Tropical Forest Carbon Cycles: Two Modeling Challenges

Abstract

Dry season length plays an important role in determining competitive dvantageousness of different phenological strategies in tropical forests. What remains unclear is at which dry seasons lengths deciduous species or evergreen species will dominate, and at which dry season lengths coexistence between the two is possible. In this paper, I modify a previous criteria developed by Adams et al. (2007) to create a model for conditions for evergreen invasion of a deciduous monoculture and vice versa, based solely on the physiological parameters of both. The competition criteria for deciduous and evergreen trees reduces to three values: the ratios of understory success of the invader species to the monoculture under the canopy of the monoculture, UEE/UDE and UED/UDD, and the ratio of lifetime reproductive success of an open-grown deciduous canopy tree to lifetime reproductive success of an open-grown evergreencanopy tree, rD/rE. Together, UED/UDD, rD/rE, and UEE/UDE form the conditionsfor evergreen dominance, coexistence, and deciduous dominance, moving from one state to the next with increasing dry season length. This theory of competition between deciduous and evergreen trees differs from past explanations of deciduous and evergreen competitive dynamics, which have focused on strategies and success of canopy trees, as it also reveals the importance of understory dynamics in shaping competition, bringing to light the competitive imbalances produced when deciduous and evergreen trees compete in the understory.Carbon storage, particularly above ground biomass (AGB) of tropical forests varies by biogeographic region, with palaeotropical forests in particular having higher AGB than neotropical forests. This regional difference in biomass could be driven by environmental factors, such as rainfall, temperature, soils, or disturbance, or could be driven by regional differences in plant traits, in particular height-diameter allometry. Feldpausch et al. (2010) developed regional allometric equations for African tropical forests that reflect the higher height of African tropical trees compared to Neotropical trees at any given diameter. Current allometric equations in the GFDL LM4-PPA assume pan-tropical height-diameter relationships; if height-diameter allometries indeed vary significantly by biogeographic region, this introduces global bias into tropical biomass estimates. The LM4-PPA has been parametrized in the past for a neotropical forest (Barro Colorado Island, Panama; BCI), but never before for a paleotropical forest. The purpose of this study is to parametrize and test the LM4-PPA for the first time in a palaeotropical forest. To do this, an African-specific plant functional type was created using data from Ankasa, Ghana, and the Feldpausch et al. (2010) regionalheight-diameter allometries, and simulated in both Ankasa and BCI. The LM4-PPA withAfrican allometry underestimated NPP in Ankasa, yet overestimated biomass. This biomass overestimation was due to the fact that the allometric relationships designed both for the BCI location and the Feldpausch et al. (2010) regional allometry vastly overestimated tree height at every diameter. This work is an indication that there is a need for the LM4 model to be tested in more paleotropical locations, and that more variation in PFT allometry needs to be included in the LM4 PFTs, even beyond the regional level.