We modified the ecosystem process model BIOME-BGC, as a novel application for Paleozoic plants (pBGC) for seven dominant plant types (lycopsids, medullosans, sphenopsids, cordaitaleans, peltasperms, tree ferns, and Walchian confiers) of the Late Carboniferous with simulated input meteorology at a daily time-step. Process modeling was based on defining individual plants types based on foliar characteristics that could be derived from fossil material: 1) leaf conductance, both stomatal boundary layer (as a function of leaf morphology), 2) foliar carbon to nitrogen ratios (C:N_leaf), 3) mesophyll pathlength (D_m).  In addition, specific leaf area (SLA), which is an intrinsic character of foliage that represents the allometry of leaf mass to surface area, was defined for these plant types from C:N_leaf from a proxy function between SLA and C:N_leaf values developed from data for modern New Zealand podocarps.  For this project, a new constraint on stomatal conductance was added to the pBGC model in which the parameter D_m, was used to subsume the influence of venation and leaf thickness, factors known to have differed in non-angiosperm plants of the Paleozoic.  Simulations at specific points and globally showed minimal influence of oxygen on ecosystem function such as productivity, water use, or nitrogen cycling despite early suggestion that higher pO₂ would increase atmospheric pressure reducing overall inherent evaporation rates.  Other scenarios assessed including varying pCO₂ between 200 to 600 ppm increased productivity, though reduced water availability and water runoff.  However, much of these effects were due to coincident change in simulated temperature and precipitation associated with these CO₂ changes derived from meteorological data derived from the GENESIS model by Chris Poulsen, University of Michigan.  Rather, much of the change in ecosystem functions simulated (i.e. productivity, water budgets, and nitrogen cycling) were due to differential water use by the different taxa.  In particular, lycopsids, as the primary plant type of the ?coal? swamps of the Carboniferous, were found to have very high stomatal conductance based on the fossil morphology, that coupled with high D_m simultaneously resulted in high water loss with sensitivity to reduced leaf water resupply.  This result is different than previous conceptualization of this plant type which has been represented as being very water conservative based on inadequate comparisons with extant lycophytes. This leads to one of the main conclusions of our study which is that reconstruction of plant function of taxa from deep time are best approached by mechanistic modeling informed from direct fossil evidence, rather than inference of functional limitations based on nearest living relatives. For global simulations of these plant types, we found distinctive geographical distribution limits that we refer to as the physiological landscape.  High correspondence was found between locations on Pangea where our simulated plants were modeled to have grown compared to locations of major fossil assemblage sites.  The major find of global simulations was that the total terrestrial land area covered by Carboniferous vegetation ranged from 20 to 60% based on our assessment of the physiological landscape of plants that existed during that time.  This is compared to modern land area coverage which is approximately 88%.  Limitations to distribution are associated with water management by these taxa.  From a global climate perspective, this provides us our second major conclusion which is that plants of the Carboniferous had limited capacity to affect climate through either carbon or water exchange capacity, rather influenced by their presence or absence across the large Pangean supercontinent through both planetary albedo and interception of precipitation affecting runoff and direct soil erosion. Thus, the land area to support plant growth constrains climate interactions as a function of evolutionary physiological limitations.

Last Modified: 04/25/2019
Modified by: Joseph D White