Assessing the impact of mid-to-late Holocene ENSO-driven climate change on toxic Macrozamia seed use: a 5000 year record from eastern Australia

Journal of Archaeological Science 40 (2013) 471e480

Contents lists available at SciVerse ScienceDirect

Journal of Archaeological Science journal homepage: http://www.elsevier.com/locate/jas

Assessing the impact of mid-to-late Holocene ENSO-driven climate change on toxic Macrozamia seed use: a 5000 year record from eastern Australia Brit Asmussen a, b, *, Paul McInnes c a

Archaeology, Cultures and Histories, Queensland Museum, South Bank, Post Office Box 3300, South Brisbane, Queensland 4101, Australia Archaeology Program, School of Social Science, The University of Queensland, St Lucia, 4072, Australia c Research and Innovation Division, The University of Queensland, St Lucia, 4072, Australia b

a r t i c l e i n f o

a b s t r a c t

Article history: Received 11 March 2012 Received in revised form 11 June 2012 Accepted 13 June 2012

Palaeoenvironmental and palaeoclimatic data indicate that during the mid-to-late Holocene eastern Australia became significantly drier and experienced more intense and more frequent droughts. These changes, driven by the re-emergence and intensification of the ENSO climate phenomena, have been argued to have had considerable impact on Aboriginal societies, although there is uncertainty as to the exact nature, timing and magnitude of this impact. This paper analyses changes in the utilisation of toxic Macrozamia (cycad) seeds at seven archaeological sites in eastern Australia, identifying an extremely close correlation between the intensity of seed use and two proxy ENSO datasets, and a weaker correlation with a third ENSO dataset. Given the ecological attributes and resource potential of these plants, it is argued that these correlations are best explained as an intensified exploitation of a lower-ranked resource in direct response to the increased subsistence risks and lower productivity created by ENSO-driven climatic conditions. It also suggests that by 3000 BP the intensification of the ENSO system was driving changes in human subsistence behaviour on a sufficient scale to have considerable impact on other aspects of the wider cultural systems. Ó 2012 Elsevier Ltd. All rights reserved.

Keywords: ENSO Holocene Aboriginal Australia Macrozamia Toxic plant Seeds Backed artefact Palaeoclimate

1. Introduction 1.1. ENSO El Niño e Southern Oscillation (ENSO) is a global-scale climate phenomenon where an irregular cycle of fluctuations in the sea surface temperatures in the Pacific Ocean influence air circulation and rainfall in many parts of the world. Warm sea surface temperatures in the eastern Pacific (El Niño conditions) bring reduced rainfall to eastern Australia and wetter conditions to coastal Ecuador and Peru. Cooler sea surface temperatures bring the opposite conditions (La Niña). Interaction with other climatic systems extends ENSO’s reach to many other major regions (e.g. North Pacific, north tropical Atlantic and Indian Oceans, Alexander et al., 2002: 2205). Extended or intense El Niño or La Niña events are associated with extreme climatic conditions including floods, * Corresponding author. Archaeology, Cultures and Histories, Queensland Museum, South Bank, Post Office Box 3300, South Brisbane, Queensland 4101, Australia. Tel.: þ61 07 3840 7604; fax: þ61 07 3846 1918. E-mail addresses: [email protected], [email protected], [email protected] (B. Asmussen), [email protected] (P. McInnes). 0305-4403/$ e see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jas.2012.06.005

droughts, fires, and changes in rainfall and storm patterns. Modern accounts and early written records attest to the significant socioeconomic impacts wrought by past El Niño conditions in the ENSO region including crop failures, malnutrition, famines and population decreases (Grove and Chappell, 2000). In Australia, palaeoenvironmental evidence indicates that climate variability associated with the ENSO phenomenon was largely absent during the first half of the Holocene, with climates relatively stable, and warm and wet (Donders et al., 2007; Gagan et al., 2004; Kershaw, 1995: 667; Shulmeister, 1999). The ENSO cycle re-established itself by c. 5000 BP (Haberle, 2000: 69). Between 4000 and 5000 BP climate conditions altered rapidly (Kershaw, 1995: 669) and there was a further marked intensification of ENSO between 2000 and 3700 BP, causing significantly drier conditions, more frequent droughts and increasing climatic variability (Donders et al., 2007). Several palynological datasets covering the mid-to-late Holocene indicate a reduction in precipitation, reductions in canopy cover, lower lake levels and lower river flow velocities, while increased climatic variability resulted in increases in sclerophyll and eucalypt-dominated communities, heath and peatlands, sand-dune reactivation, massive dust-storms (Kershaw, 1995: 669, 670) and large intense fires (Black et al., 2007).

472

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

This two-step intensification of ENSO is reflected in a set of highresolution pollen records from different regions in Australia, which indicate an early Holocene moisture optimum starting around 8000e9000 BP, an initial drying phase starting around 5000e5500 BP and an intensified drying phase starting 3000 BP (Donders et al., 2007: 1634, 1635). After 2000 BP the conditions ameliorated with the frequency and intensity of ENSO events decreasing towards modern-day levels, although within this period conditions were notably wetter during the Medieval Climatic Anomaly 800e1200 BP and cooler during the Little Ice Age at 200e600 BP (Williams et al., 2010). In Australia, there has long been a recognition of the potential impact of Holocene climate change on Aboriginal societies (for example, David and Lourandos, 1998; Hiscock, 1994; Hiscock and O’Connor, 2005; Mulvaney and Kamminga, 1999; Rowland, 1983, 1999). However, with the emergence of more detailed data on the nature and timing of ENSO-driven climatic change it has become possible to examine possible impacts in more detail. For example, Smith and Ross (2008) have used summed radiocarbon probabilities to argue that periods of increased ENSO activity corresponded with significant declines in the population in the arid zone. Turney and Hobbs (2006) have used similar methods to argue that there was a dramatic and sustained increase in landscape activity at inland sites from 4860 BP driven by intensified ENSO activity. Haberle and David (2004: 165) have argued for a climatic driver in the adoption of innovative subsistence strategies and technologies, tied to social fissioning and regionalisation of populations. Williams et al. (2010) have gone further and argued that changes in site use, rock art, territoriality, and the use of plant processing, often considered the result of population change or economic intensification, can all be considered as a series of responses to climatic variability (Williams et al., 2010: 832). A number of researchers have emphasised the increased subsistence risks created by the harshness and unpredictability of the environment (Attenbrow, 2004: 218; Attenbrow et al., 2009; Hiscock, 2002, 2006, 2008; see also Cosgrove et al., 2007; Veth, 2005; cf. White, 2011 and comments thereon). Lithic evidence has been used to argue that risk mitigating strategies may have been widespread across Australia in the mid-to late-Holocene (see Clarkson, 2007: 52 and Hiscock and O’Connor, 2005: 59). For example, in south-eastern Australia the production of backed artefacts, a type of stone artefact, increased substantially during cooler, drier conditions of the mid-Holocene c. 3500e4000 cal BP, and were produced in lesser quantities after c. 1400e1500 cal BP (Attenbrow et al., 2009: 2769; Hiscock, 2002, 2006). It is argued that the timing of the proliferation of backed artefacts suggests a strong climatic trigger (Attenbrow et al., 2009: 2769). Recent residue and use wear analysis indicate that these were general purpose tools, used in the working of wood and bone, and a variety of other tasks (Robertson et al., 2009; see also Attenbrow et al., 2009). Hiscock argues that these uniform, multi-purpose tools would be most cost effective in “circumstances with low resource predictability, induced by either high mobility and/or unfamiliarity with the environment, in which systematic scheduling of activities was difficult” (Hiscock, 2006: 85). Similar arguments have been made for the emergence and rapid continent-wide spread of the tula, another multi-purpose stone tool (Veth et al., 2011). 1.2. Macrozamia This paper explores the rate of deposition of toxic Macrozamia seeds in archaeological sites to provide a new line of evidence for assessing the nature of possible ENSO-driven subsistence changes. Macrozamia is a genus of cycad with c.41 species (Terry et al., 2008: 321), all of which are endemic to Australia (Jones, 1994: 14; 230).

