LULUCF-Based Assumptions

Surprisingly, the presentations suggest that livestock agriculture has significantly higher land-related emissions compared to crop agriculture. This contradicts common agricultural practices, where emissions from livestock are expected to be lower than those from crops. Furthermore, there seems to be no consideration for the maintenance of CO2 soil stocks in grasslands.

The source study appears to be a mixture of actual data (land areas and yields), assumptions, and projections regarding climate and environmental impacts, which are fraught with inaccuracies and uncertainties. Scaling these results to a global level is misguided; if the initial results are flawed, amplifying them will only magnify the errors.

In the following sections, I will explore these issues in more detail, starting with OWiD’s assumptions:

1. Land Use Change

Livestock agriculture is characterized by stability. Most grasslands have evolved through centuries of consistent use, and herd sizes typically remain constant year to year, meaning land use rarely changes. Significant land use changes, such as converting forests or peatlands to grasslands, are only necessary if there is a substantial increase in herd size. While such changes might occur in regions expanding their livestock, emissions from land use change in most places are generally negligible.

2. Land/Pasture Use and Management

In both versions of the presentation, we must question the relevance of the extensive land areas linked to ruminants in terms of climate impact. Instead of merely asking about the number of hectares, we should consider the various functions this land serves and their respective impacts on the climate, environment, and food security.

Much of the grassland used for livestock is marginal land, unsuitable for crop production, meaning it does not compete with the cultivation of plant-based foods. Additionally, grasslands feature deep root systems that have historically sequestered considerable amounts of carbon.

When managed appropriately (e.g., moderate grazing), pastureland can act as a carbon sink. For instance, meadows used for hay might only be tilled every several years, thus sequestering CO2 with minimal disruption. Similarly, permanent pastures, which are not arable, are likely to remain untended. Grasslands, second only to forests, are among the planet’s largest terrestrial carbon sinks, containing approximately 12% of the terrestrial carbon stock. However, OWiD’s presentation does not acknowledge this sequestration or the preservation of significant carbon stocks in grassland soils, instead depicting them as net sources of emissions overall. While variations in sequestration outcomes can arise from multiple factors, especially within the topsoil layer, such claims warrant scrutiny. Evidence suggests that well-managed grazing can enhance carbon sequestration in grassland soils.

Although crop agriculture also sequesters CO2, primarily in the topsoil, this process is disrupted annually due to practices like tilling, which releases stored carbon back into the atmosphere. Additionally, every tilling operation contributes further CO2 emissions, such as those from diesel fuel. Weed and pest control practices add to the emissions associated with crop agriculture, yet OWiD’s presentations appear to overlook these factors across all listed crops.

In summary, the gross emissions from livestock land and pasture management are usually minimal when compared to crop agriculture. For permanent pastures, gross emissions are likely close to zero under good management practices. Net emissions are expected to be negative for both pastures and grasslands, owing to carbon sequestration and less intensive land management. OWiD’s presentations appear to conflict with established agricultural practices and their related emissions and climate impacts. If its figures align with the stated sources, I remain skeptical of their accuracy.

3. Liming, Fertilizing, and Irrigation of Pastures

Because meadows and pastures are rarely tilled, they are less intensively managed, resulting in liming and irrigation activities occurring less frequently than in crop agriculture. Fertilization of meadows generally utilizes manure instead of chemical fertilizers, which generate significant CO2 emissions during production. Permanent pastures primarily receive nutrients through natural grazing by livestock. The assumptions in the presentations seem misaligned with typical agricultural practices, which usually see higher emissions associated with crop agriculture than with livestock.

What about Slaughter Waste and Food Waste?

The presentation also mentions emissions linked to slaughter and food waste. These will be discussed further in the section titled Additional Thoughts on the Source Paper.

Summarizing My Thoughts on LULUCF

OWiD’s presentations seem to exaggerate certain LULUCF aspects while neglecting other crucial elements. Overall, emissions associated with land use changes in livestock appear overstated, while those linked to crop agriculture seem underestimated or ignored. Systematic omissions of carbon stocks and sequestration in grasslands further skew the narrative. The resulting portrayal does not align with typical agricultural realities.

