Are Earth system models representing forest regeneration well enough?

The Science
Forests will only persist where future trees are able to reproduce, disperse, germinate, and grow into mature trees (i.e. “recruit”). These critical regeneration processes are generally not well represented in the models that ecologists use to predict future forests. We critically reviewed how regeneration processes are represented within models that strive to predict forest demography in Earth system models. We found a need to improve parameter values and algorithms for reproductive allocation, dispersal, environmental filtering in the seedling layer, and tree regeneration strategies adapted to wind, fire, and anthropogenic disturbance regimes. These improvements are needed so that models can capture forest responses to global change.

The Impact
Vegetation demographic models represent forest dynamics in the Earth system. They provide the opportunity to more fully integrate ecological understanding into predictions of future climate and ecosystem states. This paper identifies critical areas where models are not prepared to Forest regeneration is a series of environmentally sensitive processes that culminates in tree recruitment. Environmental variables act as filters (depicted as varying sieve mesh sizes) that impose constraints on each process capture future forest responses to global change variables such as changing precipitation and disturbance regimes. This review helps model developers identify what kinds of improvements are needed. It also helps field ecologists understand what types of data are best suited to supporting the improvement of models. Improving these models will advance our ability to predict the role that forests will play in sequestering and storing carbon, providing habitat for biodiversity, and provisioning critical natural resources for people.

Summary
Earth system models must predict forest responses to global change in order to simulate future global climate, hydrology, and ecosystem dynamics. These models are increasingly adopting vegetation demographic approaches that explicitly represent tree growth, mortality and recruitment, enabling advances in the projection of forest vulnerability and resilience, as well as evaluation with field data. To date, simulation of regeneration processes has received far less attention than simulation of processes that affect growth and mortality in spite of its critical role maintaining forest structure, facilitating turnover in forest composition over space and time, enabling recovery from disturbance, and regulating climate-driven range shifts. Our critical review of regeneration process representations within current Earth system vegetation demographic models reveals the need to improve parameter values and algorithms for reproductive allocation, dispersal, seed survival and germination, environmental filtering in the seedling layer, and tree regeneration strategies adapted to wind, fire, and anthropogenic disturbance regimes. These improvements require synthesis of existing data, specific field data collection protocols, and novel model algorithms compatible with global scale simulations. Vegetation demographic models offer the opportunity to more fully integrate ecological understanding into Earth system prediction; regeneration processes need to be a critical part of the effort.

Contact: Adam Hanbury-Brown, University of California, Berkeley, ahanburybrown@gmail.com

Funding
Funding was provided via Next Generation Ecosystem Experiments-Tropics (NGEE-Tropics), funded by the US Department of Energy, Office of Science, Office of Biological and Environmental Research. The lead author also received support from the National Science Foundation and the National Aeronautics and Space Administration during this research.

Publication
Hanbury-Brown AR, RE Ward , LM Kueppers (2022), Forest regeneration within Earth system models: current process representations and ways forward. New Phytologist. https://doi.org/10.1111/nph.18131

Researchers found that degraded forests (burned and logged) do less photosynthesis during the
dry season, but recover productivity in about 4 years.

The Science
Large areas of the Amazon forest are being degraded through fires and logging. Degraded forests may suffer more water stress than intact forests during the dry season, but it remains an open question. A team of researchers used multiple remote sensing data to test this possibility. They used laser sensors on an aircraft to see changes in forest structure caused by degradation. They also used a satellite sensor that can see how much photosynthesis plants are doing. They compared both data sets to check how long it takes for forests to recover structure and photosynthesis after disturbance.

The Impact
The team found that fires cause much more damage to forests than logging. They also noticed that recently burned forests did less photosynthesis than intact forests. To their surprise, in just 4 years after disturbance, burned and logged forests were already doing as much photosynthesis as intact forests. However, the structure of burned forests remained very different from intact forests even after 14 years. This is important because each forest characteristic may take a very different time to recover from degradation.

Summary
Human cause disturbances that degrade tropical forests. Forest degradation from selective logging and fires alter forest structure and function. These changes also impact the ability of forests to uptake carbon. This study used airborne laser scanning data over the Amazon to investigate how forest structure varies across burned and logged forests of different ages since disturbance. In addition, the team used solar induced chlorophyl fluorescence (SIF) data from the TROPOMI mission. SIF is a proxy for photosynthesis, and the TROPOMI data provide information on how photosynthesis varies across seasons and across degraded and intact forests.

