Technical Comments

Temperature Changes During the Younger Dryas in New Zealand

See allHide authors and affiliations

Science  05 Feb 1999:
Vol. 283, Issue 5403, pp. 759
DOI: 10.1126/science.283.5403.759a

The conclusion by Christiane Singer et al.(1) that there was no significant temperature decline associated with the Younger Dryas (2) in New Zealand is not supported by the pollen records from northwest Nelson, New Zealand, that they present. In reviewing recent late glacial pollen records from New Zealand, McGlone concluded that poor chronological resolution and apparent conflict between the various types of evidence for climate change make detection of an unambiguous Younger Dryas cooling signal in New Zealand problematical (3). In addition, most New Zealand pollen records have suffered from coarse sampling resolution, and their precise climatic signals are inevitably blurred by uncertainties surrounding pollen representation and provenance. In their report, Singer et al. do not account for these well-established limitations, yet all are apparent in the data they present.

First, the chronological control of the three records presented is not strong. Two records have just one basal date, but there is no indication of what time is represented by the rest of these records. Core CV1 has the Younger Dryas constrained by two dates, and therefore is the only relevant record presented.

Second, although it is stated (1, p. 813) that they “resampled and counted one core at 1-cm intervals through the key time phase,” implying that a fine sampling resolution was achieved, this is not apparent in the CV1 profile presented. Rather, figure 2B in the report (1, p. 813) shows a sampling resolution of just eight samples across a sediment thickness of ∼20 cm, which spans the ∼1000 year Younger Dryas interval.

Third, and perhaps most critically, their assertion that there is no evidence for cooling through the Younger Dryas period is not appropriate. Singer et al. attribute a marked oscillation in the Halocarpus (incorrectly referred to as a “warm indicator”) curve to “site hydrology effect,” apparently because it “coincide[s] with a period of inorganic sedimentation at that site interpreted to be the result of the site drying out.” Singeret al. do not offer an explanation for this unconventional interpretation, yet a change from organic to less organic or inorganic sedimentation is frequently observed in records showing Younger Dryas cooling in the Northern Hemisphere (4). Also, Singeret al. do not attempt to reconcile this interpretation with their later statement that “the pollen imply data … that precipitation increased during the deglacial in New Zealand” or with the fact that “drying out” of pollen depositional sites would be expected to result in a hiatus or degraded pollen assemblages. They state that “[c]old-loving taxa … show no increases through the period,” yet the Halocarpus oscillation (and presumably the inorganic layer which is not illustrated, but which Singer et al. state is coincident) coincides with a rise in grass pollen and a decline in lowland forest (which is, in fact, a warm indicator) in core CV1.

Rather than concluding from these data that there was no significant temperature decline associated with the Younger Dryas in New Zealand, one could conclude that the reverse is just as plausible: that is that a minor, short-lived, cooling episode was registered at northwest Nelson. However, given the problems in interpreting pollen records referred to above, a more justifiable conclusion would be that the existence of Younger Dryas cooling cannot unequivocally be confirmed or refuted from these data. Therefore, it is inappropriate to discuss various models of Younger Dryas initiation in the context of pollen evidence from these northwest Nelson records or, for that matter, from elsewhere in New Zealand, and it will remain so until we obtain detailed late glacial pollen records with unambiguous pollen-climate indicators, strong chronological control, and fine sampling resolution at climatically sensitive sites. The ambiguities in inferring climatic signals from the type of data discussed here underpin McGlone's (3) conclusions regarding the likelihood of ever detecting (or rejecting) Younger Dryas cooling in New Zealand from pollen evidence alone.


Response: We thank Newnham for allowing us the opportunity to clarify some points in our report (1). Newnham's most important questions relate to our assertion that there is no evidence for cooling from the New Zealand site. In particular he raises the issues of the “site hydrology effect” and our use of the term “warm indicator” for Halocarpus.

The “site hydrology effect” does require elaboration. The inorganic sedimentation in question is a yellow-brown silty-clay. Unlike the stiff, olive-gray clays associated with the glacial conditions at the base of the core, these yellow-brown silty clays were rich in pollen. Three of the Younger Dryas samples were counted from this clay, and there was an increase in Nothofagus fuscatype pollen at the expense of Halocarpus. BecauseHalocarpus grows on the bog, and mountain beech (N. fusca pollen type) around it, at the present day, it can hardly be attributed to climate change, especially because there is no discernible rise in cold climate types. We take issue with Newnham's attribution of significance to a <5% variation in grass pollen and a slight depression of the lowland forest types. Because this bog acts as a focus for local drainage, we interpret the silt as a flood deposit washed off the surrounding moraine. We have two possible explanations for the high concentration of beech pollen. Either it simply reflects what was washed in with the flood, for which highNothofagus values are entirely predictable, as, coincidently, are the lower lowland forest types. Alternatively, beech may have temporarily expanded onto the bog, on the inorganic sediment surface, after the flood. This latter possibility led us to refer to the “site drying out.” Retrospectively, the term “flood deposit” might have been clearer.

With regard to our use of Halocarpus as a “warm indicator,” we were careful in our report to use “warm” in parentheses. The point that we were making was that a cold Younger Dryas, in a site where valley glaciers would have re-expanded, should have a cold climate pollen record for the period. Newnham is incorrect in asserting that the other records are irrelevant. The identification of a dated Last Glacial Maximum flora in core CV3 specifically tells us what the flora at this site should revert to, or at least toward, if the Younger Dryas really were a significant thermal event. The herb,Phyllocladus, and grass association is distinctive, and even the high-altitude (1300 m) site shown in figure 2C of our report above mean sea level shows no suggestion of a significant increase in these taxa after the start of the deglaciation.

The apparent discrepancy in the core depths and the number of samples in CV1 is the result of decompression of a Livingstone core. Samples were taken at 1-cm intervals on the original core. We would, of course, like to apply more radiocarbon dates to these cores, but the dating is perfectly adequate for the purposes that we have used it.

Finally, this set is just one of a whole series of deglacial pollen records. In totality, the records are unambiguously contrary to a Younger Dryas with a large thermal event in New Zealand. The results of our study do not preclude other types of Younger Dryas signals (for example, small thermal changes or precipitation changes), where our data cannot resolve the issue. We hope that our report will generate more work into the Younger Dryas signal in New Zealand. There must be some Southern Hemisphere response, and New Zealand is a key location to look for it.



Navigate This Article