Madagascar corals reveal Pacific multidecadal modulation of rainfall since 1708

Pacific Ocean sea surface temperatures (SST) influence rainfall variability on multidecadal and interdecadal
timescales in concert with the Pacific Decadal Oscillation
(PDO) and Interdecadal Pacific Oscillation (IPO). Rainfall
variations in locations such as Australia and North America are therefore linked to phase changes in the PDO.
Furthermore, studies have suggested teleconnections exist
between the western Indian Ocean and Pacific Decadal
Variability (PDV), similar to those observed on interannual timescales related to the El Nino Southern Oscillation ˜
(ENSO). However, as instrumental records of rainfall are
too short and sparse to confidently assess multidecadal climatic teleconnections, here we present four coral climate
archives from Madagascar spanning up to the past 300 yr
(1708–2008) to assess such decadal variability. Using spectral luminescence scanning to reconstruct past changes in
river runoff, we identify significant multidecadal and interdecadal frequencies in the coral records, which before 1900
are coherent with Asian-based PDO reconstructions. This
multidecadal relationship with the Asian-based PDO reconstructions points to an unidentified teleconnection mechanism that affects Madagascar rainfall/runoff, most likely triggered by multidecadal changes in North Pacific SST, influencing the Asian Monsoon circulation. In the 20th century we decouple human deforestation effects from rainfallinduced soil erosion by pairing luminescence with coral
geochemistry. Positive PDO phases are associated with increased Indian Ocean temperatures and runoff/rainfall in
eastern Madagascar, while precipitation in southern Africa
and eastern Australia declines. Consequently, the negative
PDO phase that started in 1998 may contribute to reduced
rainfall over eastern Madagascar and increased precipitation
in southern Africa and eastern Australia. We conclude that
multidecadal rainfall variability in Madagascar and the western Indian Ocean needs to be taken into account when considering water resource management under a future warming
climate.


Introduction
Tropical Indian Ocean warming in the 20th century has accelerated since the late 1970's affecting rainfall patterns and intensity across much of the Western Indian Ocean and adjacent landmasses of Eastern and Southern Africa (Richard et al., 2000;Funk et al., 2008).Warming of the South-Central Indian Ocean (0-15 • S, 60-90 • E) is thought to reduce the moisture flux towards sub-Saharan Africa promoting droughts in austral summer and fall (Goddard and Graham, 1999;Richard et al., 2000;Hoerling et al., 2006;Funk et al., 2008).As Eastern and Southern Africa heavily depends Figures

Back Close
Full on regular rainfall for food production and ecosystem sustainability (Fleitmann et al., 2007), the uncertainty in the rainfall response to accelerated warming of the Indian Ocean is a serious socioeconomic issue of global importance (Funk et al., 2008).However, to fully assess this response it is necessary to identify the long-term natural rainfall patterns, yet currently we lack an understanding of the major drivers of natural decadal rainfall variability in the Indian Ocean and what the regional synergy is with global warming (Cane, 2010).
There is some evidence indicating that multidecadal South African rainfall is associated with ENSO-like interdecadal variability, due to the shifting tropical temperature troughs in response to large-scale changes in Indo-Pacific SST and sea level pressure (Reason and Rouault, 2002).Since rainfall patterns are sensitive to sea surface temperature (SST) change, which includes both natural internal variability and anthropogenic forcing, here we investigate natural multidecadal modulation of Indian Ocean rainfall in response to multidecadal SST variability.Massive corals such as Porites sp.generate precisely dated century-long, highly resolved and continuous proxy records of changing land-ocean interactions (Lough, 2007;McCulloch et al., 2003;Fleitmann et al., 2007).Here we present 300 yr of monthly resolved proxy records of soil erosion from four giant Porites sp.colonies growing in two coastal marine catchments of Eastern Madagascar (Fig. 1).
The PDO is a major internal mode of ocean-atmosphere variability (Mantua et al., 1997).Positive PDO phases are characterised by low SST in the Central Midlatitude Pacific and warm anomalies along the northern and eastern margins, and south of 30 • N. The PDO is remotely forced from the Tropics in part (Schneider and Cornuelle, 2005), and responsible for strong multidecadal (50-70 yr) (Minobe, 1997) and interdecadal Pacific oscillations in SST (IPO; 17-28 yr) (Meehl and Hu, 2006).It is considered the leading mode of North Pacific SST defined by instrumental data for the past 120 yr (Mantua et al., 1997) kilometers to the Indian Ocean (Cole et al., 2000;Crueger et al., 2009).The positive PDO phase corresponds to warm Indian Ocean SST anomalies (Deser et al., 2004), thought to exceed anomalies associated with ENSO (Krishnan and Sugi, 2003), particularly in the Southwestern Indian Ocean (Fig. 2a) (Meehl and Hu, 2006).While it is evident that changing rainfall patterns over Australia respond to the PDO (Arblaster et al., 2002), links to rainfall in Southeastern Africa and the Western Indian Ocean have only been suggested (Deser et al., 2004;Zinke et al., 2008).

