Nonequilihriurn Regolith Thickness in the Ouachita Mountains

Tonothan D. IJlzilli~,s,Daniel A. Mrrrion,' r(enl2etl.1Lzlclto~r,' rzl2d Icristin R. Adllms

ABSTRACT

Intcrl~iet,~t~onsof iegol~th anel so11 th~ckncss111 the context of landscape evolutlon ;Ire typlc,llly hascd o~ithe liotion that thl~ki7ess1s co~~t~ollecIby the Iliterdctloll of wcatlier~ngrates anct erosion '111d tu11ed to topography On siJcslopcs oi the Otldchita Mount,ilns, Alltansas, how eve^, there 1s ,I high clcgrcc of local spat~~lvarlah~l~ty th,lt 1s ldrgcly u111c1,ttcd to topog~aphy T111i iriel~cates~~onccl~~il~l~r~ti In the sc~lsc that there 1s 110 cv~dcnccof '1 h,il'~ncc between ~atcsof weatl~crlngaid ~et~~ovdl,as 1s ~~o\t~~l~ited111 so111e co~~ce~tu'll17iodels 111 gco~norphologyalld pedology Johnson's sol1 tl~~clzne\srilodcl 1s ,~pplledas an ;~lternat~veto lnteiprct local v'1ii~tionsIn regol~thtl1icl

Introduction Several models of landscape evolution postulate products, the wcatlzering- or tr,lnsport-limited na- the development of '1 steddy state ecl~~il~hri~~i~reg- ture of lanctscdpe evolution, and storage of water, olith thiclziicss, implyiilg '1 iuorc or lcss spc~tially carbon, ilutrlcnts, and other elements. Thiclzilcss uniform regolith cover withi11 reas as with siil~ilar is '11so a functamental property relevant to the use environillental controls 111 some situations, how- and man,lgement of land '111d soil resources. ever, rcgol~th,lnd soil thiclzness exhibit 2 high de- In th~sstudy, ccluilibriuill (a terin that is vari- gree of loc,ll spatial vari~hility,111l~lyi11g ilonequi- ously aild poorly defined in geology) refers to a lihriuill T11c purpose of this study is cxailliile the ste,~dystate in the case of rcgolith thiclzncss, 1111- spatial v~ri~lb~lityIII regolith thiclzness in the plyii~gthat the production of weathered debris is Ouachita Mountains, Arlz,ulsas. In adciition to 'lt- approxiin'~telybCllancec1 hy thinning processes such tempting to ~rilderst~lndrcgolith and I'lndscape evo- AS eros~oll~~liemov'~1 to illdiiltain a r~l~~tivelyCO~I- lut~on,as well as soil vL1riLlhilityat our sttes, we statlt thiclzi~e~~.111 the sl~tialdoin'~in, this wo~11~1 11opc to 'llso address morc gener'll theoretical clues- be manifesteci ;is iniiliill~llyv'1riC~ble regolith thielz- tioils of (noi~)ecluiI~b~i~ii~~regolith. The thiclzncss of ness w~thinany clr~aof negligibly vc~riablegeology soil ai~dregolith is ,I fund,lmeilt,ll property related ~ndtopogr'lphy that has heen subjected to the same to tlzc inass h'11,lnce c~i~clallo~ati~ii of ~e~~tlierii~ghistory of e~~viroi~i~~e~~t~llch~lilgesand dist~lrh~~nccs. Following Kenwiclz (1992),we distiilguish between disecluilihnum ,lnd nonequilihrium. A discquilih- 2004 rium systeill IS OIIC that IS MOVIII~ towarc1 ste,ldy ' Southcln licse,~r~li5t,1t1on, U5I)A rorest Serv~cc,1'0 lhx statc but that 112s not 11'1~1s~ifficiei~t tiille to achieve 1270, 1 lot 5prlnq5, AkCln5,ti 7ILjO2, U 5 A oLlde~ltJN3tlolldl Fole5t U51)A Selvlce Iic,t it-in this case, a regolith cover tll~tis thlclzen~ng \pl lnp, Arlz,~n\,~\71 902, (J ? A or thinning tow'lrd some steady state th~clzncss. jTlrc ~ourii.11ot (;i~~logy,LOO5, voitiiirc 11.3, 1,. ~12i-.34O]LOOi hy Tlic Uiiiver\lty 01 (:lilc:igo All rigl~t\reservcd 007L-17h/L005/l 1.30.3-OOOiSl5.00 Nonecluilibrium systems are inherciltly dyiiaimi- (1997, 1999, 2001), who present methods for testliig cally ~iiist~ihleor dominated by treclueiit cfistur- ecluil~hriumthiclzness. Tliey presciit radioiiucl~de bance and do iiot develop a steady state eclu~ltb- data that lndlcdte a11 expo~~ei~ti~~lcfecl~ne of soil rium. Noiiecluilihriuiii regol~thtli~clzncss would be proctuction with depth (consisteilt with the reriew'il ch~~racterizedby thiclzeiling or thinil~iigtliLit can of exposure concept) and show tli~tsoil thiclzness conrtinue until Iimitect by external f'ictors, with iio varies iiivcrsely wit11 slope curvature. Carter and particuldr tc~idencyto m'iintain constant thick- Ciollzosz (1991)touiic1 that the tliiclziiess of surfici,il ness. In the sp,iti,~l ciomarn, this is in'mifestcd 2s (0,A, met E) liorizons was inr related to slope 111 variable tliiclzness withln areas of s~mil'lrgeology, soils eleveloped on s'it~dstonein Pen~lsylv~~iiiamet topogr,zplly, aiid other environmental coiltrols concluded tli'it the rate of soil formation exceeds The concept of cc~uilihriumregolitli or soil thiclz- the rate of crosloil This 1s tinplicitly based on a ness is expl~citor ii~lplicit111 iilost of the best- 11otio11 of sol1 thiclzness as 'I function of rcites of lznown models of l'iiidscapc evolution. At ledst as foriiiatioi~versus removal, altliough the authors far b'iclz as Davis 11 892) ,inel Gilbert (1909), it has dlso speculate that effects of erosion oil steep slopes been suggestect thdt weathering r,itcs decline with are inaslzed by laterC1l movement of throughflow soil thiclziiess and that hillslope transport flux 1s and by the effects of tree throw. Siiliilar conclusions proportional to slope gradient. Penclz (1924)more were re'1chcd by L~sctslzi~(1999) on the basis of explicitly cfescrihed such ;I situation 111 111s discus- studies of dated surfaces in the Russ~~tnsteppes. sion of the "renewal of exposure" concept, whereby Heliiisath and others (1999)postulate that if the the production of weathered debris is inversely re- rate of bedrock conversion to iiiobile regolith is a lated to the th~clznessof the regolith cover. Where fuilction of local soil thiclzness, then on ~ui~forni conditions allow regolith to accumulate, eventu- bedroclz, a h~llslopeapproaching dynamic ecjuilib- ally a steacty state coilditioil will be reached wliere riuiil should have a un~forinso~l/regolltl~ mantle. debris proctuct~oii,lt the weatl~er~ngfront is hal- Conversely, they iilaiiitaiil that variations in soil 'iiiced by erosional rei-~~ovdlsat the surf'lce (Penclz depth would produce variations 111 so11 production 1924).Several hillslope dild regol~thevolutioi~ mod- and a non- or d~secluillhr~ui~~coi-~clition. The notion els arc baseel on the production of i~l~~terialhy of feedbCiclzsbetween regolith thiclzness and weath- weathering, a decline in wcatlier~ngratcs ,is rcgo- cilng rates produc~ngtor rooting ,lnd burrowing. Uphullding iiiecliaizical pioccsscs iii adclitloiz to weathering 111 is coizsidcrcd to be potcntl'11ly l~iiltedto surface sed- regol~tlzdeepeiiliig 2nd other aspects of rcgolttli iiizerzt c~ccietion(A), org'lnic m,lttcr ~lccui~~ulatloiz cvol~~t~o~i(I'liillips ,~iidM,~iioii 2004, 2005) Thus, (O),'lnd volume cxp,rnsion ( V)due to I~io1ogic~llpro- we '~doptcdA conceptual model ~iicorpor,ltingbi- cesses sucli '1s loots 'iiicl h~trrows Our consider- ologlc,~lprocesses more explicitly, whicli might ac- ntioii of rcmov'lls 'illows for surfdce removals due co~uiitfor noiiccluilibr~~~iithicltncss to trailsport by erosion (waterbeing tlic most lllzcly Soil Thicl~nes~Model. The soil th~cltncssmodel agent 111 tlie stucfy 'lred) a11d 111'1s~ w'lst~iig (E)dnd of Johnsorr (1985, Joliiison ct '11. 2005) conceptual- consu~izpt~oiihy fire, ~pt~~lte,~11dhC~rvesting (C,,,,,). joui 11a1 oi Geology NONE~~UILII3IIIUMREGOLIT13 THICKNESS 129

