Paratropical floral extinction in the Late Palaeocene–Early Eocene

The Palaeocene–Eocene Thermal Maximum (PETM) at c. 55.8 Ma marks a transient (c. 100 ka duration) interval of rapid greenhouse warming that had profound effects on marine and terrestrial biota. Plant communities responded rapidly with major compositional turnover. The long-term effects on tropical vegetation communities that stem from the brief period of global warming are unclear. We present pollen data from the paratropical US Gulf Coast (eastern Mississippi, western Alabama and Georgia), which had background Palaeogene mean annual temperatures of 26–27 8C. Sporomorph data (pollen and spores) demonstrate that taxonomic diversity increases over c. 1 Ma in the Late Palaeocene but this trend is replaced, with the first occurrences of taxa that mark the Early Eocene, by a pronounced extinction into the Early Eocene (c. 20% of the palynoflora). Taxonomic diversity also decreases by up to 38% in the Early Eocene. The timing of the extinction is not clearly resolved but may be restricted to the earliest part of the Early Eocene. Two richness estimators (Chao 2 and Jackknife 2) both demonstrate that Late Palaeocene samples contain significantly more taxa than those in the Early Eocene. Extinction on the US Gulf Coast proves that ancient tropical ecosystems were highly susceptible to changes in diversity mediated directly or indirectly by environmental change even during equable greenhouse climates in the early Palaeogene. The impacts of rapid climate warming and changes in atmospheric carbon dioxide on modern tropical plant communities are the subjects of intense study (Bazzaz 1998; Clark 2004; Cowling et al. 2004; Körner 2004). Research interest centres on the role that tropical forests play in the carbon cycle because tropical forests are both significant sources and sinks of CO2 (Bazzaz 1998; Mahli & Grace 2000). Changes in tropical forest structure, composition and biodiversity caused by factors such as deforestation or autogenic community responses may have significant negative feedbacks on the atmospheric carbon cycle (Bazzaz 1998; Bush et al. 2004; Clark 2004; Cowling et al. 2004; Wright 2005). In the future, extinction of selected plant species and changes in the geographical range of many other plants are expected (Bazzaz 1998; Bush et al. 2004; Cowling et al. 2004). However, the long-term impact of warming is unknown and wholly speculative (Hughes 2000; Woodruff 2001; Clark 2004). The geological record preserves intervals of past rapid climate warming in which the response of plant communities can be studied in detail and on different time scales. Specifically, predictions on geographical range changes and diversity dynamics as a result of rapid warming can be explored using the fossil record of plant organs such as pollen. The Palaeocene–Eocene Thermal Maximum (PETM) is foremost among these warming events (Zachos et al. 2001, 2003; Nunes & Norris 2006) and marks the beginning of the Eocene Epoch at 55.8 Ma (Gradstein et al. 2004; Zachos et al. 2005). During the PETM global land and sea surface temperatures increased by 5–10 8C (Kennett & Stott 1991; Fricke et al. 1999; Zachos et al. 2001; Fricke & Wing 2004) over c. 10–20 ka before returning to warm background levels again over the succeeding c. 100 ka (Röhl et al. 2000; Bowen et al. 2001; Farley & Eltgroth 2003; Zachos et al. 2003, 2005). Climate modelling predicts that .4500 Gt of light carbon was released into the atmosphere and oceans during the initial stages of the PETM (Zachos et al. 2005). Major transient floral and faunal composition changes are recorded at high latitude (Clyde & Gingerich 1998; Wing et al. 2005) as a result of the ensuing warming. Despite a paucity of floral localities from within the brief PETM interval (e.g. Wing 1998; Jaramillo 2002; Harrington 2003a, b; Wing et al. 2005), several regions globally have excellent plant records both immediately before and after the PETM. For example, in North America, plant and pollen records are documented from different vegetation types from the warm-temperate Bighorn Basin in Wyoming (438N, 1078W), which had a mean annual background temperature of c. 16–17 8C (Wing 1998; Wing & Harrington 2001), to the paratropical eastern US Gulf Coast (328N, 878W) (Frederiksen 1994, 1995; Harrington 2003a; Harrington et al. 2004), which had a Late Palaeocene–Early Eocene background mean annual temperature of c. 26–27 8C (Wolfe & Dilcher 2000; Ivany et al. 2004). The changes of the US Gulf Coast floras provide an example of the responses of a warm-adapted, taxonomically diverse vegetation type to the impact of rapid climate warming. Hence, we use pollen records from the eastern US Gulf Coast to explore whether there are significant diversity changes in paratropical vegetation types during the Late Palaeocene– Early Eocene. A previous investigation (Harrington et al. 2004) focused on the changes in composition that resulted from immigration of Eocene taxa but lacked sufficient sampling coverage to thoroughly tackle questions on changes in diversity. Thus, we present sporomorph (pollen and spore) data from an expanded series of sections over an extended time period (.1.5 Ma) from the eastern US Gulf Coast, which are used to analyse changes in taxonomic diversity and composition.