A systems perspective on what I think this conversation still needs
A fascinating debate is unfolding in the regenerative agriculture world. John Kempf, in his recent Acres U.S.A. piece “The Take-Half, Leave-Half Fallacy,” argues that this foundational grazing principle has failed farmers who’ve practiced it for decades without building soil organic matter. His solution: graze grasses hard to favour forbs, which produce more root exudates and therefore more stable soil carbon.
Laura Paine and Jason Cavadini responded with a defence of take-half, leave-half, arguing it was never about soil carbon in the first place—it’s about maximising cool-season pasture productivity by keeping plants photosynthesising continuously.
Both make valid points. But I believe we’re still missing something fundamental.
This debate reveals a tendency that runs through much of the regenerative agriculture conversation: even when we’re thinking in systems terms, it’s easy to slip into optimising for single metrics and universal rules. Whether we’re focused on soil carbon (Kempf) or pasture productivity (Paine & Cavadini), there’s a risk of the same reductionist thinking that got industrial agriculture into trouble in the first place.
The Limits of Single-Metric Thinking
Kempf’s argument hinges on Christine Jones’s research suggesting that over 90% of carbon sent out as root exudates remains stable in soil long-term, compared to just 10% of vegetative biomass carbon. From this, he concludes that forb-dominant pastures—which allocate roughly 35% of sugars to exudates versus 5% for grasses—are the key to building soil carbon.
Paine and Cavadini counter that soil carbon accumulation is constrained by soil type and climate more than management, that mineral-associated organic matter (MAOM) tends toward saturation, and that ecosystems move toward equilibrium. You can’t force a soil to store more carbon than it needs to function.
Both are right, as far as they go. But carbon is just one metric among many.
I’ve spent nearly 30 years working with farmers across diverse landscapes, and I’ve come to believe that our obsession with soil carbon—important as it is—has become a distraction from what actually creates functional, resilient grazing systems.
What We're Actually Managing For
When I teach regenerative grazing, I use a framework based on the four elemental conditions that govern landscape function:
Earth (Soil Health): Yes, this includes carbon and organic matter. But it also includes soil structure, aggregate stability, microbial diversity, and the biological activity that converts exudates into humus. A herbal ley with an open structure that allows frost penetration in winter will damage that microbial community—regardless of how many exudates the forbs are producing.
Water (Hydrology): Walter Jehne’s work on the soil carbon sponge reminds us that water dynamics may matter more for climate function than carbon sequestration per se. As Jehne argues, the hydrological cycle governs 95% of Earth’s heat dynamics. Restoring the soil’s capacity to infiltrate, retain, and slowly release water—through good structure and ground cover—is arguably more important than any specific carbon number.
Air (Atmosphere): Good soil structure enables gas exchange—nitrogen fixation, CO₂ release from decomposition, oxygen availability for aerobic biology. The insulating layer of residual plant material protects soil surfaces from temperature extremes that kill microbes.
Energy (Fire/Primary Production): This is about maximising the capture of solar energy through photosynthesis—something both Kempf and Paine agree on, though they disagree about how to achieve it.
The point is: we’re not managing for carbon. We’re managing for whole-system function, of which carbon is one indicator among many.
The Toolbox, Not the Rulebook
The debate between “graze hard to favour forbs” and “take half, leave half to maintain productivity” presents these as competing systems. But any experienced grazier knows they’re both tools—and good management requires knowing when to use which.
My methodology, which is underpinned by Holistic Planned Grazing, provides three core tools:
- Rest: Total rest from trampling and grazing, allowing plants to recover fully and complete their growth cycles
- Animal Impact: Trampling plants and chipping soil surfaces to stimulate growth, improve water infiltration, and cycle nutrients
- Minimising Overgrazing: Avoiding repeated grazing of plants before recovery, which weakens root systems
The critical insight is that these tools are applied to achieve specific goals in your specific context. There is no universal rule that works everywhere, all the time.