Subsurface reinovals can be siinmldrly categorized roots and w~th'11ii1nal burrowing. Root conccntra- '1s a result of transport such '1s lcach~ngor prpc t~onsand pore space arc ~nd~cationsof th~s process. croslon (L) or consumptton such as volatil~zatioi~ Surface rernovals by croslon and inass wasttng or ~lptake(C,,,,,). Note that 's~~bsurf~zceinass ren1ov- (E) woulci he topogral?l-irc~zllycontrollecl and asso- '11s w~llnot res~lt111 voluine or tl11c1zness reduc- ci'lted w~thsteeper slopes. Regolith tliiclzness that tlons unless they 'zre ciss~jciatedwith settl~ngor 1s srgn~ficantlyinfluenced by eroslon should be sys- collapse. Th~sIS not neccssdrlly the cdsc because tematlc,zlly related to slope gradic~itsor curv~zture. ~sovol~unctricwcatlzer~ng can occur, where the loss Recent or rap~deroslon indy also he ~ndicatedby of ln'iss 111 sol~tioi~is not accompan~cdby surface features such as erosi011 pCzve~ne~its,rills or g~lll~es, lower~ng(CleLzves et dl. 1970; Cleaves 1993; Ollier and truncated so11 profiles. and Parn 1996). Symhol~cally, Subsurface removals (L)by lcach~ngtyp~cally re- sult In tsovoluinctr~cweathering 111 hum~dsub- tropic'il cllinC~tes(Cleaves 1993; Pa111 ancl Olller 1996). The bulk delislty of the lower rcgolrth (C 'ind Cr horizons) IS typ~cdllyabout 1 35-1 7 g el11 ' in C and 1.7-1.9 g cnl ' 111 Cr horizons (So11Survey Staff 2004). Bulk dens~tyof the u~lderly~ngroclz, Assessing Thicliiless Processes based on 46 speclfic gravlty tests on saniples of the Atoka and 23 of the J~iclzforkSandstone formations The bas~capproach IS to link the terms in ecluatlon (two of the three formatiorls underly~ngthe study (5) to field ohservat~ons.Because the processes sites), IS 2.57 g c1n ' (Klinc 1999). The difference tlieinselves cannot he d~rectlyohscrved, the lzey is ~ndicatcssoine renloval that 1s assumed to he dom- to detcrin~ilewhat type of signatures or funct~onal ~n~zntlyassoclatcd with weathcr~ngand leachmg. relationsh~pswould be ,~ssociatedwith the opera- However, ~f the rock fahr~cIS eviclcnt 111 these layers tion of a glvcn process. (as ~t is 13y definrt~on111 Cr horizons), this suggests Regol~tl~deepening by weather~ngat the wcdtli- that the rernov,ll has not resulted in any collapse, cring front (W)IS ~ndlc,itedhy weather~ngprod~~cts that the wcather~ng1s isovolumetnc, and that this (e.g., iro11 ox~desand othcr seconciary minerals) and process does not leeid to ciecreases In regol~thth~clz- the development of saprolites or Cr (w~dtheredbed- ness. Subsurface or surf'ice consumpt~on(C) of roclz) hor~zonsDecpeu~ng due to hiot~rl-t~~t~on(13) mass hy fire, uptalze, 'ind so 011, a~pl~esto org'inic would be ~nd~c~ltcclby evidence of b~ologic,il ~n- matter. Because the org'inlc matter content of the struslons s~ichas root growth and an~m,zlburrows rn~ilcr'll hor~zoilsof soils in the st~~dyarea IS less Into CI horizons and bedroclz. than 2% in ,111 sd111ples tested 2nd less tha11 0.5'X) If ul-tbu~ld~ngctue to surfczcc ,zccrction (A) asso- 111 many c~ises,we assume that this process 1s not c~~ztedw~th scd~ment depos~t~on is a s~gil~ficantpro- s~gnificant~n the study ared. cess, there sho~ildbe a systelnatlc relCltionsh~phe- tween regol~th th~clzness dnd topography, wlth th~clzcrregol~tlzs 11-1 topogral-th~cpos~t~ons where de- Study Area poslt1o11 is lilzcly or at least possible. Bcc,lusc ,1coli~i1inp~lts 111 the study '1re,i 'Ire I~IIIO~,the suit- Study sltcs are in the Ouachrta Moui~t~z~nswithin 'iblc topog~aphlc settings woulct include dc- the 0~1'1chrtaNational Forest 111 Arlz~znsas(fig. 1) prcsslons, toeslopes, 'znd low-slope and low- '111d hdve been clescrihcd 111 detail elsewhere (Ph~l- clevcitio~~;Irecis 111 general Where such deposition 1117s dnd Marlon 2004, 2005). The Ou~ch~t'isare is r~p~clor recent, add~tlon'zl evidence such as strat- character~zedby 1~~irczlle1,~'1st-west-trending r~ciges ified surf,lcc dcpos~ts,curnulic su~facehor~zons, and Intermontane has~ns,w~th ndgetop elev'ztlons litl~ologic~zldiscont~nu~tles, sh,zip textural con- ranging from 230 to 850 in a.s.1. The hecirock 1s tr'lsts, 'z11d other evidence nldy also he present. Up- colnposccl of extensively f'zulted and foldect Palco- bu~ld~ngciuc to ,tdd~tronsof organlc m,ttter (0)can zo~csed~nlcntary roclzs originating from a v'zricty occur as l~tter1,zycr ,znd 0 horizo~lsth~clzen. Al- of in'lrlue sources (Stone ,uld Bush 1984).The strata though the coi111~1nccithiclzness of l~ttcrand 0 110- ale tyl-t~c~illy~ltern~~t~ng layers of saildstoi~cand rizotis w'zs me,isurecl, this process has no direct 1111- shale (Jorclan et dl. 1991)alot~g with lesser amounts ~dct011 the regolith tlirclzi~essdat~ bcc,r~isc the of clu,lrtzlte, novaculite, di~dchert The regio~iwds ldtter ~'1sn~e~is~~red reldtrve to the top of the SUI- uplifted and cxtenslvely tccton~cally cteformcd ficial ni~l~er'il1,lyer (Ahorizon). Volume cxpanslon from the iliidctle I'ennsylvan~~inthrough the Perm- (V)n1dy oce~iiin conjunction wlth the i~ivasioiiof id11 (Stone dilct R~sh1984). S~ihderldlerosion hds Sample Site Locations Scale for site location maps - RF 1 :55,000

