Chapter I

Concem about cl~matic change in thc future, both natural and man-made, and la

consequences on ecosystems has been glven utmost impatrncc in m t yesrs. Rsonsmrctiacl

of thc past changes panlcularly dunng the last 10 kyr years, i.c.. Hobxnc during which the esrch

has recovered from la latest glaclatlon and h u m pressures upon g l o w vegetatioa haw

become ~ncreasingly Intense (Moore eta/., 1991). Behavror of thc climate system plays MI

Imponant role on terrestrial ecosystem, particularly on vcgetatlon. 'lhe veptatlon suarssroa

~mpl~e\ d~rectlonal changes In species composttion and in community structure thrwgh t l m .

On the other hand the changes In the vegetauon can also exat a strong influence on the

dynamlcs ot temstr~al ecosystems, soils, atmospheric vace p concentrpticms, and clrmstic

change (Overpeck. 1993).

The vegetation patterns m d processes extend beyond the time span usually coruidmd

w~thin the concerned contemporary studies of gapphase (Shug~, 1984) a patch dyMmia

(Pckett and White 1985) The forest dynamrcs along with the succeslan tncoqxntes dr

vegclallon change5 that are not observable in a short period of tlmc. But long-term vegetltim

changes I e changes In locatton and dom~nancc of tree poplldons and pocesses such rs

comptltlve exclusion In response to the env~mnmtntal changes occur on a time scak of millenL

(Delcoun and Delcoun. 1987).

The pnvdmt disturbance regime and rhe changing climrtic conditions o m hundnd to

thousands of years Influence the rate and direction of the long ten vegastion chma (Jacobson

and Gnmm. 1986) Thm is an tncnascd demmd fa infomution on the history of the earth's

natural systems whlch have influenced In the past nd it is imporunt for understanding the

present day changes on the earth's system. Thus dr lrudy of the urth's history from various

proxy records will provtdc the test bed for physical models m t l y being developed to predict

future environmental changes and the consequent rrsponse of the planet's b ~ o l o g d systems.

Late quaternary cltmatc and vegetation change becomes a very nucia1 nsourcr for

undentmdtng the response and lnfluencc of vegetation on the climatic system , and in

pndtctlng the future cl~mate change (Overpcck.1993). The focus in Quatanrry studies is on

undcrstandtng how populations. communrtia and ecosystems responded to the mvironmtal

changes This pmv~des tnstght Into the rate and d i d o n of btotic changes drst may occur in

the near future I a consequence of cllmauc warming resulting from an- inprt of C 9

and ocher green house gam to the ilrmosphere (Nltional Racarch Cwncil, 1986, 1988, Davis.

1988,1990) In vtew of this, vegewlon history bscana a wry impccunt tool m rcammmng

the changes in the m h ' s hlstory. Thus the pollen in lrlrc sdimcnu rnd prz deposits rhu occur

In strar~fied sequences provide a means for mcing the history of plant cornmulubes by

constdenng whole assemblage of gratns and spaa (Moon et d., 1991). The changes of the

quaternary htsrory have k c n lnferred by uslng different proxy words globally

Quaternary ncord of past changes in a tmqkr ic Cq, p a v e d within gas hub in

polar ce (Dclmas et.al., 1980, Neftel et d., 1982, 1985). indicating the low wncmtratlons of

CQ during glacial penod. Thc polar-lcc d estaMishts the bag term evldenct of prehistoric

global changes In atmospheric C@ that has p.cadcd the cumt ampl~ficatron multing from

defonsrauon and burning of fossil fuels slnce the industrid revdmion (Oammon a d., 1985).

Reg~onal and global patterns of climatic change i n f d from the Quafcmary fossil

mod help identify the pnmary causes of cllmabc change and provide specific analogues f a

fum c l ~ m r conditions. Dunng the m~ddk Hdocm (i.e., Hypslthnmal Inttrval) penod ban

about 9 kyr to 4 kyr years ago there was tncrcase In the global tempemm (Wnght. 1976).

changes In posltlon of ecotones between b~orncs, and shifts In spccles distnbutlons changes In

plpnt and anlrnal cornrnunlty structure, whlch may yeld analogues for future changes to be

antxipared w~thln the blotrc communluu. The late Qmtmay record of sh~fts In the

dlstnbut~on of lndlv~dual taxa because of cl~matlc warmlng is ins@umcntal In evalmng

paaaal conquences of future cl~matic warming to nahue ptserves Wished Md managed

to paerve b~ologlcal dlverslry (Davis, 1988; Peters and Darling, 1985; SchoncwpldCox,

1988). Dunng ~hc late Quaternary penod. the earth's cllinatc system was driven by long-term

and penodlc changcs In solar radlatlon lhat generated global climatic change.

