Holistic Management: What the Peer-Reviewed Data Actually Shows

The Savory Claim and Why It Created a Decade of Arguments

silvopasture: the agroforestry extension of Savory's multi-strata land management of what would become holistic planned grazing in the early 1980s, drawing on his observations of wildlife migration patterns in Zimbabwe and the Kalahari. His central claim: biochar in arid dryland systems as a complementary intervention to holistic grazing is caused not by overgrazing alone but by periods of prolonged rest, which allow standing dead plant material to oxidise rather than cycle through animal digestion. The solution, he argued, was to mimic large migratory herbivore herds by bunching composting the concentrated manure from high-density mob grazing events, and then resting paddocks for extended periods. Done correctly, he claimed, the process could reverse decades of grassland degradation.

The 2013 TED talk extended those claims to a planetary scale: watershed-scale planning as the hydrological foundation for large-scale grazing restoration could be treated by scaling holistic planned grazing, potentially sequestering enough carbon to return atmospheric CO2 to pre-industrial levels. The talk reached 10 million views and triggered a wave of both enthusiasm and scientific pushback that has not fully resolved.

The pushback was not irrational. Savory's 1984 Zimbabwe trial, which he cited as evidence for a 400 percent stocking rate increase on recovering grassland, was a single non-replicated observation at one site with no control. It was not designed as a controlled experiment and cannot carry the evidentiary weight placed on it. Scientists who spent careers running range trials on multiple sites, with controls, took issue with treating a case study as proof of a universal mechanism.

The argument this page is making is not that Savory was right or wrong. It is that the original claim and the scientific critique were largely talking past each other, and that the last ten years of longer-term AMP trials have produced a more granular picture. The mechanism is real in specific conditions. The planetary claims are not supported. The practitioner question is which conditions predict meaningful results for an individual operator.

Briske et al.: What the Meta-Analysis Actually Found

David Briske and colleagues at Texas A&M published a synthesis of range science literature in 2008 in Rangeland Ecology and Management, followed by an extended analysis in 2011. Their conclusion was direct: across the controlled trials available in the peer-reviewed literature at that point, rotational grazing systems did not show statistically significant advantages over continuous grazing on either forage yield or animal performance when stocking rates were equivalent. The meta-analysis covered more than 15 years of published range trials across multiple regions.

This was reported in some circles as a decisive refutation of Savory. It was not. Briske and colleagues were evaluating the literature as it existed, and that literature had a systematic problem: almost none of the trials used grazing protocols that matched Savory's actual recommendations. The studies that showed no benefit compared systems with 2-4 paddock rotations and 30-60 day return intervals. Savory's protocols called for 60-100 paddocks, return intervals of 60-180 days depending on season, and densities an order of magnitude higher than those used in most trials. The meta-analysis was a valid summary of what had been tested. It was not a test of the Savory protocol itself.

Briske acknowledged this limitation explicitly in the 2011 paper. The honest summary of his findings is: simple rotational systems with short paddock numbers and moderate densities do not outperform continuous grazing. This does not test AMP. The distinction matters because most ranchers adopting "rotational grazing" use the simpler system, not the AMP protocol, and Briske's findings apply directly to them.

Texas A&M AMP Trials: What Long-Term Data Shows

Richard Teague and colleagues at Texas A&M AgriLife Research have run the most rigorous long-term comparison of AMP grazing against continuous and simple rotational systems in North America. Their 2016 paper in the Journal of Soil and Water Conservation examined 13 AMP grazing sites in the Northern Great Plains, comparing soil organic carbon (SOC), water infiltration rates, and ground cover against paired continuous grazing sites at equivalent annual stocking rates.

The AMP sites showed soil organic carbon accrual rates of 0.2-0.7 tonnes of carbon per hectare per year, compared to effectively flat or declining SOC trajectories at continuous grazing comparators . Water infiltration at AMP sites averaged significantly higher than at continuous grazing sites, a result attributed to improved ground cover maintaining soil structure and suppressing surface sealing from direct rain impact. Ground cover percentage was higher across all AMP sites despite equivalent or higher stocking rates.

The mechanism proposed by Teague is consistent with the pulse disturbance hypothesis: plants grazed briefly at high intensity and then rested for full photosynthetic recovery allocate proportionally more carbon to root growth and rhizosphere exudate production than plants under continuous moderate defoliation. Those root exudates feed the fungal-bacterial network that builds stable soil organic matter. Continuous moderate grazing suppresses this cycle without delivering the complete rest period required for root carbon recharge.