'4'0"W Geologic Formation Pam - Atoka, Middle Part Pal - Atoka, Lower Part e Sample site Pjv - JohnsValley Shale *" Stream Pj - Jackfork Sandstone ...... ------Road (gravel) Ms - Stanley Shale

Figure 1. Study alea show~nglocatloiss (4 the study plots and the undellylng gcology Journ'11 of Ccology NONE()UILIRIIIlJM RE lilzely been ongoing sriice tlie M~ddlePeniisyl- circular, with a 20-in radlus (-0.13 11,i). Most plots vanldli. have southern aspects, the exccpt~oiishe~ng the Sample sltes are withiii three lithologic ~liiits:the hardwood-dom~natcd and closed-caiiopy pine- Stanley Slidle, Jc~clzforlzSandstone, '1nd lower Atolza doiiiiiiated sta~ldsbecause these forest types arc not Forniation. All three unlts arc common 111 tlie foui~don soutlierly aspects. All are on s~dcslopes, Ouachit,i Mount,iins. They are sili1ilar 111 that they with tlic exception of oiic that is on a iiiliior 'ill consist of steeply d~pp~ng,cxtens~vely faulted, ridgetop. lnterniixed beds of fine- to rned~u~i~-gr~~nectsaiid- Three so11pits wcre excavated witli a baclzhoe dt stones c~ii~Ifiiic-grained sliales. The foriliatioiis d~f- each of the I h plots (two plots w~thhigh clegrccs fcr In age and in tlie re1,ltive proportions of each of topographic variability had four pits). Twenty roclz type (Jordanet '11. 1991; McFarland 1998).Ex- "postliolc" p~tswcrc dug by hdiid at each plot. posed shales arc deeply weathered and liiglily ero- These are 10 pdirs rel3rese1itlng ~011s~lncierlyiiig ct~hlc,whereas tlie saiidstones are i~otiee~lblyless coarse, woody debris and tminedidtely ~~djacent altcrecl 'ltict morc dur'~ble.R~dgctops are composect nondebris sltes (carbon 2nd iiutrieiit contents of of the morc resistant s'incfstoiies, cluartzitcs, diid these saiiiples will be comparcct 111 separate stud- iiov;iculttes. Sideslopes arc often undcrla~ii by ies). In this stucfy, tlie si~i~~llpits are tre~tedsi~iiply slicile, witli saiiclstone outcrops common. as paired samples. The pairs werc geiicr,illy within Soils are etescr~bcdiii detail elsewhere (Pliilllps 1 111 of each other hut occasionally slightly farther and Marion 2005).Tlie single inost coininon serlcs away to avoid roclz outcrops or trees. The pairs wcrc ~lldppedat the study sites is the Slicrless (Typ~c deliberately selected to be delit tical in slope, to- Hapludult), which is forincd in sliale-doinlnatcd pographic curvature, and elevat~on(within 5 cm). parent material. The 111ost co~~~~iioiiseries forined Racklloe pits, each a mlniiiiuiii of I m wide and in sai~dstone-dominated sites is Piruln (Typic Ha- 2 in long, were dug to or below hcdroclz. The post- pludult). Climate in the study area is liumid sub- hole pits typically consisted of approx~inatelyclr- tropical. All sanlplc sites are forested. Current for- cular pits about 30 ciii in diameter. Most pits ex- est vegetation consists of oak-hickory (LC., teiicfed to bcctroclz or '1 lithic or paralith~ccoiitact; liarctwood dominatect), shortleaf pine (pine domi- 111 s0111e cases, actditional aLlgeriiig was iiecessdry nated), 2nd oalz-pine jn~txcdpnc-l-rardwoocl) forest to sample the erit~reregolith thiclzness. Tliiclzness types. 15 iiicasured as the distalice frotn the top of the iiiiiier,il surface (A horizon) to tlie top of a R (bed- rock) or Cr (wc,ithercd l-tcdroclz)horizon. Cr 1ior1- Methods zons 111 tlie study area are essciit~allysaprolite. Al- S~lmpleDesign and Data Collectiom. The sanlple tliough Cr is weathered 'lnd softer than liltact ciesigil WCIS j>drtly deter111111ecI by the role of t111s bedrock, it retains the structure, f'xbr~c, and dom- work w~tliin'1 broader study of the silvicultural, iuaiit color,ltlon of the parent roclz. Measurements ecological, and pcdological effects of forest tnan- were 11i~icfe~~siiig 'I foldi~ig r~iler directly on tlie dc- ,~gcmc~itdncl ecosysteim restoration pr'lctices in scribed pit faces or in tlie postliolc pits. The ldrgc connectton w~tliefforts of the USDA Forcst Scrv~ce p~tswerc dcscr~beduslng staiidard USDA mctliods to restore tlie sliortlc,lf p~iic-hluestem comiiiuni- ,ind procedures (So11 Survey Llivssion Staff 1993). tles tlldt were coInilioii 112 the O~dcliit~N~ttondl In the posthole pits, the dcptli diid secluence of 110- Forest at the tlil~cof Europedn settlcmei.~t.This nzoiis wcre recorded, along wtth the texture ,~nct s,liiiple design has hceii ctescril-tcd hcfore (Phillips Munsell color of the A

Table 1. S~~nlnlaryof Kcgolit11 Th~clznessTrends by Study I'lot Plot MCJI~ SD K'III~C MPD Comments

,320Op4 82 4 11 5 58-104 7 8 Sl~ghttcndcncy fol tl-nclte~~egol~th on gentler slopes '111~1 convex~t~es,th11111cr 011 COIlC'lVltles 3826~28 61 4 14 2 18-92 12 6 Mostly s'lndstonc p~rci~trndtcr~dl 4025~40 54 4 8 4 47-7 1 7 7 ?I 1 00p 1 Xi 5 14 5 54-123 18 i Slight tendency for tliiclte~regol~th on COIlCdVltiCS 3000p2 83 9 22 5 37-125 16 5 Sl~ghttendency for th~clter~cgolith 01-1 h~gherelcvat~ons, five sc~ndstoncpedons 'Ire th~nncston slte 34281734 59 4 9 6 47-87 10 2 35141712 53 0 8 0 38- 7 1 86 Sl~ghttendency for th~ckcrrcgol~th on gentler slopcs 35 14pX 64 0 1,3 4 48 98 6 7 Sllght tendency fo~thiclzer regol~thon gentler slopcs, e~ghtsancistonc pedons ge~~erdllythinnei than others 3912p10 60 1 9 0 42-80 7 6 5cvcn sandstone pcdoni thinner tl1,in no st others 3627~~34 74 9 20 7 44- 122. 19 3 Some tendency for thicltci rcgol~thon gcntlc~ilopcs and on convcxlt~cs,thin- 11e5t 011 COIlC'lVltles Ilardwootl 1 55 5 12 4 3 5-73 5 9 Twelve sdndstone pcdons sl~ghtlyth~nner 111 gencr'11 tl1~111c~ght s11'1le Iiardwooci 2 71 ,3 14 0 40->I 18 16 4 Four th~cltestpcclons ~ncludctwo sandstone AC 1 55 9 12 4 37-85 12 0 Fl'lttop i5 9 I0 1 41-77 8 0 Gcner,~lly,thicl