Tropical landscape 1s a mosw of landform8 and deporits, thar an frag~lc and ephrmerPl

when vlewed on longer tlmc scales and they klong predominantly to the last 20 - 40 kyr of

earth's hlstory (Thomas. 1994). The palynologlcal evidence by Van der Hammen (1974)

strengthens the understanding of the Important vegetation changes In northern Scufh America.

The concern of paleocnv~ronmntal lnterpntatlons in the tmpics has bccomc senow m

the recent years and there has been an explosion of tesearch on the Quaternary climate change

In the troplcs The number of s~tes for which long-term chronological records are available IS

small, and fewer of these Ile In the hum~d tmplcs (Thomas, 1994).

The glacial cycles In the eanh history arc h e consequences of orbital vanaoons. The

oxygen-lwtope record# from ocean xd~ment cotes have confirmed the telat~onshlp between

Ice volume changes to oxygen-isotope vanauons (Shackleton and Opdyke, 1973, 1976:

Shackleton. 1987)

There has been a consldcrable dtscuss~on on the v~ab l l l t y of Afncan-Man

monsoon The climvrc vanablllty In the tropics an on the tim scaks of annual crcks to g b a l -

~nterglac~al cycle$ domcnaled by seawnal monsoons (Rell and Kuabach, 1987). The studles

from paleolakes, pollen profiles and deep sea cons have provided lnformatlon that lndlcate the

strength of the monsoon to descnbe large-scale, seasonal changes of wind and precipitation

patterns respons~ble for the climate change s~gnals. The s~mulations from General Circulation

models (GCM) appear to p d l c t weakened monsam dunng the Last Glacial Maximum (LGM)

at around I8 kyr BP, and a smngthened monsoon during the penods of rncrcascd solar radiation,

as ai 9 kyr BP (Thomas. 1994).

Studies by h e analysrs of pollen spectra (Hoogh~em~stm 1989; Hooghlernstra and

Agwu. 1989) and ev~dence from deep core off the northwest African west (Lat~tudc 1350 N)

have confirmed the drler and wetter climates In the wnsltlve Sudanlan zone lylng ktwern the

fmt$ and the open Sahellan scrublands (Ltzlne. 1991) There 1s evidence for an onset of late

Plerstocene cool~ng by 40 k y BP, and hum~d climates probably prevaled widely Pfin 26/25

kyr BP ( G ~ K PI 01.. 1989. Bonnefille et all. ,9900; Van dcr Hammen cr a/., 1992% b; Vlncents,

1991) The w~despread cooling and drylng IS recorded from both mountain and lowland

envtmments (Flenley, 1979. 1985: Strat-Rmtl et al.. 1985; Talbot er al.. 1984; Talbot and

Johannesxn. 1992) lndlcatlng the Impact of full glacial condluons on rroplcal cllmats is most

apparent between 22/20 kyr BP and 14/12 kyr BP. Coolest and dnest cond~tions

cormpondlng to d penod dunng and followrng the Last Glac~al Max~mum (LGM) centered at

I8 k y BP The rap~d change towards posr-glocral condlt~ons may have begun befm 13 kyr BP

and war certa~nly evident by I2 5 kyr BP ar most ncorded s~tcs. There followed everywhere a

5ho11 but slgn~ficant return to cooler and dner codtlons dunng the Younga Dryar Cdd Event.

11-10 kyr BP. ctfur whlch condlttons favonng the ~ N m to runforests In the hum~d m p s w m

establ~shed hy 9000 BP, when ~nter-trop~cal lake kvels were also hlgh om mon. There is some

evldence to suggest that post-glacial humtdity In Afnca, increased first In low latitudes and

moved northwards tnto the Saharan margin over a penod of some 2000 years (Thomas, 1994).

