EFFECTS OF SHIFTING CULTIVATION
ON NATURAL SOIL CONSTITUENTS
IN CENTRAL AMERICA
A.DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
Sincere appreciation is extended to Dr. L. C. Hammcnd, Chair-
man 'f the writer's Supervisory Committee, for his patient -uidance
and supervision; to Dr. F. B. Smith, Head of the Soils Department,
without whose interest field work outside te tate of Flo.rida would
have: been difficult; to members of his Supervisory Comnittee for their
part in planning and E-upervisinc the research and course work; to Dr.
A. E. Brandt for help in statistical interpretations lf the data; and
to Tr. H. C. Conklin, of Columbia diversityity for clarifying the
anthr'p.-'1 .gical picture.
Th-e writer is also indebted to many residents of Guaterala
who provided facilities and encouragement, especially to MessrE. Ch!iris-
topher and Alan HempEtead, whose farms were headquarters for most of
the field work, and who provided ver:" active assistance during the
entire investigation; to Mr. Den-i -Yoester who was generous with the
facilities at Fir.ca 'eanma,; and to Mr. Alan Probert who Frrvide- an
early stLiulus to the writer and reintroduced him to the Folcchic
Gratitude is also expressed to the personnel of the ..c: --
feller Fm'unm.t ion who arranged financial assistance which iade the
field part -f this investirc'.7 ., c:s-i.'.
Through this study the writer first met many worthy friend I.
T,J.LE OF I._j:IT'JT3
LI T OF TA L S . . .
LI "? F FI . . .
I DUCTI :: . . . .
General Descri-tion of hittingg cultivationn .
Lnvlr"nmental Factors in thiftin Cultivation .
Vppetation . .
Soil fertility .
Fitrogen an1 orpar.i. matter
>;tion exchange ra-arity and
..-il ability .
7'-tsssiur. and scn :i .
aliium and ma-ne-i'. ..
"ananege and zinc
"an1anee inc .. .
.*io!phorus . .
il structure .
_r-sosi n . .
base Patu ation
The Fe- le .
hiftir.f cultivation n .
-escrirticn nf itemss . .
Methods . .
ACN:) ALEDC aK:.C ..... ............ ..
T -ME OL CI' VALLLY DF G J.T'.L. . .
' m6;'i iru pr-'ce !.rc .
"er-i,.al an,' rh -1 1 an l:,.-es
Siolr-ical analFt .
TABLE OF CONTENS--Continued
RL'CULLT AND :ICU I . .
Chemical and Mineralogical Sil Factors .
Clay minerals . . .
Nitrogen and organic matter . .
Cation excrlange rapacity,
and base saturation .
Soil aci :ity .
Potassium and sodium .
Manganese and zinc
total exchar 'l 1. bases
.. .. .. .. .
... ... .. ... ..
..sical Soil Fact rs . . .
Eulk density . . .
Soil Biological Factors . . .
"itrification . . .
Microcrgar.i rrs . . .
a:e:atodes . . .
Soil Fertility in the Polochi' Valley . .
L-I.C'1. Y AND JCLJI .. I-,II. . . ..
LIT. ..U CIT . . .
EI:Gi I -.L SKTCH. ...
.LT CE TJ'L
1. Monthly precipitation and aximur:-miinjnim ter.peratLres
for various stati ns in 'oloc:.ic Valley, ;'Ita VJrapar,
,Guateniala . .. .. 4d
Features of Poloc!.ic Valljy sites i:.v--tigat.. in siii't-
LnE cultivation studies . 7
T. Clay minerals of Fol.chic Valley soils irterrr,?t-' from
X-ray Jiffraction patterns . 80
4. :'itrcgen and organic matter contents of Pol-chic Villey
s,-il under three stages of sL.i:'ting cultivation 82
5. relationshipp between climate and .r,.anir matter contents
of the 5-20 cm. horizon of shif.ini cultivAtion fields
in the Polochic '.'al.y, Guatencla . 85
'atir. exchange capacity and base saturation of ?-l-chic
Valley scils under three stages of shifting cultivation 89
7. Value. of pH for Pclocr.Lc Valley scils under three stapes
of shifting cultivation . .. 92
8. -ffect of shifting cultivation on exrhanpeable aluminum
in -clcchic Valley soils. .. . 99
9. ;.xc!az.z:eable calcium and per cent calriiru saturation
: the soil exchange complex under three Ftages
of shifting cultivation in the Poloctir '.'lley I'.
L. effectss of sl.ifting cultivation on ,xchang.-'able
.i-ugrneiun in ?oli-..ic Vall?yv soils . .. 113
I. xc.angeable potassium and sodium or soils uder
thre tage. of shifting cul'iv ti-n in the
:'cl ic -,- . . 11
i1. .xtractable mangan.e an. zinc of soils un dr three
stages of sr.ifrinr cultivation in the rl-ocic
Valle:. . . 117
LI3T OF T LigS--Continued
15. Lxtractable phosphorus under three stages of
shifting cultivation in Polochic Valley soils .. 119
14. Bulk density values of Polochic Valley soils
under three stages of shifting cultivation 122
15. Numbers of streptomyces, bacteria and fungi in
surface soils from four Polochic Valley fields 128
16. Plant parasitic nematodes found in four Pcloc'ic
Valley fields . . .. 129
17. chemical and ph~ical analyses of shifting iulti-
vation soils planted with corn during the summer
of 1956 in the Pclochic Valley . 132
LILT DF FIGL-.L
1. Map of Central PolccLic Valley in Guatemala showing
locations of field sites and rain guage stations 4
V. X-ray diffraction patterns for colloids from fou
r:.ifting cultivation sites in the Polochic
Valley . . 78
3. X-ray diffraction patterns for coll-iis from five
-hifting cultiv-tion sites in the Pcl-c .ic
Valley .. . ... .. ... 79
4. -.elatior.ship between climate and C:: ratios for the
20 to 40 cm. horizon of Polochic Valley soils 86
5. ,.elaticnsh.ip between organic matter and C:N ratios
in ''-olochic Valley t psoils . 87
e.lat' shipp between organic matter content and
cation exchange capacity for Pclochic Valley
soils .. . .. 88
7. Differences between nIT measured in ,istilled water
and in 0.01 M Ca,'"2 for soils of varv.inp acid-
ity in the Polochic Valley . 94
8. --elati-nshio between pi and total exchanreatle bases
( a, Mg, 1a, K, Al) of Polochic Valley soils ..... 96
9. .,ela'inship between ;H and per cent base satur-
ation (Cor, Ng, :, a) of cation exchange' cap-
ucity for F-olochic Valley soils . .
1 ,;. l--tionshir of pH to I KCl exchanc :.'. -" alumi-
num in Polochic Valley scilr ........ .. 1 1
1'. "x-hangeable aluminum per unit of pi! plotted against
excngeable aluminum for P'l-chi- '.';~---., soil-
(excluing atanas II) . .
12 .~lti-nshir between pHr an' neutral normal anon-
in acetate exi:.anee.l-e c;iclr- for I-lochic
Valley soils (e:c-lvin- Matanzas II) .. .. 1..^
LIST OF FIGURES--Continued
13. Exchangeable calcium per unit of pH plotted
against exchangeable calcium for Polochic
Valley soils (excluding Matanzas II) . .108
14. Relationship of pH to aluminum and calcium fraction
of total exchangeable metal cations (Ca, Mg, K,
Na, Al) for Polochic Valley soils with less than
40 per cent organic matter. ...... 9
15. Relationship between exchangeable aluminum and calcium
in Polochic Valley soils (except Matanzas II) i
16. Nitrification rates of surface soils from selected
Polochic Valley fielJs . . .12
17. Cumulative production of nitrate nitrogen in surface
soils from selected Polochic Valley fields .. .126
V'a!t land areas ,:f tne hunid trorics are used for shifting
cultivation. ^rr'Fs are rl-ir.ted and harvested in clearini-s that are
cut and L'rnel fror. the fore'2t. After productivity declines, the
land is abandoned to second -rowth and another forested site is
c -3-e... :.r:ughout the world there is increasing, interest in re-
rlacing F.tif'*irng cultivation with a more permanent type of agricul-
ture. A Erc. .-ing wcrli FcpuJati.in, which is grad-ually outstripFing
its food E~rl:r, needs increasing agricultural rr:..'uction on lands
now .tli--1 :'-r vwr., short r-ric..s of time.
Larg.--- ale research ce-r.ccrning this t:rr.. of land use has
t been sufficientl:, ehasi-.. Kellog (l'96) has stated the
rese h n : research ee rearchis more a.fl.e n.ei than that
to establish the precise actions cf the bush fall. I shCuld
tr.-r now the answer to'this question than that -f any other un-
S.': Lc : soil science. the results aff t ri.illions of
acre and millions of pe. ilc." 7Heggers (19.- 41 're'-sed the important
influence c:' food resources on human culture by b ta'.irn that, "The
level to which a culture can velop is*er.eneidert +Ipr. the agri-
cultural r- tentialil;,, cf the er.viro:-.er.t it -cruj: ." The Fr .-i and
.--ricu.lture -.ganizati.-r. the '.Fnited ri-s (1.. ) estimated that
..: -r.:: ..:- affe ted some ... illi: : .o;17-: 'ccr-r, ing 14 rilli:r.
% 'arL ~ile (Z- rillicr. : r .), an -t' i; (ic' .) re!- rt-.i that
approximately one-third of the total land area used for agricul-
tural purposes in Southeast Asia was under shifting cultivation.
researchers have previously examined the factors limiting
the duration of time ;ihich land can be cropped under a system of
shifting cultivation. Because these studies have yielded' conflict-
ing results, an investigation to determine the basic soil properties
and how they change during the cycle--forest, crop, second growth--
was undertaken on mountain slopes bordering the Polochic Valley in
Cuatemala, Central America. Since conditions restricting permanent
crop production may vary geographically, fundamentartl 1-noledge of
the soil changes which occur under shifting cultivation may be use-
ful to pre.-iict more efficient and productive management practices
after local limiting factors are recog-nized.
?oil samrles from thirty fields under different stages of
shifting cultivation were collected from eight sites within an area
of approximately 100 square miles in the Polochic Valley. Locations
were sel2,cted to include a range of -c-l,-gical conditions so that the
results might have wide application. Soil properties analyzed in the
laboratory were: clay minerals, nitrogen, organic matter, cation
exchange capacity, exrharngeable bases (calcium, magnesium, potas-
sium, and sodiar), bulk density and extractible phosrh-rus, aluminum,
narngne-e and zinc. Nitrification rates, micro-organisms and nematodes
were examined in soils from four fields within the same area. In
addition, extensive field data were collected on ecolcrical factors.
The results of these aralses and their implicating within the
tro ical rain forest -r,\vir.nnent are discuss: in this proesnt.tio,.
EP'.I?';. OF LITER'ATUL
General Description of Shifting Cultivation
Thrcughcut the humid tropics shifting cultivation follows
the same basic pattern of forest, cleared lar.n, and second grcv-t...
Local variations in shifting cultivation are usually limited to the
types of crops and differences in the period of forest fallow. This
change- slightly front place to place depending on the climate, soils
and nature of second-growth vegetation. Detailed descriptions of
this widesFread practice were given by Conklin (1957), Oook (1921),
De Schlirpe (1956), F,). (19E7), and Steggerda (1941).
The term "shifting cultivation" has not been universally
accer-te', but its use is more widesrread than other names like the
foll'.4-iii-: "fire agriculture" (3r.ith. 1954), "t-ush-fallcwing rota-
tion" (Larrau, lPU), "field-forest rotation" (Pelzer, 194E),
"rnonaiic agriculture" (Tondeur and PBergeroo-Ca.iragne, 1956), "swidden
agriculture" (Conklin, 19E7), "nilra agriculture" (:ook, 1921),
"r.aingi:" (Pen:-lrton, luJ5), "chen.." (..illis, 1922), and "laaang"
(. urou, 195c). "JhitinC Cultivtiorn" was the title for a cmposium
at the Ninth Pacific science Congress, hcle in Bangkok, Tahiland,
':verLber 18 to Diecemb.er 9, 19,7. F:e Focd and Agriculture Crgan-
ization of the United ':~tions has frequently used this term in
preference to others.
Shiftib,- cultivation in -_rtral America has been described
-.. several authors. *-k (1- ), .rerser. (1 -f-), He.-ter (l.4),
Mcrley (195t), Stegger.,!a (1941), and Thonpson (1'3C) discussed the
rrl: of agriculture within the Maya cE pire and compare it with rrcsent-
day crop pr-'uction in the same area. Morley pointed out that shifting
I-Lltivatinr nowadavy in Mexico and Guat2.'n.-l,, is identical to the trpe
used for the pat. three thousand years in the area. The only recent
innovation is the advent :-f steel tools--the axe and machete--which
rtclaced the stone axes used fo-rmerly. The basic rtes in preparing
a field and growing a crop are almost universal wherever shifting
cultivation is found.
The size of plots cleared for shifting cultivation each year
vary in different areas. Villa (1945) obtained data on field rise
f-om fifty-tCwo infroants. 'ie found that the average size of a
fiel the first year after clearing the forest was 5.6t acres, of a
second year lield after forest was 3.67 acres, and a clearing after
7 or 8 years cf second growth was E.15 acres. Most of the inform-
ants maJe more than one clearing in a season. 'te-gerda (1941)
fond that the average field in Yucatan varied in size depen'iing on
tne fertility of the soil. The average size of 635 field .s based
on five annual surveys was 9.8 acres.
Though shifting cultivation clearings are usually less than
ten acrs in size, the farmer needs much more land since, by necessity,
part ?f it is in second grA-th each year and other areas are unusable,
7>?1. (1'212) calulated that 1-0 to 2'C acres were n-eJded perr family in
eastern -uate.Tala. In areas of less fertile soils t!is value miiht
rise to 7 J or l,:':, acres. Hester (1iS4), wor'ring in southern Mexico,
estimated that eleven acres would support one individual under bush
fallo'wirn: ; thus Et persons could be supported Der square mile.
Villa (195-) described several ecological observations which
were used by the cultivator of -iuintana Roo, Mexico in his selection
of a Jesirable area for clearing: "He knows that black land (ek-lu-u.)
is more fertile than reddish land (kan-luum), that even better are the
black lands on which gro wine palms or similar trees, called labcah
because they are supposed to be the sites of ancient villager?."
:Occasionally one finds other reasons given for shifting
cultivation instead of the environmental factors cited in the follow-
ing pages. Holdridge (1947) described a case where social influences
The Haitian peasant's system of shifting cultivation is not
entirely of his own making. When iueotioned as to why
they do not plant coffee or other permanent crops, they re-
ply simply that if they worked hard to develop a valuable
plot it would ver likely be taken away from them by someone
of more iT'ircrtanrce to the government.
Environr.er.tal Factors in Shifting Cultivation
Since shifting cultivation is a widespread problem of tropical
regions many writers have delved into the possible limitations of
continuous agriculture within the same areas. Many conflicting theories
are found in a review of the literature. -The possibility arises that
limiting factors are not the same in all environments where shifting
cultivation is practiced.
In humid tropical areas the growth and density of invading
plant material far urpass.s that facing farmers of ter.perate regions.
Thu:, the problem can be appraised more r.ili-tically by:. -2m.pha'.izing
v:-getationUl succession rather than invasion by individual lKt..
Ti,,? rrobler. of grasp invasion in the Pclchic Valley beill be :!i.cp'sed
ldter. It has also been sirn.led out as an important detriment in
other areas of the tropics. Indeed, in ....ia, lalanLg or cogon grass
(L-.perata cy-lindrica) has long been recognize!' as the most serious
threat to a permanent syster.- f agriculture (Pelzer, 19-1; Pen:leton,
Many workers have considered that the most L r-rtant value of
second growth forest is to eliminate herbaceous weeds rather than to
restore soil properties. After a survey. of northern lucatan, Emerson
(195i) concluded: ".eed competition rather than soil depletion is
the factor rrir.arily responsible for the lessened ;.leld of the
second-year milpa (cornfield). It seems equally clear that tree
growth after abandcnrent of a milpa functions pritarlr y in choking
annual weeds rather than in re-toring depl-teid soil fertility."
A sir.ilar c1:nclu-si.n was reached by Joachim and Kandiah (194:-) after
stuy'jing the effect. of shifting cultivation on scils in eyl-n.