Note All nulncr1c.11 v,~luciIII Lcntllnctcl\ $13 = \i,lnd,lrd tlcv~,~t~on,Ml'13 = me,ln dlttc~encehctwccn pared \.tlnplc\ Jou~n,llof Geology NONC.~)UILIl3RIUMREGOLITH THICKNESS 333

in sai~dstoneparent mdter~alto be thinner thdil those formed in shale. Topographic Relationshipr. If regol~tl?thiclzness 1s significai~tly~nfluenced hy remov'ils froi-n upper slopes and deposition In depresslolls ciilcl lower slopes within plots, ,I sigilificaizt relatioi~sl-~~phe- twcciz tlziclzncss md clevatloi~would be expected. The correlation between regolith tl~iclzi~ess,~nd el- cvatlon for the entire data set shows 110 sigi~ific,lnt rel~tioilship over the roughly 200-in clcv,ition railge (not sl~owil).For each plot, regolith thiclzness w'ls regressed agaiiist elevation relatlve to the cell- 0 I 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 ter of the plot. There werc no stdtisticdlly sigiilfi- Slope c'111t relat~oi~ships. A relatloiishlp betwecii slope gradieilts and thickiless would bc expected if erosioizal or illass Figure 2. Mc~nrcgollth th~clzncsi(avclagc of pallet1 wastiilg removal is an irnportailt coiltrol of thick- s,lu~~plcs)plottcd agaln5t ilopc glactlcnt ness due to the well-lznowi~relationships between inass wdsting, water erosion, 2nd slope gractients. Although soiile individual plots showeci wcalz re- ineaii th~clznessIS 65 c111, with '1 standard deviation lationships hetween slope and regollth tli~clzrmcss of 16.8. (see table I), the highest coefficients of deterini- Table 1 shows regolith tl~iclzi~csstendencies for nation wcrc I?' = 0.47 ailci K' = 0.35. All other val- each plot. Withiii a giveil plot, regolith thiclznesses ues were lcss than 0.20. Average thiclzness for each of illeasurcd pcdons varied hy 30-88 cin wl-~ei~coin- pair of sainples IS plotted against slope gradient in p.1r1ng the tl-riclzcst aiid thinnest saiilples. This rep- figure 2; there IS no stat~sticallys~gnificant rela- resents coilsiderahlc v,irlation, consldcrtng the t~onsh~p.The saint results werc obtained using the sinall ;Ireas of the plots, tlic range relative to typrcal ~iidiv~dualsample values. The saiiie gcilcral results thiclziicsscs, .tilei the fact that some plots actually werc obtained whether using slopes ineasured in h'lvc mii~imumthiclzilcsscs of, or ,~pproachii-~g, the fielei wit11 cl~nometers(presented in fig. 2) or zero. The standarci deviatloils wcrc gelicrally 8-20 slopes calculated from the DEM. cm. Highly localized variab~lrtyin regolith tl~~clz- Slope curvature (rate of change in slope gradlent) ilcss is iiicticatcct hy the differci~cchetweeil the iilay be more closely rcl'lted to tlzc illass halance p,~lrcd s,ulil~plcs, whicll were gciier,tlly a meter or at '1 r)oiilt th,lil to stc~pi~cssor grC~diei~t,but there lcss apxt 111 idelltic'll (except with respect to was also no st~~tistic~~llysig~~ificar~t r~l~~tioiiship be- coarse, woody ctehns) settings. Tile mean differelice tween regolith thiclz~~essand curvature. The rela- betweel1 'ldlacei~t p~ts1s 11.4 cin, witli a standarct dcv~~itioi~of 10.9. In 60 (of 160)cases, the difference is iclii or less, while in 29 cases, the dlffcrci~ceis 20 cm or more. The thicltiiess dlfferctlce between ~dl~lcentp;i~rs railges from 0 to 62 cn1. Tl-te corrc- lation hetwcci~the s,linple pairs is statist~callysig- nlficaiit but surpr~siiiglyweak (li' = 0.,38), give11 tliclr ,~djacency. 111 general, the v,iriability in tl~iclzncssover short d~st~~nc~s'111~1 si~~dll 'ireas IS s~lggest~veof 11oi1- ecl~~ilihr~urn.A few gcilcr,ll trends 'Ire '1pl-t;irent from the comiiiei1ts 111 table 1 Thcrc 1s soinc tci~dcncy in some plots for th~clzcrsoils to be ,rssoclLitectwtth gentler \lopes 'ind vice versa. These iclatioiish~ps 0 1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 arc generally we~kdlld noisy, howevcr, arici iiot 13~~- Curvature sent in every plot. In some plots, tllcrc is ,rlso cv- lctcilce of we~lzrel'ltioilshlps witli slope coi~vexity/ coilcavlty, hut these dre spotty ~11di11coizsiste~tt. A Figure 3. Ilcgol~th tll~clz~~csiplottcd dg~111itslope second trci~d1s '1 general te~iciei~cyfor ioils for111ed cLlrvcltLIre Table 2. Lkpth to Rcclrocl< {cm)for Sample 1'1t I'air5 greater than that of the 82 pits where the ~areilt 1)cpth values material 1s domiizantly saizdstoiie 166 vs. 55 cm)

Slope seg~ncnt Mean SD 12 This 1s coiisistent with tlze general observation of tlziiziicr regolitlzs on sandstone-derived soils 111 the Convcx 69 7 1 9 <3 3 4 plots where both parent lnaterlals werc present Stlaigh t 62 i 14 8 X 2 Conc,tve 66 4 17 1 44 (though note the except1011of the hardwood 2 plot). The difference wo~~lclbe even more pronounced if outcrop sltes ~rcresdrnpled kecause these were prc- doizlinantly sandstone (otliers were quartz or otller tionship for the entire data set (usirlg the average metamorphicsJ. tliiclziiess value of each pit pair) IS show11 111 figure Sailcistone surface fragiiicnts were coininon at ev- 3 Tlziclzness versus curvature regresslo11 a~zalyscs ery slte, eve11 ~f there was no sandstone 111 t11e UII- were also prepared for each ind~vidualplot, with dcrlylilg hcdroclz. This iizdicates ruass wastlng of lineal, cxpo~leiitial,power fnnct~on,and sccoi~d- sandstone from tlzc ndgctops to the sldc4opes. All order polyiioinial fits attempted In 12 cases, there soils saiziplcd 011 shale hact some degree of vertical was 110 statlsticdlly signlficaizt re1atii)nshlp I11 onc textural coiitrast, typically with loain A horizo~is of the other foul cases, the best fit lilie was a izeg- and clay loaizi or silty clay loail1 Rt horizons Tlze ative llnear trend [thlclzncss decreaslizg as curvd- C lzorizons are allnost always the saiizc texture as turc increases) and in anotlici a positive llnear B horizons and were dlstiizgulslzed from the latter treizd. I11 a thlrd case, tlie best fit trend was a pos- mainly on the basis of the prescizcc of recognizable itive exponential, and the fourth was a second-order reinnants of shale, inassive rather than subangular polynomial, with thiclzizess decreasing up to a cur- bloclzy structure, and the absence or scarcity of clay vature of about 0.15 and lncreasliig thereafter. Wltlz fil111s. the exception of the polyno~ulal(IiL = 0.7),all co- Three coiizinon features of the shale-derived soils cfficlents of determination were <0.45. are tlie presence ot clay films 111 the Bt lzorizons, Each pair of plts was also classified 111 the field incflcating translocation; a lack of texture contrast as occurring on convex, straight, or coilcave slope between the Rt and C horizons; and the prcseilcc segments. Mean rcgolith th~clznesscsainong the of sandstoile fragments at the surface aizd often three groups showed no statistically significant clif- tllrouglzout the profile even lf there 1s no sandstone ferciices according to a t-test (tahles 2, 3). 111 the uizdcrlying parent mater~al.The roclz frag- Litlzology. TThc geological framework of the ilients in the solum lack beelding or consiste~zton- study plots is show11 in table 4 (see also fig. 1).All ciztatlon, indicat~ilgthat they arc izot siinply in- the ll~appedforiziatlons (Staliley Shale, Jacliforlz herlted from the parent roclz. Together, these Sa~zdstonc,and lower Atoka formations) coiztain lncticatc a significaiit role for upbuilding because, both shale ai~dsandstolle strata. The percentage of 111 inaizy cases, the ndgetop sandstoiles arc the only tlie soils 011 tlze ~l~tsassoeiatecf wltli ~ilderly111g plausible source of the coarser surficial material shale varles froin 35% to 100'%, with only two plots aiid sanctstone fragments. Weatlzcriizg of these dominated by saiidstoi~crather than shale (3826 clasts, plus '~ssoclatedsaiidy debris, lilzely accounts p28 and hardwood 1) Dlp angles recorded on the for the loailly surface layers in sods otherwise geologic,il map reflect the oftcn steep aild liighly formed from shale. If surface-clown translocatlon of variahlc dip5 iiz tlzc regton, raiigiiig fro111 20° to 65' clay w~~~theredfroim tlie parent lndterial werc the for the study plots aild al~l~roacliiizgvertical In tlie prliiiary source of clay in the Bt horizoiis, tlic ar- stucly vi~ii11tyillore generally Strilzes are gcilerally gillic horizons sl~ouldbe sigiiificantly filler aizcl 80"-85" tlzo~ighlocdlly var~dble111 the vicii~ityof lzave higher clay cotitelit than the C horizons, but the numerous faults in the reglon The role of local thls 1s not the case I~tliolog~c~~laild structural variability clicl not he- come apparelit until late iii the data collection, and the dip or orieiztation of uiidcrlylng strata was not Table 3. Signlfi~anccTest5 for L>cpth to Bcdrocli i~icasurcclin so11 pits ,it 10 of the plots This was Test for slgnif cant iiieasured at the pine aizd hardwood plots, as well ~Iiftcrcnces as the plue-hlucstciii aiid Pote~~icoiltrol plots Ta- Slope scgillcnt i statlitlc df Significa~t: ble 5 sllows that tlze orieiit,itloiz of bedding often Convex versus 5t1alght 0594 114 No vdries ~ubstaiitial1yat tlze plot scale Convex vcrius concave 4353 76 N o Mean regolith tliiclzness for the 238 pits In solls St~~lightversus ~onc~~vc2125 124 N o forlilect 11-1 clolniila~itlyshale parelit material was Note cii = degree5 of tlccclom fourn'11 of Geology NONL(>UILIliRlUM lIEGOLITH 71-11cKNE55 ,id 5