These palm-like plants grow in the subtropical and temperate regions of the continent, growing in open forest and woodland and often form part of the understory shrub layer from the coast to inland gorges (Hill, 1998; Jones, 1994: 232). Macrozamia are slowgrowing plants that can survive for centuries and often appear in small colonies, making them stable features of the economic landscape. The woody seed shells (sclerotesta) survive comparatively well in archaeological deposits, and they have been excavated from a number of archaeological sites around Australia with the oldest specimens deposited prior to 15,680 years bp in Western Australia (Smith, 1982, 1996) and significant assemblages spanning the last 5000 years in eastern Australia. Mature, female plants bear large seed cones (strobili) (Jones, 1994: 44) holding from 80 to 300 seeds, depending on the species (Asmussen, 2009). Seed production is extremely energy demanding (Tang, 1990: 371e373). Seed production is generally quite variable from year-to-year especially in areas of lower rainfall. On occasion mast seeding occurs, where a high proportion of the plants in an area seed at the same time. Mast events are usually followed by an interval of several years where few seeds are produced, allowing starch reserves to be recouped by plants (Ballardie and Whelan, 1986: 100). Large, robust, older female plants, and those growing in well-watered environments can produce several strobili in a reproductive event, while others in less fertile environments may produce a strobilus once every four years, with gaps of between 10 and 15 years observed in some Macrozamia species (Baird, 1939: 154; Jones, 1994: 15; Ornduff, 1991: 206). From a human point of view, Macrozamia seeds are a toxic, low energy food with an unusually large window for exploitation and whose future production can be monitored a year in advance. Due to the long period of cone development (18 months for one species as documented for Brough and Taylor, 1940: 496e497), strobili are visible on plants for approximately 12 months (Brough and Taylor, 1940: 496e497, pers. obs. Macrozamia moorei), allowing foragers to identify productive plants well in advance. Fresh kernels are highly toxic, and contain several carcinogenic and neurotoxic compounds (see Brenner et al., 2003; De Luca, 1990). Known methods of processing include roasating, leaching in water and combinations of the two methods. Seeds were leached for between two days to a number of weeks, with the longer periods of leaching functioning as short-term food storage (Asmussen, 2010a). Available data indicate that the seeds have comparatively low energy value (362 kJ per 100 g, and 7.3 g of protein for Macrozamia communis (L.A.S. Johnson, 1959), Miller et al., 1993). Leaching reduces this further by removing starch from the kernels (Beaton, 1977). Seeds can remain in the environment in an edible form for over a year, extending the window during which the seeds could be exploited. Despite earlier suggestions that seeding could be synchronised via fire (Beaton, 1982), there is little ecological evidence that Macrozamia production can be reliably manipulated in this way and regular fires may harm the plants and reduce future seed production (Asmussen, 2009). 2. Materials and methods 2.1. The sample A literature review identified several archaeological sites in eastern Australia containing Macrozamia spp. These included Kenniff Cave, The Tombs, Cathedral Cave, Rainbow Cave and Wanderer’s Cave in the Central Queensland Highlands (CQH); and Botobolar 5, Noola, Capertee 3, Angophora Reserve, Durras North, Currarong 1 and Bomaderry Creek in New South Wales (see Fig. 1). Published counts of the number of identified specimens (NISP) per excavation unit were available for Botobolar 5 (Colvill, 1995;

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

Fig. 1. Location map of sites included in the analysis.

Pearson, 1981) and Angophora Reserve (Beck and Webb, 1992; McDonald, 1992). Assemblages from Capertee 3 (McCarthy, 1964), Bomaderry Creek (Lampert and Steele, 1993), Cathedral Cave, Rainbow Cave and Wanderer’s Cave were analysed by Asmussen (2005; see also 2008, 2009), and NISP per excavation unit calculated for these sites. In most sites, Macrozamia specimens were distributed evenly throughout excavated layers, indicating that the assemblages were accumulated from multiple events over time. The exceptions to this were Bomaderry Creek, which apparently included one 2 cm layer of seed shells in Square A13 (Lampert and Steele, 1993: 60), and the specimens from Capertee 3, which were located in a hearth, possibly representing a single depositional event. However, it is not clear whether the whole of the hearth or a portion was excavated. In a published account McCarthy wrote that “a mass of Macrozamia shells and kernels ... were uncovered in a hearth 600 thick” (c. 15 cm) (McCarthy, 1964: 199), however a handwritten note by McCarthy found with the Capertee 3 Macrozamia specimens states that at Capertee 3 “a portion of the Macrozamia hearth was excavated” in section 6 between 4 and 1200 inches (c. 10e30 cm) depth. Macrozamia assemblages from Kenniff Cave, The Tombs, Durras North and Noola were unable to be included in this analysis.