Sometimes I need to rest pastures for an entire summer to build biomass, allow plants to set seed, and break parasite cycles. Research shows that grass maintained at 15-25cm supports twice the invertebrate numbers of grass grazed below 10cm, with significantly more web-building spiders (Wakeham-Dawson et al., 1998). Those spiders help regulate ticks and face flies—reducing the need for insecticides, which in turn supports healthier soil biology.
Sometimes I need to graze hard in spring to prevent tussock grasses becoming dominant and to stimulate tillering in the main platform pastures I’ll rely on through the growing season.
Sometimes I’ll deliberately take grasses low when I feel cocksfoot or other coarse species are taking over.
Sometimes I’ll stockpile for winter grazing, accepting that the “rules” say I should have grazed that growth.
The person and mindset managing the system—their observation skills, their understanding of their specific landscape, their clarity about what they’re trying to achieve—matters far more than any grazing formula.
What About Species Diversity?
Kempf argues that take-half, leave-half favours grasses over forbs because grasses store energy in roots and recover faster, outcompeting forbs whose energy is in exudates. His solution is periodic hard grazing to reset the competitive balance.
But this oversimplifies the complexity of plant community dynamics—and misses a crucial point about how diversity actually works.
Yes, we want diverse swards including forbs, legumes, and grasses. But the research shows it’s not just species diversity that matters—it’s functional group diversity. Having grasses, legumes, and forbs together creates complementarity effects that none of them can achieve alone.
Fornara and Tilman (2008) found that high-diversity mixtures stored 500% more soil carbon than monocultures of the same species. But here’s what’s critical: the presence of C4 grasses and legumes together increased soil carbon accumulation by 193% and 522% respectively. The mechanism? Legumes fix nitrogen, which C4 grasses use efficiently, and their activity is differentiated in time—cool-season legume nitrogen fixation followed by warm-season grass uptake. It’s the combination that works.
Research from the Pyrenees (2021) confirmed this at landscape scale: soil carbon storage was maximised when grasses and forbs were co-dominant at optimal legume proportions (7-17%). Forb dominance alone actually reduced the positive effects. The niche complementarity of all three functional groups—different root depths, different growing seasons, different nutrient strategies—is what drives carbon storage.
The Jena Experiment in Germany—one of the world’s longest-running biodiversity studies, now over 20 years old—has tested mixtures from 1 to 60 species and found that higher plant diversity increases carbon storage through a mechanism neither Kempf nor Paine fully addresses: it’s the soil microbial community. Higher diversity increases rhizosphere carbon inputs, which drives increased microbial activity, which in turn drives carbon storage. The diversity effect strengthens over time as the soil biology matures.
So when Kempf suggests we should graze hard to favour forbs, he’s potentially undermining the very functional group diversity that creates the carbon sequestration he’s after. Managing for forbs at the expense of grasses isn’t the answer. We need all three functional groups working together—and the grazing management to support that balance, not disrupt it.
And there’s the practical problem: achieving diversity through overgrazing is a gamble. As Paine notes, you might get beneficial forbs—or you might get thistles, docks, and burdock. The outcome depends on what’s in the seedbank, soil conditions, timing, and factors outside your control.
More importantly, in a diverse sward, plants will always be at different reproductive stages at different times. Some will be vegetative, some at boot or bloom stage, some setting seed. Suggesting we should manage the whole sward to a particular growth stage is frankly impossible—and probably undesirable anyway.
What we want is a range of plants producing different root exudates with different chemistry, at different times, feeding a diverse soil food web. That diversity may be what creates disease-suppressing soil biology—yet another function that neither the carbon-maximising nor the productivity-maximising framing accounts for.
But here’s the thing: we can’t force diversity by grazing hard and hoping for the best. Diversity emerges from healthy soil conditions. Which brings me to what I think is the missing dimension in this entire debate.
Managing for Soil Succession, Not Just Species Composition
I’m less interested in managing for specific plant functional groups than I am in advancing overall soil succession. Dr Elaine Ingham’s work on the soil food web demonstrates that as ecosystems mature, the ratio of fungi to bacteria shifts—from bacterial-dominated early successional soils (which favour weeds and annual plants) toward fungal-dominated soils (which support perennials, shrubs, and eventually forest). Grasslands sit somewhere in between, with healthy productive pastures typically showing a fungi:bacteria ratio approaching 1:1.