Table 4. Geology of thc Stucly I-'lots Lltllology Plot Formation DI~ postholes ('XI) Pits 3200~4 Ms 100 s21 3826~28 Ms 70 ss, 30 sh 4025~40 Ms 90 s11, 10 ss 3100~1 Ms 100 sh ,3000p2 Ms 75 sh, 25 ss 3428~34 Ms 80 sh, 21) ss 3514p12 Pi 85 sh, 15 ss ,3514~8 Pi 55 sh, 45 ss 39121310 P j 65 sh, 35 ss 3627~34 PI 65 sli, 35 ss Hardwood 1 I' j 65 ss, 35 sh Hardwood 2 Ms 60 sl-1, 40 ss AC 1 Pi 95 sh, 5 ss Fl'tttop Pi 95 sh, 5 ss Pine-blucstein Pal 90 sh, 10 ss Potcau control Pal 55 sh, 45 ss SOLIILC M,ipped forin~tioni'~nd d~p ,rngles from lloles, N~inrodSE, ai~dParon SW <)ucidr~nglcs(C:eoIo,yyy) 1995 Little Rock, Ark Ccol Comm , scale 1 24,000, La~nbcrtconformal coi~i~projection Ui~p~~hlishedCoCeoMap Project field nl,ips by C Stone dnci B Hdey On file at Ark'lnsai Geologi~alCon~mission, 3815 West Roosevelt Rodd, Little Kocli, Arlianias 72204 Note Tlte study ~ncludesm,ippcd formation, dip angles, and the lithology of posthole sClmple\(percent of samples wlth dominantly shale or sC~nctstoneCr 01 R horizons) and full-iizc soil points (l~thologyas observed in p~ts)Intcrhed\ arc l~stcdwlth the domlilant hthology first (e g , sh w/ss = shale wit11 interhcddcd sandstone) Ms = Staltley Shale, PI = fackfork \andstone, Pal = lowcr Atolila, sh = shale; ss = sandstorte; clz = cl~lartz

Analysis of rock fragment distributions iiidicatcs controlled by pedogen~cprocesses in the soluiii or tliat vertical inixiiig of the soil is extensive, with upper regolitl~. tree throw and the downward rnoveilleilt of mass Rioturbation. There is no strong evidence of Into tree stunip holes bemg pC~rtlcularlylillportailt plot-scale effects of vegetation cover on regolith (Phillips and Marlon 2004). Th~scan account for thickness. The iilean th~clznesses for the two the presence of enough sand and s~ltin Bt and C hardwood-doimiiated, four pine-dominated, and 10 horizons to give a clay loain texture. mmxed pine-hardwood plots (63.4, 62.3, and 68.3 Pedogenic Development. The depth to the top of cn~,respectively) dld not dlffer significantly froin B horizon ( = thiclzi~essof A and E lior~zons)is plot- cach other or the overall meail value (65 cm). ted agaiiist regolith thickness in figure 4 A positive The sample plots contained 21 tree throws, de- relationship would suggest a strong role for up- scribed in inore detail by Phillips and Marion building and/or for processes sucll as vertical trans- (2005). Thc thiclziless of the rootwdd call be as- location or faunalturhation that would tend to in- suined to represent the zone contaming the malor- crease A and E horizoii thickne~sesas significant ity of the coarse root illass and thus the depth in cteterininants of regolith thiclzi~ess A negative re- which nlost rooting occurs. Exaiii~nationof the tree lati~nshir)w~~uld indicate erosloli (profile trunca- throw pits indicates that material clown to a Cr or tion) However, there is no ~igiiificaiitrelationsh~p R horizon was typically removed. Thiclzness of the Lisetskii (1999)suggests that the ratlo of total B rootwads ranged from 19 to 100 cm (iileai~= 45). horizoi~thickness to that of total A and E horizon Although this is less than tlie mean regol~ththick- tl~icknessis an index of pedogenic development ness, the tree throw data reflect the fact that When th~sindex is calculated for the study area uprootings are no re common in shallower soil. soils and plottect against regolith thickness (fig. 5), 1)eseriptions of tlie full-s~zesoil pits noted roots the relationsh~p is statistically significant but iii the horizoii just above a Cr or R horizon (e.g., weak withln a C or lower Rt horizon) in every case. These The development of I3 horizoiis and their depth ctata indicate root peiietration to the base of the is likely to be ~lifluencedinost strongly by eluvi- rcgolitll. atioi~/illuviationprocesses and/or by mixing in a Faunal activity and fauiialturbation of the rego- surficial biorliantle The weak relat~onship he- lith 1s coininoil in the study area, but there was no tween regolith thiclziiess diid peetogenic develop- systeinatic evidence of burrowing at the base of the iuent suggests tliat varlatioiis In thiclzness are not regolith. As we have noted elsewhere, 57'% of the Table 5. DI~Angles 111 Ilegrees l'lot Man PI~ Commcnts Hardwood 1 40, 17, 0, 0 Hardwood 2 55, 32, 17 In one p~t,bedd~ng vdr~ed from 10" to 90" on varlous p1t faces I11 other, chps ranged from 0" to 17" AC 1 46, 41, 15 Flattop 48, 30, 22 1'1ne-bluestem 19, 0, 0 One p~t11~1s sd~~dsto~~e fr,igments 111 up- per rcgol~thor~ented at 50" 1'0 teau control 40, 42, 42 011e lxt has s'lnditone fr'1g117ents 111 ~113- per regol~thor~enteil at 70'

