Miocene/Pliocene shift: one step or several?

Abstract

Cerling et al. reply — Köhler et al. suggest that phenomena other than floral change may be involved in the late Miocene global vegetation change, such as monsoonal dynamics or unnamed “other factors”. Citing evidence from Spain and Pakistan, they do not believe that there is necessarily a synchronicity or a causal link between faunal and vegetation change in the late Miocene epoch. However, on the contrary, it seems highly unlikely that a vegetation change on the scale documented¹ would be uncorrelated with faunal change.

Main

Widespread faunal change in the late Miocene epoch was recognized⁷,^10,11,12 long before the carbon-isotope shift was identified; our work was the first to attempt to link these widespread faunal changes to global vegetation change¹,¹³. For example, in North America, Webb et al.¹⁴ state that the “boundary between the Early and Late Hemphillian (about 6 Myr ago) records a mass extinction event for equids, when about ten of the existing 18 lineages vanished”. Although it is difficult to ‘prove’ causality in historical events, it seems likely that widespread faunal changes are linked to widespread vegetation changes.

The data from the Siwalik sediments in Pakistan are especially informative, because only from this region are there coeval data on faunal turnover, isotope palaeoecology, and upwelling related to monsoon dynamics (Fig. 1). Smoothed palaeosol data for carbon-13 content (δ¹³C) show a sharp change starting about 7 Myr ago and continuing to about 5 Myr ago, denoting the shift from C₃- to C₄-dominated vegetation.

Figure 1: Data from Pakistan's Siwalik sediments show the two biggest events occurring at about 7 and 8.5 Myr ago.

a, δ¹⁸O data from Silawik palaeosols, representing a trend determined by taking a 10-point running average of the roughly 200 palaeosols from the interval 16 to 0 Myr ago¹⁶. Also shown is the fraction of Globigerina bulloides from the Arabian Sea, an indicator of upwelling related to monsoon dynamics¹⁷. b, δ¹³C data for palaeosols and for mammals' tooth enamel¹,^18,19,20 in the Silawiks, representing a trend determined by taking a 10-point running average of the 200 or so palaeosols from the interval 16 to 0 Myr ago¹⁶. c, Faunal change index from the Siwaliks, represented by the number of first (n_f) and last (n_l) occurrences, including only occurrences (n_o), normalized to species richness (n_sr). Data from ref. 7. The index is normalized to 1.0 for the total data set.

The δ¹³C data for tooth enamel show that the dietary change, which enhances the C₃ or C₄ signal by selective feeding, can be seen somewhat earlier than in the palaeosols, a result to be expected. Smoothed δ¹⁸O data from palaeosols indicates a change in soil waters that precedes the δ¹³C shift and which is correlated with increased abundance of upwelling indicators in the Arabian Sea at about 8.5 Myr ago.

Therefore the isotope record in the Siwaliks records two signals: a change in monsoonal dynamics at about 8.5 Myr ago and a pronounced vegetation change at about 7 Myr ago. Detailed faunal collections from the same region document several important turnover events. The two biggest events are at about 7 and 8.5 Myr ago (Fig. 1) and correspond to the two periods of change recorded in the isotope record.

Although the record is indeed complicated, the stable isotope record documents two important events affecting faunal change in the Siwaliks: one starting about 8.5 Myr ago that is related to the monsoon intensification, and a slightly later event related to expansion in C₄ biomass. Earlier faunal changes, such as those before 10 Myr ago as mentioned by Köhler et al., are unrelated to the global expansion of C₄ biomass.

C₄ photosynthesis is an adaptation to low atmospheric CO₂ levels. Because CO₂ gain and water loss both occur through stomata in C₃ plants, we expect that C₃ plants adapted to aridity would prosper in periods of lower atmospheric CO₂. We would therefore expect that global changes within C₃ flora accompanied the C₄ expansion at the end of the Miocene epoch. Changes within C₃ ecosystems can be related to changes in atmospheric CO₂ levels (for example, the Pleistocene/Holocene transition¹⁵).

So, although C₄ plants did not flourish in Europe or in other high-latitude regions, it is likely that floral change occurred in those regions within C₃ ecosystems through the Miocene/Pliocene transition. The absence of evidence for C₄ expansion in Europe should not be taken to mean that floral change did not take place in Europe at the end of the Miocene; the isotope record is silent on that issue.