Coral sampling and analysis
Three corals MAS1, MAS3 and ANDRA were drilled in March 2007 from Antongil Bay, NE Madagascar, dating back to 1904, 1880 and 1914, respectively (Fig. 1) (Grove et al., 2010).The core MASB ( 15• 30,566 S; 49 • 45,437 E) was drilled in October 2008, dating back to 1708.Three of the corals used for this study, MAS1, MAS3 and MASB are influenced by a major river draining into the Bay, named Antainambalana (Fig. 1).Its source lies 1450 m a.s.l. and its watershed covers an estimated 4000 km 2 .As well as being influenced by the Antainambalana, a third coral ANDRA is located 30 km south of MAS1/3/B, and is influenced by a much smaller river called the Ambanizana, which has a watershed of 160 km 2 .The average growth rate of the three short coral cores was approximately 12 mm y −1 .Growth laminae were visualized by X-radiograph-positive prints, and the growth axis of the coral slab was defined as the line normal to these laminae.All cores were sectioned into 7 mm slabs, cleaned with sodium hypochlorite (NaOCl, 10-13 % reactive chloride; Sigma-Aldrich Company) for 24 h to remove residual organics that would quench luminescence, and subsequently scanned under UV-light to measure continuous spectral luminescence ratios (G/B).Luminescence in banded corals is indicative of past humic acid runoff from river discharge (Isdale, 1984;Barnes and Taylor, 2005;Lough et al., Figures Back Close Full 2002; Grove et al., 2010).Indeed, correlations of MAS1 G/B and regional rainfall are statistically significant (Grove et al., 2010).Luminescence images of the MAS3 core revealed dark stains, likely organics, in the older sections of the core which could not be removed by bleaching, therefore as a precaution luminescence data ends in 1930.Laser-Ablation Inductively Coupled Plasma Mass Spectrometry (Laser-Ablation ICP-MS) profiles were taken to analyse the trace element ratios of Sr/Ca, Ba/Ca and Mn/Ca at 40 µm intervals on the coral cores MAS1 and MAS3 at ANU Canberra (Sinclair et al., 1998;Fallon et al., 2002).Profiles cover the entire age of MAS1 (1906MAS1 ( -2006) ) and since 1935 for MAS3 (1935MAS3 ( -2006)).We use Sr/Ca ratios as indicators of SST (Corr ège, 2006;Alibert and McCulloch, 1997), whereas suspended sediment runoff is reconstructed using Ba/Ca ratios (Alibert and McCulloch, 1997;McCulloch et al., 2003;Sinclair and McCulloch, 2004).Ba/Ca in the coral cores analysed here showed a high temporal correlation with spectral luminescence ratios (Grove et al., 2010).Mn/Ca is used as an indicator of ash fallout from slash and burn deforestation (Abram et al., 2003;Lewis et al., 2007).As luminescence and Laser-Ablation data have a sub-weekly resolution, interpolation to a monthly time-series provides a high level of accuracy.