The glac~atton of troptcal mountains and the comspond~ng sh~ft of the vegetation belts

has been a major wrce of rnference about c l l w change in the trop~cs, even tf ~t 1s acccpcd

that tmponant changes to the laps rate dtd not occur at times of cookr cl~matc, fhe presence

of snow and ~ce do create local cl~mates stgntfmdy colder than those of ne~ghboring non-snow

area History of thca glac~al advances and meeats In the tropics IS w m d by Hwemath

(1988) The temperaturea at hlgh elevat~ons In the trop~cs betng about 8 to 12T c w l a dunng

the Last Glacial maxlmum (LGM) are Inferred fmm the ice corc records from Huascaran, Prm

('lhomp\on er ~1.1995 1.

Model> \how that global changer In temperature frequently only d~splay sea surface

temperature (SST) for whtch mtcrofosstl assemblages (forarnln~fera) and daOdafacan k used

(CLIMAP. 1976. 198 1. Man& & Hahn, 1977). On th~s bass ~mperahlm depmsston (SST)

at the g l ~ ~ ~ d l maxlmum (LAM) IS thought to have been 6-IOT In northern regons affected

by u a Ice But In tropical arcas the SST ckpmlon shows a range of 0-2T over most of thc

lower lat~tudea. These conclustons however appear to confl~ct the evldence on land for

temperature depre\s~on In tmp~cal h~ghlands and the supposed causes of andq In troplcal

lowlanda (Rind & Petect. 1985. Crowley and North, 1991)

Reumh &red largely on palynolog~cal stud~es of montane s~tcs tndiiates the deprrss~on

y &W& (1000-lsoom) ;.. L tkolr:(ol ","t"',, v y , Q . of vegcriuion z m s h b and th~s qu i res

temperatun rcductlonb of 5 - 9 T (Flenley.1979 ; Hamilton .1982). Many studies on the

equatorial peat bogs from South Africa have been camcd out successfully on different

paleoclimatic tndtcators such as pollen (Bonnefille er al.. 1990, 1992, Bonnefille and R~ollet,

1988: Bonnefille and Mohammed. 1994). a " ~ , and pollen (H~llure-Marcel, er a[.. 1989,

Mohammed et a l , 1995), and pollen and carbon ratlo (Bonnefille ct al.. 1995). The study of

quatonal sites in Burundr. (Bonnetille and Rtollet 1988) estimated a bmpmhm fall of 4-693.

bur In more recent stattstical study of pollen time-senes data from the same am thts estimate

wa, modtficd to 4 i 2 T (Bonnefille ct a/. , 1990). S~mllarly, a 30 kyr m o d of "C and 180

changes In organtc matter from pat bogs from the same reglon have also been reporled

(Aucour er a / . 1993). In East Afnca. lakes. Malawt and Tangany~ka me thought to have been

lowered by 250-SOOm dunng the U;M (Scholz & Roundahl. 1988: Gzssc ctd.,I989: Baltzcr.

1991. Vtncens. 1991. Vincens n a l . 1993). and then IS much evldence to indicate the htgh lake

levels In the early Holocene

Available cvtdcnce for the andity of southwest Indta from 22 to 18 kyr BP, assoc~aud

w l h the cooling of thc M ~ a n Ocean and weakening of the monsawr whKh graduaIly extended

I& influence after 16 5 kyr BP, culm~natrng tn very humid conditions by around I I kyr BP (Van

Camp. 1986) It also suggests that global cool~ng before and dunng the LPSt Glscial Max~rnum

(LGM) accted all tropical sites so far analyzed In terms of the~r pollen and phytolrth words.