Multi-Purpose Upcycling

Both sequestration and upcycling appear to be absent from the presentations and the source paper. Ruminants function as bioreactors powered by solar energy, efficiently converting inedible plant materials and crop waste into nutrient-rich human food with superior bioavailable proteins compared to other sources. A staggering 86% of their feed is inedible by humans, and 77% of the land dedicated to livestock feed consists of grasslands and pastures that are non-arable (Mottet et al., 2018). Professor Frank Mitloehner’s brief video presentation provides a quick reference on land areas and their characteristics.

Without livestock, we would face significant food security challenges, as there would not be enough arable land to replace the protein lost from meat production through increased crop cultivation. Furthermore, the accumulation of crop waste currently used as livestock feed would pose a substantial environmental problem, exacerbating methane emissions from landfills. By converting plants and crop waste from marginal land into human-edible protein, ruminants help address these environmental challenges and play a vital role in global food security.

All Land is Not Created Equal

The preceding discussion raises an essential point: there is a significant difference in the meanings and implications of "land use" between arable and marginal land. Is it fair to treat both types equally? Should we evaluate the use of arable versus marginal land with different criteria?

These two primary land categories serve distinct purposes and have unique limitations, making direct comparisons misleading. Plant-based foods necessitate arable land, a finite resource requiring extensive management and preparation, resulting in higher CO2 emissions. Using arable land for crops means it is entirely dedicated to food production and cannot simultaneously serve other functions.

In contrast, marginal land, which is non-arable, cannot be utilized for growing plant-based crops. Without grass and ruminants, it would yield no valuable contributions to human nutrition or food security. As discussed in this article, marginal land generally acts as a considerable carbon sink. However, when associating "land use" with meat production and grazing, it does not imply that this extensive area is being "spent" like arable land. Instead, marginal land fulfills multiple essential functions simultaneously, closely linked to the grazing of ruminants and the maintenance of grasslands.

Using ruminants to convert marginal land plants into human nutrition exemplifies effective resource management, as food produced on non-productive land does not compete with land used for plant-based agriculture. Moreover, while ruminants produce this food, they provide critical services for biodiversity and climate that cannot be performed on arable land. The deep roots of perennial grasses help maintain carbon stocks in meadows and pastures, which depend on grazing for their long-term preservation.

Additionally, the ecosystem services provided by ruminants must be recognized. Many endangered species inhabit meadows, pastures, and rangelands. The removal of ruminants would jeopardize these species and their habitats. To preserve the diversity and richness of ecosystems, ongoing grazing by ruminants is crucial. Eliminating them would not only lead to the loss of numerous species and ecosystems but also risk diminishing carbon stocks and food production, threatening global food security.

Research is increasingly showing that the use of ruminants, along with grazing and even hunting, does not harm ecosystems; rather, it supports them. For instance, the vast grasslands of the Serengeti have been shaped and sustained by humans and ruminants over millennia. Likewise, human activities such as hunting and agriculture have influenced rainforest ecosystems for tens of thousands of years. The diversity of nature and its ecosystems is largely a product of human intervention, making it unrecognizable without the presence of ruminants and agriculture.

What, then, about the fairness and representativity of the land area figures used to calculate footprints per unit, such as the carbon footprint of a kilogram of various foods (Figure 3) or their protein content? These figures appear reductive, lacking distinction between the two primary land categories. The presentation fails to consider the multifunctionality of marginal land, which provides numerous benefits simultaneously, including food production, biodiversity, and carbon sequestration. It also neglects the differing emission profiles of the two land types.

Given that arable land is a limited resource dedicated solely to food production, would it not be more beneficial to assess arable land use per unit of food product, rather than lumping arable and marginal land together as OWiD has done? Focusing on arable land would highlight the significantly smaller areas utilized by ruminants. Additionally, it could facilitate discussions about whether more arable land can be allocated solely for crop production instead of also being used for livestock.

