Which Terrestrial Biome Has the Most Net Primary Productivity?

Tropical rainforests produce more new biomass per unit area each year than any other terrestrial biome, and the gap is substantial. Their net primary productivity, the amount of carbon that plants fix through photosynthesis minus what they burn off through their own respiration, routinely exceeds 1,000 grams of carbon per square meter per year, roughly double or triple what temperate forests manage and many times what tundra or desert scrub can sustain. The reasons trace back to a simple alignment of heat, water, and sunlight that no other biome can match, but the full story involves some surprises about which ecosystems punch above their weight and how quickly the global productivity map is shifting.

What Makes Tropical Rainforests the Productivity Leaders

The core advantage of tropical rainforests is that they never stop growing. There is no winter dormancy, no months-long freeze, no dry season harsh enough to shut down photosynthesis entirely. Temperatures hover in the range that enzymes work most efficiently for carbon fixation, rainfall stays high year-round in most equatorial lowland forests, and the sun angle is steep enough to deliver strong light in every month. Those three inputs, warmth, water, and light, are the raw materials of plant productivity, and tropical rainforests get all three simultaneously and continuously.

One factor that often gets overlooked is the sheer density of leaf surface available to capture sunlight. Leaf area index, the total leaf area stacked above each square meter of ground, is a key driver of how much photosynthesis an ecosystem can perform. Direct landscape-scale measurements in tropical rainforest confirm that LAI values are among the highest on Earth, giving these forests an enormous photosynthetic surface per hectare compared to more open or seasonal ecosystems.¹ When you have that much leaf tissue operating nearly every day of the year, the cumulative carbon uptake dwarfs what biomes with shorter or less active growing seasons can achieve.

The relationship between a forest’s total carbon uptake and how much it retains as new growth is not fixed, though. Environmental conditions in the previous year, especially drought and temperature stress, influence what fraction of total photosynthesis ends up stored as net production versus burned for maintenance. A severe drought year can lower overall photosynthesis enough that the ratio of new growth to total carbon fixation actually stays stable or even rises, because the trees also cut back on respiratory costs.² Tropical forests generally maintain a high absolute NPP because their photosynthetic machinery rarely encounters the kind of extreme stress that forces prolonged shutdowns.

How Tropical Forests Split Their Carbon

Not all the new biomass in a tropical forest ends up as towering trunks. When researchers measure where NPP actually goes, the allocation across different plant parts is more evenly distributed than you might expect. On average, roughly a third goes to the canopy through leaf growth and litterfall, about two-fifths goes into wood production including trunks and branches, and the remaining quarter or so goes underground into fine roots.³ That split matters for understanding carbon cycling because wood stores carbon for decades or centuries, while leaves and fine roots turn over much faster, cycling their carbon back into the soil and atmosphere within months to a few years.

There is also substantial variation from site to site. Some tropical forests invest heavily in wood at the expense of roots, while others channel far more carbon belowground. Soil fertility, moisture availability, and species composition all play roles. One practical finding from this research is that litterfall alone turns out to be a strong predictor of overall NPP in tropical forests, whereas stem growth is only moderately useful and fine root production is a poor guide to the total.³ That matters for scientists trying to estimate productivity on a budget, since counting fallen leaves is far cheaper than excavating root systems.

Savannas, Grasslands, and Temperate Forests

Below the tropical rainforest tier, the ranking gets more interesting than most textbooks suggest. Tropical and subtropical savannas are often dismissed as moderate-productivity systems, but their grass layers can be powerhouses. In mesic (relatively wet) savannas, grasses using the C4 photosynthetic pathway contribute around 40% of total ecosystem carbon uptake, and their light-use efficiency exceeds that of the co-occurring trees, particularly during the wet season.⁴ C4 grasses are exceptionally good at converting light and CO₂ into biomass under warm, high-light conditions, which lets savannas accumulate carbon at surprisingly fast rates when water is available.

Eddy covariance measurements from Brazilian tropical savannas illustrate the range. Gross primary production in these systems spans from about 7.5 to over 18 megagrams of carbon per hectare per year depending on vegetation type and method, with total NPP ranging from roughly 1 to 9.5 megagrams per hectare per year.⁵ The upper end of that range overlaps with some temperate and even tropical forest values, which underscores the point that biome labels hide enormous internal variability.

Precipitation is the master control in savannas and grasslands. In one high-rainfall tropical savanna, rising grass productivity over recent decades tracked almost entirely with precipitation changes, which accounted for about 84% of the upward trend in grass carbon uptake.⁶ Rising atmospheric CO₂ added only a small additional boost to the grassy understory. Trees in savannas are somewhat more responsive to CO₂ than grasses, but the overall message is that water availability is the throttle on savanna productivity.

Temperate forests sit solidly in the middle of the biome productivity spectrum. They benefit from moderate to high rainfall and warm summers that drive vigorous growth, but they lose several months of productivity to winter dormancy. Temperate broadleaf forests tend to be more productive than temperate coniferous forests at similar latitudes, mainly because broadleaf trees photosynthesize more efficiently during their shorter active season. Still, even the most productive temperate forests rarely approach the per-area output of lowland tropical rainforest.