473

Mulvaney and Joyce (1965: 168, 198) reported that Macrozamia was present throughout the top 18 inches (c. 46 cm) of Kenniff Cave, and the “upper horizons” of The Tombs, but not present below this. Unfortunately Macrozamia was not retained following excavation and quantities were not described in published accounts. However, phytolith analysis of the sediments from Kenniff Cave has confirmed the presence of Macrozamia in the upper 46 cm of the deposit and not below this depth (Bowdery, 2006). Macrozamia remains from Durras North (Lampert, 1966) and the assemblage between 10 and 15 cm depth from Wanderer’s Cave (Beaton, 1991a) were unable to be analysed as they were unable to be relocated and quantities were not described by the excavators. The assemblage from Noola was not available at the time of this analysis and has not been included in these figures. The single specimen excavated from Currarong 1 was not included in this analysis (Lampert, 1971) for reasons of sample size. From the seven selected sites, care was taken to ensure methodological comparability in current and previous collections of NISP data. For the sites analysed by Asmussen, each Macrozamia specimen (including sclerotesta, sporophylls and kernels and strobili pieces) was analysed under strong bi-directional light using a 30 hand lens. Only sclerotesta (woody seed shell) was used in the calculation of NISPs as it is directly related to the extraction of the edible kernel. The sclerotesta of each specimen was examined for presence of sarcotesta (fleshy seed coating), rodent toothmarks and insect damage. Evidence of human seed processing was identified using fracture characteristics identified in previous replicative seed processing experiments (Asmussen, 2008, 2010b; see also Beck, 1989). Seeds with evidence that they were humanly fractured were included in the analysis (using correlates outlined in Asmussen, 2010b), while those breached by rodents or insects were not. In Beck and Webb’s analysis of Angophora Reserve, the breakage pattern of each sclerotesta specimen was analysed to confirm human processing. Only one specimen exhibited gnaw marks, suggesting animal disturbance was not a major factor affecting the remains (Beck and Webb, 1992: 180). In her analysis of Botobolar 5, Colvill (1995, Table 1 and Appendix 2) presented raw data in an appendix and along with a tally of the part of plants represented, made several observations on the preservation of each specimen. Colvill identified 23 specimens with insect damage (present in different layers rather than a group). No rodent gnawing was identified on Macrozamia specimens. In the majority of NSW sites, Macrozamia were identified by previous researchers as M. communis, with the exception of Botobolar 5, which was “tentatively” identified as Macrozamia secunda (C. Moore, 1884) by L.A.S. Johnson (then of the NSW Herbarium) (Pearson, 1981: 130). The CQH assemblages were identified as M. moorei (F. Mueller, 1881) (Beaton, 1977). 2.2. Quantification of Macrozamia To enable direct comparisons between sites, estimates were calculated for the number of Macrozamia deposited for each 100 year period for the last 5000 years. Excavation units were dated using ageedepth curves. For Capertee 3, both the published, calibrated depth and age of strata (Hiscock and Attenbrow, 2005) and the date for the hearth obtained by Johnson (1979) were utilised. Ageedepth curves were constructed for Angophora Reserve (see McDonald, 1992), Bomaderry Creek (Lampert and Steele, 1993), Botobolar 5 (Pearson, 1981) and the CQH sites (Beaton, 1977, 1991a, 1991b; Asmussen, 2005, 2009). All radiocarbon dates were calibrated using Calib 6.0 (Stuiver and Reimer, 1993). All dates used in this analysis were derived from terrestrial charcoal samples, and the SHCal04.14C terrestrial dataset was used. After calibration, the median probability for each

474

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

radiocarbon determination was used and plotted at the depth of the sample on an ageedepth curve. The depth of each XU was drawn and the start and end of each XU determined graphically. Between one and eight determinations were used to construct ageedepth curves (see Table 1). No additional dates were obtained for this analysis. Chronologies are considered reasonably secure for the CQH sites (which had a date for each XU), Capertee 3 and Botobolar 5 (dated hearth events) and Angophora Reserve (multiple dates with problematic samples identified and excluded by the original researchers). Bomaderry Creek is less secure with a single date available from the middle layer of the site. The number of specimens belonging to each 100 year period was estimated by assuming a constant rate of deposition within each excavation unit. For example, an excavation unit spanning 175e350 BP would have 14% of its specimens allocated to 100e200 BP, 57% allocated to 200e300 BP and 28% allocated to 300e400 BP. A Macrozamia deposition index (MDI), measuring the relative intensity of deposition through time, was then calculated for each 100 year period at each site by dividing the number of specimens in each period by the largest number of specimens at the site in a single period, creating a score between zero and one (Supplementary Figs. S1 and S2). The mean MDI was calculated for each 100-year period, equal to the total MDI for the 100-year period divided by the number of sites in the sample occupied at the time. This provided a smoothed, aggregate measure of broad trends in Macrozamia use giving equal weight to trends in each site (Fig. 2).

BP. The data from Conroy et al. (2008b) were graphed and a graphical mean value calculated for each 100-year period by measuring the area under the curve. Makou et al. (2010a) analysed the concentration of two molecular organic geochemical proxies in a sediment core from the Peru Margin. Elevated levels of cholesterol were argued to be the result of upwelling of nutrient rich waters characteristic of La Niña conditions and elevated levels of dinosterol the result of dinoflagellate abundances and red tides associated with past El Niño events. This provided a time-averaged record of the overall prevalence of La Niña and El Niño conditions over time. They identified a close correlation between the frequency of El Niño and La Niña events through time, and an increase in both La Niña and El Niño conditions in the last 2000 years. A total of 65 data points were listed (Makou et al., 2010b), spanning the period 494e3868 BP. The data were graphed and a graphical mean value calculated for each 100-year period by measuring the area under the curve. Moy et al. (2002a) estimated the frequency of ENSO events per 100 years by analysing distinctive inorganic layers deposited by flooding in Laguna Pallcacocha, southern Ecuadorian Andes. Based on a comparison with historical data they inferred that only moderate-to-strong El Niño events created identifiable layers in their sediment cores. Analysis of these cores identified an increasing frequency of these events between 7000 BP and 1200 BP then a decrease towards the present. The number of moderate-tostrong events per 100-year period used in this paper was taken from Moy et al. (2002b).

2.3. Palaeoclimatic proxy data

2.4. Taphonomic cross-checks

Three palaeoclimatic datasets were used in the analysis, providing information on two different aspects of ENSO-driven climatic change. Data from Makou et al. (2010a) and Conroy et al. (2008a) were used as measures of the overall frequency of El Niño and La Niña events regardless of their intensity, a proxy measure of the inter-annual climatic variability experienced in the study region. Estimates of the number of moderateto-strong El Niño events through time from Moy et al. (2002a) provided a measure of the frequency of major droughts in the study region. Conroy et al. (2008a) analysed the sediments in El Junco Lake in the Galapagos Islands in a continuous record dating back to 9200  160 cal years BP. The islands are located in the middle of the ENSO system, with rainfall largely determined by El Niño conditions. Based on historical data they argued that the proportion of silt in their cores reflected overall rainfall events associated with El Niño events. Sediment analysis identified an increase in silt abundance and decreasing C/N values indicative of increased rainfall with a two-step transition at 3200  160 and 2000  100 cal years

An overall increase in the mean MDI through time was identified (Fig. 2), raising the possibility that the trend reflected progressive loss of specimens through time rather than changing human use. However, the patterns of deposition at individual sites are not consistent with simple constant decay (see Supplementary Figs. S1 and S2) and preservational issues do not appear to be responsible for the trends. Taphonomic research conducted by Asmussen (2005) on the effects of burial, trampling, root-etching, insect damage, rodent gnawing, water submersion and pH was used in a general assessment of the physical condition of the archaeological Macrozamia specimens. This indicates very few cases where older specimens had significant physical degradation. Detailed checks were possible for the Central Queensland Highland sites where detailed faunal data were also available. Analysis of the bone at Wanderer’s Cave shows that Macrozamia deposition is higher in the upper levels and bone deposition higher in the lower levels, indicating that good overall preservation of organic material and different trends in bone and Macrozamia. Cathedral Cave displays excellent preservation of bone in lower layers where there are few Macrozamia. At Rainbow Cave the deposition of bone and deposition of Macrozamia peak in the middle levels of the site, not a pattern that can be explained by simple loss of specimens with time. While there are undoubtedly taphonomic factors at work in these sites, the overall trend in the CQH is not explained by progressive taphonomic loss and it is inferred that trends in mean MDI reflect changes in deposition of Macrozamia in the sites over time. Assessments of taphonomic integrity provided by other site investigators were also considered. The Botobolar 5 occupational sequence was described as a lower phase containing stone artefact sequence only, a climatically related occupational hiatus (1500e4000 years ago), and an upper 20 cm of deposit in which lithics, Macrozamia and faunal specimens were present. Pearson