The researchers found that forest fires suffered the largest changes in the vertical distribution of foliage and canopy height compared to logged and intact forests. Moreover, SIF in recently burned forests were significantly lower than in intact forests. In contrast, within 4 years after the disturbance, SIF values were higher in regenerating forests than in intact forests. This was surprising because regenerating forests still had lower leaf area. These findings highlight that degraded forests recover photosynthesis rates faster than they recover forest structure. The results also indicate that degraded forests can accumulate large amounts of carbon during recovery from disturbance.

Contacts: Ekena Rangel Pinagé, College of Forestry, Oregon State University, rangelpe@oregonstate.edu
Marcos Longo, Lawrence Berkeley National Laboratory, mlongo@lbl.gov

Funding
This research was funded by the Australian Government Research Training Program Scholarship, the USDA Forest Service Pacific Northwest Research Station and International Programs, and by the Next Generation Ecosystem Experiments-Tropics, funded by the U.S. Department of Energy, Office of Science, Office of Biological and Environmental Research. The research carried out at the Jet Propulsion Laboratory, California Institute of Technology, was under a contract with the National Aeronautics and Space Administration.

Publications
Pinagé, E. R., D. M. Bell, M. Longo, S. Gao, M. Keller et al. “Forest structure and solar-induced
fluorescence across intact and degraded forests in the Amazon”. Remote Sensing of Environment
274, 112 998 (2022). [DOI: 10.1016/j.rse.2022.112998]

A synthetic review leads to a new hypothesis framework that sheds light on how and why plants are dying more under a warming climate.

The Science
Increasing rates of woody-plant mortality, including trees and shrubs, presents a large scientific challenge because we do not yet understand what is causing the increasing loss of plants. This study reviewed the literature to identify the key mechanisms underlying warming-induced mortality, and presented a testable framework that yields insight into the drivers of plant death as well as how to better model these processes. Ultimately, mortality under drought, rising temperature, and rising carbon dioxide results from depletion of water and carbon stores, leading to irreversible dehydration and the inability to maintain metabolism. Warming exacerbates these storage declines, while elevated carbon dioxide has mixed impacts. The net result of the increasing rate and severity of warming and drought overwhelms the benefits of elevated carbon.

The Impact
Plant mortality is rising globally, leading to negative impacts on ecosystem services of value to society including economic, aesthetic, and ecological consequences. Plant mortality reduces carbon uptake and increases carbon loss, promoting a decline in terrestrial carbon storage. Despite these consequences, our ability to predict plant mortality is limited by a lack of knowledge of the underlying mechanisms, their response to climate, and their integration into models. Here, scientists reviewed the literature to generate a synthetic hypothesis framework that pinpoints the key mechanisms driving mortality under a changing environment. The result is a roadmap for future research, including the provision of a set of testable hypotheses that will rapidly increase our knowledge, and identification of key mechanisms that should be included in process models to enable more accurate representation of the impacts of climate change on plant survival.

Summary
Plant mortality is rising in concert with increasing droughts, warming, and carbon dioxide, but the mechanisms driving the increased mortality are poorly known. This knowledge gap leads to large challenges for prediction of the future of terrestrial ecosystems, including their role in water, carbon, and nutrient cycling. Here we integrated the literature on plant mortality and subsequently generated a synthetic and testable hypothesis framework describing the mechanisms underlying plant death in a warming and drying world. The stores of carbon and water are depleted under changing climate, with some amelioration due to rising carbon dioxide. The decline in these stores leads to failure to maintain hydration and metabolism, and can promote death outright or through failure to defend against attacking biotic agents. Acclimation can promote survival to an extent. Determining the net impacts of rising carbon dioxide versus drought and warming remain a major science challenge.