Pelzer (9l-C.) cite- grasses an weed as the main cause f-r abandon-
ing lanL! in the Asiatic tropics, an'i Freeman (1957) believd:
* *..:ee i growtY more than any other single factor, is re.r.cn-
si'Llr for crop failures" on shifting cultivated land of the I'an
The controversy on the effects of invasion of grasses when
rain forest is cleared for crops has raged for many years. Bartlett
(1l5s) described the different patterns in the tropics based on a
survey of the literature and his personal experience. In addition,
his annotated bibliographies (Bartlett, 1955, 1957) brought together
much of the Iridely scattered information available on changes in the
vegetation brought about by periodic or sporadic burning. Commenting
on whether grassland or trees will occupy an abandoned cultivated site
in the forest, he said:
Vegetational history shows that, in general, deserted
tropical agricultural clearings not burned over after
abandonment, if surrounded by damp forest, quickly
become seeded and reforested by quick-grcwing, light-
loving trees. The most rapid deflection from
normal ecological succession occurs in clearings
which border regularly burned grassland. Here de-
flection of the succession to grassland is almost
In centrall America many authors hypothesized that shifting
cultivation was limited by the length of time the land could be kept
free of grasses. They argued that once grasses became established in
clearings, labor for eradication was too arduous and consequently the
st.ifting cultivator moved on to a new forested site. Forest may then
eventually crowd out the grass. Morley (1956), one of the foremost
authorities on M1ayan civilization, was a strong proponent of the theory
that the Mayan Empire collapsed when grass crowded out the bush in
lands under shifting cultivation.
One of the earlier and more vociferous proponents of the
hypothesis that grass was the main limiting factor in shifting
cultivation in Guatemala was Cook (19C0, 1921) who maintained that the
replacement of forest by grassland set a natural limit to the period
of agricultural occupati n in a system of shifting cultivation an-!,
corsequently, that grasslcj was a foregone conclusion of rcpeateJ
sh-i'ting cultivation. LL'J.:11 (1937) agreed with Cook that fires
Destroyed many large forest areas in Guato.nala. After watching fires
evt nd the savannas of Central Peten during the very dry year of 1933,
he c included that such extreme climatic conditions coupled with shift-
ing cultivation led to the denidation of forest and the subsequent
f.:ration of grasslands in that region. However, I undell (1940)
in another paper described large areas of ancient terraces in British
-iorn :ras which he concluded were indicative of continued occupation of
lan under at least a semipermanent agriculture. It would appear that
unier these conditions a grass succession did not force the population
fro? the area.
.Argrrier.ts against the theory that the Mayan system of shifting
cultivation finally failed because the land was invaded by grasses
were rrrvi.e' by Thompson (1i54). As evidence he said:
It is true that grass will appear in cleared forest
lan. if those patch.: are kept free of trees and shrubs
for several year, but the Maya abandoned their clear-
infj after one or two season' use," and in that short
ti'-ne grass can not e-tatlish itself. I have noticed
that verges of roads out in the forest and used for
several years to extract mahogany are often of grasr,
but when those roads a aban !one they revert rapi .il
to f-rest. 5ome :.-ears ago I was in Chicchanha, 'i
i-.rcrtant Ma-.a wn in southern uLntana r.- until its
a.ar.Jonment in l r.. Luring its occupation the main
'laza and the streets must have been under grass, as
in any other Maya town. Yet, when I visited the town,
it was entirely covered with icee forest, to a la:.-an
indistinguishable fror the surrounding virgin forest.
r.ere are no savanna lands around the great
concentration of ceremonial centers in the northern
Fetrn, or around Quirigua with its deep soil, or
along the Usumacinta.l
Further support to Thompson's arguments was given by Hester (1954)
There are no perpetual grasslands in the lowland Maya
regions occupying soil- which otherwise would be well
suited to maize agriculture. It cannot be said for
this part of the world that grasses have invaded, and
become ineradicably established upon agricultural
lands as a consequence of shortcomings of primitive
General exhaustion of soil fertility has been cited by many
writers as the primary reason for the abandonment of fields under
temporary cultivation. obviously y this depends on whether one is
dealing with old, highly'-leached soils or relatively young soils
derived front basic volcanic or sedimentary parent materials. The
staff of F. 9 (19C7), in a general article on shifting cultivation,
emphasized the point that soil fertility is the most important de-
terminant of the lernth of time that land can be cultivated. They
It seems to be established that the factor which keeps
people at the lvel of chi'ir:ng cultivation is in the
first <1 ice the rapi lity with which tropical soils
lose their fertility i.., their lack of r-t .irnin'
capacity cf ;lant nutrients) and undergo undesirabil-
1Thorrsr. ur -e'te- that the collapse of the Mayan Lrr-ire
was due to revolts L; the peasant classes against the non-rro ucing
hie ra rch''.
charges in ph'.'slral conditions. These ilfflrulties can be
srlit into several factor?--low absorpticr. anasrlty for
exrhangeatle bases of the soils' rlay fracti.-'ns, the ten-
dency of these -las to imnirili.ze phosphates, the heavy
percolation rate ?f tropical rains through generally pcr-
Ous soils and the resulting leaching of plant nutrients,
the rapid destru.rirn of organic matter by bacterial
action under conditions of high tnrmFiratures, and so on.
i.ccr.-rJingl:', the fallow then restors soil fertility. In areas of rich
stable soils a continuous cultivation is possible.
A dinmal agricultural picture was painted of Africa by Harroy
(19;'9) wo 'alaned intentional firing of land for the dehydration of
surface soil colloids, leacing out cf salts, destruction of organic
matter, notification of soil p., and alteration of microfauna and
microflora. :.artholor.ew et al. (1953) hypothesize. that "depletion of
ava:ilal-? nutrients may well 7.-ng the most important" causes for
the ,icline in productivity of lands under shifting cultivation.
icnards (.1:''? in a general discussion, related the entire shifting
cultivation r'.-cle-forest cleared field forest--to the impoverish-
nt and s' ir.uent restoration of soil fertility. q Fiquier (1953), in
'adaJgascar, '-ncluded that two or more seasons of cropping so exhaust
the -:ils of hearingsgs devoted to shifting cultivation that reforest-
ation or re':-r.3ration is of u.-'btful success.
Man:-y writers, after careful stu ly, have arrived at the opposite
concluzicr: lenletion of soil fertilit:,- is not of paramount importance
ir. sr.iftl.- c.ltivation. ..- stated earlier, Joachim anr' Kandiah (1942),
after an 1:.'.t.Tr.iv- 1 .vr4i 1.'i-rn of the eff'.rt on soil. b;y -..if'Lng
'?Ltivivti:. ir. 'e -l:., c--r.rl ej that the main "L itinF factor was weds,
and not the e-rle ti n r. -: fertility. T-r conclusions were:
These investigations have indicated that while there
are some losses of ore nic matter, nitrogen, an' mineral
nutrients as a result of chenaing land, these would
occur under any system cf rotational rropring on new
jungle land and are not such as would render the land
unsuitable for further profitable cultivation with a
succession of annual crops.
From work in Malaya, Cculter (19iC) concluded that physical rather
than cner-ical factors contributed to the degradation of tropical soils
that were cleared of rain forest.
In centrall America, much shifting cultivation occurs on very
fertile soils and many investigat-rs have attributed declines in crop
yields to causes other than lack of soil fertility, although one
worker (Lundell, 1' ) wrote that apparentlyy continued rotation leads
to contlete soil exhausti.in" in northern Guatemala. However, Emerson
(19E3), after investigations in Yucatan, concluded that: ")ne with an
agronomic bac:grcrun3 finds it difficult to believe that milpas, after
two crcps of maize, have been abandoned because of soil depletion,
and equally difficult to conceive of soil fertility, once depleted,
being restored by a few :,ears of tree growth." Morley (1956), in
essential agreement with Lnerson, said:
That soil exhaustion is not the chief factor respon-
si.le for this decrease in the yield of Maya cornfields
today has been demonstrated in the Carnegie Institution's
experimental milpa at :.icn:en Itza (Yucatan). After
the harvest each successive year, specimens of soil
have been taken from this cornfiel. and over a period
of ten teh, at; nuae anal .vr.s cf these specimens
have shown no appreciable decrease in the amount of
necessary nitr-.genous salts, nor a sufficient amount
of deterioration in the chemical composition of the soil
to account for the IJi~ini.Ling yearly yiel'.
More recentl:r Hester (i.l4) after stlying soils in the Maya area,
c.'ncl.ded that: there was no pron.-unced indication of soil
x:.:au."tior.. )n the u:i-is of soil analyseE, present agricultural
te3cr-niques co.ulJ b continued in'c finittely." The soil sampling
technique was descriLued by He-ter in the f:ll.-wing words: ".amjles
were gather.! wherever nr,.itle at intervals of ten kilonete-, meas-
ured along :"igh.ways radiatinF from MeriJa, .C'ther sam]l'Is were
taken at Flacez away fro-i paved roads, Derth of sar-ling and
whether single or composite sar-.les were used is not stated. Fifty
soil samples were anal.,zed at the University of .alifornia. Soils
were extracte. with ten per cent sodium acetate at pH P,.O Appar-
entl:., Dhosph-ru-, potasslur-. calcium and maFnesiur were determined
in the extract. In ad:iti'r., soluble salts, nH, moisture capacity,
total nitrogen, aTnonia nitrogen and nitrate were also measured.
Tn.hrpson (19.5) pointed out that: "uuiripua was one of the earlier
cities to cease functioning in the Maya Empire although the alluvial
soil there is ver, rich." Actially the soil of Quirigua is probably
not as good for crop production as Thompson would have us believe.
The writer is of the orininr that Charter (1940), after surveying
soils in "ritish Hcn.!urac, g v, the most accurate summation of the
Lasic liritations of shifting cultivation in the Maya region.
:e sai k.
.e'lirn frer soil fertility and competition from weeds
are equally LP-'r.ant factor" in the decline in yields
'f and under shifting cultivati .n. The sil is actually
rich en-ugh to -u -'rt an intensive agriculture with
weed control and a smnall ar.-urt of fertilizer (which
is ;n:ce7sF Y- urndeIr intensive ag-rirulture).
Literature can be drawn from many sources, not directly re-
lated to shifting cultivation, on the effects of burning and forest-
clearing on particular soil properties. Pertinent literature for each
specific property ,will be discussed under separate headings.
IIitrogen and organic matter.--3ne common misconception of
tropical soils is that all nitrogen and organic matter contents are
low. Originally, Jenny (193C) hypothesized that soil organicc matter
contents decreased with increasing temperatures, rainfall remaining
constant. His survey of the United States showed that the soils high-
est in organic matter were in the northern portion and values decreased
progressively southward. several writers, among which were Mohr and
Van Baren (1954), extraFclated these data and concluded that organic
matter contents for tropical areas must be even lower than in temper-
ate climates. This misconception was probably responsible for the
erroneous ideas that prevailed for a long time, that all tropical
soils are low in .rgn;ric matter. However, Smith et al. (lE51), after
investigations in -'u--rt.. Pic.o, proposed that the concept of Jenny
be limited clinatically to temperate regions by the frost-free boundary.
They sug-ested that the absence of killing frosts favor the building
of orgar.ic material b green plants much more than it favors destruc-
tive nicrobiol.o ,ical activity within the soil. Data from many areas
supporrt their assumption.
High soil organic matter contents in the tropics correspond
to areas of ?.igh rai;:fall with no marked dry season and to elevated
re,-i,-ns with cool t :-e-ratur.--. Very high soil nitrogen values in
the ..estern Herisrhere have been reported from Hawaii, r:.itish Honiuras,
'.osta Rica, -'olormia and F"aerto Pico (Dean, 1930; Hard:, et al., 1935;
T.--rny et al., 1949; Luar-.l e C'astro and r drigue:. 1'_,11; Schaufeloer-
ger, 195'; 3r.ith et al., 1'3O). A few :-tui' wher- an extreme pau-
city of _rcanic mtter has been reported in the troFis were from
areas especiallyy in Africa) either with less than 21..hity inches of
rainfall, or with sever-l very dry month! during the year.
elaticnships between soil organic matter accunulation and
cl iate and elevation have been thoroughly studied for several areas
of the tropics, and general levels of organic matter were much higher
than would be exTected from sir.ilar studi.- in tennerate zones. Soil
crhan.ic matter increased with an increase in rainfall if temperature
r- ined constant ("raig and Halais, 1934; "irch and Friend, 1956;
.e.r., 1-7 ; c'.afelbereer, 1~n'). Jennrr et al. (194E) suggested that
t..i n increase was logarithmic. 1.il organic matter also increased
-ith a decr-a.se in temperature, which generally occurred at higher
l 3.-v9ti-ne (:-a.irez, 191E; L-,an, 1937; Schaufelberger, 1956; Jenny
et .-i., 14c). Hcw.'ver, 'ir'P- an; Frienc (1956) found the effect small
c-rrr.ar-- with that of rainfl.l.
The -.ih rat- of rr-' u.ti n of tropical vegetation is undoubtedly
one f the most important factors in the high levels of soil organic
matter r-"-: :, i;n the hiwid trrFics. :easurements of organic matter
rrc !ucti are avai l.ble fr-.-r. several t.r-ri"al areas and indicate that
races are much .iricer th in temper'-..' clL-ates. 'uroa de _stro and
-*'r -uez ('""= alculetd t.e li't-r r t r:-..- .-. ll to th-e -il in
Colomiian coffee plantations was between 4.6 and 13 tons per hectare.
Jenny et al. (1949) estimated the annual production of organic matter
in the form of leaves and twigs in Costa Rican and Colombian rain
fcrests was 9.4 to 13.2 tons per hectare and that the time to reach
near equilibrium of the forest floor was less than a decade. Laude-
lout and Meyer (1,l4), for the Belgian Congc, reported organic matter
Froducticn was from 12.3 to 15.5 tons per hectare and the quantities
of mineral nutrients involved in the annual cycle were similar to the
quantities involved in temperate regions, excert for nitrogen which
was 6 to 10 times greater in the equatorial forest. Orchard and Darby
(1956) estimated the enrichment of veld soils under leguminous wattle
trees was 12C + 4C pounds of nitrogen per acre per annum over a period
of 50 years. Tvi-ently, additions of organic matter are of sufficient
magnitude in the humid tropics to maintain soil humus at a high level
although decomposition is rapid.
The clearing or cultivation of forest soils appeared to result
inevitably in lower organic matter values. On forest soils of Trinidad,
Duthie et al. (ICZ,) found that organic carbon, nitrogen, and the C:N
ratio immediately decreased to a ierth of two feet after forest land
had been cleared& ichards (19' ) interpreted the results of Duthie's
investigation as indicating that effects of forest clearance alone on
-rganic matter are ve-ry transient but effects of cultivation over a
long duration of time are much more lasting. In a study of 22 locations
in Georgia, Gi Hrid-n et al. (?17) reported that cultivation lowered the
vs-race content -f .rrg-anic matter from 3.29 rer cent in forest soils
to 1.43 per cent for cultivated areas. Coulter (1950) reported that
the average nitrogen content of virgin rain forest scils in Malaya
was consi ierably higher than that of cleared soils, H,-wev'r, not all
rain forest soils were !ilh in organic matter, nor all learned fields
lo.. The average nitrogen content of forest soils was 0.240 per cent
whereas the average value for r.lintati r. soils was C.155 per cent.
Coulter concluded that; "It wod appear that, with l-r.itationr, the
nitrogen content of the soils cOul.i be utilize] as one criterion for
f-llowing the rejuvenation of a degraded soil." In Ghana, forest
soils after clearing lose organic carbon in the C-12 inch layer at a
rate of 3 per cent per year (Nye, 1955P). In Madag scar, forest
clearance appeared to destroy organic matter and humus and to check(
its rubsequnt accumulation without necessarily reducing available
nitrogen ('.i luier, 195 ).
'es:-ite almost universal agreement about the effects of clear-
ing on soil organic matter, the effects o'f burning appeared to be much
more c2ntrc.versial. He:,-ard and Barnette (1934), Klemmedson et al.