Table 1 Summary of the assemblages included in the analysis. Site

Date of first Period with NISP # Dates for Species occupation Macrozamia ageedepth (cal BP) (cal BP) curves

10,000 Capertee 3a Angophora Reserveb 2650 Bomaderry Creek 3350 Botobolar 5c 6840 Cathedral Cave 3780 Wanderer’s Cave 6517 Rainbow Cave 4261 a b c

c. 1940 0e1775 0e3350 0e268 0e2334 0e4850 0e1800

204 121 3051 465 2870 1517 113

1 (hearth) 3 1 (middle) 1 (hearth) 4 6 8

M. M. M. M. M. M. M.

communis communis communis secunda moorei moorei moorei

Capertee 3 date of first occupation after Hiscock and Attenbrow (2005), p. 29. NISP data from Beck and Webb (1992). NISP data from Colvill (1995) and Pearson (1981).

475

1-100 101-200 201-300 301-400 401-500 501-600 601-700 701-800 801-900 901-1000 1001-1100 1101-1200 1201-1300 1301-1400 1401-1500 1501-1600 1601-1700 1701-1800 1801-1900 1901-2000 2001-2100 2101-2200 2201-2300 2301-2400 2401-2500 2501-2600 2601-2700 2701-2800 2801-2900 2901-3000 3001-3100 3101-3200 3201-3300 3301-3400 3401-3500 3501-3600 3601-3700 3701-3800 3801-3900 3901-4000 4001-4100 4101-4200 4201-4300 4301-4400 4401-4500 4501-4600 4601-4700 4701-4800 4801-4900 4901-5000

Mean MDI

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

Fig. 2. Mean MDI per occupied site, 0e5000 cal BP.

(1981: 148) suggested that the trends in the occupation of the site were most likely related to cultural phases and climatic variability resulting in people using the site and landscape differently throughout the last 4000 years. At Bomaderry Creek, Lampert and Steele (1993: 63) conducted an in-depth site formation and taphonomic analysis, including considering skeletal part representation and differential preservation, effects of shelter morphology on organic preservation (e.g. dripline effects), bone weathering, site ‘clean-up’ activities and the effect of camp dogs on faunal evidence. They concluded that there was very good preservation of organics, no major taphonomic decay processes, and that the only significant factors were clean-up activities and trampling causing downward movement of lithics. The specimens at Angophora reserve were excavated from an alkaline shell midden matrix (pH 8e9.5) that offered an exceptional preservation environment for plant remains, including paperbark and seeds from Macrozamia, Banksia spp. and Casuarina spp. (McDonald, 1992: 39, 118).

3. Results As indicated in Figs. 3e6, trends in Macrozamia deposition closely track the ENSO proxy data regarding overall levels of ENSO activity from Conroy et al., (2008b) (Fig. 3) and Makou et al. (2010b) (Figs. 4 and 5). There is a far weaker relationship between MDI and the frequency of moderate-to-strong ENSO events from Moy et al. (2002b) (Fig. 6). To further assess the relationship between Macrozamia deposition and ENSO climate change, Spearman Rank Correlation coefficient was calculated between the mean MDI and each ENSO proxy dataset (Table 2 and Supplementary Figs. S3eS10). The key results are: 1. There are strong correlations between mean MDI and palaeoclimatic proxy data measuring inter-annual variability from Conroy et al., (2008b) (rs ¼ 0.853, p < 0.001) and Makou et al. (2010b) (dinosterol rs ¼ 0.893, p < 0.001; cholesterol rs ¼ 0.824, p < 0.001).

0.5

95

0.45 85

0.4

% Silt

0.3 65

0.25 0.2

55

Mean MDI

0.35

75

0.15 0.1

45 35

1-100 101-200 201-300 301-400 401-500 501-600 601-700 701-800 801-900 901-1000 1001-1100 1101-1200 1201-1300 1301-1400 1401-1500 1501-1600 1601-1700 1701-1800 1801-1900 1901-2000 2001-2100 2101-2200 2201-2300 2301-2400 2401-2500 2501-2600 2601-2700 2701-2800 2801-2900 2901-3000 3001-3100 3101-3200 3201-3300 3301-3400 3401-3500 3501-3600 3601-3700 3701-3800 3801-3900 3901-4000 4001-4100 4101-4200 4201-4300 4301-4400 4401-4500 4501-4600 4601-4700 4701-4800 4801-4900 4901-5000

0.05

%Silt

0

Mean MDI

Fig. 3. Mean silt% from El Junco Lake cores superimposed on mean MDI for eastern Australian archaeological sites, 0e5000 cal BP, silt data from Conroy et al. (2008b).

0.5

550

0.45

500

0.4

450

0.35

400

0.3

350

0.25

300

0.2

250

0.15

200

0.1

150

0.05

100

0

501-600 601-700 701-800 801-900 901-1000 1001-1100 1101-1200 1201-1300 1301-1400 1401-1500 1501-1600 1601-1700 1701-1800 1801-1900 1901-2000 2001-2100 2101-2200 2201-2300 2301-2400 2401-2500 2501-2600 2601-2700 2701-2800 2801-2900 2901-3000 3001-3100 3101-3200 3201-3300 3301-3400 3401-3500 3501-3600 3601-3700 3701-3800 3801-3900

600

Mean dinosterol

Mean MDI

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

Dinosterol (µg/g OC)

476

Mean MDI

Fig. 4. Mean dinosterol levels from the Peru Margin superimposed on mean MDI for eastern Australian archaeological sites, 500e3900 cal BP, dinosterol data from Makou et al. (2010b).

2. There is a weak correlation between mean MDI data and palaeoclimatic proxy data measuring the frequency of moderate-to-strong ENSO events from Moy et al. (2002b) (rs ¼ 0.236, p ¼ 0.098). 3. There is a somewhat stronger correlation with dinosterol concentrations (El Niño) than cholesterol (La Niña) from Makou et al. (2010b). 4. Overall, the changes in MDI correlate more closely with interannual variability (overall frequency of El Niño and La Niña events) rather than by the frequency of moderate-to-strong ENSO events during this period.

