Geographic range of plants drives long-term climate change

Atmospheric CO2 is a primary driver of Earth's long-term temperature, with Phanerozoic fluctuations attributed to tectonic and biotic events. While stabilizing processes like silicate weathering prevent runaway climates, the relative contributions of individual processes remain debated. Existing Earth system models struggle to reproduce Phanerozoic climate change accurately. Vegetation plays a crucial role, influencing carbon, energy, and water exchange. Plants act as both carbon sinks (photosynthesis, weathering enhancement) and sources (respiration, biomass burning). These interactions vary with local climate conditions, and changes in plant community structure affect climate by altering carbon exchange rates. Current long-term models often oversimplify plant influences, using global values instead of accounting for spatial heterogeneity driven by local temperature, moisture, and light regimes. This simplification reduces the reliability of carbon cycling predictions. This study advocates for incorporating detailed plant geographic range dynamics into models to better understand vegetation's influence on biogeochemical cycles and climate over geological timescales.

Existing models like GEOCARB and SCUUM attempt to represent plant-climate feedbacks, often incorporating plant fossil data. Early land colonization is hypothesized to have drawn down CO2 and cooled the climate in the Paleozoic, although this remains uncertain. However, these models often lack spatial resolution in representing vegetation, neglecting the heterogeneity of plant interactions with local climates and biogeochemistry. This study aims to address this limitation by incorporating detailed plant geographic range dynamics.

This research combines the FLORA (Fast Land Occupancy and Reaction Algorithm) vegetation model with the SCION (Spatial Continuity Integration) biogeochemical model. FLORA provides a spatially resolved and dynamic representation of terrestrial vegetation, while SCION incorporates abiotic parameters like runoff and weathering rates. The coupled model assesses how plant-climate feedbacks affected climate regulation over geological timescales. SCION uses pre-run general circulation model (GCM) climate simulations to calculate continental weathering rates on a 2D grid, considering local temperature, erosion, and runoff. Previously, vegetation in SCION was represented by a single global biomass value. The integration of FLORA allows for a dynamic spatial biosphere, mapping biomass distribution across continents over time (Triassic to present). The analysis focuses on this timeframe because FLORA currently lacks the representation of major Paleozoic plant evolutionary changes (roots, vasculature, wood). SCION-FLORA model outputs are compared to proxy data for atmospheric CO2 concentration and global average surface temperature. Proxy data includes paleosols, stomatal ratios from fossil Ginkgo, and paleotemperature records from climate-sensitive lithologies and oxygen isotope records. The model's sensitivity is analyzed across 1000 runs, with successful runs used for comparison against proxy data. Biotic enhancement of silicate weathering in the model is calculated using a linear relationship with FLORA net primary productivity (NPP) and an abiotic factor, reflecting the assumption that higher NPP leads to greater weathering enhancement. Biomass death in FLORA considers the effects of increased atmospheric oxygen, with higher oxygen leading to increased turnover. Leaf respiration in FLORA is calculated using daylight hours, assuming an average of 12 hours per day across all latitudes.

The coupled SCION-FLORA model shows improvements in predicting atmospheric CO2 concentration and global average surface temperature compared to the original SCION model (without spatial vegetation). The model demonstrates that limited plant geographic range over the arid Pangaea supercontinent resulted in lower carbon fixation rates and up to double the previously predicted atmospheric CO2 concentration. The Mesozoic dispersion of continents increased model plant geographic range, leading to amplified global CO2 removal. The terrestrial biomass distribution in SCION-FLORA closely follows the distribution of continental forest, influenced by water availability and temperature. The model highlights a 20% increase in plant-habitable area between the Triassic and early Cretaceous. The breakup of Pangaea resulted in more humid conditions and reorganized tropical circulation, which expanded vegetation and enhanced silicate weathering despite a relative drop in plant productivity. This contributed to a 2-3°C global temperature decrease. The model indicates that even with high Mesozoic plant biomass, the biogeochemical and climatic effects were limited by plant distribution. During the Triassic and Jurassic periods, the spatially dynamic vegetation resulted in increased model CO2 concentrations and surface temperatures. During Pangaea, aridity limited plant habitable area, and restricted biomass to tropical regions. The breakup of Pangaea increased habitable areas and promoted a global spread of vegetation. The model suggests that assessments of post-continental weathering rates should consider the habitable area for plants, which may be especially crucial for the Paleozoic expansion of plants.

The results address the research question by demonstrating the significant impact of plant geographic range on long-term climate change. The findings highlight a previously under-explored mechanism influencing Phanerozoic climate regulation. The improved model predictions compared to previous models underscore the importance of incorporating spatial vegetation dynamics. The study's findings are relevant to the field of paleoclimatology and Earth system modeling, providing insights into the complex interplay between biotic and abiotic factors in shaping long-term climate patterns. The model's limitations, such as the lack of consideration for angiosperm radiation and dynamic feedback between the GCM climate model and the local plant biosphere, do not appear to alter the primary conclusions.

This research demonstrates that plant geographic range is a significant, yet often overlooked, driver of long-term climate change. The coupled SCION-FLORA model provides improved predictions of atmospheric CO2 and temperature compared to previous models that lacked spatial vegetation dynamics. Future work should incorporate evolutionary events like angiosperm radiation and dynamic feedback between the climate model and local plant biosphere to further refine the model and improve our understanding of plant-climate interactions over geological timescales.

The model's limitations include the lack of consideration for evolutionary events such as the angiosperm radiation and the absence of dynamic feedback between the GCM climate model and the local plant biosphere. The model also omits large external perturbations to the carbon cycle, such as large igneous provinces and mass extinction events. While these limitations do not appear to significantly alter the main conclusions, future improvements to the model should incorporate these factors.