use in Mai Son district, Son La province, Vietnam, using the Gendered Feed Assessment Tool (G-FEAST). CGIAR Research Program on Livestock. Available at https://hdl.handle.net/10568/111524
Atieno M, Bùi VT, Phan HC, Tran TB, Mai TT, Duncan A, Douxchamps S, Peters M (2023) Feeds and forages intervention strategies for improved livestock nutrition and productivity in the Northwest Highlands (NWH) of Vietnam. In ‘Proceedings of the XXV International Grassland Congress (IGC 2023)’, 14–19 May 2023, Covington, Kentucky, USA. pp. 1297–1301 (International Grassland Congress 2023). Available at 10.52202/071171-0316
Baltenweck I, Thinh NT, Nga NTD, Hung PV, Nhuan NH, Huyen NTT, Lapar ML, Teufel N (2018) Assessing competitiveness of smallholder pig farming in the changing landscape of Northwest Vietnam. International Livestock Research Institute. Available at https://hdl.handle.net/10568/98904
Baumüller H (2018) The little we know: an exploratory literature review on the utility of mobile phone-enabled services for smallholder farmers. Journal of International Development 30(1), 134-154.
| Crossref | Google Scholar |
Benavidez R, Jackson B, Maxwell D, Norton K (2018) A review of the (Revised) Universal Soil Loss Equation ((R)USLE): with a view to increasing its global applicability and improving soil loss estimates. Hydrology and Earth System Sciences 22(11), 6059-6086 Available at.
| Crossref | Google Scholar |
Blanchard M, Le Thi Thanh H, Ickowicz A (2019) Multifunctions and contributions of grassland-based livestock systems in North-West Vietnam. Proceedings of the Nepal Veterinary Association. Available at https://agritrop.cirad.fr/594772/
Croft MM, Davis V, Longley K, Ferris S, Templer N (2021) The role of seed vouchers and fairs in promoting seed market development: opportunities and limitations. A Feed the Future Global Support Seed Systems for Development Activity (S34D) Report. Available at https://hdl.handle.net/10568/130910
De Jager A, Nandwa SM, Okoth PF (1998) Monitoring nutrient flows and economic performance in African farming systems (NUTMON): I. Concepts and methodologies. Agriculture, Ecosystems & Environment 71(1–3), 37-48.
| Crossref | Google Scholar |
Douxchamps S, Tùng BV, Chương PH, Atieno MO (2021) Nutrient flows in crop–livestock systems of the North-West Highlands of Vietnam. Alliance of Bioversity and CIAT. Available at https://hdl.handle.net/10568/116063
Enahoro D, Mason-D’Croz D, Mul M, Rich KM, Robinson TP, Thornton P, Staal SS (2019) Supporting sustainable expansion of livestock production in South Asia and Sub-Saharan Africa: scenario analysis of investment options. Global Food Security 20, 114-121.
| Crossref | Google Scholar |
Engidawork B (2018) Artificial insemination service efficiency and constraints of artificial insemination service in selected districts of Harari National Regional State, Ethiopia. Open Journal of Animal Sciences 8(3), 239-251.
| Crossref | Google Scholar |
Estacio D, Lejarde G, del Mundo-Angeles AN, Kiester V (2021) Agricultural land conversion: an environmental management approach. International Journal of Engineering Technology Research & Management 5(10), 34-48 Available at https://www.researchgate.net/publication/355147772_agricultural_land_conversion_an_environmental_management_approach.
| Google Scholar |
Fan S, Rue C (2020) The role of smallholder farms in a changing world. In ‘The role of smallholder farms in food and nutrition security’. (Eds S Gomez y Paloma, L Riesgo, K Louhichi) pp. 13–28. (Springer International Publishing) 10.1007/978-3-030-42148-9_2
Fennessy PF, Byrne TJ, Proctor LE, Amer PR (2018) The potential impact of breeding strategies to reduce methane output from beef cattle. Animal Production Science 59(9), 1598-1610.
| Crossref | Google Scholar |
Foley JA, Ramankutty N, Brauman KA, Cassidy ES, Gerber JS, Johnston M, Mueller ND, O’Connell C, Ray DK, West PC, Balzer C, Bennett EM, Carpenter SR, Hill J, Monfreda C, Polasky S, Rockström J, Sheehan J, Siebert S, Tilman D, Zaks DPM (2011) Solutions for a cultivated planet. Nature 478(7369), 337-342.