4. Discussion As indicated in Fig. 2, there are significant changes in the intensity of deposition of Macrozamia over the last 5000 years. Based on the results of the taphonomic analysis, it is argued that

the changes in Macrozamia deposition reflect genuine changes in the intensity of human exploitation of this resource over this period. Between 3000 and 5000 cal BP there was a low intensity of exploitation, a tenfold increase between 2000 and 3000 cal BP, followed by a generally consistent high level of use for the last 2000 years. The magnitude of the increase in deposition between 2000 and 3000 cal BP indicates that there was a very significant shift in the way that this resource was used through time. The change cannot be explained by an increase in seed availability during the last 3000 years: the mean MDI increases during a period with drier conditions. Given what is known about the determinants of seed production in Macrozamia species, drier conditions reduce overall seed production. As indicated in Table 2, and Figs. 3e6, correlations between mean MDI and palaeoclimatic data suggest that Macrozamia exploitation was strongly and directly influenced by fluctuations in ENSO activity. It is argued that given the ecological attributes of this

0.5

450

0.4 0.35

350

0.3 300

0.25 0.2

250

Mean MDI

Cholesterol (µg/g OC)

0.45 400

0.15 0.1

200

501-600 601-700 701-800 801-900 901-1000 1001-1100 1101-1200 1201-1300 1301-1400 1401-1500 1501-1600 1601-1700 1701-1800 1801-1900 1901-2000 2001-2100 2101-2200 2201-2300 2301-2400 2401-2500 2501-2600 2601-2700 2701-2800 2801-2900 2901-3000 3001-3100 3101-3200 3201-3300 3301-3400 3401-3500 3501-3600 3601-3700 3701-3800 3801-3900

0.05 150

Mean cholesterol

0

Mean MDI

Fig. 5. Mean cholesterol levels from the Peru Margin superimposed on mean MDI for eastern Australian archaeological sites, 500e3900 cal BP, cholesterol data from Makou et al. (2010b).

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

0.5

35

0.45

30

0.4

25

0.35

20

0.3 0.25

15

0.2

10

0.15 0.1

5

0.05 1-100 101-200 201-300 301-400 401-500 501-600 601-700 701-800 801-900 901-1000 1001-1100 1101-1200 1201-1300 1301-1400 1401-1500 1501-1600 1601-1700 1701-1800 1801-1900 1901-2000 2001-2100 2101-2200 2201-2300 2301-2400 2401-2500 2501-2600 2601-2700 2701-2800 2801-2900 2901-3000 3001-3100 3101-3200 3201-3300 3301-3400 3401-3500 3501-3600 3601-3700 3701-3800 3801-3900 3901-4000 4001-4100 4101-4200 4201-4300 4301-4400 4401-4500 4501-4600 4601-4700 4701-4800 4801-4900 4901-5000

0

Mean MDI

# El Nino events

477

# El Niño events per 100 years

0

Mean MDI

Fig. 6. Estimated number of moderate-to-strong El Niño events at Laguna Pallcacocha superimposed on mean MDI for eastern Australian archaeological sites, 0e5000 cal BP, El Niño data from Moy et al. (2002b).

reduce yields based on natural spoilage, and seed predation by animals and insects. 3. As a cross-check, ethnographic data on seed processing effort for other species (Blurton Jones et al., 1994; Jones and Meehan, 1989) were used to create a general estimate of the postencounter returns for Macrozamia. Returns are estimated at around 400 kcal per hour for M. communis and 700 kcal per hour for M. moorei (see supplementary text for details). Comparing this with return rates for Australian plant and animal species listed in O’Connell and Hawkes (1981) and Bird et al. (2009), Macrozamia offers returns around the level of grass seeds or yams but lower than most other resources (Supplementary Fig. S12).

resource these correlations are best explained in terms of an intensified utilisation of Macrozamia in direct response to the altered availability or reliability of other resources created by the ENSO climate phenomena. Drawing on the concepts of the prey choice (PCM) model from behavioural ecology (Bettinger, 1991; Bird and O’Connell, 2006; Winterhalder, 1981) it is argued that Macrozamia functioned as a lower-ranked resource in terms of the energy it provided per unit time of handling and processing. While detailed behavioural ecological data does exist for some Australian plant resources (for example, Bird et al., 2009; Gould, 1969; O’Connell and Hawkes, 1981, 1984) no such data is available for Macrozamia. However, three lines of evidence strongly support this general characterisation as a lower-ranked resource: 1. The seeds have a low energy value compared with other resources. M. communis seeds have a relatively modest energy value of 362 kJ per 100 g. This ranks M. communis 23rd out of the 25 species with edible nuts or kernels and 194th out of 304 in the full list of traditional Aboriginal plant foods listed in Miller et al. (1993) (see Supplementary Fig. S11). There is no specific empirical data for M. moorei. Leaching of the seeds to remove toxins would further reduce energy returns. 2. Processing the seeds for consumption requires non-trivial amounts of effort. Ethnohistoric data indicate that seeds were extracted from their hard woody shells, often pounded, and then leached for between two days to a number of weeks (Asmussen, 2010a). Well-aged Cycas seeds can be eaten without processing (Beck, 1985; Beck et al., 1988) and the same may have been true of Macrozamia. However, this would

Two scenarios were considered in explaining the correlations of mean MDI with the palaeoclimatic data (Table 2). First, that the increased deposition of Macrozamia reflected its use as a specialised emergency resource during extended droughts. Second, that the increased deposition was the result of a general intensification of the use of Macrozamia as a lower-ranked resource not specifically tied to major droughts. While cycads have been used as emergency foods in some parts of the world (for example, Africa, Pacific Islands, New Guinea, Ryukus, Guam; Thieret, 1958; Whiting, 1989), the available ethnohistoric information in Australia suggests that despite being toxic, Macrozamia was routinely used as part of the diet in the form of a somewhat unreliable plant resource (Asmussen, 2008). The very weak correlation between mean MDI and the frequency of moderate-to-strong ENSO events (Table 2; Fig. 6) is inconsistent with the exclusive use of Macrozamia as an emergency food during

Table 2 Results of statistical comparisons between mean MDI and palaeoclimatic proxy datasets. Dataset

Data analysed

El Junco Lake, Galapagos Islands (Conroy et al., 2008b) Peru Margin (Makou et al., 2010b)

% silt

Laguna Pallcacocha, southern Ecuadorian Andes (Moy et al., 2002b)

Dinosterol concentrations Cholesterol concentrations # moderate-to-strong El Niñ o events per century

Period covered by dataset (cal BP)

