An early onset of ENSO influence in the extra-tropics of the southwest Pacific inferred from a 14, 600 year high resolution multi-proxy record from Paddy's Lake, northwest Tasmania
Introduction
The tropical Pacific El Niño-Southern Oscillation (ENSO) is an important component of the global climate system, influencing climatic, physical and ecosystem processes both within and beyond the Pacific Ocean basin. Over the historical period, changes in the intensity and frequency of ENSO events are correlated with a range of important processes; such as, trends in rainfall (Hoerling and Kumar, 2003, Moy et al., 2002, Ropelewski and Halpert, 1987), biomass burning (Fletcher et al., 2015, Kitzberger and Veblen, 2003), mammal population dynamics (Lima et al., 2002, Ogutu and Owen-Smith, 2003) and plant phenology (Asner et al., 2000, Dech and Nosko, 2004, Fletcher, 2015). While longer-term changes in ENSO are implicated in major ecosystem transformations (Fletcher et al., 2014), cultural shifts (Sandweiss et al., 1999, Sandweiss et al., 2001, Turney and Hobbs, 2006, Williams et al., 2008) and both terrestrial (Cao et al., 2004) and marine biogeochemical cycling (Karl et al., 1997). Indeed, the “switching on” of ENSO – the shift from a muted to ‘modern’ ENSO system that occurred in the mid Holocene –is implicated in changes in climate-driven processes across the entire Pacific Ocean basin (Fletcher et al., 2015, Fletcher and Moreno, 2012, Sandweiss et al., 1996, Turney et al., 2004). Despite the widespread recognition of the impact of the mid-Holocene amplification of ENSO, uncertainty exists over both where and when the effects were felt in natural systems (Conroy et al., 2008, Moy et al., 2002). Here, we use a high resolution pollen, charcoal and geochemical analysis of lake sediment from northwest Tasmania, Australia, to assess the influence of millennial-scale ENSO variability over climate variability and consequent landscape dynamics at a site located in the mid-latitudes (extra-tropics) of the southwest Pacific.
The dominant role of hydroclimatic variability over long-term Holocene terrestrial ecosystem dynamics in the mid to high-latitudes of the South Pacific Ocean basin is well described (e.g. Fletcher, 2015, Fletcher et al., 2014, Fletcher et al., 2015, Fletcher and Moreno, 2012, Kilian and Lamy, 2012, Lamy et al., 2010, Mariani and Fletcher, 2016, Moreno, 2004, Pesce and Moreno, 2014, Rees et al., 2015, Stahle et al., 2016); and is understood as a dynamic interaction between zonally symmetric shifts in the extra-tropical southern westerly winds (SWW) and the meridionally asymmetric hydroclimatic signature of the ENSO system (Fletcher and Moreno, 2012, Shulmeister, 1999). In other words, millennial-scale shifts in the SWW during the early to mid-Holocene (ca. 12,000–6,000 years ago – 12-6 ka) drove synchronous in-phase vegetation and fire regime changes across the South Pacific, while the mid to late-Holocene (ca. 6-0 ka) was characterised by sub-millennial scale hydroclimatic oscillations across the region that reflect the influence of long-term phasing of tropical ENSO variability (Fletcher and Moreno, 2012). Critically, while it is clear that ENSO variability approaching ‘modern’ appeared first at ca. 6.7 ka (Moy et al., 2002), the timing of the onset of ENSO inferred from proxy records within the Pacific Ocean basin varies widely (usually between ca. 6-3 ka). In east and southeast Australia, a region in which ENSO variability, in particular the warm (El Niño) phase, is a key determinant of rainfall variability and fire activity (Mariani et al., 2016), a consensus of ca. 5 ka appears to have been reached from a range of proxy types (Donders et al., 2007, Fletcher et al., 2015, Rees et al., 2015, Turney and Hobbs, 2006). An initial onset of the effects of ENSO at ca. 5 ka is consistent with a substantial increase in the frequency of El Niño events recorded in the tropical east Pacific (Moy et al., 2002), but it post-dates the initial amplification of ENSO by ca. 1.7 kyrs, suggesting a spatiotemporal complexity of how ENSO influences the climate of the Pacific Ocean basin.
Tasmania is a cool temperate continental island located in the extra-tropics of the Southern Hemisphere (41–44°S), at the extremity of the ENSO zone of influence in the southwest Pacific and positioned at the climatic interface between the tropical ENSO system and the extra-tropical SWW (Mariani and Fletcher, 2016). Tasmania is, thus, critically located to assess how these two important systems interact in space and time (Fig. 1). Evidence for the influence of ENSO over the climate of Tasmania reveals substantial heterogeneity in where and when the impacts of ENSO were first felt in this complex landscape. Recent studies of lake sediment macroscopic charcoal sequences from southwest Tasmania, a region in which ENSO has little explanatory power over modern rainfall variability (Mariani and Fletcher, 2016), reveal that the prominent spike in El Niño frequency at ca. 5 ka (Moy et al., 2002) was sufficient to teleconnect hydroclimatic anomalies to sites located in western Tasmania that are outside the modern ENSO area of influence (Fletcher et al., 2015, Rees et al., 2015). Initial burning of rainforest vegetation at ca. 6 ka in southern Tasmania, where both ENSO and the SWW are important in governing rainfall anomalies (Hill et al., 2009, Mariani and Fletcher, 2016), is synchronous with a peak in tropical El Niño frequency and reveals an influence of ENSO almost 1 kyrs before the initial ENSO impacts reported in southeast Australia and western Tasmania (Fletcher et al., 2014). Finally, a recent long-term vegetation and fire record from northwest Tasmania on the cusp of the SWW-ENSO zone, reports the first influence of ENSO at ca. 3.5 ka (Stahle et al., 2016), synchronous with a prolonged phase of amplified El Niño activity that is implicated in substantial changes in both natural and cultural systems across the entire Pacific Ocean basin (Sandweiss et al., 1999, Sandweiss et al., 2001, Williams et al., 2008), revealing a substantial lag at that site compared to elsewhere in the region.