Research area and climate setting
Coral cores were taken from Antongil Bay in NE Madagascar, which is surrounded by one of the country's largest remaining rainforests (Birkinshaw and Randrianjanahary, 2007).Air temperature and rainfall in Antongil Bay was monitored for the period 1992 to 1996 (Kremen, 2003).Antongil Bay is characterised by an August-December cold-dry season and a January-July warm-wet season.Air temperatures peak in December and January and are lowest between July and September.Highest rainfall occurs between January and April, while lowest rainfall occurs between September and November (Kremen, 2003;Jury et al., 1995).The annual average precipitation at Andranobe (coral site ANDRA) was 6049 mm (1 SD = 979 mm) between 1992 and 1996.Highest river discharge occurs between February and April, one to two months after Figures

Back Close
Full peak rainfall (Gerten et al., 2008).Runoff decreases but continues until September then reaches lows in October and November.

Results and discussion
We measured soil-derived humic acids in the coral skeletons by spectral luminescence scanning (Green/Blue ratio; G/B) to determine seasonally resolved runoff resulting from hinterland rainfall (Grove et al., 2010).Our longest coral G/B record dating from 1708 to 2008 (MASB; Fig. 1) was compared to the NE Asia tree ring based PDO reconstruction (D' Arrigo and Wilson, 2006) to investigate multidecadal variability in rainfall (Fig. 2a, grey box), since the instrumental PDO index (Mantua et al., 1997) only dates back to 1880.Both climate records show near identical changes in amplitude and timing for over two centuries (Fig. 1c) then diverge after the 1920's.Cross spectral (Fig. 3) and wavelet coherence analysis (Fig. 4) of the PDO tree ring index and MASB G/B confirm the clear relationship between rainfall and the PDO on multidecadal time scales since at least 1708 and until the 1920's (Appendix A).Coherent temporal changes in signal amplitudes and timing between both records show that positive phases of the PDO correspond to positive rainfall anomalies (Fig. 2c).
To further investigate post 1920 PDO modulation of Eastern Madagascar soil runoff, we also analysed the G/B records of additional corals in combination with high resolution geochemistry.We used Laser-Ablation ICP-MS to determine coral Ba/Ca as a proxy of past sediment runoff (McCulloch et al., 2003;Fleitmann et al., 2007), Sr/Ca as a robust proxy for SST (Abram et al., 2003;Zinke et al., 2008) and Mn/Ca as an indicator for ash fallout from slash and burn deforestation (Abram et al., 2003;Lewis et al., 2007).Together, they allowed us to decouple the three major components influencing Eastern Madagascar soil runoff; i.e. human land-use changes and natural decadal climate variability interacting with Indian Ocean warming.Long-term changes in runoff appear in the 10-yr running mean of both G/B and Ba/Ca in each coral (MAS1, Fig. 5a,c; MAS3, Fig. 5d,f).Most pronounced is the continuous increase in humic acid Introduction