The diagram presented (Figure 3) lacks such nuances and fails to acknowledge the benefits provided by marginal land. Without differentiating between these two land categories, the presentation propagates the misleading notion that the sole purpose of extensive marginal land is meat production. Such oversimplifications can lead to misconceptions regarding the interchangeability of different land types, perpetuating inaccuracies in the food discourse for decades.

Failing to recognize that ruminants enable the multifunctionality of marginal land and the differing emission profiles of land categories means that presentations like these cannot accurately represent the true climate or area footprint of various food products.

Reflections on the Source Paper

Did OWiD make a wise choice in using Poore and Nemecek’s paper, Reducing Food's Environmental Impacts through Producers and Consumers as a reference? I have highlighted errors stemming from overlooked, misinterpreted, or exaggerated assumptions in the paper. Did the GWP100 error distort more sections of the study than just methane estimates? These and other issues will be explored in the following sections.

Failure to Show the Climate Footprints of Foods

The source paper’s reliance on the faulty GWP100 metric assigns enteric methane a warming impact equivalent to CO2 emissions, inflating its significance by a factor of 28. However, using the more recent and improved GWP* metric (2015–20) provides a more accurate assessment of short-lived greenhouse gases, revealing that the study’s estimates are incorrect by orders of magnitude. When herd sizes remain stable, methane emissions from ruminants have negligible warming effects.

Using a GHG metric that greatly overstates the warming impact of ruminants is a fundamental error that propagates misinformation throughout the paper. Overestimating land use change emissions linked to livestock while underestimating those associated with crop agriculture further amplifies bias by misrepresenting land use and management figures. Ignoring carbon sequestration in deep soil stocks linked to livestock agriculture exacerbates the issue. These fundamental flaws lead the source paper to fail in accurately depicting the climate or carbon footprints of various foods.

Grazing, Sequestration, and Alternative Uses of Grasslands

According to the paper, "Improved pasture management can temporarily sequester carbon." This statement obscures the fact that while enhanced management might increase sequestration potential, much of today’s practices have long served as effective mechanisms for carbon storage, contributing to one of the largest existing carbon stocks.

The belief that grazing-induced sequestration is temporary may stem from studies focusing predominantly on the volatile topsoil. This upper layer, sustained by decaying plant material, is an inefficient mechanism for carbon capture. The significant role of the deep root systems of perennial grasses is often overlooked.

Research has shown that good grazing practices can positively impact sequestration outcomes. Instead of debating whether to graze or not, the focus should shift to the importance of effective pasture management in maintaining the unique sequestration dynamics of grasslands. The primary purpose of grazing is not merely to maximize carbon capture in the topsoil but to sustain the grassland ecosystem itself and its deeper-rooted carbon dynamics. Grasslands that are poorly managed due to inadequate grazing will ultimately transition into forests. Therefore, the net value of grasslands, considering their diverse uses, should not be viewed in isolation but evaluated against alternative land uses.

Research suggests that grasslands may serve as larger and more resilient carbon sinks than forests in many regions. This is partly due to the fact that outside the boreal forest zone, vast forest areas may not sequester carbon in the soil. The common notion of "rewilding" or reforesting such regions may not be favorable. These forests primarily sequester carbon in their trunks and are thus vulnerable to natural decay, drought, or pest infestations, which can release stored carbon back into the atmosphere.

Additionally, rapid turnover rates in tropical forests can diminish the effectiveness of their carbon sinks. In a protected forest, where carbon is not utilized for long-term applications, the value of the carbon sink may approach zero over a 30–40 year period. This indicates that transitioning to grassland or other agricultural uses with a net zero carbon balance may not be the climate catastrophe often assumed.

Slaughter, Processing, and Consumer Wastage

What about emissions associated with wastage from slaughter, processing, and consumption? The paper claims significant levels of wastage at the slaughtering stage and throughout the supply chain, stating, “Wastage is high for fresh animal products, which are prone to spoilage.” Additionally, it notes that “consumer waste, not assessed elsewhere in this study, is 2.5–9% higher in animal than in vegetable proteins.”

However, while the animal is more than just meat, the byproducts generated from slaughter are not wasted; they serve as raw materials for various industries and products. Ultimately, only about 1% of the animal is discarded as waste.