(-l'4[-), Meir:le 'hn (j5E1 and 'uarez de Castro (1957) f-und that burn-
inp either increased scil nitrogen or had little effect. Meiklejohn
believ-,J that burr.ing the vegetation might kill the nitrifying bacteria
in the soil below the burn nd so lead to a c..nnervLtion of soil nitro-
ger. 'nl -isi et ;.-. (ji ), uring pnt test-, re prted that the nitrogen
surrl.inr nT'-wer of the soils irvertigated was increased up to eight-
fl" by warning. .wv-r, J.achi. ian Kan iian (1.-lk) re-ortc. losses
of nitr rrcn zan-: .r';ranic matter as a re-uii t .f .urninp in ?e'.1.
The main effect of forest clearance on soil organic matter
is through the loss of a constant source of fresh litter which con-
tinuously accumulates on the soil surface and is rapidly attacked
by the soil organisms. Furthemnore, the removal of shade probably
leads to a faster oxidation of soil organic matter as a result of
higher surface temperatures from increased sunlight. The mainten-
ance of a large supply of organic matter in most tropical soils under
continuous cultivation for a long period of time is doubtful. Lugo-
Lopez et al.(195) found that cultivated latosols in Puerto irvo
seemed to reach an couilibrium at a maximum of 3 per cent organic
matter irrespective of treatment with organic materials. Tuarez de
Castro and (od5rigue- (19'5b) reported that the response of coffee
soils in Colombia with 3 per cent organic matter to additions of organic
materials was of the same magnitude as temperate climate soils con-
tainin- 1 per cent crganio matter. They suggested that soils contain-
ing les- than 3 per cent organic matter were deficient in this respect.
The ratio of carbon to nitrogen is often useful to predict
either the amQunt of immobilization, or rate of release, of mineral
nitrogen. Much i lp-rtance is attached to the values of C:N ratios as
indicators of v.hether organic additions to soils will release or tie
ur mineral nitrogen. Th: latest views were summarized by Russell (1950)
who said that easily decomposable -lant material when added to the soil
..ill release inor-.r.i- nitr.e-r, if the C:N rjtio of the added material
is less than 2. and rill remove inork7aici nitro-en if it is much above
50 in tropical r ni .nr. Black (19'7) placed the critical C:N limits
between 15 and 33, and Harmsen and Van Schreven (1955), in a review
of literature, placed it between 20 and 25. Hardy (1946a) concluded
that for comparable C:'J ratios more nitrate was produced in the wetter
'..est Indies than in drier ,ueensland. Thus far, for tropical areas,
the 'i: ratio has been of little value or use in predicting responses
to applied nitrogen, though Hardy (1946b) successfully demonstrated
that ':N ratios, in the West Indies, below 10.5 produced bolting in
cotton. Above that value, less nitrogen was mineralized and a normal
cotton crop was produced.
Investigators in many tropical areas have found interesting
relationships between soil C:N ratio and environment. In Mauritius,
scils cf wetter areas contain more organic matter with a wider C:N
ratio than thcse of dry areas (Craig and Halais, 1934). Dean (1957)
C-rnlu.ed that the C:N ratio of soils in Hawaii increased with in-
creasing rainfall but was not significantly related to elevation.
However, :!ohr and Van Baren (1954) assumed that the C:I1 ratio would be
higher at higher altitudes than near sea level. They cited H. J.
Harion's wrk in Indonesia, which indicated a positive correlation
'itr. temperature, a negative one with pH, and no correlation with
moisture and soil type. Jenrn et al. (194-) confirmed Hardonse
observations with investigations in Cclmbia; but Ramirez (19E1),
working in the same country, found diminishing C:'J ratios with in-
Nitrification under the favorable temperature and moisture
conditions which are present in the .ropics usually proceeds at a
hiEh rate. However, the ver, large quantities of nitrate nitrogen
found in many tropical regions has baffled investigators, for it w-uld
appear then that sources of nitrifiable nitrogen are inexhaustible.
Greenland (195r) probably correctly diagnosed the situation when he
,.Uhile precise data on soil conditions under tropical forests
are lacki:.., it is almost certain that they are such that
absor ti-n of nitrate by micro-orgar.isms, denitrification
and lachi:,g .ill all proceed quite rapidly. The high levels
of nitrate most probably represent an equ'iliorium level
typical of a rapid cycle in which relatively large addi-
tions of nitrifiable organic material and of newly fixed
nitro:-n-, and ccrre.spondingly large losses of nitrate by
ljaching and Jenitrification occur.
Tropical workers for many years have noted the seasonal re-
lease of nitrates in soils exposed to wet and dry seasons. Generally,
at the onset of the rainy season, a very rapid increase in nitrifica-
tion occurs which lasts approximately six weeks (Griffith and Manning,
1?49; Greenland, 1958; Hagenzieker, 1957; and Birch, 1958). Greenland
(1hI5t) has suggested that nitrifiable organic matter increases during
the dry season when crop residues and dead micro-organisms accumulate
in the dr-. soil. This organic matter is then available for nitrifi-
cation processes at the beginning of the rainy season. Thus, the
rro-Juction of nitrate nitrogen is usually very high immediately after
an extensive ir, season. One right conclude that crops should be
r'-:',rn as early as possible after rains commence in areas with a marked
dry season to coincide with the period of maximum nitrate production.
In the humid tropics with a mild .'ry season, a slightly dif-
ferent picture can be drawn which might be exemplifi,.-! by the data .-.f
hardy (19-16t). In Trir.ilaac .ith a rainfall of 68 in-he?, Harly, found
that axirani-, accujulati,.! of nitra~'s occurred during g the milii iry
season. T. accurulati -n of nitrates bewan when the amount of evap-
orsti-n exr'.eded the rainfall; u-uall. when the rainfall of the pre-
vious 28 .a:: period was below 4.? inches. It is r-ossi'-le, as Green-
land (195t) hypothesize !, that the soil moisture Lurin. the dry season
:.. be L.ig. enough so that nitrification is not impaired (Greenland
:'unri. conideralle nitrification at moisture contents as low as 4
r-.r cent'. Gre-rland concluded that nitrates accumulate under these
c-r. 'itions 3aring the dry season and are ra-i `ly leached at the
be.-i.:ing of the et season. Since nitrification has continued
luring the dr:. season, little nitrifiable organic matter is left to
s~ r..rt a high ra.e of continued nitrification during the rainy
Investigators in the tropics have presented conflicting evi-
dence -f the effects of shade on mineralization of soil nitrogen,
I.ar.-m n and Van Cchreven (1955) concluded from a literature survey
that smaetes too 1-igh temperatures may suppress mineralization.J
H!ewever, Cr:ffit. and Manning (1949), in Uganda, reported that during
the rainy season nitrate arriL-ulation on bare soil was much higher
trian *,nder a grass ulch cn medium textured soils. Mills (1953), in
the sa"c area, recorded nitrate accur.lati.ns up to 20G ppm. on bare
fallow -il, and only:, i ppm on shaded an'/or .nulched soil. Green-
lin- (1.5 ,. in contrast to Griffith.s an ~d ill's work, found ver-.
lit*e .'f-:-rence in nitrate lev.:lf between shaded and unshaded bare
fallow plots on sandy soils. He believed that in Griffith's work the
unshaded plot would have been baked hard by the sun and leaching would
have been reduced, so that much higher levels of nitrate were found.
Since Greenland's work was on coarse, sandy soils the effect of sun on
soil structure would have been small. Jacquerin and Berlier (1956),
however, found that nitrification was almost nil in a bare soil on the
Ivory Cast. They indicated that the low nitrifying power of forest
soil is increased by clearing but will soon drop to zero if the soil
is kert bare. Apparently, some of the controversy on the effect of
shade on mineralization of soil nitrogen is caused by the effect of
shade on soil structure and infiltration capacity occurring at the
Cation exchange capacity and base saturation.--Cation exchange
capacities of tropical soils vary widely but are generally considered
to be low. A range from 2 to 64 me. per 100 g. has been reported for
Puerto Fican soils (onrmet et al., 1951), and a range of 2.8 to 127 me.
per 100 g. has been found in Hawaii (Kanehiro and Chang, 1956).
Crganic matter may contribute a large part to the exchange capacity
of tropical soils. Abruna-lodrlguez and Vicente-Chandler (1955)
reported values for the exchange capacity of organic matter of some
Puerto Rican soils generally between 100 and 150 me. per 100 g. and
ac-ounted for about 25 per cent of the total exchange capacity of
kac-linitic s-il:- with widely airing pH values.
Pase (or nation ) saturation has been related to several en-
vironmental features. Mik-.liTh (1?lI) was able to show that the
relationshipr between base unsaturati.:n and pH was characteristic of
the type of clay mineral present, with the exception of montaorillon-
ite and hy'irous mca which had similar curves. Kanehiro and Chang
(1:SC) founj an cverall I.-rrease in base saturation with increasing
rair.fall for Hawaii. Laui.elr.ut and Meyer (1954) reported that the
caitijn conpo7ition of Flant material returned to the soil was related
to "the legr-..e of cation saturation of the soil adsorbing comnl:x" in
tae ielgian Congo,
Oclemar. et al. (195E) and Coleman et al. (19E9) discussed the
concert of permanent charge and pH-dep'ndeint charge components of
the exchange capacity, Permanent charge was defined as the sum of
exci.angeable cati:,ns, incl'idint aluminum displaced on leaching with
a neutral salt solution. The pH-dependent charge was regarded as the
acnt of exchange aciity remaining after neutral salt leaching
(mair.nl ori-inatinr with weakly ionized, acidic radicals). They felt
that cation saturation based on permanent charge was preferred to
saturation based on exchange capacity as a concept of soil fertility.
Soil acidity.--W'ell drained, mature soils of the humid tropics
ar-' highly acid because most cf the exchan,;eatle bases have been
l,-aich-eJ fr.r-. the profile by the heavy rainfall. Nevertheless, many
:.Triters have reported that tropical crot are abl- to grow well at
pH values .ar.' ilerably below those re'euirel fr temp-'prate crop-.
Sicnard-rr (191i7) -ail that in areas ',rth an appr. xirrate rT of 5, no
tropical or .*;trcpcal crrs other than pear.uts are lil-1y to need
line. -:?I (l-9) btel,:-'.'d that crops t.l-rate aci it-, better in
moist than in dry climates. He gave as a reason: for a given
soil pH as measured by shaking up in water, the acidity of the soil
solution will rise as the soil becomes drier, and the root system of
the crop may suffer accordingly."
The acidity of soils under a heterogeneous forest cover is
usually quite variable, )vington and Madgwick (1957), in England,
were able to grade a large number of tree species according to their
effect on soil acidity. Kellogg1 cautioned that the pH under tropical
forest trees is highly variable. He reported that soil pH values in
the same mixed forest in ifrica may vary from 4.5 to 7, depending on
whether the samples are collected under sulfur accumulators or under
Burning generally increases the pH of the topsoil because ash
from vegetation contains various amounts of the oxides of calcium,
magnesiu-., potassium and sodium. The reaction of these cations with
the exchange cormlex increases the percentage base saturation and
reduces soil acidity (increases soil pH). Investigators, in widely
separated areas have reported the decrease in surface acidity which
accompanies burying of vegetation (Evers, 1954; Heyward and Barnette,
7lJ4; Isaac and r -ir. l'Z7; Tsn-E et al., 19E5; Suare- de Castro,
1957; Vine, 1955, 19C4; Duthie et al., 1956).
1Prs,.ral communication from Dr. C. F. KellDgg, Assistant
Administrator for --il u'-v *, United States epartmert if ;.gricul-
ture .oil 'or.s-.'tion ervic--, 1'. shingtor., D. C., dated October 31,
Several investigators (Maher, 194'; .leln~an. i'.19; Mii.lelLurg,
19L2; Peel +, 194L) have fund that the correction of soil acidity by
li-iing is Df doubtful merit in the troFicF because it promotes trace
element deficiency, ht-nus deomposition and structural det-rioration
cf latosols. :ildelburg (l;r) and Schuffelen and Middelburg (1954)
fond that liming decreases permeability rf latosols in Indonesia
until a pH of 7 is reached !, above which permeability increases again
with further additions of line. The;' also found that liming reduced
inor element availability ex-ept molybdenum and increased nitrogen
loss. n the 'eril cla. lear'. in the southern United States, appli-
cations of caliL. in the form of carbonate, chloride or hydroxide
d.-r.easedr saturated pernealility, (Peele, 1940). In Colombia, although
i:- ir. increased soil nitrate content, Fuarez de Castro and Rodriguez
(1 .') sugg :t. d tat growth results indicated it was inadvisable to
use 1"-I, ._:-.-iciliy at heavier rates. Apparently, since many tropical
cr-n do tolerae v.r-,- acid conditions, living soils to decrease soil
acidity should be rr.-crined with great caution.
..lr.i: .-Many of the negative effects of extreme soil
aci it;.- on cr-r rrcth can probably be attributed to the large amount
*.f al wir. ich comes into solution at low soil plT. Soil pH values
be-c,' 5.5 indicate aluiir.un is present in the exchange cor-:.lex, and
at vaiu.:.s el-w 5.0 aluminum is one of the ma.-or exchlanreable ions
(" lear.;', i '). Arnon (---tted by Allaway, 1957) denonstrate.i that
plants, .r- n.n in s-luti .n. are not greatly affected '-o nH values
frr,. 4 to 9. :!owe/r, "-leran et al. (i956) found that root -rr':-th
was largely inhibited if soil pH was below 5.5 and if the soil con-
tained more than about 90 por. rf exchangeable aluminum. 'Usualy,
aluminum toxicity was indicated by a stunting of the root system.
Grasses, such as corn and sugar cane, are sensitive to high amounts
of exchangeable alur:inun, whereas tea, coffee and possibly cacao can
withstand conditions associated with high aci.lity (Hardy, 1956).
Magista. (1925), wc-rking -rith solution cultures, reported that aluminum
exerted a pronounced toxic action when present in concentrations of
50 to 1 pl' ppm. in cultures at pH 4.2 and 4.0, respectively. Corn was
more sensitive than other crops and gro..th was seriously affected at
:as.s of aluminum toxicity are often reported on highly
lachu acid sc.ilc of the humid tropics. Duthie and Bourne (1939)
found that aluminum toxicity, caused by the use of acid fertilizers
on sugar cane in British Guiana, was particularly noticeable at pH
values below 4.7. They recommended flood fallowing to remove soluble
salts and raise the available calcium content of the soil. Nguyen-song-
Vien (1l93) suggested that the beneficial action of both mulch and
calcium ;hsc'Sphte on sterile patches cf soils in Indochina could be
attributed to the iniobilization of free alwr.inun i-ns. Coarse-
textured soils in iU.-nda with r' values near 4.0 contained as much as
460 rpr. availa.ie aluminum (Chenery, 1954). ?uch acid soils were
found to occur in regions with annual rainfall in excess cf fifty
Potasiu.m and sg:e.lru.--In general, potassium deficienoios are
not widespread in the tropics. Although levels in the -xchange complex
nay not bp !.igh, the rari I rate of weath-ring which occurs in the
haid tropics suffires to release -n-uph nrtassiur fromr potash-hearing
r.inerals to supply lrant needs. l ven ... -h slightly weather, rh:.o-
litic soils of Sunatra ma', -crntain cnly 7.- to 195 ppm. exchangeable
r,"tassiur,, crops do not rez:-on. to applications of potash (Van der
Mar-l, 1947). Likewise, y and Ltephens (195e) reported that savanna
soils of G`iana gave no resp.n-e- to potassiu. fertilization during
tight years cf crcpping thouk-h they cont.iirne as little as 40 ppm.
eY.xchangeabl.? potassium. The lack of resp,.'nS* to potassium in Suma-
tra an-.i Gharn has been attributed by these authors to the release of
s9:fficient p.-.assium from the rapid weathering'of minerals in a hot,
;nre-ll:., the po.tassi'an content of topsoil will increase
aftr 'urni:,- (. aarez de Castro, 1957; Focan et al,, 1950) because
catio-s are r-l-a'qed from oxidation of vegetative material. In soils,
rclKaed I-. 'r.in, nctassiun is not leached as rapidly as either
cr i. i or .a-'n'-iu". (Hardy, 1956; Mehlich and Peed, 1945; Suarez
': .a~tr,: and :':dricuez, 195&). One reason is that a decrease in
exhang"- -:-acit. (h-.ich clearing produces) fav-r E the adsorption of
rnnv;,l ..: ions at the expense of the divalent ions ('iklander, i'55).
Iur')r.'~r''r, as a scil becomes mor'. aci (by le -:hing), potas3i r.. ions
are held i livnl;- m_.re :.tr.ngly n calcium ins (Rusell, i.: ).