The mrgnttude of that cooling may have only lad stgntficancc. but an incrrastng number of

records indtciue slgnlficant cooling of 41?T (Vlncens cr al., 1993) The ttmlng of the onr t of

cooler conditions In the tropics remuns a maacr of &bate (Bush et al., 1990), and temperaturn

may have been depressed for much of the last glacial cycle. However, they probably dipped

seriously after 40 kyr BP (Bonnefille et d.. 1990) and reached heir maximum values e t t k si&

of the LGM around I8 kyr BP

If Quaternary climatic changes in I d a

Many Quaternary cl~matlc studies have been carned out In Indla also by uslng a wlde

nnge of proxy records In the Indian subcont~nent and the oceans sumndtng the peninsula

the cllmatlc events we largely influenced by the monsoons The change in cl~mat~c signals on

longer t~me scales inferred from the study of past cl~rnatcs has rtveakd that the changes are due

to the ~ntenslty of the monsoon system 'Fhe studles arc ma~nly canicd out on the deep sea corn

md vanow other tem\tnal deposits, uslng dlffercnt proxy lndlcaton such as pollen ncords.

oxygen ~sotopo. carbon so tops, forarnlnlfera dlatoms, etc. (Van Campo, 1986).

The monsoon flucruat~onb on the Indian subcontinent dunng the late Quaternary 1s

monnmrted from the manne sedlrncnt of the southwest coal of Indla by uslng a"0 The

polkn rnd~catcs rwo extreme periods.

A very and pcnod obscrved ktwccn 22 -I8 kyr BP and

* A very humld penod culrnlnatlng ;ll I I kyr BP ( Van Campo. 1986 )

The pollen mord and the oxygen ~socopc from the deep sea c m from nonhern

A h a n u a has rev& that summer monsoon was stronger dunng glacial pcnods ( Van camp

et.al., 1982). Thc study by Duplessy (1982) using foraminifera have indicated salinity pattern

In the northern Ind~an ocean and that southwest monsoon dunng Holocene was weaker than

today and northeast monsoon was stronger during LGM. This was also shown by (Sarkar er

a/., 1990) from sediment core of Arab~an sea. The studles on manne cores of Arabian Sea and

Bay of &ngd have evidenced the incrcasc in salinity dunng LGM d d#rea~ dunng I2 - 15 kyr BP, ~nterpreted based on the fauna that are sensitive to saline d t ~ o n s (Cullen.1981).

Many other studies were camed out on the land deposits that have been used for the

nconstructlon of vegetation and climat~c history In the tropics. The alluvial formations from

Son valley in north central lnd~a has reported wet phase a n d 10 kyr BP ( Williams and

Clarke. 1984)

Palynolog~cal Investlgatlons have ken camed out in different regons in M a Pollen

studies of three salt lakes from Rajasthan. north western India (S~ngh cr ul., 1972; 1974) have

~mhcated fresh water condltlons dunng Holocene penod ( I0 kyr BP). In another study from lake

Lunkaransar (Bryson and Swam. 1981. Swam era/. , 1983). anfall history was nconslnrcted

from pollen profile. wh~ch lndlcated the higher lake levels dunng Holocene (10 8 -10 kyr BP)

lnfmng h~&r summer monsoon. also supponed by a study In south west lndra(Van Campo.

1 986 1

Hyperand cllmate wa\ ~nfemd from pollen profiles of Rajasthan lakes around 20 to 13

kyr BP by the presence of Chenopod~aceae/Amaranth~ccoc. grasses. Anemrrra Md Ephcdra

tndlcatlve of steppe vegetation (S~ngh e t a / . 1990). Pollen profiles from manne sediments of

Arablan sea have shown an increase in tax& which indtcates enhanced monacan during 13 to

12.5 kyr BP and 10 to 9.5 kyr BP. These evidences have been conobwated by other proxy data

(Overpeck et ul, 1996) In a study from eastern Himalaya, In Mirik lake in Dajeeling, the

pollen profile was found to ~ndicatc open gramlands anxlnd 20 kyr BP wtrh pndominant grasses

and sedges which was subsequently replaced by mixed brod leaved forests towards mtd-

Holocene indlcatlng change from cold and dry climate to warm and motst ( S h m and Chauhan,

1 994 )

A manne core study uslng a'" and manne pollen In the Inner shelf off Kanvar of

Western Ghats, western lndla has lndlcated a change around 3.5 kyr BP by showing decnase

In forest and mangrove pollen and comspondlng Increase in savanna pollen and higher a " ~

values suggesting leqs hum~d climate (Carattn~ er al.. 1994).