Spearman rank correlation

0e5000

rs ¼ 0.853, p < 0.001

500e3900 500e3900 0e5000

rs ¼ 0.893, p < 0.001 rs ¼ 0.824, p < 0.001 rs ¼ 0.236, p ¼ 0.098

478

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

extended droughts. Its use as a general lower-ranked resource is supported by a modest bias towards the exploitation of Macrozamia during the more arid El Niño conditions: there is a somewhat stronger correlation between mean MDI and with dinosterol concentrations (El Niño) than cholesterol (La Niña) in the data from Makou et al. (2010b), and a modest decline in mean MDI broadly coincident with the wetter period associated with the Medieval Climatic Anomaly at 800e1200 cal BP (Veth et al., 2011; Williams et al., 2010). The general subsistence use of Macrozamia is entirely consistent with the nature and limits of seed production in these plants. Seed production is unpredictable from year to year, suggesting it would be used opportunistically rather than as a staple. With long-term seed production constrained during long dry periods (Asmussen, 2008) it would have limited scope as an emergency food during extended droughts. At the same time, the ability of foragers to monitor seed production up to a year in advance, and utilise seed a year after it fell to the ground, would give foragers considerable strategic flexibility in when and how to use Macrozamia compared with other resources. For example, choosing not to exploit the seeds during a good season when higher-ranked plant and animal resources were in greater abundance would act as a kind of “passive storage” strategy. If we accept that Macrozamia was used as a general lowerranked resource, a number of key inferences can be derived. Behavioural ecological models of prey-choice predict that foragers incorporate lower-ranked resources in response to the reduced availability of higher-ranked resources (Winterhalder, 1981; Winterhalder and Smith, 2000), suggesting such a change may have occurred as the result of the intensified ENSO climate phenomena. The magnitude of the increase in exploitation implies that the changes were almost certainly part of a far wider shift towards other lower-ranked resources in response to permanent or periodic reductions in the availability of higher-ranked plant or animal resources. More broadly, a substantial increase in the exploitation of lower-ranked resources is consistent with the predictions that increased foraging risks were a key outcome of the emergence of the ENSO system, requiring significant reorganisation of subsistence and adoption of different strategies to mitigate the risks created by unpredictable inter-annual climatic variability. It also suggests that whatever other cultural dynamics were at play, by 3000 cal BP the intensification of the ENSO system was driving changes in human subsistence behaviour on a sufficient scale to have considerable impact on other aspects of the wider cultural systems, potentially including patterns of landscape use, territoriality and demographics. The Macrozamia data provides direct archaeological evidence of a number of cultural developments predicted by other researchers. The close correlation identified here between ENSO and Macrozamia utilisation supports claims for ENSO-induced changes in the use of plant processing (Williams et al., 2010). The close correlation may also be seen as evidence for adoption of innovative subsistence strategies as an ENSO response (Haberle and David, 2004: 165), in this case taking advantage of the “passive storage potential” provided by the longevity of the seeds in the environment. The overall trend in Macrozamia exploitation supports available lithic evidence suggesting risk mitigating strategies may have been widespread across Australia in the mid-to late-Holocene (see Clarkson, 2007: 52; Hiscock and O’Connor, 2005: 59). The visibility of Macrozamia in the landscape may also have encouraged changes in landscape activity (Turney and Hobbs, 2006) and site use at certain times and places (Williams et al., 2010). In addition, the results indicate that societal responses to ENSOdriven climate changes were complex and multi-dimensional, with different strategies adopted at different times and for different

reasons. For example, there are different trends in subsistence and technological responses e the south-eastern Australian production of backed artefacts substantially increased 3500e4000 cal BP (Attenbrow et al., 2009: 2769; Hiscock, 2002, 2006), while the ‘proliferation’ of Macrozamia occurs 3000-2000 cal BP. The deposition of backed artefacts decreased in the last 1500 years (Hiscock, 2002; Robertson et al., 2009) but the intensity of Macrozamia exploitation remained high until European colonisation. It suggests that some economic changes were directly driven by the need to cope with the general unpredictability of rainfall, independent of magnitude of the ENSO events themselves. 5. Conclusions It is concluded that changes in Macrozamia deposition in archaeological sites in eastern Australia over the last 5000 years provide direct archaeological evidence of a substantial intensification in the exploitation of a lower-ranked resource in direct response to the challenges created by ENSO-driven climatic variability. This provides an important new line of evidence for understanding the forces influencing societies during the mid-tolate Holocene. The data suggest that that the changes were complex and multi-dimensional, while supporting the overall hypothesis that risks created by ENSO-driven climate changes were a significant factor influencing Aboriginal societies during this period. The magnitude of the changes suggests that by 3000 BP the intensification of the ENSO system was driving changes in human subsistence behaviour on a sufficient scale to have considerable impact on other aspects of the wider cultural systems. Acknowledgements Thanks to Richard Robins (then Queensland Museum) for access to the CQH collections during the PhD research and to Leanne Brass and Val Attenbrow (Australian Museum) for access to NSW collections and laboratory equipment at the Australian Museum. Thanks to the reviewers of this article for their helpful comments. This paper was largely written while an honorary researcher in the Archaeology Program, the School of Social Science, the University of Queensland. Appendix A. Supplementary material Supplementary material related to this article can be found online at http://dx.doi.org/10.1016/j.jas.2012.06.005. References Alexander, M., Blade, I., Newman, M., Lanzante, J., Lau, N., Scott, J., 2002. The atmospheric bridge: the influence of ENSO teleconnections on airesea interaction over the global Oceans. Journal of Climate 15, 2205e2231. Asmussen, B., 2005. Dangerous Harvest Revisited: Taphonomy, Methodology and Intensification in the Central Queensland Highlands, Australia, Unpublished PhD thesis, School of Archaeology, the Australian National University, Canberra. Asmussen, B., 2008. Anything more than a picnic? Ceremonial cycad feasting and mid-Holocene socio-economic change in Australia. Archaeology in Oceania 43 (3), 93e103. Asmussen, B., 2009. Another burning question. Hunteregatherer exploitation of Macrozamia spp. Archaeology in Oceania 43, 142e149. Asmussen, B., 2010a. “There is likewise a nut.” a comparative ethnobotany of Aboriginal processing methods and consumption of Australian Bowenia, Cycas, Lepidozamia and Macrozamia species. Technical Reports of the Australian Museum, Online 23 (10), 147e163. Asmussen, B., 2010b. In a nutshell: the identification and archaeological application of experimentally defined correlates of Macrozamia seed processing. Journal of Archaeological Science 37 (9), 2117e2125. Attenbrow, V., 2004. What’s Changing: Population Size or Land-use Patterns? The Archaeology of Upper Mangrove Creek, Sydney Basin. In: Terra Australis 21. Australian National University, Canberra.