Although some plants are able to use sodium as a partial
substitution when potassium is deficient, there is considerable var-
iation among plants in their behavior. Corn tends to exclude sodium
and shows no response to fertilizer applications even when potassium
is deficient (lack:, 1957).
Calcium, ani magnesium.--A large part of the literature per-
taining to calcium investigations in tropical soils is concerned with
the practice of lining. No effort has been made in most cases to
separate the effects of calcium nutrition from the effects of soil
pH changes. For this reason, literature on liming has been included
in the section on soil acidity.
?vidences of manresium deficiency have been reported from
several tropical areas (Kandiah and Rodrigo, 1954); Ferrand, 1957;
He:.-itt and Bull, 1956; -.-binson and Chenery, 1958). However, com-
paratively very. little work has been done on nutritional responses
to calcium and magn-:siur in the tropics.
'anganest an,: zinr.--Analyses of exchangeable manganese in
soils are available from several tropical regions (Biswas, 1951, 1957;
-r-:ins et al., 1''4; Koch, 194E). Vlamis (19E3) compared toxicity
of marannese with aluminum in solution cultures and found that alumin-
um toxicity occurred at 1 to 2 arts per million (npn.) whereas man-
':.nes' was toxic at concentrations of 10 to 15 nor.. -lark (1957) said
that corn may tolerate over 15 rr.rr. manganese in displaced soil
Thorre (1957), Ln a review of zinc nutriti-,n, stated that
zinc in the soil is usually:. associate i with residues of organic
nat'rials ani greatest c',:rcentraticin are found in leaf litter.
Iiiv rtigators have generally; associated zinc deficiencies with soils
that have a pH of 6.0 or higher and a san:, texture. Fruit crops are
ost commorrly affected but corn probably suffers the most of annual
crops from lack of this nutrient. Gall and Barnette (194C) reported
toxic levels :f replaceable zinc for corn on Florida soils was between
2f. aniJ ik' ppr. on Norfclk sand, between 24E and 572 ppm. on )range-
buLr fine sandy loam, and between 528 and 733 ppm. on a Greenville
shiftingg cultivation is perhaps a good way of avoiding zinc
Jeficier.cy' without the use of chemical fertilizers. Rogers et al.,
(l..i) finJ that native weds, on zinc deficient soils of Florida
were far better collectors cf zinc than planted crops. Native weeds
containeJ on the average 140 ppr.. of zinc in their dry matter whereas
a cover crop of crotalaria contained 4 to 11 ppr.. They suggested
that the best way of preparing land for corn is to allow native weeds
to cover the land, and then plow them in before corn is planted.
The forest fall-w in shifting cultivation would pr'-bably perform
thre mme f~Lction.
:.isphcrus.--Fhosphorus has always been one :f the most dif-
ficult *.l--r.nts to analyze in tropical -oils. xtra'ticn methods,
devel-ip.e: for less acid anr less highly l.'ach,-L sils .f teprate
clinate-, hav not :j-en correlate'- v.-r;, rucce-silly with cror,
responsr3-E in the tropics. Birch (1952), in Ken:.a, found a more sig-
nificant correlation between phosphate responses and percentage sat-
uration of the cation exahange capacity than with amounts of acid-
scluc.l adsorbed and ,rater-soluble phosphate in soils. neverthelesss,
in Ghnis- Tye and Bertheux (1957) obtained a better indication of phos-
phate responses on sorghum, corn and peanuts with the Bray "quick test"
extraction than with more complete types of phosphorus extractions
such as the acetic acid-soluble, the sodium hydroxide-soluble inorganic
forms, and the crgar.ic form. Lopez (1956) reported a very significant
correlation between the strong Bray, extraction (0.035 NH4F in O.lN HC1)
and corn yields in fertilizer trials in Colombia. The method of Bray
and ur-tz gave the most significant correlation with field responses
to phctprhorus fertilization on acid soils in a survey of methods used
y laboratories in the United Ctat+. (Anonymous, 1956).
.hosphcrus, next to nitrogen, has been found to be lacking
the most in trorical crop production in general, although Nye and
Bertheux (19,7) said that in -ha-r.a and other parts of West Africa,
phosphorus is the most serious nutrient deficiency. Nye (1958),
a1sing Era;,'s rapid extraction procedure, found an average of 5 ppm.
:.f phosphorus in the 0 to 15 cm. layers of eleven undisturbed forest
sites in Ghana, but pcinte I out that on sites disturbed by habitation
v:ery high values, sometimes over 1~ ppm. rhosrh'-rus, were obtained.
Lopez (19c.5) extracted 17 to r" rrr. phr'srhrru- with the strong Pra-
extraction procedure on soils in i-_n:'ia to which no o1i-,.phorus had
been added; heavily, fertilized -mils had values as high as 70 ppm.
Usually in well-lached., trorFial forest soils the amount of
ari -spclut! 1p.osphcrus is much higher in the topsoil and decreases
ri i ly with .leth (.':ll-ggp and Dav.l, 1949; 'ndredy and Montgomery,
l:f-i; Nye anI DBerthl ux, l',"''. This is in contrast to less well-
l,'-ir-he. te-irerate soils where the phossphrrus generally increases with
jer:- (,,'e and 3ertheu-x, I '). 7ince phosphorus is closely linked
to the orgaLic matter cycle, the greatest concentration in highly
weathered soils will rre-r.nd? to the zone of accumulation of
7r~'aic matter. In ihana scils, the acid-soluble phosphorus is pre-
domiant in surface soil h-rizons and is definitely correlated with
the total nitrogen content (..ndrJely and I-!ontgomery, 1954; Nye and
Nye and .tephens (195t) p,-inted out that in forest soils
much of the available phc r..hcrus for crops was maintained by release
:f 1h -sphates fror. organic phosphorus during the decomposition of
:cr-aric matter. The:. suggest<' that the rate of oxidation is about
the sme as soil or.::ic matter. In continuous fertilizer trials they
rf:1, the rate --f loss -f -r .rnir carbon to be 3 per cent per year
in forest for the 0 to ~ cr. norizon (the horizon from which they
Sworn obtained 9' per cent of its nhO-h'.-ruj- requirement).
7.e:;, ~al.31ed that in forest soils, with abrut 400 pounds per acre
:f .-'r-aric ihosph-rus in the top 3 cm., .he release per :.ear was
2 .-'.i? --f .phosFh.-.r' or the equivalent of l1 ir..is of single
si;-er-.- : .ate f:.rtilizr. M.'ch f the value of re'zting all-.; was
attri:.ted to i.s r.-' in r.plenir.-ii- 'these -u:-1.s f available
phos[r:-a in the -r .-!1.
lhen vegetation is burned to clear land for croppin-, much
soluble phosphorus is added to the soil. ,lamris et al. (195E), in
pot tests, fcu n- that the phosphorus content of burne- soil increased
so as to give ten times higher yields unless partially fixed b., the
soil. In oolombia, the phosphorus content of corn leaves was increased
by birr.ing (-uare= de Castro, 19E7), and, in California, the phosphorus
content of burned pastures was double that of unburned pastures (Hart
et al., l~1j3). Turned plots of green manure (Ztizololobiur.) in
Tligeria always gav_. better growth of corn during the first few weeks
than plots where the green manure had been dug-in. This was attri-
buted by Vine (19?) to the fact that the phosphorus in the green
manure was made more readily available by burning.
.c.ii structure.--The importance of soil structure in crop
production has been well-recognizc i by a few writers. Baver (1956)
said: ":oil structure is the key to soil fertility!" Martin (1944)
concluded from field observations in Jganda that: it was the
physical rather than the chemical condition of the soil that deter-
::e and Steph.e-.s (195, ) described the effects of shifting
cultivation on soil structure in the follcring way:
Under a forest fallow the Eurface horizon develops
an exc--ll -nt .hy-rical condition. Th,3 surface of the
soil is r tecte b;. the litter, the fauna keep the
soil open, and even on the light soils a weak crumb
structure v 1ps.
The weak structure develcr-~ ] under fally..: is
rapi ly de-tr..-,;d b:. cultivation, and to preserve it
the less the soil is cultivated the better. The
traditional far-.ing practice in the forest, which
inv-lves a mir.L-un of soil iisturbance and a gorod
soil cover preserves structure and organic
Most tropical scils are very well structured, primarily due
to the influence of ferric hydroxide (which dries almost irreversibly)
on the aggregation of soil particles (Baver, 1956). Lutz (1934)
s-ggested that the role of iron in aggregation was probably that of
fl-cculaticn and cementation of soil aggregates. Pussell (1959)
mentions this in describing tropical soil structure and its implica-
tion.-.. He related soil structure to
the binding together of soil particles into
fc.irly: sponse-like crumbs by precipitated and fairly
dehd.r?,te.l iron oxides. This iron oxide gives a red
color to the soil, but not all red soils have this
very desirable structure. When it does occur, as it
does for example, in parts of the Kikuyu reserve in
Kenya one can practice an intensive system of nearly
continuous croiring in which little use is made of
resting crops or farr.yard manure without the complete
collapse of soil structure or crop yields; a system
which would fail cor.-letely for example, on soils
derived from granite 1-v in iron.
The effects of burnir:g on soil structure have not been ex-
tensively investigated and no generalities can be derived from the
few results available. -'-ri.-uez (1952) found that texture was un-
altered I 1 rring in Colo.'' a, but that structure was markedly and
cunilatively improved with -cnsequent improvement in noncapillary
rcrcsit:-, aeration and permeability. Burning increased the percentage
of aggre-ga's larger tha 0.5 mm. fro- 35 per cent to 42 per cent and
decreased tie amut f bairegates smaller than 0.25 Im. from 5. to
42 per cent and decreased the amount of aggregates smaller than
0. mrm. front 50 to 43 per cent. .'hen twice as much brush was burnt
on the soil these differences were more marked. He suggested that the
large increase in yields he obtained on burned plots was probably due
to the improved physical condition rather than any increases in fertil-
ity. --ryess and Youngberg (1957), studying the effects of logging
and slash bJurrning in Oregon, found that structure was adversely affected
rnlr in the most severely burned areas where there was a significant
decrease in the amount of clay present and a 20.6 per cent decrease
in the degree of aggregation. They concluded that since only 8 per
cent of the soil surface was severely burned, detrimental effects were
not arpreciable. :.After investigations in Ceylon, Joachim and Kandiah
(1948) reported: here are no appreciable changes in soil structure
as a result of chenain: (clearing) land."
Effects of burning on infiltration have been more thoroughly
studied than effects on soil structure per se. Several studies (Burgy
and :.cott, 1952; .cott, 1956; Scott and Burgy, 1956) have shown that
burning increased infiltration rates of the soils studied except in
the case of intense turningg of a wet soil surface. However, Burgy
and ?cott (1955) pointed out that run-off from a burned area exceeded
that frror. an unburned area for a period of two years. They proposed
that this was caused by a lessened friction factor to overland flow
whenever rainfall intensity exceeded infiltration capacity. Suarez
de Castro (1953) found Lurnin-, increased infiltration on two plots
studied in Colombia. Houe v.-r, u_;o:s:i (1956) quoted F. W. Freise,
who worked in Brazil, as writing that infiltration rates were arrrox-
ir.ately twic. as slow after forest had been felled and burned than
The :iulJ density (apparent density) of soil is a good measure
of ':il structure. The bulk densities of clay, clay 1Tar, and silt
l.?a-a surface soils normally may range from 1.00 to as high as 1.60 gis.
pe- cc,, ierending on their condition (Lyon et al., 1952). However,
forest soils with high organic matter contents in the surface horizons
ma:. ce much lower. Values for the Ao horizon of forest soils are usu-
all- about C.2 and those of the A1 horizon are commonly less than 1.0
(L.': and .'lanjl.dr, 1946). Many tropical soils, because of their
.i1- state cf aggregation m'ay also have low values. On Hawaiian sugar
cane lan.Is, bu densities cf surface soils average around 1.0 and
Valj-- cf 0.4 and 0.5E have been reported (Shaw and Swezey, 1937).
..;.' iri h.ic latosols of ji.&aii generally tend to have bulk densities
rangir.ig fr'rn C.I to -.7, t;.h:-rh they average approximately 0.5 (Sherman,
l"*'. Trcuse (quoted by saver, 1956) found bulk densities ranging
fr:.-i 0.4 to 1.- on Hawaiiar soils derived from volcanic ash.
Tlie effects of bulk densities and soil compaction on plant
gr.-,th have been studies -. several workers. Viehmeyer and Hendrick-
son (19 ) related limiting~ :.'ulk density values to different textures.
The f?'.: chat thresholdd densities above which sunflower roots did
not enter seeie to be about 1.75 for the sands and varied from about
1.;. to l.'. for the cays." Investigations in Louisiana (.carsbrook
et ._l., 1l."2? in ;icat-: that compact' plo.? Crn- with bulk densities
of 1.35 to 1.40 seriously restricted the penetration of cane roots.
-rl1: (1954) suggested that plowsole pans in Florida with densities
higher than 1.8 should be given special attention. Bertrand and
Kohnke (1957) found corn roots did not penetrate a subsoil compacted
to a bulk density of 1.5, but they grew profusely in subsoil with a
bulk density of 1.2. For the tropics, Lugo-Lopez and Acevedo (1956)
showed that compaction by heavy tractor traffic increased bulk den-
sities of Puerto Rican soils from 1.14 to 1.35 and reduced quick
rrosion.--nIormally, with comparable slopes, soils in the
tropics are less erosive than those in temperate regions. But among
the most productive soils in the tropics are those on steep slopes,
kept fertile b.' natural erosion of the leached surface layer, or on
'.-ung volcanic Jepo-its, which are often hilly. Indeed, Pendleton
(1-56) ascribed one of the major difficulties of large, flat humid
tropical areas to too little erosion of the highly leached surface
s.il. Ives (19.1) recognized leaching as more important than erosion
in Costa .ica and recommended that practices for the establishment of
:.rganic matter to minimize leaching in surface layers would be more
fruitful than erosion control. S-ith and Abruna (1955) thought that
>jerti 'ican soils appeared to be two to three times as resistant to
erosion as soils in the temperate zone. Although tropical soils are
much more resistant to erosion than temperate soils, the relatively
higher rates of rainfall within the humid tropics make erosion an ever-
present menace on ste.-cly r: in- lands.
Detailed studies of te effects of various soil covers on
erosion and runoff for steep slopes have been made in several areas
of high rainfall tropics. Vicente-Chandler (1953), working in Puerto
'ice, reported that runcff from plots on a 40 per cent slope with an
anr.ual rainfall of 71 inches was more than ten times less when cane
trash was left as a mulch than when it was burned. Molasses grass
was also much more efficient than burned cane trash in erosion control.
r.ith and Abrua (1955), working with red clay latosols of 40 to 53
per cent slope in Puerto Fico, found that natural (bare) fallow soil
lost an average of 26 inches of water and 2553,000 pounds of soil per
acre per year over a six year period, whereas plots with cultivated
crops lost an average cf 5. 7 inches of water and 4,400 pounds of soil
in the same period of time. When cane leaves were used for mulch in-
stead of burned, soil loss, which was 15,240 pounds per acre per year,
decreased to 1,370 pounds, though runoff remained the same. On red
sandy oams with a 6.6 per cent slope in Tanganyika, Van Rensburg
(1~5'i f:uni that the average runoff from storms for eight seasons
decreased from 31 per cent on plots with sorghum to 15.5 per cent on
lots with narrow grass belts across the slope. Runoff was further
decrease. to 11.5 per cent when the top half of the plots were
cul'Lva'el and grass was left on the lower half. The least runoff
of 4.9 er cent occurred under a full grass cover. In an experiment
by F'are: de :astro (19E,) in Colombia, burned plots had only 44 to
4. per cent f the runoff of cut t b utunurned plots, though the soil
losses. on r?-e unburned areas wre from 76 to C- per cent of the burned
field i;. In another experiment conr.aring erosion losses on bare roil,
pasture and coffee under an annual rainfall of 100 inches, Suarez de
Castro and Rodriguez (1955a) reported that nutrient and soil losses
were much greater than those observed for agricultural soils of tem-
Literature on the effects of shifting cultivation on erosion
is scarce and factual data is almost nonexistent. Freeman (1957) ob-
served in Sarawak that when virgin forest was cultivated for one season
only, erosion was slight. He suggested that erosion might become
serious when land is farmed for two or more years in succession or
when the cycle of cultivation becomes too short. Cater (1939) and
Hardy (194:) described the accelerated erosion on the hillsides in
Trinidad aggravated by the large increase in the amount of shifting
cultivation. For Central America Cooke (1931) listed some of the
factors which he considered caused the decline of the Maya Empire.