| Crossref | Google Scholar | PubMed |
General Statistics Office of Vietnam (2019) Kết quả toàn bộ Tổng điều tra dân số và nhà ở năm 2019. General Statistics Office of Vietnam. Available at https://www.gso.gov.vn/du-lieu-va-so-lieu-thong-ke/2020/11/ket-qua-toan-bo-tong-dieu-tra-dan-so-va-nha-o-nam-2019/
Government of Vietnam (2013) Decision No. 899/QĐ-TTg approving the Agricultural Restructuring towards raising added value and sustainable development dated 10 June 2013. Government of Vietnam, Hanoi. Available at https://vanban.chinhphu.vn/default.aspx?pageid=27160&docid=168000
Government of Vietnam (2018) Law No. 32/2018/QH14 of the National Assembly: law on animal husbandry dated 19 November 2018. Government of Vietnam, Hanoi. Available at https://vanban.chinhphu.vn/?pageid=27160&docid=206100
Government of Vietnam (2020) Decision No. 1055/QĐ-TTg approving the National Climate Change Adaptation Plan for the period 2021–2030, with a vision to 2050 dated 20 July 2020. Government of Vietnam, Hanoi. Available at https://vanban.chinhphu.vn/default.aspx?pageid=27160&docid=200561
Government of Vietnam (2021) Decision No. 255/QĐ-TTg approving Agricultural Restructuring plan 2021–2025 dated 25 February 2021. Government of Vietnam, Hanoi. Available at https://vanban.chinhphu.vn/default.aspx?pageid=27160&docid=202717
Government of Vietnam (2022) Decision No. 150/QĐ-TTg dated 28 January 2022 approving the Sustainable Agriculture and Rural Development Strategies for 2021–2030, with a vision toward 2050. Government of Vietnam, Hanoi. Available at https://vanban.chinhphu.vn/?pageid=27160&docid=205277&tagid=6&type=1
Hammond J, Caulfield ME, Mai T, Teufel N, Van Wijk MT, Douxchamps S (2021) Farming systems and accessibility in North-West Vietnam: a baseline survey from the Li-Chăn project. ILRI. Available at https://hdl.handle.net/10568/115890
Hegarty RS, Passetti RAC, Dittmer KM, Wang Y, Shelton SW, Emmet-Booth J, Wollenberg EK, McAllister T, Leahy S, Beauchemin K, Gurwick N (2021) An evaluation of emerging feed additives to reduce methane emissions from livestock. A report coordinated by Climate Change, Agriculture and Food Security (CCAFS) and the New Zealand Agricultural Greenhouse Gas Research Centre (NZAGRC) initiative of the Global Research Alliance (GRA). Available at https://hdl.handle.net/10568/116489
Ho N-N, Do TL, Tran D-T, Nguyen TT (2022) Indigenous pig production and welfare of ultra-poor ethnic minority households in the northern mountains of Vietnam. Environment, Development and Sustainability 24(1), 156-179.
| Crossref | Google Scholar |
Honan M, Feng X, Tricarico JM, Kebreab E (2021) Feed additives as a strategic approach to reduce enteric methane production in cattle: modes of action, effectiveness and safety. Animal Production Science 62, 1303-1317.
| Crossref | Google Scholar |
Hung NT, Quyen NH (2017) Effect of the change of climate indicators on agricultural yields in Son La province. Vietnam Journal of Science and Technology 55(6), 6.
| Crossref | Google Scholar |
Hung PV, Nga NTD, Huyen NTT, Pho NV, Thinh NT, Unger F, My PK, Nhuan NH (2019) Challenges for sustainable development of smallholder pig value chain in Vietnam. Vietnam National University of Agriculture. Available at https://hdl.handle.net/10568/101911
Huyen LTT, Herold P, Valle Zárate A (2010) Farm types for beef production and their economic success in a mountainous province of northern Vietnam. Agricultural Systems 103(3), 137-145.
| Crossref | Google Scholar |
Huyen LTT, Van DTT, Markemann A, Herold P, Zárate AV (2012) Beef cattle keeping by smallholders in a mountainous province of northern Vietnam in relation to poverty status, community remoteness and ethnicity. Animal Production Science 53(2), 163-172.
| Crossref | Google Scholar |
Ives S, Lane P, Bonney L, Cuong PK, Cuong VC, Hai DT, Duteurtre G, Hung PV, Khoa MA (2013) Overcoming technical and market constraints to the emergence of profitable beef enterprises in the north-western highlands of Vietnam. Australian Centre for International Agricultural Research, Final report. Available at https://www.researchgate.net/publication/263582782_Overcoming_technical_and_market_constraints_to_the_emergence_of_profitable_beef_enterprises_in_the_north-western_highlands_of_Vietnam
Karimuribo ED, Batamuzi EK, Massawe LB, Silayo RS, Mgongo FOK, Kimbita E, Wambura RM (2016) Potential use of mobile phones in improving animal health service delivery in underserved rural areas: experience from Kilosa and Gairo districts in Tanzania. BMC Veterinary Research 12, 219.
