Orinoquia Climate smart livestock and Climate

Last registered on May 11, 2026

Pre-Trial

Trial Information

General Information

Title
Orinoquia Climate smart livestock and Climate
RCT ID
AEARCTR-0018562
Initial registration date
May 05, 2026

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
May 11, 2026, 8:54 AM EDT

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Locations

Region

Primary Investigator

Affiliation
Universidad EAFIT

Other Primary Investigator(s)

PI Affiliation
Universidad EATI
PI Affiliation
World Bank

Additional Trial Information

Status
Completed
Start date
2024-06-05
End date
2025-09-30
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
The Orinoquia region of eastern Colombia (comprising Meta, Arauca, Casanare, and Vichada) is one of the country’s most important livestock frontiers and a landscape of high ecological value, characterized by savannas, wetlands, riparian forests, and floodplains that provide essential ecosystem services. While cattle ranching remains central to rural livelihoods, employment, income generation, and food security, continued agricultural expansion and climate variability pose increasing risks to the sustainability of livestock systems. Extensive production practices, pasture degradation, limited tree integration, and growing exposure to droughts and floods threaten both ecosystem resilience and the long-term viability of livestock-based livelihoods.

Within this context, the Orinoquia Climate-Smart Livestock and Climate Project (OGSC) was implemented as part of the World Bank and BioCarbon Fund–ISFL agenda, in partnership with FEDEGAN, CIPAV, CIAT, and Universidad EAFIT. The project aimed to evaluate whether structured rural advisory services could support a transition from extensive livestock production toward climate-smart livestock systems. The intervention was grounded in the premise that adoption of climate-smart practices depends not only on information provision, but also on participatory learning, experimentation, repeated reinforcement, and producers’ perceptions of feasibility, risk, costs, and expected benefits.

The OGSC intervention delivered a Rural Advisory model over a 12-month period, combining collective learning spaces (such as participatory planning, technical workshops, field days, and group discussions) with individualized on-farm advisory visits. Demonstration farms and producer-to-producer learning mechanisms were used to reduce perceived adoption risks and provide visible local evidence of practice feasibility. The advisory strategy prioritized management-based and incrementally adoptable practices with potential co-benefits for productivity, pasture condition, tree cover, water protection, forage availability, conservation, and farm-level decision-making.
To generate credible evidence, the project used a counterfactual evaluation design with treatment and control groups, drawing on 2,133 observations that combine baseline and endline measurements from more than 1,000 livestock producers. This design allowed the analysis to distinguish changes attributable to the intervention from broader regional trends. The empirical findings show that OGSC generated meaningful increases in the adoption of climate-smart livestock practices, particularly those related to pasture–tree integration, live fences, forage management, conservation-oriented behaviors, and managerial capacity.

The results indicate that treatment farms were more likely to increase tree cover in grazing areas, with estimated effects of approximately 7.4 percentage points in the most complete model specification. They also showed a substantial increase in live fence adoption, with effects of approximately 12.7 percentage points. These practices contribute to improved microclimatic conditions, biodiversity, landscape connectivity, and ecosystem service provision while maintaining productive capacity. The intervention also improved producers’ capacity to manage forage scarcity: treatment farms were more likely to establish forage banks, with an estimated effect of 7.4 percentage points, and those with forage banks were more likely to report that these effectively covered feed requirements during scarcity periods, with an estimated effect of 8.1 percentage points.
Beyond specific production and conservation practices, OGSC strengthened managerial resilience. Among producers who kept farm records, those in the treatment group were more likely to analyze this information and use it for decision-making, with an estimated effect of 4.8 percentage points. The project also increased the likelihood of allocating specific farm areas to conservation by approximately 5.1 percentage points, reinforcing the environmental co-benefits of climate-smart livestock systems. However, effects on credit access and credit approval were limited or statistically insignificant, suggesting that advisory services alone may be insufficient to overcome structural barriers related to finance, market access, and institutional support.
Overall, the findings demonstrate that well-designed rural advisory systems can play a central role in shifting extensive livestock landscapes toward climate-smart trajectories, especially when they combine peer learning, individualized technical support, demonstration-based evidence, and continuous monitoring. The results also suggest that low-cost, management-based practices are particularly responsive to advisory-driven behavior change, making them strategic entry points for scaling. Sustaining and expanding these gains will require deeper analysis of heterogeneity across departments and production systems, continued follow-up, and stronger linkages between advisory services and enabling conditions such as finance, institutional support, and market incentives. In this sense, OGSC provides practical and evidence-based lessons for advancing productivity, climate resilience, and environmental sustainability in one of Colombia’s most important cattle-ranching regions
External Link(s)