The-" were: (1) erosion of the soil and the consequent scarcity of
arable landL, (2) silting c"f the lakes and the destruction of water
transportation, and (3) diminution of the water supply during the dry
season. H-wever, charterr (1940), who did a soil survey of British
Honduras, denies emphatically that large-scale erosion ever took place
in areas occuried by the ITaya.
TL rP'LjDCHIC V.:LL.Y OF GUATEMALA
The Folochic Valley is one of two main river systems which
!rain the eastern slope of the highly elevated Guatemalan uplands
int. the '-aribbean Sea. It joins the northern Department of Alta
'erapaz with Lake Izabal to the east. Lake Izabal is a large body
of fresh water, 27 miles long and 12 miles wide, which is accessible
I- river boat from the seaport of Livingston. Throughout the river's
lergth of 65 miles it twists and turns, dropping sharply from 4,000
feet to near sea level in the first quarter of its course. During
the brief dry spell in March it is a pleasant, peaceful stream, pro-
vi lig water and transportation for the residents along its banks.
duringg the suner rainy season, it becomes a savage giant, gouging
anr. chewing away vast pieces of land, and recklessly spilling out
cf its bed onto the floodplain where it rapidly devours cornfields,
laboriousl;.- cleared from the forest by patient farmers. Afterwards,
it races on to Lake Iabal. There the turbulent waters are calmed
and iare .e is of silt are deposited, rapidly expanding the swampy
-elta. The river is younri and evidence of the dynamic state of its
ev l.uti-.r: is reflect! in the rapidly changing landscape.
7'.e topograph:- the area surrounding the Polochic Valley is
excerti n;l;:,y r'oug:h; ry few people have scaled the peaks of the
ruege' i:rra de las Mins which forms the southern boundary of the
valley and separates it from the larger valley of the Motagu- River
to the south. The mountains north of the Polochic Valley are less
rugged; and much coffee, as well as corn, is raised in this area.
Since most of the corn is produced by shifting cultivation, the result-
ing effect on the landscape is a patchwork mosaic of vegetation, con-
sisting of forests, cleared fields, and young second growth.
Some semipermanent agriculture is found in the valley bottom
along the rich floodplain of the Polochic River. Pasture, corn,
sugar cane and many other minor crops are grown almost continuously.
Further up the slopes are the coffee farms, clinging precariously to
the rteep lands. Surrounding them on the poorer soils are the corn-
fields interspersed with forest and second growth.
The soils of the valley bottom consist of rich, deep allu-
vium, derived from limestone and shale. The soils of the steep foot-
hills are more shallow, but in many cases are very fertile due to the
ever-reneving influence of mild erosion. The hill soils clearly
reflect the influence of parent material: soils overlying limestone
at a very shallow depth are calcareous in nature while other soils
are much more acid.
The heartland of the Department of Alta Verapaz in Guatemala
is an elevated area of rugged karst topography which Stuart (1950)
called the Meseta de Coban. Iear the southern end of this tableland
lies the town of Tactic at an elevation of 4,954 feet. The Polochic
River begins just east cf Tactic and, after plunging sharply down a
steerp-sidei gorge, meanders along a flat valley flor which is more
than 6 mileF wide at its eastern en]. The Polochic River at its
western extremity separates two parallel mountain ranges which have
an easterly, trend--the Sierra de Pansal on the south and the sierra
de Xucaneb on the north. The Sierra de Xucaneb extends eastward into
the 'ierra Tzalamila which reaches elevations of about 6,50': feet.
Facing it on the southern side of the valley is the large Sierra de
las :inas, towering up to 9,000 feet.
The Sierra de Pansal consists of limestones of Permian age.
T.ese mountains were the result of folding and faulting along an
east-west axis during the Pliocene orogeny. Much of the range was
mapped by ?oberts and Irving (1957) as Chochal limestone and dolomite
vlc. they assigned to the middle Permian. This formation is equi-
valent to .:apper's ";arbonkalk" which he (1937) had earlier classified
as _arboni:'erous. In a few places the underlying Santa Rosa formation
(.2arly Pernian) crops out, exposing strata of interbedded limestone,
congl-.erate, marl, shale and sandstone.
The lower slopes of the Sierra Tzalamila are composed of
un'jifferentiated Pre--ambrian metamorphic rocks such as schists,
gneisse-, rr.yllites, quart"ites, and marbles. The range is topped
by 'hoc .al limestone which forms a characteristic, sharply-defined
escaromeint al-ng the southern margin, which can easily be seen from
the ?cl-c..ic Valley bel-w.
The climate of Alta Verapaz is determined primarily by the
effect of the mountains on the trade winds and occasional norther-
lies. The "trades", coming from the northeast, bring much precipi-
tation during the summer months. The mountain slopes facing north
and east receive the greatest amount of rain, whereas those facing
south and west are in the rain shadow and consequently are drier.
duringg the winter months, occasional cold air masses move down from
North America, bringing a few days of lowered temperatures and drizzly
Since the Polochic Valley has an easterly trend and is pro-
tected on the north by the high Xucaneb and Tzalamila ranges, the
winter months are relatively dry, because here the northerlies can no
longer produce much rain. During the summer months most of the rain
is produced locally by fast-moving convectional storms which move up
the valley from the east. The rains generally have a high intensity
and are frequently accompanied by violent displays of lightning. The
lower slopes of the mountains receive more rain than the valley floor,
though the orographic effect Jiininishes on the upper slopes.
Almost the entire Polochic Valley has a rain-forest climate
except for some of the smaller protected valleys and leeward slopes.
Precipitation values in Table 1 are for stations in close proximity
to areas sanrled in this stu.:-. ,De data has been compiled from
appearr (1932) and from unpubliJ.he rainfall records collected by the
Institute Meteorc lc'gico de Guatemala through the cooperation of owners
SN CNNN W C' LMr
So 1 t Cji C
*-. r4 o i) car-4 C
--I r-4 r-4
co w Ln *r I
f-4 H r-4 r4 r-4
w o' ") 4 U N ID N Cm ONC
N 0 0 cj 0) m IN 0) tO 0
NODC" V) WV -q t O0)O
rH r Nq wiV) :; ':NrW
0D %f u) ') o- 0 0D to 0) LO
LO) LI) N C V ) t- w
H Ncv CO N 4-i
0 CO C4CIia N tO 0) COC) t % CO
1-4 r-1 M cr ( e-4 r 4
c 4 rO4e to r a ll
u qr U) N Co N4 H
I0) '0 r -4S ) C) N C t
2 N K) CJ U) to NC C- r D
*P S 0
Co n (0 to to L O rO rS td
t I ,1 4I I I I
C, rN, aN r4 j N W
m CV CvN r4 Cv i H H
of the various farms. The geographical relationships of the stations
may be seen on the map in Figure 1. All stations have an annual rain-
fall of more than two meters and as much as five meters falls in some
areas. Months with less than 50 mm. precipitation are extremely rare.
Large variations occur between stations because of the orographic
effects of the mountainous topography. Trece Aguas has an average
annual rainfall of 5,617 mm. whereas leas than half that amount
(2,256 mm.) falls on Westfalia.
According to Sapper (1932) the annual mean temperature at sea
level is approximately 26 to 26.50 C. and the temperature lapse rate
for Alta Verapaz is 0.6100. Thus for each 130 meter rise in elevation
the average annual temperature decreases by 0.610C. At elevations
higher than approximately 1,400 meters, occasional frosts are reported.
The coldest months occur during winter. Data in Table 1 are unpublished
values of maximum-minimum temperatures collected by the Hempsteads for
three of their farns in the Polochic Valley. At sea level the differ-
ence between winter and summer minimum temperatures is approximately
60 C., but at higher elevations the variation is reduced.
Since the lowlands of the Polochic Valley have higher temper-
atures (and also less clouds and more sunshine) than the surrounding
7l-pes, corn growth in the bottomlands is much faster than at higher
elevations. As previously mentioned, two crops of corn may be har-
vested in the valley proper during a growing season, but at higher
elevations twelve months may be r.eeded for one crop of corn to mature.
This is especially, true in the mountains above Tactic.
The dominant vegetation type of the Polochic Valley is the
tropical rain forest, as one might predict from the climate. How-
ever, changes in topography and soils produce local modifications in
the plant cover.
Stuart (1950) notified Sapper's original classification (1903)
of the Alta Verapa: vegetation into five zones: (1) lower broadleaf
forest, (2) lower savannas, (5) u-per broadleaf forest, (4) upper pine
forest, and (E) cloud forest. Of these, only the lower savanna zone
is not rrcperly found within the Polochic Valley.
Pine forest (Pinus caribea and P. oocarpa) occurs in scattered
patches throughout the Polochic Valley, generally on the more sandy
soils or on serpentine found near Lake Izabal. At one time pine may
have been more widely distributed as Cook (1909) indicated:
In eastern Guatemala the secondary character of
supposedly primeval forests is shown by the fact that
pine roots are often found in the ground in districts
from which living pines have been completely driven
out by the more luxuriant tropical types of vegetation.
The Indians dig up the pitchy roots of the extinct
pines and use them for torches. Such roots are found
in the alluvial bottom lands of the Polochic Valley,
near Panzos, almost at sea level, and also in the
coffee district to the north of Senahu at altitudes
of about 3, 70'j feet and upward. Both localities are
distant several miles from any living forest of pine.
.incf no areas under pine were included in the present study, the dis-
cussion shall be restricted to the three broadleaf groups listed by
The extent of the lower brca.leaf forest coincides with the
range of the Corozo or Cohune r;L (Attalea cohune), which is restricted
to -levati-. bel-. '- ters. 'an:.. _. -':* r.i: r'r-,:.,l and
r .:.er (' "illeia *el:s.ti are addiir:nal charact-ristic species.
I':i. coerc. aln c.cc.'l :-- a !i:-.i:znnt position either cause of its
I-t, r a;* t tion ithin the -ir.ax form-.i .., or because it is selec-
t.!'ily r-" :tei !:,' he 1 .- itants cf the area who value it because
: e ht, -- c-r.tr t the seed an o protect it wi-en they make
their cl.'i :-s in the reat. ror.inraji- of the corozo palm has not
been ir.t-.:" i:v.l,'.' T .-'--, -.' .'tevensron (1928) believed it was a
.X-, r' i -r. wv, Hw rtlett (1936) believed it to be sub-
-'.L.a<, 'te Lr.a.i: -?ir,.- :,. -3h'gany-sapoJilla forest. At present,
lir.;- as are now cove red :Inost pure stands of the paL-.
.:'.r-' it" n~ rn, aK:r of '. ia-. among the forest trees in this forma-
tion ar r .: ie same as described by record d and Kuylen (1926)
"'-r te "-" '-,- .'.i: wich runs parallel and ad;'acer.t to the Polo-
. ... *.- s :
S .erer are no -- stan's of nin-gl .~[ecies, but
tn tree- w ich ~r:-..1-:.I rise above the others are
l. t. .*- :.. in number. Dr. !. N, -.Thitford, who
visited the r'r?: in i 1919, estimated that three-
'- stand was composed of five rpecies,
nel r, itarindo (i ilL.. divaricoatur.), narsano
( *l ~ia obovata. asico ("r,.ai:.. terrab'anur;),
-' i-l1 rahbui) .ar.n-! ceiba "eiba
'* .t : --':!- f 'rest extends up to the lower limits cf
i;-.'e:, e f'--e c'ul.,'atin a:d at one time was r,-.bably covered with
rain fore. ':..-*i:r.7 c' '.'.'-ion has "on1n.iderrly modified the vege-
taton and at :r-r,-".' *.*r little true virin :-rest exists. ,
r.: ... has -,..r,:. ,ic. .j -.-,.--. <4nts, :-; : t w.ere.
ground water is near the surface, for the dry season, though short,
is quite sever-. In areas of rocky limestone outcrops, the formation
assumes a much more xeric arnearance.
'nderr 'rth! in the lower broadleaf forest is scant. )ne can
Lasaallr walk a:-.:ng the trees unhindered by ground cover. A profusion
:f cr;.epers and herbaceous plants usually indicates that the forest
has been cleared rece-tly. ELi-hsytes are plentiful though not as abun-
dant as at high-ier elevati.-ns. lians, though not frequent, are well
re-r.-e:i. ted. T..e trees have the usual buttresses and smooth bark '
characteristic of rain forest trees.
The re.-i n of the upper broadleaf forest is commonly referred
to as the "coffee zone." Since much of the better land is used for
coffee, very little mature forest persists except for areas where it
has been preserved along streams to protect the watershed. There
doesn't -ppear to be a very distinct difference between the upper and
lo wer -roa.lejf forests except for the absence in the upper zone of
several ultratrro. 1i1 species as Attalea cohune, Swietenia macrophylla '
and ?3-tillea elastic. Record and Kuylen (1926), for the Motagua
.'-lley, said: "'':he higher Ir.lands merge into the foothills and while
the same speci,:- r.a- be found, their relative abundance changes, and
r,a.r.n~o and tamarindo become more common. Still higher up, the char-
acteri-tic tree is the "-inta ':aria (Cal.phyllum Calaba)."
The hurid cloud forest is found in the subtropical zone. The
lower bcun r:-. coincides with the upper limits of coffee-growing and
the froFt-fr e zone. This -.-;11i-- occurs above the 4,000 foot contour.
More virgin forest is found in this area since the land has little
value for coffee and the growing season for corn is much longer than
at l,-wer elevations.
The general aspect of the clouJ forest is one of dampness.
.:-ually the trees are env-'loped in a heavy fog or mist, since the
ncisture-la ien winds, si;?c-ing in from the ocean, are lifted and cooled
by the mouAntinr, con !ensin on the upper slopes. Since the air is
saturated with moisture and evaporation is low, the vegetation is well
E:rte. to an environment -'f high atmospheric humidity. Epiphytes
abound in much greater amounts than at lower elevations. Orchids,
br'-.eliads, mosses anJ hyprorhilous ferns grow in profusion on the
trees and fallen logs. Tree ferns are also very abundant. This area
possesses one of the richest floras of Guatemala. Standley (1945) said
trportant tree species were: Liquidambar st, raciflua, Hedyosmum mex-
ic~r.u, and .ngelharJtia gu.,temalensis.
'.Te'-tational successions after primary tropical rain forest
has r reen cut have never been studied adequately. Richards (1952) in
hi. classic bok called this the most serious gap in our present knowl-
e.!ge of the tropical rain forest. He says:
T:cu;-. many tyDes ,f secondary vegetation derived from
I'r'rical Rain Fore t have been -escribeiJ, they have
r. I-m been clsel.- stuJied and very few systematic
-:servatinnE have been made on the successions of which
I.h: ar otag's. It is unfortunate that this should be
so, because no asr'rt of the ecology of the Tropical
'lin 'Frest is 61f greaterr rracti'al value or promises
r-' its of more 'r..eretical ir.r-rtance.
An ir.v' -1.iicr. on f-re t reenerati.-n after clearing might rcpssibly
r-:i i: answers to such ?etr.ns as: (1) .:at are the various
succefifdnal stages from cleared field to climax rain forest? (2) What
is the nature of competition in a tropical rain forest? (3) What are
the pr.oc.sses involved in determining the great heterogeneity of species
found in the climax rain forest?
lot enough time was available in the field to make more than
limited observations of secondary successions following the abandon-
ment of cultivated lands. Early plant species invading recent clear-
ings in the Pclochic Valley were cana brava (Gynerium sp.), bracken
fern (?teri'i'i a.quilinur) and Heliconia sp. It is felt that in the
future a detailed successional study would yield much useful infor-
mation on the rate of rejuvenation of soils after establishment of
second growth and also on the competition among different forest
species. The Polochic Valley would be ideal for such a study since
shifting cultivation leaves a pattern of old forest, young second
Fr-irth and cleared fields lying in close proximity. Thus a large
variety of seed sources are available which provide plant material
for colonizing old fields. In the following paragraphs the literature
available from Central America has been used to indicate the informa-
tion obtained so far which may show the sequence of events possibly
ta':ing place in the Polochic Valley.