| Crossref | Google Scholar |
Kong I, Lee D (2014) Establishment of priority forest areas based on hydrological ecosystem services in Northern Vietnam. Journal of the Korean Society of Environmental Restoration Technology 17(5), 29-41.
| Crossref | Google Scholar |
Lal R (2016) Soil health and carbon management. Food and Energy Security 5(4), 212-222.
| Crossref | Google Scholar |
Lukuyu BA, Van Eerdewijk A, Kinati W, Sultana N, Mulema AA, Duncan AJ (2019a) Gendered Feed Assessment Tool (G-FEAST) focus group discussion guide. International Livestock Research Institute (ILRI). Available at https://hdl.handle.net/10568/100243
Lukuyu BA, Van Eerdewijk A, Kinati W, Sultana N, Mulema AA, Duncan AJ (2019b) Gendered Feed Assessment Tool (G-FEAST) individual farmer interview questionnaire. International Livestock Research Institute (ILRI). Available at https://hdl.handle.net/10568/100244
Mukiri J, Notenbaert AMO, Van der Hoek R, Birnholz CA (2019) CLEANED x-version 2.0.1 technical manual and user guide. International Center for Tropical Agriculture (CIAT). Available at https://hdl.handle.net/10568/107238
Mulisa F, Dida, Dida M (2017) Review paper on determining stocking rate in tropical countries by the use of tropical animal unit month (Taum). Available at https://www.researchgate.net/publication/349477648_Review_Paper_on_Determining_Stocking_Rate_in_Tropical_Countries_by_the_Use_of_Tropical_Animal_Unit_Month_Taum
Notenbaert A, Lannerstad M, Herrero M, Paul B, Fraval S, Ran Y, Mugatha S, Barron J, (2014) A framework for environmental ex-ante impact assessment of livestock value chains. In ‘Paper presented at the 6th All Africa Conference of Animal Agriculture’, 29 October 2014, Nairobi, Kenya. (International Center for Tropical Agriculture) Available at https://doi.org/10.13140/2.1.5011.3287
Notenbaert A, Groot JCJ, Herrero M, Birnholz C, Paul BK, Pfeifer C, Fraval S, Lannerstad M, McFadzean JN, Dungait JAJ, Morris J, Ran Y, Barron J, Tittonell P (2020) Towards environmentally sound intensification pathways for dairy development in the Tanga region of Tanzania. Regional Environmental Change 20(4), 138.
| Crossref | Google Scholar |
Omondi I, Baltenweck I, Kinuthia E, Kirui L, Njoroge-Wamwere G, Bett B, Munene A, Onle S, Dida D, Kiara H (2021) Mobile veterinary clinics in the drylands of Kenya: securing pastoralists’ livelihoods by bringing services close. Development in Practice 31(5), 561-579.
| Crossref | Google Scholar |
Onyimonyi AE, Okeke GC (2007) Assessment of the practical potential of bambara (Voandzeia subterrenea Thouars) waste for weaner pigs. Pakistan Journal of Nutrition 6(3), 264-266.
| Crossref | Google Scholar |
Pingali P (2015) Agricultural policy and nutrition outcomes – getting beyond the preoccupation with staple grains. Food Security 7(3), 583-591.
| Crossref | Google Scholar |
Smit LAM, Heederik D (2017) Impacts of Intensive livestock production on human health in densely populated regions. GeoHealth 1(7), 272-277.
| Crossref | Google Scholar | PubMed |
Soussana JF, Tallec T, Blanfort V (2010) Mitigating the greenhouse gas balance of ruminant production systems through carbon sequestration in grasslands. Animal 4(3), 334-350.
| Crossref | Google Scholar | PubMed |
Tran TBN (2020) Utilization of crop by-product options in beef cattle production in northern Vietnam. Journal of Animal Science and Technology 62(2), 123-134.
| Google Scholar |
Tuan NP (2012) Contract farming and its impact on income and livelihoods for small-scale farmers: case study in Vietnam. Journal of Agribusiness and Rural Development 26(4), 147-166 Available at http://www1.up.poznan.pl/jard/index.php/jard/article/view/586.
| Google Scholar |
Waha K (2020) Feed-based dairy system intensification scenario development and national-level biophysical impact assessment. Commonwealth Scientific and Industrial Research Organisation (CSIRO). Available at https://hdl.handle.net/10568/111512