Regarding food wastage, the FAO estimates indicate that no food product category contributes less to global food waste than meat (Figure 4). While meat has a low wastage footprint of approximately 4%, cereals (>30%), vegetables (25%), fruits (16%), and starchy roots (19%) contribute significantly more to global food wastage.

An Important Note on Wastage Emissions Estimates

It is crucial to differentiate between actual food waste volumes and their climate footprints. The assumption made by the FAO (Figure 4) and many other studies that the carbon footprint of meat waste is nearly equivalent to that of vegetables is based on the flawed GWP100 metric. The improved GWP* metric reveals that this assumption is incorrect. Furthermore, the 2015 FAO estimate includes emissions related to feed provision (like manure management) in the meat waste footprint, even though such emissions are already accounted for by GWP* as causing minimal warming.

Clearly, Poore and Nemecek should have recognized the GWP100 bias not only in their food emission estimates but also in other derived estimates throughout the paper, including food wastage figures.

The high emissions attributed to low meat waste estimates (FAO, Poore, and Nemecek) stem primarily from enteric methane, which is grossly overstated due to GWP100. The true carbon footprint of this waste is likely closer to zero than the claimed 20%. Despite their paper being published three years after the initial GWP* paper, Poore and Nemecek ignored the opportunity for precise GHG accounting facilitated by the GWP* metric. They did not acknowledge the possibility that their choice of GHG metric could render their estimates of actual footprints and climate effects inaccurate. For a study aimed at estimating the true climate footprint of foods, this oversight represents a significant failure. The claims regarding meat wastage and emissions exemplify how methodological choices can undermine calculations throughout an entire paper.

Even the paper’s Erratum, which acknowledges the authors’ misunderstanding of carbon sink mechanisms, still contains errors stemming from derived assumptions. While grasslands are recognized as carbon sinks in the Erratum, the original paper fails to adequately address this. The assertion that removing livestock would enhance carbon sinks in grasslands is unfounded. Eliminating livestock would lead to the degradation of grasslands, giving way to shrubland and forest.

Basing the "corrected" values of potential carbon sink increases on the still erroneous GWP100 metric means that any errors introduced by GWP100 may only compound. A significant portion of the CO2 equivalents claimed to be sequestered would likely involve enteric methane, which, as demonstrated by GWP*, does not accumulate in the atmosphere and has minimal warming impact. The carbon cycle involving photosynthesis and ruminants typically does not contribute to warming or increase atmospheric CO2.

Thus, increasing the potential sink uptake figures (based on GWP100) resulting from reduced meat consumption would largely perpetuate the bias introduced by GWP100 from the outset. Given the existence of a scientifically sound and precise metric like GWP*, using an outdated and flawed metric to derive climate figures is misguided.

Additionally, the oversimplified view that transitioning from grassland to shrubland and forest would significantly enhance carbon sinks should be approached with caution. As previously discussed, grasslands may be larger and more robust carbon sinks than forests, and many forests may not possess substantial net carbon sink values. The Erratum fails to acknowledge these nuances or consider the multitude of factors determining the actual effectiveness and value of carbon sinks across different land types and regions.

Ultimately, the errors throughout the paper that arise from the use of GWP100 underscore the necessity for climate science and policy to adopt the improved and precise GWP* metric. Failing to do so perpetuates misinformation in the climate discourse, a problem that GWP100 has contributed to regarding enteric methane for over three decades.

Scaling Up Inconsistencies

Finally, a word of caution is warranted. Simply comparing averages and scaling up data and assumptions, as this paper does, introduces its own potential for errors and misinterpretations. While land areas, inputs, products, and yields provide relatively straightforward data, the climate and environmental data are based on modeled estimates and assumptions, making them susceptible to inaccuracies. For instance, a specific agricultural practice may yield varying climate and environmental outcomes across different farms within the same locality. What may result in a net zero climate impact in one area could lead to net emissions in another. Discrepancies may even occur within different sections of the same farm or fluctuate from year to year.

When extrapolating results across countries or regions, outcomes from similar practices can vary significantly, complicating comparisons. Any data derived from these types of basic estimates or assumptions depend on the quality of the models and the underlying data.