Allen (19-.), for Costa Pica, sair.:
A considerably greater number of species invade the
hillside clearings than are found on the flat land
which rar, reflect the better drainage conditions, or
indicate an inherent preference for clay. !illsi-e
cleariib--s in particular seem to :-n through a long cycle,
usually invi-lving a succession of three or more dis-
tinct r -* Ijations before 3:cetl.i'. approximating the
ori.'inal climax ..-sociati.n can be re-estal'liche !.
A :'i:rly tric.-l1, if somewhat -implifie ', succes-
si-rl might befin with a rari 'ly dovelopin,- herba-
ceous ccv'7r, cr:,.-.-:!c of :eurcla,~na and vjri-,u:'
Helicnr.ias and r:ithe-ar, f-llowe for a few years
b:1.- *'.crt-livd tr-'! qsuch as rhr_-Mra lagorum,
'-i:tlniia calabura, and Trema m icrantlha. The-e
t +ri; t b et'r:i 'lai.. replaced by 3oethalsia meiantha
ar:'n i !inonanax :yr- -tcni or an association of
Jc..- .'-i Tm pFarah..-- i and .1ieror;nla tectissima, or
:early ure rEnzcrciati-ns of vari-us species cf
.:r:..ia, or sometimes 5rosinur. utile. There last
:ill probaLly reach the eight of ie general canopy,
and persist fcr nany .ears, being slowly invaded
thereafter by other less specialized forest elements.
,t,.venson (192-) gave t:.-pical species of the early stages of
second grc-'h in British icnu-'ras as )chroma bicolor, Belotia Camp-
,r!.I i, Heli-carpus Dcnnell-'-rithii, Schizolobium parahybum, Cecropla
mexi,- na, rdia alliodora, 3uaz'jna ulmifolia, Miconia M., Inga spp.,
-.a rer.'.r, 'ra, and Trerna s. Hardy et al. (1955) add that the later
.'.-ij ;.' -.rcarndJary forests include only isolated specimens of these
r---: the more recent secondary growths contain Cedrela mexicana
:ad :-r'..:-' lica.tru,, which is typical of secondary forests on
ecies and rate cf growth of secondary forests in northern
'.:'a',nala re given by Lundell (19Z7) as follows:
r. southern Peter. r.ilna clearing abandoned for about
s years re -vert r-erj with secondary forest which
:-.-*ra~l about meters in he:.'".t. 'one of the trees
were as -u.r as cr. in diamter, but the mnairrity
were miler'. r T.-.: e't tres were Vite.y paumeri,
t.. "'- 3r '.t..l.n. ... *', --Ira si-.an a"-, recroria
n, n. "r- v t. :-" -".ee ertF which
". -rat r~it i!, r r P ne il -s for the
: -st f' .- had ee. choked out 1. the forest
:-,'a 'i..r.. : a "i.-a wjichr had been -anted the
previous season, I collected three herbaceous weeds.
Hyptis capitata, Eupatorium pycnocephalu.-, and
e-urolae na I bata.
In Monte Chicbul in Central Pete'n a five-year second-
ary growth after milpa averaged about 8 meters in
height. The dominant species was bitze (Inga punctata ?).
However, Cecrpia sp., Nectandra s.., Matayba oppositi-
folia, Casearia javitensis, and the tall shrubby Cor-
doncillo (Piper sp.) were all abundantly represented.
The floor was entirely free of grasses and other weeds
which typif:, the first phase after abandonment. These
had been smothered by the taller rank secondary growth.
Matayba oppositifolia and Casearia javitensis, dominant
and subdomin.-nt respectively in the adjacent high forest,
will be the important trees in this plot also when the
succession again reaches the sub-climax stage.
Many rain forest trees grow very fast. Richards (1952) said
Ceiba pentandra will grow 12 meters high in 5 years whereas Ochroma
grows to a height of 24 meters and a girth of 2 to 3 meters in 15 to
20 years. Secondary forest in the Polochic Valley was observed to
grow at a very rapid rate. Many trees in four year second growth
reached a diameter of 6 inches and a height of 25 feet. The rate
of growth varies widely depending on the soil.
The conclusions by Har;y et al. (1936) on the relationship
between soil types and vegetation for south-central Trinidad have equal
application elsewhere in the humid tropics and are included here since
they hold true for the Polochic Vlley. They summarized their discus-
sion with the following remarks:
This brief review of the relationships between environ-
mental conditions and the distribution of the different
fore Ft-tyvpe- of south-central Trirniiad indicates that
the main far tor ler i !i -' their occurrence is soil mois-
ture, whose na,'nitude and 6vailabillty depends, not
only on total rainfall and 1.'.rpraph-, but also on the
physical features of the soil-types that are represented
it' in the revi.n. When these three factors operate
in the same direction to confer a high water-surnly-
inv' ability on the soil (heavy rair.fall, flat or de-
rr,.-sed topography, a suitable degree of permeability,
cor, ictance and retentivity for water in the srnil),
the weter t:rpes of forest predor.inate. When they
corIer a low water-supplying ability on the soil (small
r:arifall, slnping topography, imrermeable, non-conduct-
irn,, nnn-reter.tive soil), a dry type of forest develops.
V.ri.-,u interr.e!iite types between these two extremes
occur whenever the three factors act in opposite direc-
tions, or when their respective maMnitudes iininish or
increase .jisproportionately. All the observed varia-
tions in vegetation may be attributed to the interplay
of these three factr,-. Soil-type appears to be sig-
nificantly important mainly with respect to those
physical features that decide its moisture relations.
Its chemical properties and attributes appear to exert
little or no influence, except in-so-far as they affect
the behaviour of soil water. Among the chemical fea-
tures, line and maeneria contents and humus content
alnne appear to be important in this connection. Either
chemical factors (nitrogen supply, phosphate content,
potash content, etc.) seem to be quite subsidiary in
deri .inp the broad distributionn of vegetation types,
alti.-.ugh they may assume important roles when forest
lands are utilized for the growing of commercial crops.
The ropulatir. of the Polochic Valley is very roughly estimated
by the writer to be near 25, -. It is mainly comprised of three
groupF', na: .-l', (1) indigenr,-us inhabitants of Maya ancestry (by far
the most nBuerous), (2) la inos (nixed In Lian and Spanish blood, rap-
i I/ inr-. sin,-), an;' (3) the few owners cf large coffee fincas
(farr-), sometimes of iauc ..-ian ancestr.,. Although most land belongs
to e la'-t grour, any memFer of the other two groups may also be
Two linguistic groups of Maya stock are present in the Polochic
Valley--Kekchi and Pocono. The Kekchis are much more numerous. Kekchi
women characteristically wear a white cotton blouse (hui;il) which in
same places is replaced by rayon of various colors, though the style
remains the same. The Pocono women can be distinguished by a red tur-
ban and heavily embroidered red blouse.
Many ladinos have filtered into the Polochic Valley from the
Guatemalan highlands in recent years, following the completion of a
road in 1948 and improved malaria control. The term "ladino" is often
used loosely in Guatemala to refer to anyone of mixed blood or to those
who do not strictly adhere to the customs of their tribe. One informer
said that a "ladino" is anyone wearing shoes. The ladinos are employed
as skilled labor (truck drivers, overseers, etc.) on the coffee farms,
or are self-employed in small businesses.
Owners of large tracts of land represent a very small minority
of the population. Formerly many were Germans, but deportation of Nazi-
sympathizers and expropriation of their lands during World War II de-
creased their numbers. Many of the German-owned fincas were nation-
alized and are now operated by the Guatemalan government.
Large Indian villages are few. Instead, the Kekchis live in
i-..ely scattered houses. The family is the basic unit and most settle-
ments consist of one house only. Large families may occupy two or three-
ne i A'hbr.ring houses. occupation n of houses is permanent--no large shift-
ing of population occurs. A person may be born and die in the same
house. However, an expranr1ing population in the last few years (possibly
attributable to the intr. ''-tl:-r of modern medicines) has caused some
pcu.1lr.tioi. r-ssurc anT ': n-J.:untly, some Indians have migrated to
less crowded areas.
La:,.. near the houses is used fcr growing a wide variety of
docrvard crops, ,-rne r-:ular garden pl .nts are: achiote, beans, cas-
sava, cha.,'yte, corn, T:t.'s-tears, pigeon poas, squash, sugar cane, sweet
potatoes, taro and Xanthosons. Cropped land further away from the houses
is Joninat. by corn. Trec .:-ership, often found in the South Pacif-
ic, is not iart of the Keklchi ruilture. V.hen a house site is abandoned,
the trees r-a:. be destroyedd by the former occupant. Another cultural
characteristic of Kekehi agriculture is that women almost never work
in the fieli-; their chores are restricted to household and domestic
T;.r. main t:.yes of shifting cultivators are found in the Polo-
chic VaII.:, na-mel., (1) landowners, (2) coffee finca workers occupy-
ing land in return for' services, and (5) squatters. Indians who have
ownership of the land they far are few and are found in more isolated
areas. Squatters may or rma. not have rights to other land in addition
to that which is -ccupi'e by s.latting.
he -argest amount of land under !. iftini' cultivation in the
Folcchic valley is occuri-.! by Indians who have part-time employment
on the coffee fincas. T.i land is owned by the large fincas and in
return fr use of the lan the cultivat-r must work part-time. This
type _f cul'.iatr .ay be divi !ei intt three grnupo--l.-cally called
"ccl:r.-.=," ",:e.rr," and "r.*-adrillal"--on the basis -f the amount of
time spent working for the finca and the type of remuneration in return
for their services.
The "colonos" work about half-time for the finca and receive
no payment in return. Instead, they occupy enough finca land to sat-
isfy their needs when used for shifting cultivation. Boundaries of
assigned land are defined by usage and are well-known (the upper Polo-
chic Valley has no land survey). The "colonos" are mostly concentrated
in the lower rain forest zone but some occupy land in the coffee and
upper cloud forest zones, though less land is available there. All
sites studied in the present investigation were farmed by "colonos"
with the exception of the highly fertile site, Vega. The available
land which can be used for shifting cultivation is near saturation
and the fincas discourage colonyo" occupation.
The ",,eseros" reside in the coffee zone. They work the greater
amount c f time for the finca and in return receive small pay, rations
(2E pounds of corn per week per couple), house, and a small piece of
land for corn and gar-den crops. The "mesere" works six days per week
for the finca but is allowed two weeks vacation for clearing land, one
week for planting, nine days for weeding, and nine days for harvesting.
In addition, he has five national holidays and three finca holidays per
"Cuadrillas" occupy finca-owned land in isolated areas, far
removed front coffee production areas. They travel to the finca in
large groups, when labor is most needed, and work a few weeks per year.
The remainder of their time is spent farming land, assigned for their
use, with traditional shifting cultivation methods. Land occupied in
this manner is mostly in the upper cloud forest or coffee zones. The
type of farr.ing used by the "cuaJrillas" probably differs less from
ancestral m,-thods than the type used by "colonos" and "meseros."
The :writer believes that food production by shifting cultiva-
tion in the Folochic Valloy is a limiting factor now in population ex-
paxi-ion there. If meiir'l improvements continue to increase the pop-
ulitiAn grc.4t'., more e'firient methods of agriculture will be needed.
The shifting cultivation cycle in the Polochic Valley is
relatively shrt and is t-pified by the use of a single cultigen--
corn. The land is cropped for as little as one year, following clear-
ing and burning of the f.-rcst, before abandonment to second growth.
rirst year corn yi3las are reported to be about ten to twenty bushels
p.jr acre wile fields s for a s.-'cnd crop, if planted, are often reduced
by -ne-hal:. Second grr.t.-t is allowed to develop for an average of
f-ur years bef..re it is cl-'ared for another crop of corn. The average
size of clpare fields varies from one to six acres, though much more
lan,! is c'-urie-?d by second growth and old forest at any one time.
r'ield size is governed by the fertility of the land, previous use of
the area. t;.pe of fall-w v.-ietation, and number of persons in the cul-
tivatr's h.u-eh-Jl. .'in.- the fields are usually small, several
states of shifting cultivation may .ioin in an area of relatively uni-
forrr. soil, -i.ich is a -nr.venient situation for a tu.idy, of the effects
of 2hiftr.; c.ltivai.r. on soil properties and veretation.
Usuall:r, land which is selected for clearing supports a good
stand of second growth or old forest. The nature of the vegetation
is important for several reasons: rate of second growth rejuvenation
and type of vegetational succession indicates inherent soil fertility;
fast-growing, woody vegetation crowds out many of the more tenacious
herbaceous weeds, requires less labor for clearing, and burns with a
hot fire. In this respect, cania brava (Gyneriurn s.) is one of the
most desired types of second growth vegetation by the shifting
Land to be cleared is usually marked off at the beginning of
the dry season. The selection of a site by the shifting cultivator
is usually limited to the confines of land owned by him or belonging
to the large landowner for whom he may work. Usually clearings are
located near the home, but some Kekchis use land that is three hours
walking distance from their residence. Cook (1921) described one in-
stance of Kekchis going sixty miles to farm and covering the distance
on foot. In such cases, small huts are erected in the fields and the
farmers live there while the crop is planted and harvested.
'nce the site has been selected and the boundaries marked,
the next task facing the indigenous farmer is that of felling and
clearing the forest. Large trees are usually cut with an axe, leaving
about one meter of exposed stLir. The smaller trees are cleared with
the machete. Much -learing is done in the Polochic Valley during the
month of March at the height of the dry season. This allows the brush
to dry out thoroughly for burning before the onset of the summer rains.
However, Steggerda (1941) found farmers of the drier Yucatan Peninsula
clearing in the rainy season because, "The trees, full of moisture ire
easier to cut and, covered with leaves, insure a better burning when
dried." The farmers, when clearing forest, probably have a selective
effect on the vegetation by preserving the more valuable trees.
Lundell (1937) found that when old forest was felled in northern Gua-
temala, the large valuable trees such as corozo and thatch palms, za-
pete, and raon were spared. Firewood and usable lumber, in most cases,
is carried from the field before burning.
The felled brush on cleared land is burned shortly before the
first rains in Arril or Ma;,. This clearing when planted is called
"r-ilpa de fuego" (fire cornfield) in Spanisi, in contrast to the "milpa
de verano" (iry season cornfield) which may be planted afterwards in
part of the same fiell. During the time of burning the atmosphere over
extensive areas becomes darkened with smoke, seriously limiting visi-
bility and irritating the lungs and eyes. The burn must be well-timed
for, if too early, weeds will sprout before enough rain falls to ger-
minate the corn; if too late, new growth and wet wood seriously hamper
an adequate fire.
The winter saw very little evidence in the Polochic Valley of
fires escaping from the confines of clearings into the fire-resistant
rain forest. All burned areas, vi.th the exception of one, observed at
the end cf the ir. season appeared to follow the straight lines of
fiel boundaries, Nev rtheless, several local residents wre familiar
with eccasi-r.s wen fires had escaped, causing great damage to the
adjacent forest. In the nearby region of Peten, where the annual rain-
fall is less than in the Polochic Valley, escaped fires are much more
frequent (Lundell, 1937),
Great emphasis is placed by the natives on securing a hot,
uniform fire. Thus, they want the brush well distributed over the
land and as dry as conditions will permit. On the Pacific coast of
Guatemala, if enough brush is not available for a good hot fire, com-
bustible material nay be brought in from adjoining pieces of land.
Apparently, the beneficial effects of a good fire are threefold. First,
seeds and vegetative material which could give rise to a large crop of
Seeds are partially eradicated as, no dcubt, are some animal and in-
sect pe-ts. Second, soil structure is improved by a hot fire (Budow-
ski, 1'56; Burg~ and Ccctt, 1952, 1955; Scott, 1956). Third, plant
nutrients are converted into readily available soluble salts from the
oxidation of organic matter.
Planting in the Pclochic Valley commences immediately after
the beginning of the rainy season. Throughout the Maya area of Cen-
tral Acerica and Mexico, corn is the main staple and is the crop com-
monly planted in f7-rest clearings. occasionallyy other crops may be
grc.rwn with the corn, especially when -lanted near houses. Use of a
single cultigen, corn, simplifies the task .f measuring effects of
shifting cultivation on the soil. Cook (1921), one of the earliest
men to work on the problem of shifting cultiv-ition in central -nmerica
said: "Tie rroble- of tracing relations between aEri'riture and nat-
ural conditions in Central Are-ica is much uimrlifle' by the fact that
one sta-le crop is rroru over the whole area."