Dr David Johnson’s research at New Mexico State University reinforces this. His work with the BEAM (Biologically Enhanced Agricultural Management) system has documented that fungal-dominant soils utilise five times more carbon in plants than bacterial-dominant soils—and that maximum productivity and carbon capture occur when the fungal:bacterial ratio reaches approximately 1:1. In his field trials, Johnson observed 10.27 metric tons of soil carbon sequestered per hectare per year over 4.5 years—a rate 20-50 times higher than conventional no-till soils. The key wasn’t a specific grazing rule or plant composition, but restoring the microbial community structure that enables efficient carbon cycling.
Allen Williams of Understanding Ag and the Soil Health Academy teaches a similar principle: that we should be managing toward higher successional plant communities with the soil biology to match. As Williams puts it, regenerative agriculture requires adaptive stewardship—flexing and adjusting according to what nature presents rather than following fixed formulas.
If I can build good soil aggregation and healthy community dynamics in my soil-plant system, I create the conditions for natural succession. Better germination conditions emerge for higher-order plants—species with more favourable physiology like wider leaf blades and more leaves before going to seed. The soil itself begins to signal for these plants.
This is where quorum sensing becomes relevant. Research increasingly shows that soil microbial communities communicate through chemical signalling molecules, coordinating behaviours based on population density. As ATTRA’s work on quorum sensing explains, when environmental conditions shift fungal and bacterial communities toward greater complexity, these microbial signals can create germination stimulants that were previously absent—potentially triggering the return of native perennials whose seeds have been present in the seedbank all along, simply waiting for the right biological conditions.
Managing aggressively for forbs—while potentially increasing exudate production in the short term—may not allow this natural progression of soil succession toward a more favourable fungal:bacterial balance. We might be optimising one variable while inadvertently disrupting the deeper biological dynamics that create the conditions for long-term pasture improvement.
This is emerging science, and I hold it with appropriate humility. But it points toward why simple prescriptions about which plant functional group to favour may miss the bigger picture of what makes soil-plant systems thrive over time.
The Whole-System View
Here’s what’s missing from the debate: the recognition that we’re not just managing pastures—we’re managing whole farm systems embedded in whole landscapes.
A grazing plan that maximises soil carbon but doesn’t produce a sod that holds up to winter poaching is useless to a farmer who needs to outwinter cattle.
A forb-dominant pasture that’s excellent for carbon but wrong for the livestock enterprise—dairy cows need different sward composition and maturity than Highland cattle—fails at the practical level.
And none of this accounts for the carbon cost of what happens when grazing systems don’t work: making hay, feeding hay, bedding cattle, feeding concentrates, carting muck. If your grazing plan falls apart in year one because it doesn’t fit the farm’s infrastructure, livestock classes, or the family’s capacity to implement it, then zero carbon gets sequestered regardless of what the theory says.
I think about whole-system design as optimising across multiple dimensions simultaneously:
- The soil system: structure, biology, hydrology, carbon, nutrient cycling
- The plant system: productivity, diversity, resilience, seasonal availability
- The animal system: our livestock health and performance, wildlife habitat, pest regulation
- The whole ecosystem: woodland, hedges, ponds, wetlands—the functional landscape that regulates pests and disease
- The human system: the family running the operation, their skills, capacity, and wellbeing
A grazing “system” that optimises one dimension while ignoring the others isn’t a system at all. It’s a theory waiting to collide with reality.
The Case for Simple Rules—Strategically Applied
Here’s something neither side of the debate acknowledges: simple rules like “take half, leave half” have enormous value precisely because they’re simple.
When I’m working with a farmer new to managed grazing, I don’t start with the complexity. I start with something like: Take half or less, leave half, aim for a mix of forbs, legumes and grasses, and move livestock before they regraze recovering plants.
That’s not the whole picture. But it’s a foundation that gets farmers started, builds their observation skills, and creates early wins that build confidence. Over a year or two, we layer in complexity: monitoring, adjusting for conditions, understanding when to break the rules and why.