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480 Attenbrow, V., Robertson, G., Hiscock, P., 2009. The changing abundance of backed artefacts in south-eastern Australia: a response to Holocene climate change? Journal of Archaeological Science 36, 2765e2770. Baird, A., 1939. A contribution to the life history of Macrozamia reidlei. Journal of the Royal Society of Western Australia 25, 153e174. Ballardie, R., Whelan, R.J., 1986. Masting, seed dispersal and seed predation in the cycad Macrozamia communis. Oecologia 70, 100e105. Beaton, J., 1977. Dangerous Harvest: Investigations in the Late Prehistoric Occupation of Upland South-east Central Queensland, Unpublished PhD thesis, Research School of Pacific and Asian Studies, The Australian National University, Canberra. Beaton, J., 1982. Fire and water: aspects of Aboriginal management of cycads. Archaeology in Oceania 17 (1), 51e58. Beaton, J., 1991a. Excavations at rainbow cave and wanderer’s cave: two rockshelters in the Carnarvon Range, central Queensland. Queensland Archaeological Research 8, 3e32. Beaton, J., 1991b. Cathedral cave, a rockshelter in Carnarvon Gorge, Queensland. Queensland Archaeological Research 8, 33e84. Beck, W., 1985. Technology, Toxicity and Subsistence: a Study of Australian Aboriginal Plant Processing, Unpublished PhD thesis, Division of Prehistory, School of Humanities, La Trobe, Melbourne. Beck, W., 1989. The taphonomy of plants. In: Beck, W., Clarke, A., Head, L. (Eds.), Plants in Australian Archaeology. Tempus 1, pp. 31e49. Beck, W., Webb, R., 1992. Appendix III. Report on Macrozamia breakage patterns. In: McDonald, J. (Ed.), The Archaeology of the Angophora Reserve Rock Shelter. Environmental Heritage Monograph Series. National Parks and Wildlife Service, Hurstville, pp. 179e183. Beck, W., Fullagar, R., White, N., 1988. Archaeology from ethnography: the Aboriginal use of cycad as an example. In: Meehan, B., Jones, R. (Eds.), Archaeology with Ethnography, an Australian Perspective. Australian National University Department of Prehistory, Research School of Pacific and Asian Studies, Canberra, pp. 137e147. Bettinger, R.L., 1991. Hunteregatherers: Archaeological and Evolutionary Theory. Plenum Press, New York. Bird, D., O’Connell, J., 2006. Journal of Archaeological Research 14, 143e188. Bird, D.W., Bliege Bird, R., Codding, B.F., 2009. In pursuit of mobile prey: Martu hunting strategies and archaeofaunal interpretation. American Antiquity 74 (1), 3e29. Black, M., Mooney, S., Haberle, S., 2007. The fire, human and climate nexus in the Sydney basin. The Holocene 17 (4), 465e478. Blurton Jones, N., Hawkes, K., Draper, P., 1994. Foraging returns of !Kung adults and children: why didn’t !Kung children forage? Journal of Anthropological Research 50, 217e248. Bowdery, D., 2006, Phytolith Analysis Report on Seven Sediments from Kenniff Cave, Queensland, Unpublished Report to Dr. Peter Hiscock. Brenner, E.D., Stevenson, D.W., Twigg, R.W., 2003. Cycads: evolutionary innovations and the role of plat-derived neurotoxins. Trends in Plant Science 8 (9), 446e452. Brough, P., Taylor, M.H., 1940. An investigation of the life cycle of Macrozamia spiralis Miq. Proceedings of the Linnean Society of New South Wales 65, 494e524. Clarkson, C., 2007. Lithics in the Land of the Lightning Brothers. In: Terra Australis 24. Australian National University, Canberra. Colvill, M., 1995. Off the Shelf-out of the Museum. The Retrieval of Plant Material from the Australian Archaeological Record, Unpublished Bachelor of Arts Honours Thesis, School of Archaeology, Australian National University, Canberra. Conroy, J., Overpeck, J., Cole, J., Shanahan, T., Steinitz-Kannan, M., 2008a. Holocene changes in eastern tropical Pacific climate inferred from a Galápagos lake sediment record. Quaternary Science Reviews 27, 1166e1180. Conroy, J., Overpeck, J., Cole, J., Shanahan, T., Steinitz-Kannan, M., 2008b. Excel Data, Holocene Changes in Eastern Tropical Pacific Climate Inferred from a Galápagos Lake Sediment Record. Retrieved June 30, 2011. From: http://www.ncdc.noaa. gov/paleo/metadata/noaa-lake-6109.html. Cosgrove, R., Field, J., Ferrier, A., 2007. The archaeology of the Australia’s tropical rainforest. Palaeogeography, Palaeoclimatology, Palaeoecology 251 (1), 150e173. David, B., Lourandos, H., 1998. Rock art and socio-demography in northeastern Australian prehistory. World Archaeology 30 (2), 193e219. De Luca, P., 1990. A historical perspective on cycads from antiquity to the present. Memoirs of the New York Botanical Garden 57, 1e7. Donders, T., Haberle, S., Hope, G., Wagner, F., Visscher, H., 2007. Pollen evidence for the transition of the eastern Australian climate system from the postglacial to the present day ENSO mode. Quaternary Science Reviews 26, 1621e1637. Gagan, M., Hendy, E., Haberle, S., Hantoro, W., 2004. Post-glacial evolution of the Indo-Pacific warm Pool and El-Nino Oscillation. Quaternary International 118, 127e143. Gould, R.A., 1969. Subsistence behaviour among the Western Desert aborigines of Australia. Oceania 4, 253e274. Grove, R., Chappell, J., 2000. El Niño e History and Crisis. Studies from the AsiaPacific Region. White Horse Press, Cambridge. Haberle, S., 2000. Vegetation response to climate variability: a Paleoecological perspective on the Enso phenomenon. In: Grove, R.H., Chappell, J. (Eds.), El Nino, History and Crisis: Studies from the Asia-Pacific Region. White Horse Press, Cambridge, UK, pp. 66e78.

