Silhouettes of tropical staple trees with glowing sunset on horizon. / Alamy Stock Photo

Lands converted from annuals to perennial staples sequester, on average, 1.9 tons of carbon per acre every year for decades.

Perennial Staple Crops

Perennial staple crops provide important foods, such as bananas, avocado, and breadfruit. Compared to annual crops, they have similar yields but higher rates of carbon sequestration.

Support SinksLand SinksShift Agriculture Practices
16.34 to 32.87
CO2 Equivalent
96.53 to 209.53
Billion US$
Net First Cost
To Implement
-0.98 to -2.11
Trillion US$
Lifetime Net
Operational Savings
1.70 to 3.65
Trillion US$
Lifetime Net Profit
Research Fellow: Ariani Wartenberg; Senior Fellows: Mamta Mehra, Eric Toensmeier; Senior Director: Chad Frischmann


Perennial staple crops currently grow on approximately 50 million hectares, mostly in the tropics. Their rate of sequestration is high at 3.34 metric tons per hectare per year. Expand this area to 128.80–213.34 million hectares by 2050 and they can sequester 16.34–32.87 gigatons of additional carbon dioxide. Our analysis assumes that expansion only occurs on existing grassland, with no forest clearing. Because their yield is 2.4 times higher than annual staples—at 60 percent of the cost—lifetime net profit is US$1.70–3.65 trillion, with an establishment cost of US$96.53–209.53 billion and lifetime net operational cost of US$0.98–2.11 trillion.


Project Drawdown’s Perennial Staple Crops solution involves the production of trees and other perennial crops for staple protein, fats, and starch. This solution replaces conventional annual crop production in humid and semi-arid tropics in nondegraded grasslands and croplands.

Most of the world’s cropland is used to produce annual staple crops like maize, wheat, potatoes, and soybeans, and these crops are a major source of greenhouse gas emissions from agriculture. Annuals are not the only crops producing staple food, however. In the tropics, many perennial staple crops are widely grown, and yield as well or better than their annual staple crop competitors. These perennial staple crops also sequester impressive carbon in soils and above-ground biomass. Their sequestration rates are much higher than any annual cropping system. Tropical staple tree crops in particular can reverse erosion and runoff. They can grow on steep slopes and in a wide range of soils and require lower (if any) inputs of fuel, fertilizer, and pesticides. They also can thrive under conditions that annuals cannot—vital in a warming world.

One critical assumption of this study is that all perennial staple crops adoption would be on nondegraded grassland and cropland, with no forest clearing, despite the current situation in which much forest is cleared for staple tree crops like avocado and oil palm. If forest (particularly peatland) is cleared for tropical staple tree planting, the result is net emissions regardless of sequestration.

This solution has received very little attention in the climate change mitigation literature. Though perennial staple crops present trade-offs and challenges, their high sequestration rate, high current adoption, and rapid growth rate indicate impressive potential.


Total Land Area

To evaluate the extent to a Food, Agriculture, and Land Use sector solution can reduce greenhouse gas emissions and sequester carbon, we need to identify the total land area available for that solution. To avoid double counting, we use an integration model that allocates land area among all of the sector’s solutions. This involves two steps. First, we classify the global land area into agro-ecological zones (AEZs) based on the land cover, soil quality, and slope and assign AEZs to different thermal moisture regimes. We then classify the AEZs into “degraded” and “nondegraded.” Second, we allocate the solutions to AEZs, with the solution most suited to a given AEZ or sets of AEZs assigned first, followed by the second-most-suited solution, and so on. Because it’s hard to predict future changes, we assume the total land area remains constant. Total land areas represent both the implementation and functional unit.

Total available land for the Perennial Staple Crops solution is 330 million hectares. Current adoption (the amount of functional demand supplied in 2018) is 50 million hectares, according to the Food and Agriculture Organization Statistical Service (FAOSS). Current growth is very high for crops such as oil palm and avocado, but those are often planted on land cleared for the purpose. This solution assumes that the rate of growth will continue, but planting will exclusively be on degraded lands with no forest clearing.