The danger of the Kempf argument isn’t that periodic hard grazing is wrong—it’s a useful tool. The danger is that inexperienced farmers might apply it universally, without understanding their specific context, and damage pastures they don’t yet have the skills to recover.
The best graziers I know—the ones who’ve been doing this for decades—use all the tools. They monitor constantly. They make real-time, place-based decisions. They hold “rules” lightly, as starting points rather than destinations.
That adaptive capacity, not any particular grazing formula, is what regenerative grazing is really about.
Beyond the Debate
So where does this leave us?
Both Kempf and Paine are contributing valuable thinking to the conversation. The research on root exudates and MAOM stability matters. The understanding of plant physiology and recovery curves matters. The recognition that soil carbon has limits matters.
But the conversation we need isn’t “which grazing rule is correct.” It’s: How do we develop farmers who can read their landscapes, clarify their goals, and apply the right tools at the right time?
That’s a question about mindset and capacity more than technique. It’s about observation, humility, and adaptive management. It’s about understanding that in complex living systems, there are no universal rules—only principles to guide context-specific decisions.
The take-half, leave-half debate is a symptom of our collective desire for simple answers to complex questions. Understandable, but ultimately limiting.
The real work is helping farmers become whole-system thinkers, capable of navigating complexity rather than seeking refuge from it.
That’s what regenerative grazing means to me.
Caroline Grindrod is a regenerative agriculture consultant and systems thinker with nearly 30 years of experience in environmental land management and regeneration. She is the founder of Roots of Nature, co-founder of Roots to Regeneration, and leads Wilderculture's work on bioregional transformation.
References
ATTRA (National Center for Appropriate Technology). Soil Social: Quorum Sensing, Part 2. Sustainable Agriculture Program.
Fornara, D.A. & Tilman, D. (2008). Plant functional composition influences rates of soil carbon and nitrogen accumulation. Journal of Ecology, 96, 314-322.
Gómez, D. et al. (2021). Positive Effects of Legumes on Soil Organic Carbon Stocks Disappear at High Legume Proportions Across Natural Grasslands in the Pyrenees. Ecosystems, 25, 960-975.
Ingham, E.R. (2000). Soil Biology Primer. Soil and Water Conservation Society.
Jehne, W. (2017). Regenerate Earth. Healthy Soils Australia.
Johnson, D.C. (2017). Biologically Enhanced Agricultural Management. New Mexico State University.
Johnson, D.C. Interview with Acres U.S.A. “Compost & The Promise of Microbes.”
Lange, M. et al. (2015). Plant diversity increases soil microbial activity and soil carbon storage. Nature Communications, 6, 6707.
Weisser, W.W. et al. (2017). Biodiversity effects on ecosystem functioning in a 15-year grassland experiment: Patterns, mechanisms, and open questions. Basic and Applied Ecology, 23, 1-73.
Paine, L. & Cavadini, J. (2025). It’s not about soil carbon — it’s about keeping cool-season pastures healthy and productive. Acres U.S.A.
Kempf, J. (2025). The Take-Half, Leave-Half Fallacy. Acres U.S.A.
Skinner, R.H. & Dell, C.J. (2016). Yield and soil carbon sequestration in grazed pastures sown with two or five forage species. Crop Science, 56(4), 2035-2044.
van Klink, R. et al. (2015). Effects of large herbivores on grassland arthropod diversity. Biological Reviews, 90(2), 347-366.
Voisin, A. (1959). Grass Productivity. Philosophical Library.
Wakeham-Dawson, A. et al. (1998). Influence of grazing management on biodiversity of grassland. British Grassland Society Conference Proceedings.
Cotrufo, M.F. et al. (2019). Formation of mineral-associated organic matter versus particulate organic matter. Nature Geoscience.
Meyer, S.T. et al. (2019). Land-use components, abundance of predatory arthropods, and vegetation height affect predation rates in grasslands. Agriculture, Ecosystems & Environment, 270-271, 84-92.