Note Ke~ortledon geolog~~~ilITIA~E (M'q>I r~~id111 lower i011 p~t~[I'ltl, me,asureci ni tlev~,itionifrom hoi~zo~~t'il(Ool 'tt ~IXitudy plot', posthole pits hat1 ;I suhs~irfdcestone liiie or zone li17r1~1111)is, of course, an cillergc~it aiid scale- (at least 70'X rock frc~ginciitsby voluine; rock con- contingent property (Reiiwiclz 1992).It is assuimed tent at least 20% greater than adlacent horizons; here that the general weatheriiig '~iidpedogenetic Phillips and Marion 2004). However, no subsurface regime has hecn present througliout the Holocene stone lines or zones were identified in the 58 full- and that the biolog~caleffects we believe arc cr~t~cal size soil pits. Th~ssuggests that the posthole pits here operate on the scale of multiple gencreltiolis were frecluently pelletrating local stone concentra- of forest (e.g., centuries). tions that are not laterally extensive enough to be Deepen~ngby weatheri~igis essentially ubiclui- recognized as a stone line or zone In a soil pit. This tous, as iild~catedhy an abundance of weatheriiig is consistent with point-centered processes sucli as products in the lower regolith and the development roclz deposition in stuinp holes rather than areally of saprolitcs aid Cr horizons. Regolith deepening exteilsive processes such as fauiialt~irl~atioi~,which due to faunalturbation does not appear to he wide- can create stone liiles in some situations (Johnson spread. The general correspondence of rooting 1990; Ralek 2002). depths w~thregolith thick~icssis coiisisteiit with Eveii though sawn stui~ipswcre excluded, there deepeniiig processes ,~ssoclatedwith trees, but it 1s was CI Ineail of iline stuinps aiid 8.9 stand~ngdead difficult to prove cause and effect. Tlic presence of trees per plot (Phillips 'II~C~Mario11 2004). Because oxidation around roots and observation of roots tree throw usually involves living trees, the >18 penetrating fractures and bedding planes below the stuinps and standing dead trees per plot conipared regolith suggests that loc'llly enhanced weathering w~tlithe incan of 1.3 tree tl~rowsi~id~cdtes that 'lround roots pcnctr,it~ngthe parent roclz III'IY be "stancling death" rcs~lltiiigfrom harvestii~g,trunlz ,In important process. brcalz, disease, fire, '~iict so 011, is inore coininon The lack of associatioii between regol~tllthick- than uprooting. iicss and topographic variables suggests that de- Roots and pores wcre observed, aloiig w~thoc- position'il upb~ulding'ind er~sioil'iltruncation are casional evidence of faunal burrows, in tlic so1~1111 not in'llor controls of v,~ridtions111 regolith thick- aild C horizoiis of every full soil pit. This iiidicates ness. This 1s not to s~ythat these processes arc some volume exp~lnsiondue to biological activity. insign~ficant.Wlth the exccpt~onof the Poteau con- Root peiietratioii into saprolites aiici bedrock, typ- trol site, '111 liave hccn logged. In the Ouach~tare- ~callyaccompanied by oxidation around the root gio~i,loggiiig IS often 'issociated wit11 periods of ac- channel, was also observed 111 sotme pits. celerated erosion, but these periods are short because of rdl?id vegetdtioil recovery 111 the sub- tropical cliiiiatc. There is ~lsono liistoric~~lor 'IT- Discussioil and Interpretations ch'ieological ev~dencethat the sideslopes typ~fying Regolith thickness in tlic study area is noncquilib- our study site werc ever cultivated, and there is 1x0 riuili; that is, there arc no iiictic~tionsof the rela- field evidence of recent cultivation (e.g., reillilaiit tively uniform cover that would be expected wrthi~i furrows and agricult~~r'ilartifacts). No active ero- small, relatively I~omogencousareas where pro- sion paveinents or gullies werc noted in the field duction of regolith by wcc~tlicn~~gof hedroclc is ap- except 111 the immetIi,~tev~c~~lity of loggiiig sk~d proxiinatcly balanced hy erosional removals and trails In a few i~istai~ces. the regolith tliickness is tuned to slope grditieiits However, 6011 111i)rphology ~Ioesitidi~~~te tll'lt col- ,1nd curv,ltures. Ecjuilibr~~ii~i(aiid dis- or noncciui- luvi,tl deposition of m,iterial derived from upslope Jo~~rn,ilof C:cology N0NEf)UILIIIIIIUM IIEGOLITH TIHICKNESS 33 7

