Site 1115 is in a low-latitude (9°11.4´S), warm-water environment (average surface water temperature = 25°-28°C today [Garrison, 1996]), with no reports of upwelling. Thus, it is not surprising to find that warm-water conditions dominated here throughout the Pliocene, because the site has moved northward less than about 1.5° of latitude since the beginning of the Pliocene (Goodliffe, 1998). The temperature trends on Figure F4 show only two divergences from the generally warm-water conditions that prevailed throughout the Pliocene, until the final sharp decline which began ~2.7 Ma and continued into the Pleistocene.
It is instructive to compare the temperature trends at this site, based on nannofossil proxies, to those of other areas. The geographically closest study is that of Jansen et al. (1993), who investigated Pliocene climatic trends by oxygen-isotope analysis at a site on the Ontong Java Plateau. Their timescale uses slightly different ages for some datum tie points, but recalculation of their ages using my datum points shows an average difference of <3% between their timescale and the scale I have used here. Jansen et al. (1993) interpreted their results as reflecting expansion of Antarctic glacial ice between 4.6 and 4.3 Ma, which corresponds rather well with the minor cool peak centered on 4.5 Ma shown in Figure F4. Similarly, pronounced cool-water events centered on 3.2 Ma occur in both the Solomon Sea (Fig. F4) and on the Ontong Java Plateau (Jansen et al., 1993). This brief cool interval has also been reported in other studies (based on diverse analytical techniques), ranging in (recalculated) ages between 3.5 and 3.0 Ma (e.g., Keigwin, 1987 [3.1 Ma]; Raymo et al., 1987 [3.4-3.2 Ma]; Rio et al., 1990b [3.1 Ma] [ages in brackets are the originally published ages]). Siesser and de Kaenel (1999) showed the same general cooling trend between 4.5 and 4.2 Ma in the Mediterranean for the early Pliocene but did not show a definite mid-Pliocene cool interval at 3.2 Ma.
Surface waters warmed appreciably in the Solomon Sea after the 3.2-Ma cool peak with a substantial warm peak from ~3.1 to 2.8 Ma. This may correspond to the "mid-Pliocene warm interval" described in several recent papers (e.g., Crowley, 1996; Dowsett et al., 1996; Siesser and de Kaenel, 1999). Crowley (1996), Dowsett et al. (1996), Raymo et al. (1996) and others have stated that the mid-Pliocene was the last time when global average temperatures were greater than temperatures of today. The time of this last warm period is estimated by Raymo et al. (1996) to be around 3.0 Ma. Siesser and de Kaenel (1999) found this to be a recognizable warm peak at several sites in the western Mediterranean; the age of the warm peak there ranges from ~3.0 to 2.6 Ma.
After this warm interval at Site 1115 in the Solomon Sea, temperatures began to decline in Zone NN16b (starting at ~2.7 Ma). I believe this signals the beginning of the well-documented cooling related to the onset of Northern Hemisphere glaciation. King (1996) has summarized the evidence for the timing of the onset of glaciation at between 2.8 and 2.5 Ma. The decline beginning here at 2.7 Ma fits this time frame well. Wei et al. (1988) found that cooling began after ~2.5 Ma on the Galicia Margin, and Siesser and de Kaenel (1999) recorded the beginning of cooling around 2.5-2.6 Ma in the Mediterranean. Jansen et al. (1993) found temperatures progressively declining on the Ontong Java Plateau after 2.85 Ma, with a short but dramatic cool peak occurring just before 2.5 Ma superimposed on the longer-term cooling trend.
An unexpected event occurred in the middle of Zone NN18 (at ~2.3 Ma) in the Solomon Sea, when the overall cooling trend was interrupted by a brief warming event (Fig. F4). This has not been reported elsewhere. However, inspection of the expanded oxygen-isotope plot of Jansen et al. (1993, p. 357) also shows a slightly warmer interval at ~2.25 Ma, interrupting the overall cooling trend on the Ontong Java Plateau.