1.4 Veget~tion d its change from tbe montane r e g h d Nigiris

The montane ecosystems are very fragtle ecosystems Around the globc hey usually

have d~stlnct btolog~cal communities and show htgh levels of endemlsm due to their

b~ogeographtc history featuring Ojtltud~nal m1grattons of vegetPtlon zones In mpnsc to climattc

change Conurvat~on of these montane ecosystems should be glven pnority as they have

dirtlnct biolog~cal communlttes and high level of endemlsm (Gentry. 1993) Montane

ecosy\tems illso play an Imponant role In matntarnlng the hydrology of low land ecosystems

The plant dlverslty of the near-pnstlne montane forests IS lower than other comparable sites In

the ocotmplcs. The osc~llatlng climate and vegetatlon has lnflucnccd rhe (itwture and

cornpalon of the montane ecosystem (Sukumar, 1995). Anthmpogenlc effects on the natural

vegetatlon, such a\, convenlon of grasslands to monocultures may , however. suppress the

natural success~on that would mum render many of the endemic s p c m homeless and mmsc

the nsk of the11 extinction. Thus the adverse effects on natural ecosystems In the montane

reglons have to be restored through lmplementlng actlve and effecuve conservation m w u n s

In these area\

Ihe natural vegetat~on of the hlghaelevaoon (IKHLWX) m as.1.) of the Nilpns plateau

compnsa of a mlxture of degraded pockets of montane evergreen forcsts confined to folds and

depnssions ~n the mounmn ranges locally termed as 'Sholas' and the grasslands. also refemd

to as 'savannas', found along the vast stretches In the open anas dong the slopes of thcx

mountmi ranges (Blacco, 197 1. Meher-Homj~, 1984). Thex evergreen forests occur on the

Ntlg~ns. Anamrld~. Palan1 and other hill ranges of southern Mla Sholas an g e n d l y found In

the valkp or folds of the= mountain ranger where there 1s abundant so11 motsntrr. The shola

tncs ulc stunled shon boled trees are c2.5 m.

The gra5hlands expenencc ground frosts dunng some nlghts between the months of

Novnnber and February, when the temperature ranges from -I to -9°C (Legns and Blasco,

1%9). Occumnce of frost and fire checks the growth of shola seedlings ~n the grassland areas

and hem they arc confined only to the valleys. The sholas art prevented by fire due to the

abundant W I I molbture that help$ In malntalncng the temperature above O°C. Thus the two

vegetation types cbexist independently as climax formations whose distribution is &mrmned

by the topography and microclimatic conditions of the repon.

Review of paleoclimatic mswuctions from pnlynologcal approach, starting from

about 40 kyr BP to reant Uma from Nilgiri hills (Gum, 1990) and from 20 k y to m n t t ~ e s

from Palani hrlls (Bera at d.. 19%) indrcates that shola forest was donunant around 15 to 7 kyr

BP The palynological investigations (Vasanthy, 1988. Sutra. 1997) indicate that the grasslands

were dominant dunng glacial penod The grassland and shola have ken existlng far thousands

of yean and they are due to effects of climauc change and not due to anthropogenic effects.

lnvestigatrons of h e vegetation and clrmstic changes spanning late Quaternary pnod have kcn

camed out using stable carbon isotopes (Sukumar ct.al.. 1993; RajappdaII ct d. 1997) and

also by using a'" (Rajagopalan, 19%) In the higher altiada of Nilgins. Vegemon changes

rn C, and C, plant types arc relakd to soil moisture md hence thc rynfall. Higher values of a " ~ ~ndrcating hrgher proponron of CI plant types suggesting humid conditions dunng Holocene

pcnod In th~\ rcglon which are In confimrat~on wrth studres from Antarcuc ice cores (Barnola

rr 01, 1987. Roblnm. 1994) Thc a " ~ s ~ g d IS interpreted to assess thc vegetation types as C,

and C,, and clirnatlc condit~ons as md and moist on a relative scale (Sukumar at 01.. 1995)