also lead ln gcizcral to systematic differences In reg- 011th tl~iclzncss,presumably ,~ssociatedwith the nzore rap~dweathering of slic~le,a conclusion sup- ported by the systcinatic differences in thickness of soils overlying shale dnct sandstone. Tree rooting IS con~eiitr~~tedjust ahove the we,zther~ngfroiit, hut roots were observed 111 so11 pits pcnctratlng bedding planes, w~thevidence of prcfercntlal weathering (oxides in root cl~ani~els).It is reasonable to spec- ~llateth'zt SUC~root penetration is facil~tatccthy more vertical heddiilg orientations, 2nd thus the local var1,1bilitics in substrate dip inay contribute o! C-C ; I 0 5 10 15 20 25 30 35 40 45 50 to local v'irlations in regolith tl~iclzncss. Depth to B horizon (cm) As d~scussed111 the "Introduction," complex lo- cal vari,lhility in regol~tlltl-riclzness could ,irise solely hcc'i~lse of tile Inter,lctions of weathering Figure 4. licgol~tlithlcktlcss plotted dgasilst ilepth to the top of the 13 horizon rates and soil thiclzness, where the feedb'~cl0.5 111) were reiliovcct from centered effects of trecs in forest env~ronments.I11 some soil p~ts111 othcrw~scshale-don~in~itcd 111~1- more coi71plex litliologles ,iild under forest cov- terial. Tlils suggests that ii~ass-wastedsaizcistone boulctcrs may function as loc,zl poclzets of sand- stoile pareilt iildterial Tlic loc'tl sp'ztial v,~riabilityin regol~thtll~clziless is appc~reiitlyassoci,itcd pr~ili,irilywith local 11th- ologic~~lvdridhilit y 'znd with tile locdlizcd pcdologic ~nf-luenccof trecs. Th~sclaini arises l~litlyfro111 the lc~clzof ~liiyotl~cr plausible source of the loc,ll var- 1'1tions but 1s supported hy field obscrv~~tionsLocal tl~~clzci~iiigot soils call be associated with root pen- etr'rtlon Into wcatliered rock 'ind with the 111filli11g of holes cicatcct by both tree three and \tump rot. Phillips diid Mdrion (2004, 2005) have l~iilzcdlocdl ...... pedod~versity111 the are'l to these effects, and loc'il v;lri,itioil\ ~n so11 thiclzness ~ssoc~~~tcilw~theffects 0 2 4 6 8 10 12 14 16 of trecs is well established from studies 111 other B thicknesslA thickness '1reas (Lutzand Griswold 1939; Ziillze 1962; Cramp- ton 1982, Schdetzl et al. 1990, V'isenev ,lnd TJI- gullydn 1995; Rdrrett 1997). Figure 5. Regolith th~clincssplotted against .ui uldcx Loc'il~zedveins, lenses, or traiisported boulder\ of pcdogciisc dcvclopmci~t,total h~iszo~~th~~lzi~e\s/to- (most commonly of s,lndstone w~tli~nshcllc) m~y t,ll A '2nd E ho~lzontl~iclznc\s er, noizecl~rilil-tri~r~zircgol~th thiclznesses arc more cause of tree throw is significant, but root pcnetra- 11lzcly. tion and assoclatect weathcr~ngfacil~tat~o~z niay he The results 'zlso indicate that the we'rthering ver- even rnore ~iizportant It is likely that the latter sus erosion framework for exalninlizg regollth ine~h~rnismis important 111 the preferential weath- th~clzizcssmay he too siii~plist~cfor ~ntcrprcting ering of fractures and bedding pl'tnes soils ancl we'tthering profiles 11-1 inany situations. A steacty state regolith clzciractcr~zcdhy coizsis- The soil tlziclzness model, which incorporates bi- tent tliiclzness within relatively l~oilzogencousa- ological influences, a11d the 1>ossibil1ty of l-toth eels ,tnd by close relationslzips between tliiclzness thiclzeniizg and thinn~izgat l-toth the surface and the 'znd topograplzic var~~~blesSLIC~I as \lope gr,tctients weatlzering front, ap~edrsto be preferable as an in- and curvatures rrray he relatively rare Results ot terpretive tool. this study suggest that an ecluilibr~uinregolith tlziclzizcss is ii~ostlilzely 111 uniform litl~ology,with a high degree of lithologrc purity, less lilzely in in- Conclusions terbedded sccl~meiztaryroclzs, and more unlilzcly Regolith thiclzness oil sldcslopes of the Ouachita still if the 1,rttcr are tllted ancl fractured Ecl~ulib- Mountains exhihits high degree of loc~lspatial rluin weathering profiles would 'rlso be more lilzcly variability tli'~t 1s largely unrelated to topography where the effects of l-tloturbation arc nzore 'treally This indic'~tes i~oiiecluilihriui~~In the sense that uniform (as opposed to the point-centered effects there is no ev~dciiccot '1 l-talaizce between rates of of individu~~ltrees) and where the bioinantlc is weathering and reinoval, as is postulated 111 sonle above the weathering front. conceptual inodels in geomorphology and pedology. Coi~ceptualizingsod, weatheriilg profile, or reg- Variability in regol~ththiclzness is not generally olith thickness as a balance between reinoval and related to topography or to pedogeii~cdevelopment proctuctioi~rates is certainly plausible in some sit- in the solum although mass wasting of illaterial uations and is no doubt a useful abstraction over froin rldgetops 1s a significant upbuildi~~gprocess. broad spatial and teinporal scales. However, inter- This iiidicates that v,zriability 111 thrclzness is re- pret'rtions of regolith tliiclzizess 111 the field arc gen- lated chiefly to processes and controls actlng in the erally better served by the so11 thiclzness model, lower regolith, below the soluiii. The prlinary con- whlch assigns a greater role for b~oturbdtionand trols of variabil~tyare threefold. First, local litho- whrch accounts for gaiiis and losses of- inass and logical variation crssoc~atcdwlth layers, lenses, or volume at all depths. transported boulders of sandstone within strongly tllted shale parent inaterial is associated with d~f- ferent~altli~cknesses, with tlz~clzcr soils on the ACKNOWLEDGMENTS more wcatlzcral-tlc, less-resistant shales Second, Tlils project was supported by USDA Forest Service the strongly tilted Paleozoic sed~lne~ltaryroclzs Coopcratlve grant SRS 01 -CA-1 1330 124-516. We coinprise 'I substrate that exliil-tits s~ibstantialvar- thank J. Enierson, J. Swafford, E. Swafford, G. Swaf- iability in resistance due to fractures and bedding ford, J. Grant Barber, R. McGrath, F. Woodral, and plalns. Preferential weathering in the lattcr helps T. Dozler of the Forest Service for their assistance. produce an irreguldr weathering front topography Z. Musselinan, L. Martin, A. Turlziizgton, and T. Third, point-centered pcdological influences of Futamura of the Uillversity of Kentucky ass~sted trees ,rppear to be responsible for some locdhzed in fieldwork, as did G. Malstaff. D. Gill-treath and vari,rtion in regolith thickness. Floralt~rb~itionat the University of Kentucky Cartography Labora- the weatl~eringfront and in the lower regolltli be- tory assisted with graphics.

REFERENCES CITED

Ahncrt, F 1976 Rncf description of a con~pichcnsivc Arlllstroilg, A C 1980 Soils and slopes In a humid tem- three-ctit~~ensio~~~il131odel of landform development Z pcr,rtc cnviront~~enta slmulatlon study C'ltend 7 Gcoinorphol Suppl 35.1-10 127-338 1987 Process-response moctels of dcnudatlon at lialel<, C L 2002 Burled artlfact5 in stable upland sttes ciiffcicnt spCltialscales Catena Suppl 10 31-50 cl~~dthe role of bioturhdtlon a review Gcoarchacology Andelson, K 5 2002 Modcl~i~gthe tor-dotted crests, bcd- 17 41-51 rock cctges, arid par'1bo11c profiles of h1gh '11p11le sur- R,lrrett, L R 1997 Podzol~z~~tionullcicr fore5t and stuillp faces of the Wlild Kivcr rd17gc, Wyoming Gco~nor- pralrlc vegetation 111 northern Michigan Gcocterma phology 46 35-58 78 37-58 Journal of Gcology N0NEC)UILIRKIUM REGOLITH THICKNESS

