Last year in a post titled “Passive Forests,” I mused about whether the significant accumulation of carbon in the world’s forests really was just the result of passive forces like carbon fertilization and climate change as argued in Allen et al. (2025). That paper suggested that carbon accounting should break sinks into their various active and passive components. My hunch was that if we did this “disaggregated accounting”, we would find a significant share of carbon accumulation wasn’t passive at all, but instead actively accumulating because of human inputs.
Well, now we have done that, for the United States at least. In a paper recently published by Eric Davis, Dave Lewis, and myself in PNAS, we present an approach for disaggregating sinks in the United States. These methods can be used by other countries that have national forest inventories taken over time. I imagine they’ll find similar results.
We find that since the turn of the century in the conterminous United States, anthropogenic factors have contributed 55% of all carbon gain in forests while non-anthropogenic factors have provided the rest (Table 1). Anthropogenic factors include forest aging, forest area, and management. Non-anthropogenic factors are carbon fertilization, temperature, and precipitation.
Table 1: Proportional contribution to net carbon gain in US forests estimated from results presented in Davis et al. (2026)
| South | North | West | All Regions
(S-N-W) |
||
|
1974-2005 |
2005-2022 | 2005-2022 | 2005-2022 | 2005-2022 | |
|
Anthropogenic |
47% | 59% | 66% | 0% |
55% |
|
Non-Anthropogenic |
53% | 41% | 34% | 100% |
45% |
We also find that the anthropogenic parts are increasing, at least in the southern US, which constitutes a large and growing portion of the overall US sink. An important reason for the large increase was management. Based on our numbers, I calculate the rate of sequestration due to management increased more than three-fold in the 2005 to 2022 period compared to 1974 to 2005, from 5.8 Tg C/yr to 19.9 Tg C/yr in aboveground C.
Forest aging is also an important part of this anthropogenic story. There will be lots of discussion about whether allowing trees to get older is anthropogenic, but undeniably in the United States private landowners make the call on whether to cut their trees or leave them and these private lands are the bulk of the gains. In the period 2005 to 2022, trees all over the eastern United States got older because people let them, leading to significant increases in carbon storage.
In the future, those same people may make different decisions, leading to carbon losses. Actually, we show just this outcome for loblolly stands in in the tumultuous 1974 to 2005 period when trade conflicts and Spotted Owl protections shifted harvests to the South causing carbon emissions in the “aging” category (see Figure S27).
I doubt you will find many people who would disagree that this increased harvesting of loblolly in the South was anthropogenic. This is basically the leakage argument under improved forest management.
So, isn’t leaving trees alone also anthropogenic?
Personally, I argue that leaving trees alone anywhere in the world – public or private land – is an anthropogenic decision, meaning any carbon change due to aging, or cutting, is due to active decision-making.
The results in this paper further illustrate the fundamental problems of the no-harvest counterfactual. There are many ways to increase carbon in forests. Aging trees is one of them, while intensifying management and increasing the area of intensively managed forests are others.
For example, here we show that more loblolly hectares increased carbon while declining oak-pine hectares reduced carbon. A reasonable approach to carbon accounting will acknowledge and attribute carbon gains appropriately. The no-harvest counterfactual doesn’t do any of this.
This approach can possibly be used in most regions of the world even without the same level of data to at least quantify the same passive factors of CO2 fertilization, temperature, and precipitation. The rest may be mostly anthropogenic. Yield functions can be calculated for many parts of the world with existing data using the approach outlined in Davis et al (2022) to identify carbon fertilization and climatic drivers.
It is useful to note that these results also present some important cautions for US forests. There does seem to be evidence that climate change, which has heretofore been a positive force for carbon accumulation, may be turning negative in places. The trends are not dramatic, but there do seem to be temperature and precipitation signals that are starting to move in the negative direction. These trends need to be watched.