Conclusions References
Tables Figures

Back Close
Full runoff since the mid-1970s and sediment runoff from the mid-1950s, towards a maximum in recent years in concert with rising South Central Indian Ocean SST (Fig. 5).Also, the longest continuous precipitation record from Madagascar (Antananarivo) is in agreement with our Ba/Ca and SST records, whereby rainfall increased from the mid-1950s until the record ends in 1987 (Fig. 5b,e).Consequently, increasing rainfall (runoff) over the catchment area appears tightly coupled to rising SST (Fig. 5).
The reduced coherence between humic acid runoff and Indian Ocean SST in the mid-20th century suggests that other factors are involved in large-scale erosion.Discrepancies between 1945Discrepancies between -1955Discrepancies between and 1966Discrepancies between -1980 occur in both cores whereby G/B increases while temperature decreases  or remains stable .
These periods are also marked by enhanced coral Mn/Ca above the seasonal background (Figs. 5 and B1; Appendix B), as found in response to ash fallout from wild fires (Abram et al., 2003;Lewis et al., 2007).Indeed, the pronounced increase in Mn/Ca testifies to the well documented intense slash-and-burn deforestation for upland rice cultivation between 1950and 1980(Green and Sussman, 1990;Harper et al., 2007), associated with the economic collapse of Madagascar and the return to subsistence agriculture.Segmentation analysis (Webster, 1973(Webster, , 1979) ) of the coral composite G/B record (MAS1, MAS3 and ANDRA) highlights these mid-20th century human deforestation periods (Figs.C1 and C2; Appendix C; see Supplement).
The coupling between increasing runoff and Central Indian Ocean warming is evident after the prominent climate shift around 1976/77 when both global mean temperatures and rainfall strongly increased (Fig. 5) (Meehl et al., 2009).As Mn is also associated with seasonal soil runoff through erosion (Lewis et al., 2007), we observe similar increasing linear trends in the G/B and Mn/Ca ratios in response to Indian Ocean warming (Fig. B1; Appendix B).As G/B is a direct indicator of soil erosion and not rainfall, variability, now increasing from the mid-1950s in agreement with the SST and Ba/Ca data (Figs. 5 and 6).Spectral analysis of the monthly instrumental PDO index (1880-present) (Mantua et al., 1997) and coral G/B-Mn/Ca show strong power in the multidecadal band (Fig. B2), in agreement with the pre-1920 frequency analysis of MASB G/B and the tree ring based PDO index (Figs.3 and 4).The tight temporal relationship with the PDO index shows that a positive (negative) phase is associated with wet (dry) conditions (Fig. 6).Interestingly, G/B-Mn/Ca correlates with typical positive PDO-like conditions in global SSTs, coupled with a positive correlation with South Central Indian Ocean SST (Fig. 6).Also, in the Sr/Ca temperature proxy record of MAS1, a positive PDO phase is associated with a warm SST anomaly (Fig. 7), pointing to a typical response to Pacific decadal forcing found in Indian Ocean SST (Krishnan and Sugi, 2003;Cole et al., 2000;Deser et al., 2004;Crueger et al., 2009).The temporal alignment of all records (Sr/Ca, Ba/Ca, G/B-Mn/Ca) with the PDO (Fig. 7) therefore argue for Pacific modulation of Madagascar rainfall on multidecadal timescales for at least the past 300 yr.Madagascar is an iconic example of the extreme environmental impacts human deforestation and habitat destruction has on soil runoff and land degradation (Green and Sussman, 1990;Harper et al., 2007).Human activity is also reported for two 200-300 yr erosion records from Kenya that show a simultaneous major shift in base level runoff at 1906 ± 3 yr and 1908 ± 5 yr (Fleitmann et al., 2007).This 1908 shift in soil to 1976) is linked with higher river discharge, and vice versa, for the positive PDO phase (Lough, 2007;McGowan et al., 2009).Correlating precipitation with the principle component time series of the IPO (Meehl and Hu, 2006), and the PDO (Felis et al., 2010) shows a negative response over Eastern Australia and Southern Africa, and a positive response in Eastern Madagascar and Eastern Africa (Fig. D1; Appendix D).Since Indian Ocean SST is sensitive to the PDO (Krishnan and Sugi, 2003) and rainfall is linked to SST (Goddard and Graham, 1999), runoff variability is ultimately controlled by Pacific Ocean multidecadal variability.During the positive PDO phase, higher mean SST is responsible for enhanced atmospheric convection over the Indian Ocean, which in turn drives anomalous subsidence over Southern Africa and Eastern Australia (Lough, 2007;Goddard and Graham, 1999;Richard et al., 2000;Hoerling et al., 2006;Mc-Gowan et al., 2009).
Long term coral data provide the first evidence that Southwest Indian Ocean rainfall is linked to the PDO on multidecadal time scales.Consequently, for the upcoming decades rainfall in Eastern Madagascar should decrease as the PDO is currently in a transition from a positive to a negative phase.Elsewhere, PDO teleconnected regions with weaker rains in recent decades should experience more precipitation, i.e. in Eastern Australia and Southern Africa.However, it remains a major milestone in future research to unravel if and when projected anthropogenic warming of the Indian Ocean (Forster et al., 2007) will dominate rainfall over the inherent multidecadal component.
Our data illustrate this interplay as an acceleration of rainfall and erosion following the prominent 1976/77 climate shift (Meehl et al., 2009), which is related to both anthropogenic and multidecadal forcing.The widespread operation of Pacific multi-decadal modulation recognised here provides new constraints for future rainfall patterns on human timescales that will assist in water management, soil conservation and biodiversity programmes throughout the tropical Indo-Pacific.Introduction