One characteristic of shifting cultivation in the Polochic
'Vlley, as in most other areas, is that absolutely no tillage of the
scil is involved. So.od is planted directly among the ashes and charred
remnants of the fell! forest. A long, pointe., wooden stick (coa in
Spanish, aulep in Y echi), the styl3 of which probably hasn't changed
in the last three thousand years, is used to make a shallow hole into
which the seed is dropped. In the flat lc.v:lands, where the soil is
more compact, points of the digging stick are reinforced with steel,
or are harilned by drying over a fire. The planter makes a hole with
the stick while standing, and then drops five or six kernels of corn
before closing the hole with his foot. The seeds are planted at inter-
vals of apnrrximat-elY three to four feet.
The amount of weeding necessary during the crop season depends
largely, on the previous history of the field. Land recently cleared
from high forest requires relatively little weeding; land cultivated
fcr several years in succession has to be almost continually weeded
tr.r:ughcut the grc.-irn season. Sometimes grasses become dominant in
a fiel- used ccntiuCu:.-ly, though crop production ceases in most fields
before tU.is stage is reached. Where grass does appear, it is quickly
era licate when the field is abandoned by the farmer to second growth.
Loservatrons of the writer while in Central America lead to the con-
clusior. at the importance of grasses as weeds in limiting crop
production. is less above 2,000 feet than at el-vs'ions near sea level.
.re of the most important ways in which invasion by grasses
liMits cror producti-n is through the accompan-in. increase in rodent
population. The deflected succession from forest to savanna by burn-
ing results in a habitat much more favorable for rodents. During the
early years of a grassland, the rodent population undergoes a very ac-
celerated growth. When grass is plentiful in cornfields, this effect
is reflected in the large number of young corn plants either killed or
injured from physical damage inflicted by rats and mice. After a grass-
land has been established for a few years, the number of rodents builds
up rapidly to a peak and then declines as natural enemies appear on
The invasion by bracken fern (Pteridium aquilinum) in shifting
cultivation lands is a more important problem than grass at elevations
above 2,000 feet. Bracken may interrupt the normal vegetational suc-
cession and, once established, reduces the usefulness of the land for
corn. Shifting cultivators and finca owners try to rehabilitate such
areas by encouraging other types of vegetation. Bracken may have some
effect on soil conditions but the actual mechanism is unknown by the
Weeds are generally cut close to the soil with a machete.
Morley (1956) blamed the machete for the large number of weeds in
Yicatan cornfields. He hypothesized that before the machete was in-
troduced, plants were pulled by hand, and scattering of seeds was less.
:'owadays, because wee competition is more severe than formerly, he
reasoned, fields remain in production a shorter period of time.
Hoes are also used for feeding in the Polochic Valley, though
much less than machetes. The hoe is employed strictly for cutting
weeds above the soil surface and, unlike drier areas of Guatemala, the
soil is never disturbed The natives hoe more frequently during the
dry season than in the rainy season. It is quite possible that hoeing
accelerates the process of erosion in high rainfall climates and for
this reason is not more intensively used in the Polochic Valley.
The Kekchis never use draft animals and have not incorporated
large livestock holdings into their agricultural system. Furthermore,
the steep topography and rocky land discourages the use of plows. On
similar land elsewhere in Central America oxen have been used to pull
wooden plows but, no douot, this hastens erosion of the soil. In some
of the more fertile floodplain soils of the Polochic Valley, where per-
manent agriculture is practiced, crop production has been maintained
almost continuously through the use of heavy machinery to till the soil.
However, these soils have a high inherent fertility and, without culti-
vation, weeds would be the main limiting factor to a permanent agri-
Cornstalk bending or doubling by the farmer, common through-
out the Maya area, is not practiced in the Polochic Valley. In other
areas the farmer believes that birds find it more difficult to eat
corn frnm an inverted ear and that, otherwise, rain water collects
beneath the husk and promotes mold. The fact that cornstalks are
almost never bent in the Polochic Valley might invalidate the rain
water hTpothesis, for the climate is certainly wet. However, the
native corn husks are approximately three tines thicker and much
tougher than introduce i varieties. Corn groim from imported seed is
much more suscec.ible to mildew during the ripening stage. Another
advantage of the local thick-husked corn is protection against weevils,
possibly the most important crop pest.
Harvestei corn is often stored in the farmer's house above the
cooking fire i.here it is kept dry. Smoke from the fire probably dis-
courages depredations by weevils. Many times corn may be left unhar-
vested for a long period of time in the field.
After corn is harvested in the lowland areas, a second crop
(milpa de verano) may be planted imru'ediately in part of the same field.
Or, the farmer may clear a second field to take advantage of the long
growing season. At higher altitudes, where the corn takes longer to
mature because the average temperatures are lower, only one crop is
harvested each year. In the mountains above Tactic one crop of corn
may take more than a year to reach maturity.
The same field is rarely ever planted more than two years in
succession. The period of fallow, during which time second growth
covers the land, lasts from three to four years. As one proceeds
northward through Peten and Yucatan, the length of fallow increases as
the country becomes progressively drier.
The shifting cultivation employed now by the natives of north-
ern centrall America and southern Mexico is essentially the same as used
by the Mayas 2,30' :.rears ago. The Maya Empire subsequently declined,
and many authors, foremost among which was Morley (1956), have suggested
that the decline was a result of the failure of shifting cultivation to
withstand the rressures of an in-reasing population. They hypothesized
that as the population increased, the time that land could be left
u:der forest fallow decreased, and eventually the land became unpro-
dutive. More recent studies by oth2r authors, of which Thompson's
(1354) is cne, have suggested that the Maya Empire disintegrated when
the peasant class, oppressed by the burden of supporting an overlarge
hierarchy finally revolted. The large number of broken monuments at
some of the major ceremonial centers (Tikal and Uaxactun) appear to
support this theory. Supposedly, the large agricultural areas of
Peten and Yucatan gradually reverted to forest after the revolt, since
a high form of civilization and organization was needed to successfully
farm rain forest land in a humid tropical climate.
The answer to the above riddle has far-reaching significance
today. Now that we are once again trying to utilize large rain forest
areas for agricultural purposes, important questions arise. Can humid
tropical lands on old soils be farmed successfully under any system?
If a failure of agriculture caused the decline of the Mayas, what were
the major limiting factors--soils, weeds, or pests? If the Maya Empire
.isintegrated as a result of a peasant revolt, why cannot the same
lan.is be made productive today with our modern technology and more
advanced civilization? The ecological basis of shifting cultivation
nust be -tulieJ before changes in the system can be considered.
Description of Sites
Shifting cultivation fields in the Polochic Valley were
selected for study to represent as wide a range of climatic and soil
conditions as were available in a limited area. The area chosen for
the present investigation was midway between the ends of the Polochic
Valley near the town of La Tinta. Sites were sampled from the valley
bottom, near sea level, to an elevation of approximately 5,000 feet on
Sierra Tzalamila. The sites included the major geological formations,
annual rainfall values varying from 2 meters to more than 4 meters,
and very acid soils with a pH near 4.0 to soils with a pH higher than 7,
Samples of soil for chemical and physical analyses were col-
lected from thirty fields in eight sites within an area of approxi-
mately one hundred square miles. At least two stages of shifting cul-
tivation adjoined in each site. The time of sampling was during the
ni idle of the cropping season. One field with an apparent high yield
potential in the floodplain of the Polochic River was sampl-ed as a
reference. 2ite locations are shown in Figure 1 and descriptions are
given in Tatbl 2. More detailed descriptions of the mode of access,
sells, stages of shifting cultivation, and other factors are given in
the following paragraphs.
The Los Flpes site is on the north side of Los Alpes-Tzalamila
trail approxirr.at.:ly one and one-half hours walking time from Los Alpes,
4. 43 1+
o4I (a a o
O 0 O
5 0a-i 0 h O
n.I I r-4 I
Cy.-4.-i ou ww 0
+) ) a)
'- 0 4 r- 0
M 0 Cd 0 0
co () ( )
0 4a 0 -4
0 > I
b 4 +3
0 3 r fiUT ~
I'ed 0 -H
I *r144 I 14 ri
CL, eN Qi a
0 :3 O
a) rri 4
r-H N c
4'Vl cI <
0' C )
0 0 4B 0
0 WO c
10 bO h
C 0 E-
0. 0) 0 10
0) r- 4
-t ** V 14
CO + 0<
11 t4 b H
Q. (-0 pi
* c B
0 SD r-i
+0 0 IH
0. 4' $4
4 ** 0
0 *r aH
C 0 W 0
0 H cy 0 *
t -o I a
i-l B t
C M 61. 0 0
r-l M "
< i :> (
and one-half km. south of the pass near La Tinta, A.V. .oils consist
of ?C cm. i' f a friable, dark reddish brown clay loam with good crumb
structure overlying a yellowish broi-n, sticky plastic clay with weak
blocky structure. Organic matter accumulation is great because aver-
age temperatures are low throughout the year. In some places leaf
nold is 25 cm. deer. Corn matures in eight or nine months because tem-
peratures are lo:. Slight nitrogen and phosphorus deficiency symptoms
were apparent on young corn plants. Tree roots were very highly con-
centrated in the top layer of organic matter; corn roots in cleared
fields appear to feed deeper in the soil.
Fields at Seamay are adjacent to the west side of Seamay-
H'sario road 17. km., from -eamay, A.V. Surface soils are friable
brown clays with good crumb structure overlying sticky; plastic, red-
dish yellow clays at 20 cm. The top soil has been mixed in many places
by leaf-cutting ants. Some nitrogen and phosphorus deficiency symptoms
were apparent on the corn plants.
Matanzas I is adjacent to the west side of La Tinta-Westfalia
trail, approximately 50 m.. west of the confluence of the Panima (lo-
cally call-d Matanzas) and Sw-dilja rivers, ind approximately 500 m.
north of bri Ige across the Panima river. Soils are friable, dark
reidish brown loams with well-developed crumb structure overlying a
sticky., yellowish red clay at a depth of 20 cm. The soil profile
appear. truncate- in some places. Decomposing limestone is found at
deptns -f 3. cm. and r.anr. IL-estone cutcroppings are present. Some
leaf-cutting arts had -illed a large part of the broadleafed vegetation
in the two year old second growth.
Fields of M:atanzas II are one-half hour walking distance west
of La Tinta-.-.:e-falia trail approximately midway between Panima and
Poloc.Lic river-. Five cm. of dark reddish brown clays with crumb
structure overlie brown clays with weakly blocky structure. Lime-
stone -utcroprings occur throughout the area. Some phosphorus defi-
ciency .ymptor.s were apparent on the corn plants, though otherwise
the plants appeared quite healthy.
Actela fields are 100 m. north of Mocca-Acteld trail and adjoin
the northeast side of the first stream (Tampona) east of Ermita San
Miguel, near La Tinta, A.V. Dark grey brown sandy loams with good
crumb structure overlie yellowish brown loans at a depth of 20 cm.
Much gravel is distributed throughout the soil profiles. Corn plants
near the forest margin appeared highly deficient in nitrogen, had thin
stalks and lod-ged easily.
The site of 'amelias I borders the south side of Mocca-Los
Alpes trail at the first bend east of the largest river (Palpeha),
approxinately midway between the two farms. Dark red brown loams with
strongly develr.ped crumb structure overlie yellowish brown clay loams
and clays at a depth of 20 cm. Gr-.t.h of corn plants was extremely
r-.or. The corn Flantr had almost no roots, a high incidence of lodging,
and were extremely: brittle. Some eighteen inch plants had produced
ears, though the yield of the "iel- was almost nil.
Fiel-s of Camelias II border the south side of Mocca-Los Alpes
trail by a spring in a cave, four km. from Mocca and 5:': m. east of
Camelias I. Soils are 5 cm. of dark brown loam overlying yellowish
brown clay loams. A strongly developed crumb structure exists through-
out the rooting zone. Most corn plants grew less than eighteen inches
and yields were exceptionally poor.
The site of Vega is approximately 300 m. southwest of La Tinta
airport in the floodplain of the Polochic River. Friable, brown clay
loams overlie sticky yellowish red clays at 30 cm. depth. Soil is very
fertile, has good physical characteristics, and supported one of the
best stands of corn (ten feet tall) seen by the writer in Guatemala.
This soil has been in cultivation for a number of years using modern
agricultural practices and machinery, though no fertilizers are ap-
plied. Corn in the tasseling stage at time of sampling appeared to
hav a yiell potential of approximately sixty bushels per acre. Occa-
sionally the land has been rotated with pasture or cover crops. The
area was sampled as a reference, since apparently soil conditions were
favorable for optinur. corn growth
The site of Cabanas I is on the west side of the Mocca-Cabahas
road, 2.2 kin. from Cabanas near La Tinta, A.V. Five cm. of dark red-
dish brown clays with good crumb structure overlie yellowish red clays
with weak blo.cky structure. The soil profiles appear somewhat trun-
cated. Many corn plants failed at tasseling stage though yield was
approximately fifteen bushels per acre. Some of the brush was removed
for firew-.-d when the field was cleared. Results from this site were
not ir.clu-ed in the discusEion on effects of shifting cultivation on
soil properties, since no old forest was nearby for comparison and none
of the fiel :3 inclu-de two year second! growth. Hrvever, :he analyses
have been used in the sections characterizing the general features of
Polochic Valley soils.
Each stage of shifting cultivation at each site was sampled at
ten random points during the middle of the cropping season, 1956. Pits
were dug and samples were collected from depths of 0-5 cm., 5-20 cm.
and 20-40 cm. by cutting into the soil profile with a large knife.
All depths were measured from the soil surface which included forest
litter in some cases. The thirty samples from each stage were com-
posited into single samples by depths and placed into cloth bags.
Thus, for each single sample analyzed from each plot, ten individual
soil samples were taken, bulked in the field, and well mixed before
subsampling in the labcrator Two of the ten pits in each field were
sampled in the 5- to 20-cm. zone with a 5 x 3.5 cm. cylinder for bulk
density. The air-dried soil samples were shipped to the University of
Florida for analysis All samples with the exception of those for
biological studies were fumigated with methyl bronide upon arrival in
the United States.
Chemical and physical analy-ses
The soil was passed through a 2 mm. aluminum sieve and placed
in ice cream cartons; pebbles were retained for geological character-
ization. Total nitrogen was determined by the A.O.A.C. (1955) Kjeliahl
methd with the modification that ammonia was distilled into a 4 per cent
boric acid solution and titrated with standard hydrochloric acid using
a mixed indicator of broncresol green and methyl red. Carbon was an-
alyzed by the wet combustion method of Walkley-Black as modified by
Walkley (194 7).
Duplicate soil samples of 25 gn. were extracted with neutral
normal annronium acetate according to the method of Peech et al. (1947)
and, after the solution was evaporated to dryness and organic matter
was destroyed at 3500 C. in a muffle furnace, the salts were taken up
in two normal hydrochloric acid. The procedure described by Black
(1955) was used to determine exchange capacity by distilling ammonia
from the soil sample into a 4 per cent solution of boric acid and
titrating with standard hydrochloric acid and a mixed indicator of
bromcresol green and methyl red. Exchangeable potassium and sodium
were jeterrdned in the hydrochloric acid solution with a Beckman Model
B flame spectrophotometer with acetylene-oxygen burner assembly.
Potassium was measured at a wavelength of 768 mu and sodium at 589 mu.
The versenate method of Cheng and Bray (1951) was used for determining
exchangeable calcium and magnesium in the hydrochloric acid solution
after first neutralizing with potassium hydroxide.
"itilleI water was added to the soil samples on a 2:1 volume
basis (1 ral. water, 5: ml. soil) and, after twelve hours, the pH of
the suwrar.ri! n wa measured with a thin glass electrode. One ml. of
one molar :alcium chlori e was then added to each sample and the pH of
the scil in the 0. 1 olar salt s'.lutior. was measured after one-half
h-ur ( l-!.rfield and Tayl-r, 1955). xchangeable aluminum was extracted
from 10 gm soil samples with 100 ml. of normal potassium chloride
and analyzed by the method of Coleman et al. (1959),
Acid-soludle and adsorbed phosphorus, extracted by the method
of Bray and Kurtz (1945), was determined colorimetrically by the pro-
cedure of Truog (1930), Zinc and manganese were extracted with tenth
normal hydrochloric acid and determined by Dr. H. L. Barrows using his
polarographic procedure (Barrows, 1959; Barrows et al. (1956).