Registration Citation

Citation
Muñoz-Mora, Juan Carlos, Mariangela Ramirez and Andres Sanchez-Saldarriaga. 2026. "Orinoquia Climate smart livestock and Climate." AEA RCT Registry. May 11. https://doi.org/10.1257/rct.18562-1.0
Sponsors & Partners

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
Experimental Details

Interventions

Intervention(s)
The Orinoquia Climate-Smart Cattle Ranching program (OGSC — Orinoquia: Ganadería Sostenible y Clima) delivers a 12-month structured Rural Advisory (RA) package to cattle ranchers in Colombia's Orinoquia region (departments of Meta, Arauca, Casanare, and Vichada). Farms in the treatment group receive technical advisory and capacity-building services designed to promote the adoption of context-specific climate-smart cattle ranching (CSCR) technologies and practices — including silvopastoral systems, improved pastures with rotational grazing, forage banks, live fences, protection of water sources, and improved record-keeping for farm decision-making.
The advisory package is delivered through six sequential and cumulative cycles, each combining collective learning activities (technical workshops and field days) with individualized on-farm visits: (1) Participatory Farm Planning and Diagnostic Assessment; (2) Forage and Nutritional Management; (3) Pasture Diversification and Improvement; (4) Conservation of Natural Resources and Landscape Management; (5) Sustainable Management of Natural Resources and Climate Adaptation; and (6) Animal Management and Herd Administration. Demonstration farms and producer-to-producer exchange are used as catalysts for diffusion, and a complementary digital extension and monitoring layer reinforces continuity of contact and feedback. Farms in the control group do not receive OGSC advisory services during the evaluation period and continue operating under prevailing conditions.
Intervention (Hidden)
OGSC is operationally structured around four interlinked components. Component 1 (implemented in 2022 by CIPAV and FEDEGÁN) identified and ranked a portfolio of context-based CSCR technologies for the Orinoquia using a multi-criteria framework (productivity, mitigation, adaptation, economic viability, and technological complexity), validated through a systematic literature review and participatory workshops with experts and producers. Component 2 — the core treatment evaluated here — delivers the structured Rural Advisory package described above through FEDEGÁN's extension network. Component 3 strengthens the digital communication and monitoring layer, lowering the marginal cost of repeated contact with producers and enabling continuous feedback. Component 4 supports knowledge-exchange forums for producer-to-producer dialogue and connects implementation learning to broader policy and institutional processes.
The Rural Advisory model is grounded in a behavior-change framework (COM-B: Capability, Opportunity, Motivation) and a sequential adoption model spanning awareness, relative-advantage assessment, learning, behavioral change, trial, adoption, and review/modification. Rural advisors act as facilitators of knowledge co-construction rather than one-way technology-transfer agents. Each cycle nests multiple activities: community-level technical workshops, demonstration-farm visits, collective field days, and at least one individualized on-farm visit per cycle to translate general guidance into farm-specific actions adjusted to biophysical, productive, and socioeconomic conditions. Implementing partners include Universidad EAFIT, CIAT, CIPAV, FEDEGÁN, and the World Bank / BioCarbon Fund–ISFL.
Intervention Start Date
2024-07-01
Intervention End Date
2025-07-29

Primary Outcomes

Primary Outcomes (end points)
Adoption of climate-smart cattle ranching (CSCR) practices, measured at the farm level via binary and ordinal indicators in the baseline–endline household survey. The pre-specified primary outcomes are:

Tree-based pasture integration — increased tree cover in grazing areas in the past year; increased vegetation islands / wooded patches in paddocks; tree/shrub corridors connecting conservation areas.
Live-fence adoption — increased live fences in paddocks (barbed wire on live stakes/trees; electric fence on live stakes/trees), and corresponding reduction in fencing configurations based on dead posts.
Feed resilience — presence of cut-and-carry forage banks, and producer-reported sufficiency of forage banks during dry-season / scarcity periods.
Conservation behavior — allocation of areas dedicated to on-farm conservation and willingness to undertake ecological restoration.
Managerial capacity — keeping livestock-system records and using recorded information as an input for decision-making.
Microclimate-management perception — producer-reported sufficiency of shade in paddocks (diagnostic/perception outcome)
Primary Outcomes (explanation)
Outcomes are measured at the farm level via a structured household survey administered at baseline (pre-intervention) and endline (post-intervention), with consistent instruments across waves. Most adoption indicators are coded as binary variables (1 = adopted/present; 0 = not adopted/absent) directly from standardized survey items. For categorical responses (e.g., fencing type in paddocks), separate dummies are constructed for each configuration of interest. The "use of records for decision-making" indicator is conditional on reporting that records are kept, and equals 1 if the producer also reports analyzing the recorded information and using it as an input for decision-making. The "forage-bank sufficiency" indicator is similarly conditional on having a forage bank, and equals 1 if the producer reports that forage bank production effectively covers feed requirements during scarcity periods. Treatment effects are interpreted as the change in the probability of adoption (or, for ordinal indicators, in the average level) attributable to OGSC advisory services, estimated via a Difference-in-Differences (DiD) specification with progressive sets of covariates capturing climate exposure, household composition, farm area, and livestock structure. Standard errors are clustered at the vereda level.

Secondary Outcomes

Secondary Outcomes (end points)
(i) Knowledge and awareness of CSCR practices and their relative advantage over traditional management; (ii) producer motivation, capability, and perceived opportunity to adopt CSCR (COM-B dimensions, captured in qualitative work and structured survey items); (iii) household and farm-level socioeconomic indicators — share of income from livestock activities, access to and approval of credit for cattle activities, use of mineral salt, vaccination, and animal identification; (iv) heterogeneity of impacts by ecoregion (Piedmont, Altillanura, floodplain), farm size, herd size, and gender of the household head; and (v) community-level outcomes related to participation in farmer-to-farmer information exchange and peer learning forums.
Secondary Outcomes (explanation)
Knowledge, awareness, motivation and barrier-related outcomes are constructed both quantitatively (Likert-style and structured items in the household survey) and qualitatively (semi-structured interviews and focus groups analyzed under the COM-B framework — Capability, Opportunity, Motivation, Behavior — complemented by a hermeneutic approach). Heterogeneity analyses estimate the DiD coefficient interacted with subgroup indicators (ecoregion fixed effects, terciles of farm and herd size, gender of household head). Community-level participation indicators are constructed from extensionist records and from a community-level survey instrument capturing the number of producers engaged in collective activities and the frequency of peer-to-peer interactions on CSCR topics.