The soil carbon data from Teague's sites overlaps with evidence from the regenerative agriculture literature on cover crop and reduced tillage systems, where soil carbon accrual rates of 0.3-0.9 tonnes per hectare per year are documented under best management conditions . The grazing contribution is additive when combined with perennial grass root architecture, which builds deeper and more stable carbon pools than annual crop systems.

Critically, Teague's AMP sites were operating with genuinely high density and long rest periods. These are not typical rancher rotations. The operations studied were running 60-plus paddock systems with rest periods calibrated to full plant physiological recovery. The results are real, but they are not evidence that any rotational system will produce the same outcome.

Dimbangombe Ranch and What the Long-Term Site Data Shows

Dimbangombe Ranch in Zimbabwe is the Savory Institute's flagship demonstration site. Located in a semi-arid savanna environment, it has been under managed holistic planned grazing since the late 1990s. The site is important because it operates in the high-brittleness environment where Savory's original arguments were developed, and it has accumulated multi-decade monitoring data.

The documented outcomes at Dimbangombe include a substantial increase in perennial grass cover on land that was previously degraded bare ground, increased water retention in the landscape (evidenced by river base flows during dry seasons where none had been measured previously), and a 300 percent increase in stocking capacity over the management period . These are meaningful numbers. They are also self-reported by the institution that has a reputational stake in the outcome.

Independent monitoring at Dimbangombe is limited. A 2014 review by Briske and others noted that while the Dimbangombe outcomes were plausible and consistent with the pulse-disturbance mechanism operating in a highly brittle environment, the data had not been subjected to peer review and lacked paired control sites. The absence of controls is a real limitation. Grassland recovery can occur over multi-decade timescales under any reduced grazing pressure, and separating the specific contribution of holistic planned grazing from general destocking is difficult without controls.

The practitioner implication is not that Dimbangombe is fabricated. It is that a single well-managed site, even with multi-decade data, cannot resolve the scientific question about whether the specific management protocol is responsible for the outcomes rather than reduced animal pressure alone. Operators looking to Dimbangombe as proof should understand what kind of evidence it is: a detailed case study in a specific environment, not a controlled trial.

What Dimbangombe does demonstrate is that the brittleness-environment version of the Savory method can produce dramatic landscape-scale recovery when implemented with full commitment over decadal timescales. That is a meaningful data point. The question of generalisability to other environments and management contexts remains genuinely open.

What the Evidence Means for Operators and What Remains Genuinely Contested

The clearest synthesis of the current evidence base is this: the AMP grazing mechanism (high density, short duration, long recovery) produces measurable soil organic carbon gains and water infiltration improvements in semi-arid grasslands when implemented with genuine protocol fidelity. The gains are real and economically meaningful, particularly when graziers are tracking soil health as part of their land value calculation. For operators considering the yield and margin comparison between management systems, soil health trajectory is the variable most likely to compound over time.

What remains contested: the landscape-scale carbon sequestration claims. Grassland systems have finite soil organic carbon carrying capacity, and the sequestration rates measured at AMP sites are likely to plateau as soils approach their regional carbon ceiling. The claim that grassland restoration alone can materially offset global fossil fuel emissions is not supported by the current evidence base, regardless of management quality. That is a different question from whether an individual operation can build meaningful soil carbon, and it is important to keep them separate.

The methane accounting question is also unresolved. Grass-finished cattle emit biogenic methane that cycles through the short-term carbon pool. At stable or declining herd sizes, this methane does not represent a net addition to atmospheric carbon because the short atmospheric lifetime (roughly 9-12 years for methane) means stock equilibrate. At growing herd sizes, the methane contribution is real and must be offset by sequestration. Most published carbon life cycle assessments, including the widely cited Stanley et al. 2018 analysis of White Oak Pastures, do not fully disaggregate this accounting.

For the operator deciding whether to adopt AMP protocols: the strongest evidence supports adoption in semi-arid brittleness-spectrum grasslands with adequate paddock infrastructure and genuine commitment to long recovery periods. For high-rainfall temperate pastures, the evidence is weaker, and simpler rotational systems may produce comparable outcomes at lower infrastructure cost. For operators interested in how virtual fencing technology changes the infrastructure economics of multi-paddock systems, that comparison is now a tractable calculation rather than a theoretical discussion.

The adoption condition that predicts success is not ideological alignment with Savory. It is environmental match (brittleness, aridity, grass type) plus management capacity (ability to monitor and adjust, infrastructure for genuine high-density brief grazing events, willingness to rest land for 60-plus days). Operators who meet those conditions should seriously consider the approach. The soil carbon and water infiltration outcomes are the most reliable return on the transition investment, and both connect directly to the long-term soil health arguments that underpin regenerative system economics more broadly.