Such cautionary principles should be considered in any research involving agriculture and climate, regardless of the models or methodologies employed. Basing studies on models that utilize the flawed GWP100 metric, as this paper does, exacerbates the distortion of estimates.

Scaling such distorted estimates skews the depiction of agriculture and its climate and environmental impacts, rendering it unrepresentative of farming practices globally or locally. It misrepresents sound local agricultural practices while making poor practices seem better than they are. Agriculture is inherently local, with practices attuned to diverse local resources, climate conditions, soil characteristics, and input variability. A global average obscures these nuances and introduces new errors.

For a more accurate and comprehensive picture suitable for policy discussions, which should occur at local or national levels rather than globally, these presentations must reflect the best available local and national data. Assessing data quality and validating them against real agricultural practices necessitates agronomic perspectives, which are often overlooked in climate science today.

Data Deficiency: A Significant Lack of Evidence

As I have attempted to illustrate, a considerable portion of what appears to be actual data in both the source paper and OWiD’s presentations consists primarily of estimates and assumptions run through models distorted by flawed metrics and methodological errors, subsequently scaled up. Consequently, these estimates carry extensive margins of error and interpretation, allowing for biases, misunderstandings, speculations, and claims.

Determining the true climate and environmental impacts of various agricultural practices across farms, regions, and countries, as well as assessing their sustainability, would necessitate significantly more comprehensive data than is currently available, along with the requisite agronomic expertise to contextualize the data. While both Poore and Nemecek and OWiD deserve recognition for their efforts, one must acknowledge that there is still no comprehensive global baseline. The shortage of reliable climate and environmental data is substantial, and the potential for error is enormous. I believe that the resulting paper and OWiD’s presentations misinform the public regarding the actual climate and environmental footprint of agriculture in any context and fail to accurately characterize agriculture on a global scale.

These concerns, along with others, were addressed in a PNAS opinion article last year titled Putting All Foods on the Same Table: Achieving Sustainable Food Systems Requires Full Accounting (Halpern et al., 2019). To conclude this lengthy discourse, the challenges outlined above are effectively summarized in the article's introduction:

> "Improving global food systems is essential to addressing climate change, mitigating biodiversity loss, and meeting both sustainability and human development goals. International assessments from the Intergovernmental Panel on Climate Change and the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services, along with business and technology innovations such as lab-grown and plant-based meat, as well as various consumer diet trends, are all rooted in studies identifying the adverse impacts of certain food systems. Yet the evidence underpinning many widely touted recommendations about what to grow and eat is remarkably sparse and generally biased."

References

Our World in Data, 2020 Environmental impacts of food production Our World in Data, 2020 The carbon footprint of foods: are differences explained by the impacts of methane? Michelle Cain, 2018 Guest post: A new way to assess ‘global warming potential’ of short-lived pollutants Poore and Nemecek, 2018 Reducing food’s environmental impacts through producers and consumers USDA, 2017 Considering Forest and Grassland Carbon in Land Management (pg 30–34) Dass, Benjamin Z Houlton, Wang, Warlind, 2018 Grasslands may be more reliable carbon sinks than forests in California Mottet et al, 2017 Livestock: On our plates or eating at our table? A new analysis of the feed/food debate Mitloehner, 2019 How Much Land is Used for Livestock? (video) Marshall et al, 2018 Ancient herders enriched and restructured African grasslands Roberts et al, 2017 The deep human prehistory of global tropical forests and its relevance for modern conservation USDA, 2011 Where’s the (Not) Meat? Byproducts From Beef and Pork Production. Infographics: The many products from cattle FAO, 2015 Food wastage footprint & Climate Change Science, 2019 Erratum (Poore and Nemecek 2018) Halpern et al, 2019 Putting all foods on the same table: Achieving sustainable food systems requires full accounting "Bondevett" is an old Norwegian term, meaning roughly "the rational mind of farmers." My writing, primarily focused on agriculture, climate, food production, forestry, and natural resources, draws on my educational and professional background in agronomy, farming, forestry, and information technology. To connect, please follow me on Twitter.