Conclusions References
Tables Figures

Back Close
Full

Coherence and wavelet analysis
We compared the long time series of the PDO and MASB directly without using an agent (CIO SST).Annual mean data of the PDO reconstruction (D' Arrigo and Wilson, 2006) and MASB for the period 1708-1920 were used for coherence analysis.The reconstructed PDO (D' Arrigo and Wilson, 2006) and MASB show significant power at decadal, interdecadal and centennial time scales (Fig. 3a, b).Strong coherence between the PDO and MASB is found at multidecadal (60 yr) and bidecadal (15-30 yr) frequencies, with the PDO slightly leading MASB (Fig. 3c).Further, wavelet coherence analysis (Grinsted et al., 2004) with the same time series supports strong coherence between the PDO and MASB on interannual, bidecadal (20-30 yr) and multidecadal bands (Fig. 4).

Appendix B Coral Mn/Ca
The MAS1 Mn/Ca record is in and out of phase with the MAS1 G/B time-series on seasonal timescales (Fig. B1).This indicates that high Mn concentrations, associated with slash and burn deforestation, are likely flushed into Antongil Bay during both the wet and dry seasons.Both the G/B and Mn/Ca have similar runoff trends as shown by their linear equations (Fig. B1).This indicates that a fraction of Mn is flushed into the bay associated with the soils or sediment, not just ash fallout (Lewis et al., 2007).This fraction is, however, far weaker in concentration than that associated with ash fallout (Abram et al., 2003).By subtracting the normalised Mn/Ca record from the normalised G/B record, we remove the deforestation effect, as well as the long-term runoff trend (Fig. B1), leaving a G/B-Mn/Ca record that shows the natural runoff variability (Figs. 6 Introduction

Conclusions References
Tables Figures

Back Close
Full  and B2).This method is conceptually similar to removing the thermal component of coral skeletal δ 18 O by subtracting Sr/Ca, leaving the salinity component δ 18 O sw (Ren et al., 2002).

Record segmentation analysis
Record segmentation analysis of the coral composite G/B record, which includes MAS1, MAS3 and ANDRA (Fig. 1), identify years within the G/B time series that correspond to phase changes in the PDO, South Central Indian Ocean SST and periods of major deforestation (Figs.C1 and C2; see Supplement).Two major shifts are detected in the PDO time series: in 1944 and 1976.The timing of these major shifts is in agreement with PDO multi-decadal changes as described in previous studies (Minobe, 1997;Mantua et al., 1997).The 1944 shift of the PDO is associated with a shift in our composite G/B record (Fig. C1); however, SST data shows a highly significant transition 2 yr later in 1946 (Fig. C2).This is most likely an artefact created by the sampling bias in observational data for this period (Gedalof et al., 2002).At the second major shift in the PDO in 1976, the South Central Indian Ocean SST shows a prolonged transition from 1976 to 1982, whereas the G/B records a sharper transition in 1982 (Fig. C1).This difference in the timing of the transition is likely a perturbation created by the 1970s deforestation period.The transitions in G/B associated with the years surrounding 1955-1958and 1970 (Figs. C1 and C2) (Figs. C1 and C2) are assigned to the enhanced deforestation marked by the highly pronounced Mn/Ca peaks (Fig. 2; Fig. B1).As the record segmentation method uses 2 × 10 yr windows, the 1970's deforestation period influences the timing of the defined G/B transition (1982) in relation to the 1976 PDO shift.