Contact: Nate McDowell, Pacific Northwest National Laboratory, nate.mcdowell@pnnl.gov

Funding
N. G. M. and C. X. were supported by the Department of Energy, Office of Science project Next Generation Ecosystem Experiment-Tropics (NGEE-Tropics). G. S. was supported by the NSFBII-Implementation (2021898). D. T. acknowledges support from the Australian Research Council (DP0879531, DP110105102, LP0989881, LP140100232). M. D. K acknowledges support from the Australian Research Council (ARC) Centre of Excellence for Climate Extremes (CE170100023), the ARC Discovery Grant (DP190101823) and the NSW Research Attraction and Acceleration Program. C. G. was supported by the Swiss National Science Foundation (PZ00P3_174068). M. M. and J.M.V were supported by the Spanish Ministry of Science and Innovation (MICINN, CGL2017‐89149‐C2‐1‐R). A.T.T. acknowledges funding from the NSF Grant 2003205, the USDA National Institute of Food and Agriculture, Agricultural and Food Research Initiative Competitive Programme Grant No. 2018-67012-31496 and the University of California Laboratory Fees Research Program Award No. LFR-20-652467. W.M.H. was supported by the NSF GRFP (1-746055). A.M.T. and H.D.A. were supported by the NSF Division of Integrative Organismal Systems, Integrative Ecological Physiology Program (IOS-1755345, IOS-1755346). H.D.A. also received support from the USDA National Institute of Food and Agriculture (NIFA), McIntire Stennis Project WNP00009 and Agriculture and Food Research Initiative award 2021-67013-33716. D.D.B. was supported by NSF (DEB-1550756, DEB-1824796, DEB-1925837), USGS SW Climate Adaptation Science Center (G18AC00320), USDA NIFA McIntire Stennis ARZT 1390130-M12-222, and a Murdoch University Distinguished Visiting Scholar award. D.S.M. was supported by NSF (IOS-1444571, IOS- 1547796). R.S.O. acknowledges funding from NERC-FAPESP 19/07773-1. W.R.L.A. was supported by the David and Lucille Packard Foundation, NSF grants 1714972, 1802880 and 2003017, and USDA NIFA AFRI grant no. 2018‐67019‐27850. R.S.O. acknowledges funding from NERC-FAPESP 19/07773-1. B.E.M. is supported by an Australian Research Council Laureate Fellowship (FL190100003. A.S. was supported by a Bullard Fellowship (Harvard University) and the University of Montana.

Publications
McDowell NG, Sapes G, Pivovaroff A, Adams H, Allen CD, Anderegg WRL, Arend M, Breshears DD, Brodribb T, Choat B, Cochard H, Cáceres MD, De Kauwe M, Grossiord C, Hammond WH, Hartmann H, Hoch G, Kahmen A, Klein T, Mackay DS, Mantova M, Martínez-Vilalta J, Medlyn BE, Mencuccini M, Nardini A, Oliveira RS, Sala A, Tissue DT, Torres-Ruiz JM, Trowbridge A, Trugman AT, Wiley E, Xu C. (2022) Mechanisms of woody plant mortality under rising drought, CO2, and vapor pressure deficit. Nature Reviews Earth and Environment. https://doi.org/10.1038/s43017-022-00272-1.

Simulating environmentally sensitive tree recruitment shows promise for helping ecologists predict where trees will grow in the future.

The Science
Forests will only persist where future trees are able to reproduce, disperse, germinate, and grow into mature trees (i.e. “recruit”). These critical regeneration processes are generally not represented in the models that ecologists use to predict future forests. The recently developed Tree Recruitment Scheme (TRS) was developed specifically to capture how changing environmental conditions will affect future trees’ ability to recruit. The TRS is shown to improve  predictions of tree recruitment rates in a tropical forest in Panama and captures how reduced soil moisture and light constrains tree recruitment.

The Impact
By improving predictions of tree recruitment using environmentally sensitive processes the TRS is well-positioned to improve predictions of future forest range boundaries, composition, and function. This is important for predicting the role that forests will play in sequestering and storing carbon, providing habitat for biodiversity, and provisioning critical natural resources for people. By representing the early stages of tree development the TRS will allow ecosystem modelers to simulate more complicated interactions between vegetation and changing disturbance regimes such as the affect of more severe fire on vegetation composition.