Paleodata provide the principal means of extending the limited insuumental and

hiqtancal record ot past changer In tcmstnal rncluding freshwater ecosystems. Fossil pollen

grains preserved In lake sediments and peat . together wtth pedological and sedimcntdogical

evldcnce prov~dc means of evaluating the ~ssoclaled change in tnrtstnal carbon storage at the

last glacial maximum (LGM) (Street-Perrott, 1994). Carbon-lsotopc sadies of terrcstnal plant

maknal preserved In lakes and swamps provide help in detectlng past sh~fts in the proportloa

of CJ and C plants on land. These C1 and G plants arc duungushcd on the well known

dlffcrence In "c/"c ratlos separated baxd on thclr pathways of carbon fixat~on, whlch

typ~cally have 3°C values In the range of -26% and -28% and -1 1% and -13%

resptlvely (Sm~th and Epteln, 197 1 ) These values reflect the disunct ecolog~cal preferences

of C3 plants (most d~cotyltdons and temperate grasses) prefemng high pnc~p~tatlon and so11

molsture and Cd plants (trop~cal gmses) prefemng low moisture. Thts signal suggests the In

wilrh of C plant matenal clnce tcmstnal plant ttuues are nch in structural carbon compounds

such ac cellulose and Itgnin. The changes observed in the d " ~ values could be compand by

the pollen Invotlgatlon whlch exh~blts marked changes in the proportion of different

palynomorph types

The earlier htud~es tnvolved the carbon-rsotope analyses ("ul?C) on peat spanmg thc

I ~ I 20 kyr BP tSukumar rr al. 1993) and 40 kyr BP (Rajag~palan et d., 1997) of thc hlgher

altltudcs in the Nllglrl h~lls. Western Ghats, southern Inda. As aconununt~m of tho study. an

attempt of complementary approach by palynological Invcstlgatm has been camed out on the

~ m c samples from the above stud~cs to reconstruct the vcgclatlon and clim~tlc changes in the

montane regions (>I800 m rs.1.) of the Nllg~n hills of southcm lnd~a from peat dcpos~ts of the

region

1.6 Aim and Objective d the praent study

The aim of the study to begin w~th concentrated on the quantltatlve study of the

vegetation composition and ecological analyses of these contemporary plant comrnurutles to help

in charactenzing vegetation response to short-term env~ronmental ranges and human impacts

(Dekwfl and Delcout. 1987). Th~s IS very imponant in establishing the prsent-day vegetation

relarlonsh~p w~th the dispets~on and distribution of modern pollen in the surface sampla. Plant

inventory data on all three vegetation types v~z.. sholas, grasslands and peat bogs were made

In the pollen lnvcrr~gated sltes

Secondly. the study lnvolves pollen lnvestlpion of modern surface sampla collected

from urface peat bogs This approach tests the relat~on between modern forest compos~tion

and surface samples conmnlng arborral pollen assemblage showing dlspersd and produaiv~ty

of polka grams. that vary among tm ma and m h ~ c region w~th d~ffcrent physiognomy

and comporition of vegetation (Delcourt and Delcoun, 1984)

Th~rdly. by uslng these appoaches the marsuuctlon of thc past vegetation is camed out

uvng foss~l pollen awmblagcs by applying Ihe relat~ve abundam of each m a In sequence of

foss~l wmblagcs Numerical approach uslng Cluster Analys~s and Dcacnded Cwrcspondcnct

Analys~r (DCA) war also attempted

h t l y , the cornpanson between fossil polkn assemblage and ai3c records 1s ottrmped

k t 0 see the similanttes In the signals obtarncd from the two proxy records to infer the

vegetation and cl~matic changes that occurred dunng the late Quaternary penod.

Overall. the man objectives of the present study an four-fold.

I To create an Inventory of thc present day dlstnbution, species d ivm~ty and compos~tlon

of the three vegetatlon types 1.e.. sholas, gratslands and peat bogs, from three study

locatlons

2 To \tudy the surface pollen assemblage of the modern surface peat samples and their

a~rn~lanties w~th the present day vegetatlon

3 To reconstruct the past vegetatlon and cl~rnav by analysing pollen that arc preserved In

the pear bog, for several thousands of years

4 To interpret h e pollen panems with thc comspond~ng values of J"c that at o h n c d

from the -me pea! core samples from montane regions.

The present study helps In cornpanng the carbon-IS- data whose s~gnals can be

clearly lnlerpreted b a d on h e type of pollen represented in the pollen p f i k s and there by

glvlng much deta~led evidence on the vegetatlon and cl~matlc shlfrs In the part In

montane region\ of the Nilgiris, southern Ind~a.