Tundra, Deserts, and the Low End

At the bottom of the productivity ladder sit the biomes where one or more essential growth inputs is chronically scarce. Tundra ecosystems are limited primarily by cold temperatures and extremely short growing seasons. Even during years when the rest of the continent is doing well, tundra productivity can be depressed by conditions that would be unremarkable further south. During 2020 and 2021, for example, growing-season NPP in North American tundra declined not because of heat or drought but because of colder-than-normal temperatures at high latitudes, essentially a cold-induced drought on plant growth.⁷

Deserts and arid shrublands face the opposite constraint. Temperature is rarely the problem; water is. The sparse vegetation in these ecosystems can photosynthesize only during brief windows after rain events, and much of the carbon fixed during those windows goes straight into root maintenance and water-seeking root growth rather than aboveground biomass. Desert NPP values typically range from 0 to about 200 grams of carbon per square meter per year, a fraction of what tropical forests produce. The carbon that does accumulate in desert soils tends to be locked tightly in mineral-associated forms rather than the free organic matter that dominates wetter, more productive ecosystems.⁸

Boreal forests fall between these extremes. They cover vast areas of northern Canada, Scandinavia, and Siberia and hold enormous carbon stocks in their soils, but their actual annual production rates are modest. Cold winters, short growing seasons, and nutrient-poor soils all conspire to keep per-area NPP well below that of temperate or tropical forests. Nitrogen limitation is a particularly important brake on boreal productivity, keeping growth rates low even when summer temperatures and light availability would otherwise support more vigorous photosynthesis.

How Nutrient Availability Shapes the Map

Water and temperature explain the broad strokes of global productivity patterns, but nutrient availability adds a second layer of complexity. The nutrient that limits growth varies by biome in a surprisingly clean geographic pattern. Boreal forests, tundra, and temperate coniferous forests tend to be limited by nitrogen. Their cold, acidic soils decompose organic matter slowly, keeping nitrogen locked up and unavailable to plants. By contrast, tropical and subtropical forests, along with Mediterranean ecosystems and many temperate grasslands, tend to be limited by phosphorus. In these warmer systems, millennia of intense weathering have leached phosphorus from the soil, leaving it as the scarce ingredient even when nitrogen is abundant.

This distinction matters for understanding productivity because it means that the same intervention, say, adding fertilizer, would have completely different effects depending on which biome you are in. It also affects how different biomes respond to rising atmospheric CO₂. Plants can only take advantage of extra carbon dioxide if they have enough nitrogen and phosphorus to build the proteins and energy-transfer molecules needed for faster photosynthesis. In phosphorus-limited tropical forests, the theoretical boost from CO₂ fertilization may be partially capped by mineral scarcity, a constraint that most global productivity models are still learning how to represent accurately.

Why Global Productivity Swings from Year to Year

If you track global terrestrial NPP over time, the line is not smooth. Some years the land biosphere absorbs noticeably more carbon, and other years it absorbs less. A surprisingly large share of that year-to-year variability traces back to a single climate pattern: the El Niño–Southern Oscillation. ENSO accounts for more than 40% of global NPP variability, with the effects concentrated in the Southern Hemisphere’s tropical and subtropical ecosystems.⁹

The mechanism is mostly about water. During La Niña phases, large parts of the tropics and subtropics receive above-average rainfall, which lifts productivity in water-sensitive biomes like savannas, seasonal forests, and grasslands. El Niño does the opposite, bringing drought to those same regions and depressing growth. The year 2011 stands out as a striking example: a powerful La Niña event that ran from late 2010 into early 2012 drove a global NPP anomaly high enough to be clearly visible in satellite data and atmospheric carbon measurements.⁹ For temperate and boreal biomes, ENSO’s influence is weaker; their variability is driven more by local temperature anomalies and regional weather patterns.

How Much Productivity Humans Siphon Off

Knowing which biomes are most productive is one thing. Understanding how much of that productivity actually remains in natural ecosystems is another. Humans appropriate a significant share of global net primary production through agriculture, forestry, and land conversion. A comprehensive global assessment put the total human appropriation of NPP at about 15.6 petagrams of carbon per year, which works out to roughly 24% of what the terrestrial biosphere would produce without human interference.¹⁰ Just over half of that appropriation comes from direct harvest of crops, timber, and other biomass. The remaining portion comes from land-use changes that reduce ecosystem productivity below its natural potential, along with a small contribution from human-caused fires.

That roughly one-quarter share doubled during the twentieth century, tracking the expansion of cropland, managed pastures, and timber plantations.¹¹ The biomes most affected are not necessarily the ones with the highest natural productivity. Temperate grasslands and savannas have been converted to cropland far more completely than tropical rainforests in absolute area terms, even though per-hectare tropical forests would represent a larger productivity loss per cleared acre. The net effect of all historical deforestation and agricultural expansion has been a reduction in global gross primary production of roughly 4 to 5%.¹²

Land-use change does not always reduce productivity. Converting low-yield cropland back to forest or grassland, as China’s Grain for Green program has done on a massive scale, can substantially increase local NPP. On the Loess Plateau, average NPP nearly doubled between 2001 and 2020, with land-use change outweighing climate as the main driver of the increase.¹³ Programs that replace degraded cropland with perennial vegetation can push local productivity closer to what the natural biome would support.

The CO₂ Fertilization Effect

One of the most consequential questions in Earth system science right now is whether rising atmospheric CO₂ is making the entire terrestrial biosphere more productive. The answer appears to be yes, and by more than many models had predicted. A reconciliation of leaf-level experiments with global atmospheric data points to a CO₂ fertilization effect on photosynthesis of about 30% since 1900, which translates to a roughly 47% boost for a full doubling of CO₂ above pre-industrial levels.¹⁴ That estimate is nearly twice the figure used in some widely cited earlier assessments, suggesting that the land biosphere has been absorbing more carbon than previously thought.

This does not mean every biome benefits equally. The fertilization effect is strongest where water and nutrients are not bottlenecks, which again favors the already-productive tropical forests and well-watered temperate systems. In nutrient-poor or water-limited biomes, the extra CO₂ has less room to boost growth because other resources run out first. And there are limits even in favorable conditions: as temperatures climb alongside CO₂, heat stress, increased vapor pressure deficit, and more intense droughts can erode or even reverse the fertilization benefit in regions that would otherwise gain the most. The net outcome for any particular biome depends on the interplay of all these factors, which is why projections of future productivity remain uncertain despite the clear signal in the historical data.

Measuring NPP from the Ground Up

One reason productivity rankings come with caveats is that NPP is genuinely difficult to measure. You cannot simply weigh a forest and check again a year later. Scientists use two broad families of methods, and getting them to agree is an ongoing challenge. Biometric methods involve physically measuring tree growth through repeated surveys of trunk diameter, collecting and weighing litterfall, and sometimes excavating roots. These approaches are labor-intensive and site-specific, but they give a direct accounting of where new carbon ends up.

The alternative is eddy covariance, a technique that uses fast-response sensors mounted on towers above the canopy to track how much CO₂ moves between the ecosystem and the atmosphere on a continuous basis. Eddy covariance captures the whole-ecosystem carbon exchange in near real time, but extracting NPP from the raw data requires subtracting out respiratory fluxes and filling in gaps when instruments are offline, each step introducing its own uncertainties.¹⁵

When both methods are applied to the same site, they tend to produce similar estimates but not identical ones. In Brazilian savanna ecosystems, eddy covariance and inventory approaches agreed reasonably well on gross primary production and total NPP, though the inventory method tended to give a slightly wider range.⁵ Some research groups integrate both approaches with continuous soil chamber measurements to cross-check the numbers and tease apart how much carbon goes to different parts of the ecosystem.¹⁶ Remote sensing from satellites adds a third layer, using vegetation greenness and light absorption data to model productivity at continental scales, though satellite-derived estimates still need to be calibrated against ground measurements. The global database of terrestrial NPP now includes records from over 450 sites spanning forests, grasslands, shrublands, peatlands, and tundra, giving researchers an increasingly complete picture of how productivity varies across the planet’s biomes.¹⁷

1
PubMed. First direct landscape-scale measurement of tropical rain forest Leaf Area Index, a key driver of global primary productivity
2
Biogeosciences. Temporal variability of the NPP-GPP ratio at seasonal and interannual time scales in a temperate beech forest
3
PubMed Central. The allocation of ecosystem net primary productivity in tropical forests
4
Global Change Biology. Is productivity of mesic savannas light limited or water limited? Results of a simulation study
5
Journal of Geophysical Research: Biogeosciences. Net Primary Production and Ecosystem Carbon Flux of Brazilian Tropical Savanna Ecosystems From Eddy Covariance and Inventory Methods
6
PubMed Central. Gross primary productivity and water use efficiency are increasing in a high rainfall tropical savanna
7
Journal of Geophysical Research: Biogeosciences. Net Primary Production of Ecoregions Across North America in Response to Drought and Wildfires From 2015 to 2022
8
Communications Earth & Environment. Ecosystem productivity has a stronger influence than soil age on surface soil carbon storage across global biomes
9
Journal of Geophysical Research: Biogeosciences. The global NPP dependence on ENSO: La Niña and the extraordinary year of 2011
10
PubMed Central. Quantifying and mapping the human appropriation of net primary production in earth’s terrestrial ecosystems
11
PubMed Central. Global human appropriation of net primary production doubled in the 20th century
12
Scientific Reports. Quantifying the impacts of land cover change on gross primary productivity globally
13
PubMed. Evaluating net primary productivity dynamics and their response to land-use change in the loess plateau after the ‘Grain for Green’ program
14
PubMed Central. Higher than expected CO2 fertilization inferred from leaf to global observations
15
Global Change Biology. A method for deriving net primary productivity and component respiratory fluxes from tower‐based eddy covariance data: a case study using a 17‐year data record from a Douglas‐fir chronosequence
16
Biogeosciences. Net primary productivity, allocation pattern and carbon use efficiency in an apple orchard assessed by integrating eddy covariance, biometric and continuous soil chamber measurements
17
Scientific Data. A global database of net primary production of terrestrial ecosystems