Adoption Scenarios

We base future adoption on a linear projection of regional data from 1962 to 2016 (FAOSS). We developed seven custom adoption scenarios based on the low and high historic adoption rates, including some with early adoption (e.g., 70 percent adoption by 2030).

We calculated impacts of increased adoption of Perennial Staple Crops from 2020 to 2050 by comparing two scenarios with a reference scenario in which the market share was fixed at current levels.

  • Scenario 1: Perennial staple crops are adopted on 128.80 million hectares (34 percent of total available land area).
  • Scenario 2: Perennial staple crops are adopted on 213.34 million hectares (56 percent of total available land area).

Emissions, Sequestration, and Yield Model

We set carbon sequestration rates at 3.34 metric tons per hectare per year, based on six data points from four sources. We assume that all sequestered carbon is re-emitted at the end of an orchard or plantation’s useful life, which we set at 37.5 years. The weighted average yield of perennial staple crops is 2.4 times greater than that of annual staples, based on analysis of data from seven perennials and 15 annuals (FAOSS).

Financial Model

All monetary values are presented in 2014 US$.

First costs were US$1,298.4 per hectare, based on meta-analysis of 12 data points from six sources. For all grazing solutions we assume that there is no conventional first cost because agriculture is already in place on the land. Net profit per hectare is calculated at US$1,025.98 per year for the solution (based on meta-analysis of 22 data points from 13 sources), compared with US$154.12 per year for the conventional practice (based on 20 data points from 15 sources). Annual operational cost per hectare is calculated at US$749.28 for the solution (based on meta-analysis of 19 data points from 12 sources), compared with US$328.42 for the conventional practice (based on nine data points from seven sources).

Tropical staple trees are not as labor-efficient as annual crops, in a mechanized context. However, 175 million hectares of the world’s farms are smallholders with little mechanization. The net profit per hectare figure shows that these crops are economically viable despite higher labor costs.


We allocate current and projected adoption of Agriculture, Food, and Land Use solutions to forest, grassland, rainfed cropland, and irrigated cropland. The Perennial Staple Crops solution was the second-highest priority for degraded grassland and the sixth-highest for degraded cropland.


Scenario 1 reduces emissions by 16.34 gigatons of carbon dioxide equivalent by 2050. The net first cost is US$96.53 billion, and the lifetime operational cost is US$0.98 trillion. The lifetime net profit is US$1.70 trillion.

Scenario 2 reduces emissions by 32.87 gigatons of carbon dioxide equivalent by 2050. The net first cost is US$209.53 billion, and the lifetime operational cost is US$2.11 trillion. The lifetime net profit is US$3.65 trillion.



Benchmarks for this solution are unavailable, and mitigation benchmarks for tree crops of any kind are rare. This suggests a need for new research. Projections for tree plantation can be used as a rough comparison. The Intergovernmental Panel on Climate Change estimates that afforestation could sequester 4.0 gigatons of carbon dioxide equivalent per year by 2030, given a carbon dioxide price of US$100 per metric ton (Metz, 2007). Our Perennial Staple Crops solution model yields much lower savings of 0.46–0.83 gigatons of carbon dioxide equivalent per year. When combined with our Bamboo Production and Tree Plantations (on Degraded Land) solutions, however, emissions reductions increase to 1.33–2.45 gigatons of carbon dioxide equivalent per year by 2030.


Additional data on financials, sequestration rates, and yields would improve this analysis. The potential adoption area could be increased to include arid regions, because many perennial staple crops like mesquite are suited to arid conditions.


Perennial cropping solutions such as Multistrata Agroforestry and Perennial Staple Crops can offer the high sequestration rates of afforestation and forest restoration while providing food. These somewhat neglected “edible afforestation” solutions are worthy of a place at the center of land-based mitigation efforts. It should also be noted that there are staple trees for cold climates, though their yields are not yet competitive with annual crops. Project Drawdown hopes to inspire further research and widespread adoption of this promising solution.


FAO. (2016, December 8). FAO Statistical Service Online. Retrieved December 8, 2016, from

Metz, B., Davidson, O. R., Bosch, P. R., Dave, R., and Meyer, L. A. (2007). Climate Change 2007: Mitigation of Climate Change. Contribution of Working Group III to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change. International Panel on Climate Change. Accessed at