479

Haberle, S., David, B., 2004. Climates of change: human dimensions of Holocene environmental change in the low latitudes of the PEPII transect. Quaternary International 118e119, 165e179. Hill, K., 1998. Ferns, gymnosperms and allied groups. In: McCarthy, P.M. (Ed.), Flora of Australia, vol. 48. CSIRO Publishing, Melbourne, pp. 514e521. Hiscock, P., 1994. Technological responses to risk in Holocene Australia. Journal of World Prehistory 8 (3), 267e292. Hiscock, P., 2002. Pattern and context in the Holocene proliferation of backed artefacts in Australia. In: Elston, R.G., Kuhn, S.L. (Eds.), Thinking Small: Global Perspectives on Microlithicization. Archaeological Papers of the American Anthropological Association (AP3A) Number 12, pp. 163e177. Hiscock, P., 2006. Blunt and to the point: changing technological strategies in Holocene Australia. In: Lilley, I. (Ed.), Archaeology in Oceania: Australia and the Pacific Islands. Blackwell, pp. 69e95. Hiscock, P., 2008. Archaeology of Ancient Australia. Routledge, Oxon. Hiscock, P., Attenbrow, V., 2005. Australia’s Eastern Regional Sequence Revisited: Technology and Change at Capertee 3. In: British Archaeological Reports. International Monograph Series 1397. Archaeopress, Oxford. Hiscock, P., O’Connor, S., 2005. Arid paradises or dangerous landscapes. A review of explanations for Palaeolithic assemblage change in arid Australia and Africa. In: Veth, P., Smith, M., Hiscock, P. (Eds.), Desert Peoples: Archaeological Perspectives. Blackwell, pp. 58e77. Johnson, I., 1979. The Getting of Data, Unpublished PhD thesis, Australian National University, Canberra. Jones, D.L., 1994. Cycads of the World e Ancient Plants in Today’s Landscape. Reed New Holland, Sydney. Jones, R., Meehan, B., 1989. Plant foods of the Gidjingali: ethnographic and archaeological perspectives from northern Australia on tuber and seed exploitation. In: Harris, D.R., Hillman, G. (Eds.), Foraging and Farming: the Evolution of Plant Exploitation. Unwin Hyman, London, pp. 120e135. Kershaw, P., 1995. Environmental change in greater Australia. Antiquity 69, 656e675. Lampert, R.J., 1966. An excavation at Durras North, New South Wales. Archaeology and Physical Anthropology in Oceania 1 (2), 83e118. Lampert, R.J., 1971. Burrill Lake and Currarong: Coastal Sites in Southern New South Wales. Department of Prehistory, Research School of Pacific Studies, A.N.U., Canberra. Lampert, R.J., Steele, D., 1993. Archaeological studies at Bomaderry Creek, New South Wales. Records of the Australian Museum (Suppl. 17), 55e75. Makou, M., Eglinton, T., Oppo, D., Hugen, K., 2010a. Postglacial changes in El Niño and La Niña behaviour. Geology 38, 43e46. Makou, M., Eglinton, T., Oppo, D., Hugen, K., 2010b. 1228D Analytical Results Appendix. GSA Data Repository 2010007. Retrieved June 30, 2011. From: ftp:// rock.geosociety.org/pub/reposit/2010/2010007.pdf. McCarthy, F., 1964. The archaeology of the Capertee Valley, N.S.W. Records of the Australian Museum 26, 197e246. McDonald, J., 1992. The Archaeology of the Angophora Reserve Rockshelter. National Parks and Wildlife Service, Hurstville. Miller, J.B., James, K.W., Maggiore, P.M., 1993. Tables of Composition of Australian Aboriginal Foods. Aboriginal Studies Press, Canberra. Moy, C., Seltzer, M., Rodbell, D., Anderson, D., 2002a. Variability of El Niño/southern oscillation activity at millennial timescales during the Holocene epoch. Nature 420, 162e165. Moy, C., Seltzer, M., Rodbell, D., Anderson, D., 2002b. Laguna Pallcacocha Sediment Color Intensity Data. In: IGBP PAGES/World Data Center for Paleoclimatology Data Contribution Series #2002-76. NOAA/NCDC Paleoclimatology Program, Boulder CO, USA. Retrieved June 30, 2011. From: ftp://ftp.ncdc.noaa.gov/pub/ data/paleo/paleolimnology/ecuador/pallcacocha_red_intensity.txt. Mulvaney, D.J., Joyce, E.B., 1965. Archaeological and geomorphological investigations on Mt Moffatt Station, Queensland. Proceedings of the Prehistoric Society 31, 147e212. Mulvaney, D., Kamminga, J., 1999. Prehistory of Australia. Allen and Unwin, Crows Nest. O’Connell, J.F., Hawkes, K., 1981. Alyawara plant use and optimal foraging theory. In: Winterhalder, B., Smith, E. (Eds.), Hunteregatherer Foraging Strategies: Ethnographic and Archaeological Analyses. University of Chicago Press, pp. 99e125. O’Connell, J.F., Hawkes, K., 1984. Food choice and foraging sites among the Alyawara. Journal of Anthropological Research 40, 504e535. Ornduff, R., 1991. Size classes, reproductive behavior, and insect associates of Cycas media (Cycadaceae) in Australia. Botanical Gazette 152, 203e207. Pearson, M., 1981. Seen Through Different Eyes: Changing Land Use and Settlement Patterns in the Upper Macquarie River Region of New South Wales from Prehistoric Times to 1860, Unpublished PhD thesis, Department of Prehistory, The Australian National University, Canberra. Robertson, G., Attenbrow, V., Hiscock, P., 2009. Multiple uses for Australian backed artefacts. Antiquity 83, 296e308. Rowland, M., 1983. Aborigines and environment in Holocene Australia: changing paradigms. Australian Aboriginal Studies 2, 62e77. Rowland, M., 1999. Holocene environmental variability: have its impacts been underestimated in Australian pre-history? The Artefact 22, 11e48. Shulmeister, J., 1999. Australasian evidence for mid-Holocene climate change implies precessional control of Walker circulation in the Pacific. Quaternary International 57/58, 81e91.

480

B. Asmussen, P. McInnes / Journal of Archaeological Science 40 (2013) 471e480

Smith, M., 1982. Late Pleistocene zamia exploitation in southern western Australia. Archaeology in Oceania 26, 117e121. Smith, M., 1996. Revisiting Pleistocene Macrozamia. Australian Archaeology 42, 52e53. Smith, M.A., Ross, J., 2008. What happened at 1500e1000 cal. BP in central Australia? Timing, impact and archaeological signatures. The Holocene 18, 379e388. Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised CALIB radiocarbon calibration program. Radiocarbon 35, 215e230. Tang, W., 1990. Reproduction in the cycad Zamia pumila in a fire-climax habitat; an eight-year study. Bulletin of the Torrey Botanical Club 117, 368e374. Terry, I., Forster, P., Moore, C., Roemer, R., Machin, P., 2008. Demographics, pollination syndrome and conservation status of Macrozamia platyrhachis (Zamiaceae), a geographically restricted Queensland cycad. Australian Journal of Botany 56 (4), 321e332. Thieret, J.W., 1958. Economic botany of the cycads. Economic Botany 12, 3e41. Turney, C., Hobbs, D., 2006. ENSO influence on Holocene Aboriginal populations in Queensland, Australia. Journal of Archaeological Science 33 (12), 1744e1748.

Veth, P., 2005. Cycles of aridity and human mobility: risk minimisation among late Pleistocene foragers of the western Desert, Australia. In: Veth, P., Smith, M., Hiscock, P. (Eds.), Desert Peoples, Archaeological Perspectives. Blackwell, pp. 100e115. Veth, P., Hiscock, P., Williams, A., 2011. Are Tulas and ENSO linked in Australia? Australian Archaeology 72, 7e14. White, J.P., 2011. Backed artefacts: useful socially and operationally. Australian Archaeology 72, 67e69. Whiting, M., 1989. Neurotoxicity of cycads. An annotated bibliography for the years 1829e1989. Lyonia 2 (5), 201e270. Williams, A.N., Ulm, S., Goodwin, I.D., Smith, M., 2010. Hunteregatherer response to late Holocene climatic variability in northern and central Australia. Journal of Quaternary Science 25 (6), 831e838. Winterhalder, B., 1981. Foraging strategies in the boreal forest: an analysis of Cree hunting and gathering. In: Winterhalder, B., Smith, E.A. (Eds.), Hunteregatherer Foraging Strategies: Ethnographic and Archaeological Analyses. University of Chicago Press, Chicago, pp. 66e98. Winterhalder, B., Smith, E.A., 2000. Analyzing adaptive strategies: human behavioral ecology at twenty-five years. Evolutionary Anthropology 9, 51e72.