Here, we aim to assess the role of tropical ENSO variability in governing fire, vegetation and sediment dynamics over the last 14.6 kyrs in part of western Tasmania where ENSO is currently the dominant driver of inter-annual hydroclimatic variability. Our study site is Paddy's Lake in northwest Tasmania, a high altitude lake (1065 masl), located immediately above the modern treeline. We use a multi-proxy approach that employs analyses of charcoal, pollen, geochemistry and radioactive isotopes to reconstruct fire, vegetation, and sediment dynamics at this site over the past 14.6 kyrs. We then compare our results to regional climate proxy data in an attempt to: (1) assess the long-term climatic framework proposed for this region; (2) determine the timing of the initial impacts of ENSO variability on the local ecosystem; and (3) ascertain the impact of changing tropical ENSO variability on the local ecosystem around Paddy's Lake.
Section snippets
Regional geography
Tasmania has a complex topography and a cool temperate maritime climate (Gentilli, 1971). The roughly north-south trending mountain range that bisects Tasmania intercepts the dominant SWW flow and orographic rainfall results in up to 3500 mm of precipitation annually on the west coast, while rainfall drops to as little as 400 mm p/a on the east side of the ranges. Inter-annual rainfall variability in Tasmania is determined by both ENSO and the Southern Annular Mode (SAM - an index that
Sediment recovery
Sediment cores were collected from a floating platform from the deepest point (20.78 m) of Paddy's Lake on November 12th, 2014. A sediment-water interface core (TAS1401 SC1) was collected using a Bolivia coring system, a modified Livingstone system (Wright, 1967), while a long core (TAS1401 N1) was retrieved using Nesje coring system (Nesje, 1992). All cores were packaged in the field and transported whole to allow for core scanning.
Chronology
Nineteen samples were analysed for radiocarbon dating (2 plant
Sediment recovery
Two cores, TAS1401 SC1 (93 cm) and TAS1401 N1 (227 cm), were collected from Paddy's Lake. Radiocarbon dating, charcoal and geochemistry were used to tie the cores into a master sequence. The composite core was 291 cm in length.
Chronology
The results of radiocarbon analysis are presented in Table 1. A maximum radiocarbon age of 12,514 ± 50 14C yrs was obtained at 255 cm. Paired macrofossil-sediment dates reveal an average offset of 223.5 yrs. This offset was applied to the entire sequence to best address
Late Pleistocene [ca. 14.6–11.5 ka]
Our results demonstrate that Paddy's Lake was ice free by ca. 14.6 ka (14,255–15,363 cal yr BP), ca. 2.4 kyrs after cosmogenic evidence of ice retreat of the Last Glacial Maximum (LGM) in Tasmania (ca. 17–20 ka) (Barrows et al., 2002). The delayed onset of organic accumulation at Paddy's Lake, which lies at 1065 masl, is consistent with other higher elevation lakes in Tasmania (eg. Tarn shelf, Upper Lake Wurawina, Tyndall Range (Colhoun, 1996, Macphail and Colhoun, 1985, Macphail, 1979)), at
Conclusions
Climate variability is the key driver of change in fire regimes, geochemistry, and vegetation change at Paddy's Lake, Tasmania. High percentage of non-arboreal pollen and dominance of Poaceae from 14.5 to 13.3 ka indicate cold conditions during the ACR. Following the ACR a significant shift in vegetation and geochemistry are confirmed by warming conditions toward the Holocene and an increase in fire activity. During the earliest Holocene, rainforest appears to have increased at Paddy’s Lake in
Acknowledgements
We acknowledge that our work was conducted on Tasmanian Aboriginal lands and thank the Tasmanian Aboriginal community for their ongoing support of our research. We would like to thank Alexa Benson, Agathe Lisé-Pronovost, Angelica Ramierez, William Rapuc, Scott Nichols and Anthony Romano for their assistance in the field. The financial support of this project comes from the Australian Research Council (award: DI110100019 and IN140100050) and Australian Institute of Nuclear Science and Engineering
References (88)
- et al.
The timing of the last glacial maximum in Australia
Quat. Sci. Rev.
(2002) Methods and code for 'classical'age-modelling of radiocarbon sequences
Quat. Geochronol.
(2010)Application of Iversen's glacial-interglacial cycle to interpretation of the late last glacial and holocene vegetation history of western Tasmania
Quat. Sci. Rev.
(1996)- et al.
Late pleistocene vegetation and climate history of Lake Selina, western Tasmania
Quat. Int.
(1999) - et al.
Holocene changes in eastern tropical Pacific climate inferred from a Galápagos lake sediment record
Quat. Sci. Rev.
(2008) - et al.
Pollen evidence for the transition of the Eastern Australian climate system from the post-glacial to the present-day ENSO mode
Quat. Sci. Rev.
(2007) - et al.
Integration of proxy data and model scenarios for the mid-Holocene onset of modern ENSO variability
Quat. Sci. Rev.
(2008) - et al.
Changes in biomass burning mark the onset an ENSO-influenced climate regime at 42° S in southwest Tasmania, Australia
Quat. Sci. Rev.
(2015) - et al.
Modern pollen–vegetation relationships in western Tasmania, Australia
Rev. Palaeobot. Palynol.
(2007) CONISS: a FORTRAN 77 program for stratigraphically constrained cluster analysis by the method of incremental sum of squares
Comput. Geosciences
(1987)