Conclusions References
Tables Figures

CPD Figures Back Close
Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | , and recognised in extended proxy time series, e.g.tree ring records of rainfall in NE Asia (D'Arrigo and Wilson, 2006).Moreover, mounting evidence indicates that the PDO has teleconnections extending over thousands of Figures Back Close Full Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | we removed the deforestation effect from the record prior to spectral analysis and filtering of the MAS1 time series by subtracting the normalised record of Mn/Ca from the normalised G/B record (Figs. 6 and B1; Appendix B).This also removed the longterm erosion trend, resulting in a G/B-Mn/Ca record that is reflecting the natural rainfall Introduction Discussion Paper | Discussion Paper | Discussion Paper | erosion was attributed predominantly to a change from traditional subsistence agriculture to intensive European land-use practices introduced by the British settlers.Yet, the Kenya coral records also indicate accelerated soil erosion between the late 1940s and early 1950s and in the late 1970s following periods of intense drought which occur simultaneously with shifts in our Madagascar coral records.These multidecadal runoff changes co-occur with the 1905, 1947 and 1976 shifts in the PDO, also suggesting a Pacific modulation of Kenyan soil erosion by rainfall.The PDO/river runoff relationship in Great Barrier Reef corals and East Australia river gauges is opposite to that in Madagascar, as the negative PDO phase (i.e.1947 Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper | Screen / Esc Printer-friendly Version Interactive Discussion Discussion Paper | Discussion Paper | Discussion Paper | Discussion Paper |

Fig. 1 .Fig. 2 .Fig. 2 .Fig. 3 .
Fig. 1.Map of the region where cores MAS1, MAS3, MASB and ANDRA were drilled.Coral locations (stars) and their corresponding rivers and watersheds (grey shaded areas) are marked accordingly in Antongil Bay.The largest river is the Antainambalana, influencing MAS1, MAS3 and MASB; the river influencing ANDRA is the Ambanizana, flowing south westward into the bay.

Fig. 4 .Figure 5 Fig. 5 .
Fig. 4. Cross wavelet coherence analysis with the tree ring based PDO reconstruction (D'Arrigo and Wilson, 2006) and MASB G/B indices, as used in Fig. 3.The 5 % significance level against red noise is shown as a thick contour.Phases are indicated as arrows, where directing right represents an in phase, and a clockwise direction indicates a lead of the PDO.

Fig. 7 .
Fig. 7.A 50-70 yr band pass filter (0.017000 ± 0.002800) applied to (a) the MAS1 Sr/Ca data (dashed) and the Mantua PDO index (solid); and (b) MAS1 Ba/Ca (dashed) and MAS1 G/B-Mn/Ca (solid).The grey bars represent the transition years of different phase changes of the PDO.

Fig. C1 .
Fig. C1.Record segmentation of the PDO vs. 3 core composite G/B (MAS1, MAS3 and AN-DRA).The top panel shows the raw G/B record with a 13 point smoothing superimposed.Middle Panel shows the change points which are above the 95 % significance level.Major change points are indicated by years (green = PDO; blue = G/B).Lower panel shows the raw PDO time series with a 13 point smoothing superimposed.The vertical dashed line marks the start and end point for reliable interpretation of the record segmentation analysis taking into account that the first and last 10 yr cannot be used for interpretation.

Fig. C2 .
Fig. C2.Record segmentation of the south CIO ERSST (Smith et al., 2008) vs. 3 core composite G/B (MAS1, MAS3 and ANDRA).The top panel shows the raw G/B record with a 13 point smoothing superimposed.Middle Panel shows the change points which are above the 95 % significance level.Major change points are indicated by years (red = CIO SST; blue = G/B).Lower panel shows the raw CIO SST time series with a 13 point smoothing superimposed.The vertical dashed line marks the start and end point for reliable interpretation of the record segmentation analysis taking into account that the first and last 10 yr cannot be used for interpretation.

Fig. D1 .
Fig. D1.Spatial correlation of mean annual averages (May to April) of the Pacific Decadal Oscillation (PDO) (Mantua et al., 1997) with global annually averaged (May to April) rainfall data produced by the Climate Research Unit (CRU) at the University of East Anglia (CRU TS3) (Mitchell and Jones, 2005).Colour shading represents confidence of 90 % and greater.Red shading indicates positive correlations and green negative correlations.Note the positive correlation of rainfall with the PDO over Madagascar and negative correlation over Eastern Australia and the Northern Rocky Mountains.Correlations were computed at http://climexp.knmi.nl/.