Experimental Design

Experimental Design
The evaluation employs a randomized controlled trial with a clustered design and pre-/post-intervention measurement, supporting a Difference-in-Differences (DiD) estimation strategy. Cattle farms in the Orinoquia region are organized into a treatment group, which receives the full 12-month OGSC Rural Advisory model (collective learning, individualized farm visits, demonstration farms, and digital extension/monitoring), and a control group, which does not receive OGSC services during the evaluation period. Sampling follows a three-stage stratified design: (i) Orinoquia eco-regions (Piedmont, Altillanura, and floodplain) serve as strata / primary sampling units; (ii) veredas (rural communities) within each eco-region serve as clusters and are assigned to treatment or control; (iii) farms within each selected vereda constitute the ultimate sampling units. Standardized farm-level surveys are administered at baseline (prior to exposure) and endline (after completion of the six advisory cycles) to enable within-farm and between-group comparisons over time.
Experimental Design Details
Treatment was assigned at the vereda (cluster) level, with Orinoquia eco-regions used as strata to ensure representation of the three dominant production landscapes. Treatment veredas were drawn from six municipalities (La Macarena, Cubarral, El Dorado, La Primavera, Paz de Ariporo, and Arauca) where OGSC could be operationally implemented through the FEDEGÁN extension network and partner organizations. To strengthen comparability and address concerns about non-random selection of treatment communities, control veredas were selected via a genetic-matching algorithm applied to two candidate samples (a) all municipalities in the Orinoquia, and (b) the subset of municipalities pre-selected by CIPAV using covariates capturing livestock production characteristics, agroecological conditions, and geographical features. The matching procedure iteratively minimized genetic distances between treatment and control units across the covariate vector, with balance assessed via standardized mean differences and balance tests on observable characteristics.
The estimation strategy combines this matched cluster-RCT structure with a DiD specification, progressively adding controls for climate exposure, household composition, farm area, and livestock structure to assess robustness. The change-score specification ΔYᵢ = α + β·Treatᵢ + Xᵢ′γ + εᵢ absorbs time-invariant unobserved heterogeneity at the farm level; standard errors are clustered at the vereda level. The qualitative component, conducted in three phases (initial diagnosis, midterm assessment, and final evaluation), uses focus groups, semi-structured interviews, and participant observation in selected municipalities (Paz de Ariporo, El Dorado, La Primavera) and is analyzed under the COM-B and hermeneutic frameworks.
Randomization Method
Randomization at the cluster (vereda) level was complemented by genetic matching to select control veredas from candidate municipalities in the Orinoquia using covariates on livestock production characteristics, agroecological conditions, and geographical features. Within selected treatment veredas, eligible farms were drawn from FEDEGÁN's national vaccination database following predefined inclusion criteria and operational feasibility constraints.
Randomization Unit
The vereda (rural community / cluster) is the unit of randomization. Eco-regions (Piedmont, Altillanura, floodplain) are used as strata. Within each treatment or control vereda, a sample of farms is drawn as the ultimate observational unit.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
22 veredas / clusters (11 treatment + 11 comparison), distributed across the three Orinoquia eco-regions (Piedmont, Altillanura, and floodplain). The treatment clusters were drawn from six municipalities (La Macarena, Cubarral, El Dorado, La Primavera, Paz de Ariporo, and Arauca).
Sample size: planned number of observations
1,100 cattle farms (550 treatment + 550 control), with an average of 50 farms per cluster.
Sample size (or number of clusters) by treatment arms
Planned: 11 treatment veredas (550 farms) vs. 11 control veredas (550 farms) → 22 veredas / 1,100 farms total.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Power calculations assumed 80% statistical power, a 95% two-tailed confidence level, 10% non-response, an intra-cluster correlation between baseline and follow-up of 25%, and covariates explaining 30% of the variance in the outcome. For the selected design — 22 veredas (11 treatment, 11 control) across the Orinoquia eco-regions and 50 farms per cluster (total 1,100 farms) — the minimum detectable effect (MDE) for farm-level outcomes is 0.20 standard deviations of the outcome variable. Alternative scenarios yielded MDEs of 0.50 SD (9 farms/cluster, n = 198), 0.40 SD (13 farms/cluster, n = 286), and 0.35 SD (17 farms/cluster, n = 374). The chosen 0.20 SD MDE is at the lower bound of effect sizes typically reported in the agricultural-extension literature (0.2–0.3 SD).
Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
IRB Approval Date
IRB Approval Number

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
Yes
Intervention Completion Date
July 29, 2025, 12:00 AM +00:00
Data Collection Complete
Yes
Data Collection Completion Date
September 30, 2025, 12:00 AM +00:00
Final Sample Size: Number of Clusters (Unit of Randomization)
22 veredas (rural communities) actually treated and observed: 11 treatment veredas receiving the full 12-month OGSC Rural Advisory model, and 11 control veredas that did not receive OGSC advisory services during the evaluation period. The 22 clusters span six municipalities across the three Orinoquia eco-regions (Piedmont, Altillanura, and floodplain).
Was attrition correlated with treatment status?
No
Final Sample Size: Total Number of Observations
1,108 cattle farms observed at baseline (538 in the treatment group; 570 in the control group), pooled with endline observations into an analytical panel of 2,133 farm-wave observations used in the Difference-in-Differences estimation.
Final Sample Size (or Number of Clusters) by Treatment Arms
Treatment arm (full OGSC Rural Advisory model): 11 veredas / 538 farms at baseline. Control arm (no OGSC services during the evaluation period): 11 veredas / 570 farms at baseline. Total: 22 veredas / 1,108 farms at baseline; 2,133 farm-wave observations in the pooled baseline–endline panel.
Data Publication

Data Publication

Is public data available?
No

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Program Files

Program Files
No
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials