Please download the Document here below:

e2003169118.full

Abstract

The responses of tropical forests to environmental change are critical uncertainties in predicting the future impacts of climate change. The positive phase of the 2015–2016 El Niño Southern Oscillation resulted in unprecedented heat and low precipitation in the tropics with substantial impacts on the global carbon cycle.

The role of African tropical forests is uncertain as their responses to short-term drought and temperature anomalies have yet to be determined using on-the-ground measurements. African tropical forests may be particularly sensitive because they exist in relatively dry conditions compared with Amazonian or Asian forests, or they may be more resistant because of an abundance of drought-adapted species.

Here, we report responses of structurally intact old-growth lowland tropical forests inventoried within the African Tropical Rainforest Observatory Network (AfriTRON). We use 100 long-term inventory plots from six countries each measured at least twice prior to and once following the 2015–2016 El Niño event.

These plots experienced the highest temperatures and driest conditions on record. The record temperature did not significantly reduce carbon gains from tree growth or significantly increase carbon losses from tree mortality, but the record drought did significantly decrease net carbon uptake. Overall, the long-term biomass increase of these forests was reduced due to the El Niño event, but these plots remained a live biomass carbon sink (0.51 ± 0.40 Mg C ha−1 y−1) despite extreme environmental conditions. Our analyses, while limited to African tropical forests, suggest they may be more resistant to climatic extremes than Amazonian and Asian forests.

Tropical forests are a critical component of the global carbon cycle because they are extensive (1), carbon dense (2), and highly productive (3). Therefore, consistent impacts on these forests can have global consequences. Their global importance is seen via atmospheric measurements of CO₂, showing a near-neutral exchange of carbon across the terrestrial tropics; hence, the large carbon losses from deforestation and degradation are offset by the significant carbon uptake from intact tropical forests and tropical forest regrowth (4). Independently, ground observations of structurally intact old-growth tropical forests also show this uptake, with forest biomass carbon increasing across remaining African (5, 6), Amazonian (7), and Asian (8) forests. Yet, unlike in Amazonia (9, 10) and Asia (8), the impact of a severe drought or a drought and high-temperature event in African tropical forests has never been documented using ground data.

High temperatures test the physiological tolerance of tropical trees. Above optimal temperatures, plants reduce their carbon uptake (11). This includes closing stomata to avoid water loss, reducing internal CO₂ concentrations, and reducing carbon assimilation in the leaf. Higher temperatures increase vapor pressure deficits (12) and alongside reduced precipitation, increase the chance of hydraulic failure (13). Individually or in combination, these impacts can slow growth and may eventually kill trees (14), although tropical seedling growth can increase with experimental warming (15). As well as reduced carbon uptake, plants use more carbon under higher temperatures; respiration rates tend to increase with short-term increases in temperature both at the leaf (16) and forest stand (17) scales, again reducing tree growth and potentially leading to tree death via carbon starvation (18). Recent analyses of tropical forest plot data showed increased temperatures over the prior 5 y were associated with lower levels of carbon uptake from tree growth and higher levels of carbon loss from tree mortality (6). Furthermore, biome-wide spatial analyses suggest the existence of a temperature threshold above which carbon uptake from tree growth declines rapidly (19). Thus, with high temperature anomalies, we expect reduced tree growth and increased tree mortality.

Drought also impacts trees as water deficits can slow tree growth and if of sufficient strength or duration, can kill trees, either via hydraulic failure or carbon starvation. Hydraulic failure of the xylem has been found across species and biomes in response to drought, while carbon starvation has been documented in some locations including one tropical site (20). Inventory plot observations before, during, and after droughts show the impacts of drought in Asia and Amazonia. In Asia, the 1997 to 1998 El Niño temporarily halted the carbon sink in live biomass in Bornean forests by increasing tree mortality (8, 21). In Amazonia, severe droughts in 2005 and 2010 elevated biomass mortality and in 2010, also significantly reduced tree growth (9, 10). The Amazon biomass carbon sink was reversed by the 2005 drought, and while it rapidly recovered, it is weaker since 2005 (7), potentially due to high-temperature impacts (6). However, while the impacts of short-term drought in their long-term context have been investigated in Amazonia and Asia, in Africa we so far lack any ground-based assessment of large-scale drought impacts due to a paucity of observations.

Although the broad responses of African tropical forests to temperature and drought anomalies might be hypothesized from first principles and the responses of other continents, there are considerable uncertainties. On the one hand, there are grounds for expecting African forests to be especially vulnerable. African forests are already remarkably dry compared with Amazonian and Asian tropical forests, with almost 90% receiving <2,000 mm y⁻¹ precipitation (22), the approximate amount necessary to maintain photosynthesis at high levels throughout the year (23). This low rainfall suggests African tropical forests may already be close to their physiological and ecological limits. Additionally, the lower temperatures African forests tend to experience—as many are situated at slightly higher altitude than forests in Amazonia—could result in limited species tolerace of high temperatures. African forests are also much less species rich than forests in Amazonia and Asia (2, 24), with a relative lack of species in high-temperature African forests (25), and this lower diversity could conceivably drive lower resistance to climate anomalies (26).

Alternatively, the relatively dry conditions of African tropical forests may, perhaps counterintuitively, confer drought resistance. African climate has oscillated between wetter conditions in interglacial periods and cooler and drier conditions in glacial periods (27), so the African pool of species present today may be more drought tolerant because some of the most mesic-adapted biodiversity has been lost over time (28, 29). Drier African tropical forest tree diversity is similar to that of the Amazon or Asia, but tree diversity does not increase with shorter dry seasons in Africa as it does in Amazonia (25), suggesting that most wet-adapted species have been lost and either the dry-adapted species remained or these lineages have diversified more, potentially conferring drought resistance. Indeed, a 40-y drought in West Africa led to an increased abundance of deciduous species in tropical forests in Ghana (30, 31). The relatively cool conditions of African tropical forests might also imply resistance as these forests are further from a potential high-temperature threshold that may limit photosynthesis. Overall, African tropical forests could plausibly be more or less vulnerable to temperature and drought anomalies than Amazonian tropical forests.

Understanding how intact African forests respond to climate anomalies is vital, not least because they have been providing a substantial long-term carbon sink, reducing the rate and magnitude of climate change (5, 6). The impacts of environmental change on African tropical forests are also important because of unique aspects of their structure. African forests typically have high aboveground biomass and so, high carbon storage per unit area—on average, one-third more than Amazon forests (2, 32, 33). African forests are composed of a smaller number of stems, ∼425 ha⁻¹ (≥100-mm diameter), compared with ∼600 ha⁻¹ in Amazonia and Asia (32) and so, are more dominated by large trees. Hence, even small decreases in growth of the large dominant trees or modest increases in the mortality of these trees could lead to large carbon stock reductions and a loss of the live biomass carbon sink.

The 2015–2016 El Niño event provided a first opportunity to assess the impact of high temperatures and strong water deficits on the ∼450 Mha (6) of African tropical forests. While three very strong El Niño events have occurred in the last 50 y (1982 to 1983, 1997 to 1998, and 2015 to 2016), only the latter occurred after a network of long-term inventory plots had been established in Africa and was poised to capture an El Niño event (5). At the onset of the 2015–2016 El Niño, we organized a specific “emergency” six-nation remeasurement program to capture the impact of the climate anomaly on African tropical forests. We therefore combine climate data with measurements from 100 African Tropical Rainforest Observatory Network (AfriTRON) long-term inventory plots that were remeasured to capture the 2015–2016 El Niño event to address the following questions. 1) Did African tropical forests experience unprecedented temperature anomalies in the 2015–2016 El Niño? 2) Did African tropical forests experience unprecedented drought in the 2015–2016 El Niño? 3) Which climate anomalies drove forest responses to the 2015–2016 El Niño? 4) What were the overall impacts on the monitored old-growth structurally intact tropical forests?

Please download the Document here below:

e2003169118.full