13irltcland, P. W. 1999. Soils aild geoi11or1711olo~. 3rd cd. , 2002. Darwin woulct be proud: bioturbation, dy- New Yorlz, Oxford Ullivcrsity Press, 430 p. namic denudation, and the power of theory in science. 13rau1-1,J.; Hciinsath, A. M.; and Chappell, J. 2001. Sedi- Gcoarchaeology 17:7-40. ment transport inecha~lisinson soil-mantled hill- Johnson, D. L.; Dolnicr, J.; and Johnson, D. N. 2005. An- slopes. Geology 29:683-686. imating soil thickness processes. C,eol11orpl1ology, Carson, M. A,, and Kirltby, M. J. 1972. Hillslope for111 forthcoming. and prctcess. Camhridgc, Callihridge University Press, Jordan, D. W.; Lowc, D. K.; Slatt, K. M.; and Stone, C. 457 p. G. 1991. Scales of geological heterogeneity of Pcilii- Carter, R. J., and Ciolkosz, E. J. 1991. Slitpe gracticnt and sylvaniail Jackforlt Group, Ouachita Mountains, Ar- aspect effects 011 soils developed from sandstol~cin Izansas: applications to field development a11d explij- Pennsylvania. C;eoderma 49: 199-2 13. ration for cleep-water sandstones. In Amcrican Cleaves, E. T. 1993. Clirnatic iinpact on isovolumetric Associatioli of l'ctrolcum Geologists 1991 Annual weathering of a coarse-grained schist in the northern Co~iventiorr(Dallas, TX, April 4-7, 199 1I. Guidcboolz piedmont pritvil~ceof the central Atlantic states. Get,- for the Dallas Geological Society fielei trip 3. Little morphology 8: 19 1-1 98. Rock, Ark. Gcol. Comin., p. 1-20. Cleaves, E. T.; Godfrcy, A. E.; and Rriclzcr, 0. P. 1970. Kansas Geological Survey. 1994. Surface 111 user's man- Geochemical balailce of a sinall watershccl and its geo- LI;I~.Lawre~~ce, Kan. Gcol. Surv., 84 p. inorphic iinplications. Chi. Soc. Am. Hull. 81:3015- Kline, S. W. 1999. Arltai~sasresources for crushed stone 3032. construction aggregate. In Howard, J. M., cd. Colltri- Crampton, C. R. 1982. Podzolization of soils under in- hutions to the geolohy of Arltansas. Vol. 4. Little Rock, dividual tree ca~lopiesit1 southwestern British Colurn- Ark. C;eol. Comm. Misc. Publ. 18-D:1-65. hia, Canada. Gcoderma 2857-61. Leigh, D. 1998. Evaluating artifact burial by eolian versus Davis, W. M. 1982. The convex profilc of bactlaild divides. hioturbatio~l processes, South Carolina sandhills, Science 20:245. U.S.A. Gcoarchaeology 13209-330. Dictrich, W. D.; Reiss, R.; Hsu, M.-L.; and Montgomery, Lisetslzii, F. N. 1999. Soil catenas in archeological land- U. R. 1995. A process-based 1node1 for colluvial soil scapes. Eurasian Soil Sci. 32: 1084-1093. depth and shallow la~lclslidingusing digital elevation Lutz, H. I., and Griswold, F. S. 1939. The iilfluencc of data. Hydrol. Processes 9:38,3-400. tree roots on soil morphology. AI~.J. Sci. 237:389- D'Odorico, P. 2000. A possible histable evolution of soil 400. thiclincss. J. Geophys. Kes. 10513:25,927-25,935. Marlzewich, H. W.; Pavich, M. J.; Jol~nsoi~,R. C.; and Dolzuchaev, V. V. 1883. K~issianCllcrnozcm. 111 Kaner, Gitnzalez, V. M. 1989. A guide for using soil and N., trans. Selected worlts of V. V. Dokuchaev (1967). weathering profile data in chronosecluence studies of Jerusalem, Iilt. Program Sci. Transl., p. 1-4 19. the coastal plain of the eastern United States. U.S. F~lrbish,D. J., and Faghcrazzi, S. 2001. Stability of creep- Geol. Surv. Bull., p. 1589-D. ing soil and ilnplications for hillslope evolution. Water McFarlaild, J. D. 1998. Stratigraphic summary of Arltan- Kesour. Rcs. ,17:2607-2618. sas. Little Rock, Ark. Geol. C;omm. Inf. Circ. 36. Cabet, E. J.; Kcichmann, 0. J.; and Seahlooin, E. W. 2003. Minasny, R., and McBratney, A. U. 1999. A rudimentary The effects of bioturbation on soil processes and sed- mechanistic model for soil production and landscape iineilt transport. Annu. Rev. Earth Planet. Sci. 31:249- development. Geoderilla 90:3-21. 273. Ollicr, C. D., and Pain, C. F. 1996. Regolith, soils, and Gilbert, G. K. 1909. The convexity of hilltops. J. Geol. latldforms. Chichestcr, Wiley, 316 p. 17344-350. Pain, C. F., and Ollier, C. D. 1996. Regolith stratigraphy: Hci~nsath,A. M.; Chappell, J.; Dietrich, W. E.; Nishi- principles and problems. AGSO J. Aust. Geol. Gco- izurni, K.; and Fiilltcl, R. C. 2000. Soil production 011 a retreating cscarprucnt ill ~o~~tl~~astcri~Australia. Gc- phys. 16: 197-202. ology 28:787-790. Paton, T. R.; Humphries, G. S.; and Mitchell, 1'. 13. 1995. . 2001. Late Quaternary erosion in southeastern Soils: a new global view. New Haven, CT, Yale Uni- Australia: a field cxa~llplcusing cosmogcilic ~~uclidcs. versity l'rcss, 213 p. Quat. Int. 83-85:169-185. Pei~clt,W. 1924. Morphological analysis of landforms. Hcimsath, A. M.; Dietrich, W. E.; Nishizumi, K.; and Czech, H., and Roswcll, K. C., trans. London, Finltel, R. C. 1997. The soil productiorr furrctioi~2nd Macmillan. landscape equilibriun~.Nature ,388:358-361. Phillips, J. D. 1993. Instability and chaos in hillslope evo- . 1999. Cosmogenic nuclides, topography, and the lutioii. Am. J. Sci. 2%:25-48. spatial variation of soil depth. Geoinorpl~ology27: . 1995. Time lags and emergent stability in mor- 151-1 72. phogenic/pedogenic systcln n~odels.Ecol. Model. 78: Johnson, D. L. 1985. Soil thicltness processes. Catena 267-276. Suppl. 6:29-40. l'hillips, J. D., and Marion, D. A. 2004. Pedological mem- --- . 1990. Bio~lla~ltleevolution and the rcdistri- ory in forest soil development. For. Ecol. Manag. 188: bution of earth inaterials and artifacts. Soil Sci. 149: ,363-380. 84-102. . 2005. Biomcchailieal effects, lithological varia- a40 J 11 I'HILLII'S ET AL

tlons, anci loc,ll pcclodivcrs~ty111 some forest so~lsof Ouachita Mo~lntains,Arlzclns,ls Little Rock, Arlz Arlz'l~~~dsGeoder~lld 124 73-89 Geol Comm , l-r 65-75 Renw~ck,W H 1992 Eclunl~briurn,discqu~lshri~im, and T'lylor, A , dild 131urn, J Ll 1095 Re1dt1o1-1hetween so11 noncclu~l~briu~n1'1ndforms 111 the 1,lndscape Gcomor- age ,~nds~licate wcatheting rates clcter~ll~~ledf~om the phology 5 265-276 chcm~calevolution ot a gl'l~lal chronoseil~ic~iccGe- S~h'letzl,K J , Burns, S F, Srilalll, T W, ancl Johl-rson, L3 ology 21 979-982 L 1990 Tree Liproot111g revlew of types dnd pdttcrlls Tw~dale,C K 1991 A motlcl of la11cIscapc evolution in- of soil d~stu~hdnccI'hys Gcogr 1 1 277-291 volv~~lg~nclcasccl 'lnd lncrcasllzg rel~ef,~mplltuclc Z Small, E E , Ailderson, K S , ,lnd Hancit~lz,G S 1999 C,eomorphol 35 85-1 09 Estlmdtes of the rdte of rcgol~tliprod~ict~on usl~ig "'Be 2002 Tlie two-stage ~oi~ceptot ld11dform ,~nd dnc1 "A1 froin '111 ,tlp111ch~llslitpc Geomorpbology 27 I,~nds~apedevelopment ~nvolvsngetching orlgirr, de- 137-1 50 velopment, 'lntl ~mpl~catlonsof an ~dcdEarth 5c1 Kcv Soil Survey L31v1slon Staff 1993 So11 survey manual 57 37-74 Wds11111gto11, DC, USIIA, Agrie~~ltur~lHandbook 18, Vasei~ev,I I , and T~I~~II~'~~III,V 0 1995 A 111oc1el for 437 p the clcvelopmcnt of 5od-pocizoli~soils by w~nclth~ow So11 Survey 5taff 2004 Ndt~on~l\oil survey characteri- Euras~~~nSo11 S~I27 1-16 zdt1011 data LII~LO~T~,NE, Six1 Surv LC1h, Natl So11 Vlles, H A 2001 Scale ~ssucsIn we,~thcr~~igstud~csGco- \urv Center http //sslciata nrcs uich gov/ 111orphology 41 63-72 Stone, C G , and 13ush, W V 1984 Sururnary of the ge- Young, A 1963 Soil movement on slopes N'ttusc 200 ology of the central ancl so~lthcrnOuach~ta Moun- 129-1 30 tans, Arlzansas 111 Stonc, C G , dlld HL11ey,I3 K , eds Zinke, P J 1962 The pattan of 11ldiv1d~1'11fi~rest tree5 A gu~debookto the geology of the ce~ltrdldnci SOU~~C~II 011 so11 propert~cs Ecology 43 130-13,1