Summary
The TRS was developed and evaluated at Barro Colorado Island (BCI), Panama where ecologists have collected a significant amount of forest demography and meteorological data since the early 1980s. These data allowed researchers at Lawrence Berkeley National Lab to parameterize TRS algorithms that represent how soil moisture and light affect critical regeneration processes such as seedling emergence, seedling mortality, and seedling to sapling transition rates. By simulating recruitment under observed meteorological conditions researchers were able to compare TRS predictions of recruitment to census observations at BCI. Compared to prior models the TRS makes significant improvements in predicting which types of trees recruit and at what rate under the current climate. By also running the TRS under El Niño, wetter-than-observed, and drier-than-observed precipitation scenarios researchers found that the TRS predicts recruitment responses to varying soil moisture and light levels that are consistent with ecological expectations.

Contact: Adam Hanbury-Brown, University of California, Berkeley, ahanburybrown@gmail.com

Funding
Funding was provided via Next Generation Ecosystem Experiments-Tropics (NGEE-Tropics), funded by the US Department of Energy, Office of Science, Office of Biological and Environmental Research. The lead author also received support from the National Science Foundation and the National Aeronautics and Space Administration during this research.

Publications
A.R. Hanbury-Brown et al., “Simulating environmentally-sensitive tree recruitment in vegetation demographic models”. New Phytologist (2022). [DOI: 10.1111/nph.18059]

NGEE-Tropics scientist Barbara Bomfim was featured in an interview on the “Tree Talk” Podcast hosted by Katie Davies. The episode is titled “Tropical Forestry,” and discusses tropical forest science and soils. Barbara also gives insight into the impact of tropical forest research through the social and economic issues surrounding the Amazon forest. Barbara is a scientist on the NGEE-Tropics project, working as a postdoc on Research Focus Area 2 (RFA2) science questions. RFA2 is primarily focused on forest structure and composition along environmental gradients. You can find Barbara’s full interview by following this link: https://pod.co/treetalkpodcast/2-tropical-forestry.

Headshot of Barbara Bomfim, PhD — NGEE-Tropics Scientist and postdoc.

NGEE-T scientists studied trunk and crown vertical volume profiles to describe the volume contained up to any height in tropical trees.

The Science
As climate changes, it is crucial to monitor the health of tropical forests. Recent studies have reported tree-level damage (i.e., branch fall, trunk breakage, and the decay caused by wood decomposition in standing trees) as one of the most important conditions preceding deaths in tropical trees. However, field-based damage assessments are very limited, in part, due to the lack of whole-tree (trunk + branches) volume equations in tropical trees. Using terrestrial laser scanning, forest ecologists studied the vertical distribution of trunk and crown (i.e., branches) volumes to provide models to estimate the proportion of volume contained up to any height in tropical trees.

The Impact
Field-based assessments of tree damage are increasingly needed to better estimate biomass losses and drivers of tree mortality. This research provides a set of models that can be used to estimate volume losses in living trees when the living length of the trunk and the proportion of newly broken branches are available.

Summary
Tree volume models are critical for forest management and for obtaining accurate forest carbon estimates. In this paper, researchers present species- composite cumulative volume profile models that describe the volume contained up to a given height in the trunks and crowns of tropical trees. They used terrestrial laser scanning (TLS) and quantitative structure models to estimate the trunk and crown volume of 177 trees (49 species) in a lowland tropical forest in the Barro Colorado Island in Panamá. The researchers found that (1) the rate at which volume accumulated with height was much higher and variable in the whole tree (trunk + branches) than only in the trunk; (2) the variability in the rate of volume accumulation was three times higher in the trunk and nine times higher in the whole tree across individuals within species than between species; and (3) parameters describing the rate of volume accumulation significantly depended on the height of attachment of the lowest branch, but not on the tree size.

Contacts
Principal Investigator: Daniel Zuleta, Ph.D., Ecologist (Postdoctoral fellow), Forest Global Earth Observatory at the Smithsonian Tropical Research Institute, dfzuleta@gmail.com
Program Manager: Brian Benscoter, U.S. Department of Energy, Biological and Environmental Research (SC-33), Environmental System Science, brian.benscoter@science.doe.gov

Funding
This project was supported as part of the Next Generation Ecosystem Experiments–Tropics and was funded by the Office of Biological and Environmental Research (BER) within the U.S. Department of Energy’s (DOE) Office of Science. Data collection was supported by the Forest Global Earth Observatory (ForestGEO) of the Smithsonian Institution.

References
Zuleta, D., et al. “Vertical distribution of trunk and crown volume in tropical trees” Forest Ecology and Management 508 120056 (2022). DOI: 10.1016/j.foreco.2022.120056