Soil samples were prepared for X-ray diffraction by the method
of Kunze and Rich (1959) after organic matter was removed with ten per
cent hydrogen peroxide and the clay had been separated after disper-
sion with 1 ml. of molar Na2EDTA. X-ray diffraction patterns were
obtained by copper radiation with nickel filter (Cuka according to
specifications set forth by Fiskell et al. (1958), with the exception
that 10 ma. current was used instead of 15 ma.
It was decided to obtain an estimate of the microbiological
activities of some representative soils within the Polochic Valley,
since many of the results from the chemical and physical analyses of
soils under shifting cultivation appeared to implicate soil biological
factors. Sampl-e for this study were collected in polyethylene bags
by Mr. C. Hencstead in November, 1957, and shipped by air in moist con-
dition to the Univ:rsity of Florida. The soils entered the United
states without fur.igation. cttal time en route was two weeks.
The samples were collected from each of four fields--forest,
second rroith and two corn fields at depths of 0-15 cm. Samples from
cleared fields were collected near corn plants so as to include as
many roots as possible for analysis of nematodes affecting crop pro-
duction. The sites used for this study were Vega, Cabahas and one
new site, Cabaias II. This last site was located 100 meters south of
the farm buildings at Finca C banas on top of a limestone bluff imme-
diately west of the main road. Vegetation was forest, though somewhat
mcre xeric than the other sites. The soil was a very shallow clay
lcam overl:ying limestone at a depth of 20 cm. The site in Vega was
near the site previously sampled for laboratory analyses, and produced
crops continuously. Corn plants were twelve days old at time of
sampling. Two fields were sampled at Cabanas, namely, one planted in
corn the preceding season, and another in two-year-old second growth.
These fields were identical to the fields previously sampled for chem-
ical and physical analyses.
The rate of release of nitrate from soil organic nitrogen was
measured on the samples in triplicate by the method of Stanford and
Hanway (1-55). Nematodes were separated and identified by Dr. W. H.
Thames1 using the technique of Christie and Perry (1951). Popula-
tion counts of soil micro-organisms were made also on the samples by
ThameF. :.ose bengal agar (Smith and Dawson, 1944) was used for plating
fungi, an.d c-il extract agar (Allen, 1949) for bacteria and streptomyces.
Plating dilution for fungi was 1:1,300 and for bacteria and streptomyces,
-_.,. H. Thamer, Jr., Associate Professor of Soil Micro-
bid-l-gy, Texas A. and MI., collegee Station, Texas; formerly Interim
Instr'-ctor in Soil Micr.-iology, University of "lorida, Gainesville.
1:100,000. Subsequently the samples were submitted to Dr. W. R.
Carroll1 for isolation of organisms producing antibiotics. Special
precautions were taken to treat all wash water, wastes, and apparatus
used in the study with formalin. Soil samples were finally sterilized
in an autoclave after the investigation.
1:r. W. W. Carroll, Fr:"'essor, Department of Bacteriology,
University of Florida, Gainesville.
- 'il T ArJT ;iISCU.'I I3N
The results of laboratory analysis of soil properties at seven
sites and inder three stages of shifting cultivation--old forest,
cle .red lan:, and two-y',ar L;--cond. growth--in the Polochic Valley are
shown under appropriate headings.
Chemical and Mineralogical Soil Factors
The X-ray diffraction patterns for the clay fraction (< 2 mu)
of the 20 to 40 cm. horizons of Polochic Valley soils are shown in
Figures 2 and 3. Interpretations of clay minerals from the patterns
for each site are shown in Table 3. The clay minerals of most of the
Folochic Valley soils are highly complicated. Dominant clay mineral
types in most cases are three-layered with some interstratification.
.'eamay soils, however, are dominated by kaolinite.
The soils with the best crystallized clay minerals have the
highest and sharpest X-ray diffraction peaks (Vega and Matanzas I and
II). Th-.s- soils are probably relatively younger than soils having
more diffuse X-ray diffraction peaks (Los Alpes, Camelias I and II,
Actala'. i.s will be shown in later sections, youthfulness of the soils,
Vega, Mattanzas I, an- II, is also indicated by their high base status.
The -.ler soils, with .egrade:. clay minerals, might be expected to have
a hi;,her fixation capacity for potassium than ;'unger soils.
2 e 0( SLI T
33 30 27 24 21' a 5 5 2
! I I I I I-
S2 e (1/2* SLIT)
Fig. 2.--X-ray diffraction patterns for the clay fraction
from four shifting cultivation sites in the Polochic Valley.
I J j I I 1 I
2 0 (ISLIT
_ I I
CAMVE -4 I
;Ah'F. ALS AS
ALNG T ''.' UNITS
Fig. 3.--X-ray diffraction patterns for the clay fraction
from five shifting cultivation sites in the Polochic Valley.
12 9 6 !
CL Y MINERALS OF POLOCHIC. VALLEY SOILS INTERPRETED
FROM X-FPR, DIFFRACTION PATTERNS
Location Dominant Clay Minerals Accessory Clay Minerals
Vermiculite, illite, inter-
Vermiculite and inter-
Well-weathered, highly de-
graded vermiculite, inter-
Well-weathered, highly de-
graded vermiculite, inter-
Well-weathered, highly de-
graded vermiculite, inter-
silicates and illite
silicates and illite
The three-larer silicates f-und in the Pcl-.chic Valley soils
on limestor:e appear to support Lawton's (1955) and Millot's (quoted
by Grim, 19?) content in that sediinents of marine origin which have
not been highly weathered usually contain high amounts of these clay
minerals. However, more recently, Weaver (1958) states: "there is no
consistent coincidence between specific clay minerals and specific
.Iitrogen and organic matter
The soils under shifting cultivation in the Polochic Valley
are essentially forest soils and contain relatively high amounts of
organic matter and nitrogen, especially in the surface layers (Table 4).
Ty-ical of forest soils, the organic matter content decreases sharply
1Nitrogen content of the topsoil is almost invariably lowered
when forest is cleared in the Polochic Valley. However, the losses
are not nearly as great as one would expect and the average for seven
sites in Table 4 falls from 1.13 per cent nitrogen to the top 5 cm. of
old forest soils to 0.82 per cent for recently cleared land& The
effects of clearing are noticeable throughout the depth of sampling
as re:lecte I by a corrc-lronding drop in the 5 to 26 cm. and 20 to 40
cr.. zoes. These results substantiate similar investigations else,-
where (:-ulter, 195C; -uthie et al., 1936; Giddens et al. 1957; Nye,
195b; iquier, 195Z). 'rganic matter losses after forest removal can
probably be sttributeo to: (1) effects of burning, (2) reduction in
IITROG.. 11ND ORGAIIC MATTER CONTENTS OF POLOCHIC VALLEY
SOIL UNDER THREE STAGES OF SHIFTING CULTIVATION
Derth Forest Cleared Land Second Growth
Location N 0.M. ':N N O.M. CiN N O.M. C:R
Cm. Per- Per- Per- Per- Per- Per-
Average of 0-5 1.05 30.2 15.9 0.74 18.6 15.0 0.76 22.7 15.2
First Four 5-20 0.49 11.4 14.5 0.55 8.5 13.2 0.52 6.5 11.7
Fields 20-40 0.25 4.5 10.9 0.19 5.4 10.5 0.19 5.1 9.4
Average of 0-5 1.13 51.2 15.5 0.82 19.8 15.1 -
All 5-20 0.53 13.5 14.9 0.43 10.5 13.7 -
Fields ?2-40 0.51 6.4 11.9 0.24 4.7 11.5 -
rate of litter repleninrU.ent, and (3) removal of shade (causing higher
scil temperatures). -stablishment of young second growth on cleared
land restores some organic matter in the topsoil without materially
altering the total nitrogen content because the newly added material
has a high C:N ratio.
Values for the 2':I ratios of shifting cultivation soils in the
Polochic Valley are also giv,.n in Table 4. Los Alpes is the only site
that has C:'I ratios in excess of 16.5, and there the ratio does not
exrced 19 for the cleared field. It can be seen from the discussion
in the literature review that these C:N ratios should be low enough
for good mineralization of nitrogen in the soil organic matter.
hearing the forest by fire reduces the C:N ratio of the soil
surface lay:,r. No doubt, burning in this instance is beneficial.
-tnerwise, the freshly cut vegetation would probably immobilize much
rf the nitrogen supply available to the young growing crop. If the
land re kept cleared for a long period of time, the C:N ratio might
be reduce still further. The soil C:N ratio rises under second
growth with the addition of fresh litter with a high C:N ratio.
The average nitrogen content for the topsoil of seven forest
fields is 1.13 per cent (Taol- 4). This value may appear to be very high
when compared with temperate climate soils, though not excessively so
when conr-ared with other areas of the very humid tropics. Values of
the sam magnitude hav- been report. from such places as Hawaii,
British '"n.ruras, Costa 'ica, "olombia and Puerto Rico (see literature
review), 71.h principal reason for nigh organic matter levels in these
areas is probably the fast annual rate of biological production, un-
interrLlted by frosts or severe dry spells. Once production is re-
duced by removal cf vegetation, soil organic matter levels generally
comparisonss between soil organic matter contents and ratios
of estimated rainfall to temperature are shown in Table 5 for each
site in the Pclochic Valley. The soil organic matter values reflect
the influence of climate, generally increasing with increasing rain-
fall and decreasing temperature. Schaufelberger (1956) related the
ratio of mean annual precipitation to mean annual temperature with
organic matter for "olombian soils and suggested that soils with a
ratio of 100 to 160 would contain 5 to 10 per cent humus, and soils
above 160 would contain 10 to 20 per cent humus. The above data for
Polochic Valley soils roughly approximate Colombian soils in this
Figure 4 shows the relationship of C:N ratios to rainfall
and temperature (elevation) for Polochic Valley soils. Data from the
20 to 40 cm. horizon were chosen since this zone is probably least
affected by vegetative cover. C:N values are generally greater at
higher elevations wheree precipitation/temperature ratios are higher)
than near sea level. C:N values are all below 11 when precipitation/
temperature ratios are lower than 110, and entirely above 11 when
ratios are higher than 140. Fi -L re 5 shows the relationship between
C:N ratios and topsoil organic matter for the Polochic Valley. As
right be predicted, the soils with the most organic matter also have
RELA71TI3NSHIP R-717:.N CLIMATEE AIMD O114A!T7 MATT7P ',)N7rFNTS
'P THE 5-20 -M. !IY'IO0N OF SHIFPTITIC,'UJLTIVJTION FIELDS
IN THE PnLOCHI, VALLEY, G'JATLMALA
Station Prec.a Temp.b Prec. Organic Matter
Tep. Forest Cleared 2nd Growth
m. or. Per Per Per
cent cent cent
1!atanzas I 2256 24.2 95 6.3 9.9 4.4
Natanzas II 2256 23.6 96 7.0 10.0 6.9
-eamay 3326 23.0 144 5.3 5.2 7.3
..ctela' 3471 .4 170 10.9 12.2
Canelias I 3C27 ZO.2 179 13.5 17.2
'anelias II 3627 2.1 180 14.7 17.8
Los :.lpes 3627 1E.5 254 14.4 8.8 18.5
aPrecipitation value is for station nearest site.
bTejperature calculated from sea level (260 C. annual mean
temrw.) assuming a temperature lapse rate of 0.610 C. per 100 meters.
the highest ":?r ratio, inJicating the relatively undecomposed state
cf forest debris in such soils.
,ati,'n exchange capacity, t-tal exchangeable
bases an:, base saturation
The exchange capacities of Polochic Valley soils are high, due
in laree measure to their high contents of organic matter (Figure 6).
The average cation exchange capacity for the topsoils under forest from
seve-n locations is 88 me. per 100 g. (Table 6). The exceptionally
large valu-3 of 187 me. per 1.1 g. for the topsoil under forest in Los
Alpes corresponds to the high value of 63 per cent organic matter for
the surface soil there. In almost all soils the exchange capacity
decreases v.ith depth w~tich corresponds to a decrease in soil huus.
90 --- 1 -- --- 1 -- --- 1 -- --- I --
6 8 10 12 14
C:N ratio (2,-4,? cm. depth)
Tig. 4.-- el:-tionship between precipitation:temper-
ature ratios and -":N ratios for the 20 to 401- cm. horizon of
Polochic Valley soils.
Fig. 5.--. :lations.ip between organic matter and
.:'N ratios h. iclochic Valley topsoils.
0 1 I I I I I
?- 50 75 100 125 150 175
Cation exchange capacity (me./100 g.)
'ig. 6.-- relationship between organic matter content
and cation exchange capacity for Polochic Valley soils.
EX:EL;NGF CAPACITY AND BASE SATURATION OF POLOCHI: VALLEY
OILS UNDLP THREE STAGES OF SHIFTING CULTIVATION
Exchange Capacity Base Saturation
Depth Forest Cleared 2nd Forest Cleared 2n.,
Location Land Growth Land Growth
cr. me./ me./ me.,' Per- Per- Per-
130g. 100g. 100g. cent cent cert
Average of C0- 94 69 75 54 59 54
First Four 5-2&- 45 42 43 52 44 46
Fields 20-40 54 59 55 46 45 50
Average of 0-5 88 63 57 42
ll 5-20 46 40 33 28
Fields 20-40 3c 56 28 28 -
The large amount of interstratified three-layer silicate clay
minerals in the Folochic Valley soils probably contributes to their
high exchange capacities. As might be predicted, exchange capacities
are lowest in Seamay, where kaolinite is the dominant clay mineral.
Changes in exchange capacity during the shifting cultivation
cycle can usually be associated with changes in organic matter contents.
Thus, the average exchange capacity of the topsoil at seven sites drops
from 8t me. per l`O g. for forested locations to 65 me. per 100 g. for
cleared land. The respective organic matter averages are 30.2 and
18.6 per cent (Table 4). Very little changes in exchange capacity
occur in deeper horizons. Young second growth has little effect on
exchange capacities. However, second growth older than two years
would probably increase exchange capacities, as the forest litter in-
creased the total amount of organic matter.
The shifting cultivation cycle does not appear to have much
effect on base saturation of the exchange complex (Table 6). Although
base saturation of some soils is increased by clearing, other sites
undergo a small decrease.
The cation exchange capacity and per cent base saturation data
are not very useful in evaluating the fertility status of Polochic
Valley soils. The exceptionally high cation exchange capacities of
these soils are associated with the large amounts of organic matter.
As such, the exchInre capacities are unstable and sensitively reflect
small changes in organic matter cus3ed by the shifting cultivation
cycle. Furthermore, organir. matter has very little permanent charge,
but an extraction at pH 7 measures large amounts of pH-dependent charge
('oleman et al., 1956), misrepresenting field conditions on normally
acid soils. As a result, there is no apparent correlation between
cation exchange capacity and total exchangeable bases. The values for
"per cent base saturation of the exchange complex," in addition to the
above limitations, give no measure of the absolute quantities of nutri-
ents available for plant growth and must be interpreted in terms of
clay mineral types. Seemingly applicable to these data is the conten-
tion of ColJman et al., (195E, 1959) that the amount of permanent
charge and its degree of saturation with nutrient bases provide a more
realistic basis than total exchange capacity and per cent base satur-
ation for assessing the fertility status of soils. Additional data
relating to this point are presented in the next section.
The pH of soil in the Polo-hic Valley was measured separately
in stillll? water and in a weak salt solution of 0.01 M CaCl2. The
data are gtven in Table 7. Since soil reaction varies with salt con-
tent (C-leman and Mehlich, 19F7), the values obtained in CaC12 are less
affected :, unknown variations in the salt content of these soils than
pH values obtained in distilled water. Therefore, the former pH values
are used in the following comparisons.
ihe average pH in the topsoil at seven sites increased from
5.0 in fore t-e lands to 5.5 for cleared areas, though little effect
was nnticea.'l at deeper derths. This is in agreement with work pre-
viously cited from many other areas. Purring may be beneficial on
VALTES OF pH FOR POLOCHIC VALLEY 3JILS UNDER
THREE STAGES OF SHIFTING CULTIVATION
Soil pH (.01M CaClI.) Soil pH (H20)
perth Forest Cleared 2nd Forest Cleared 2nd
Location Land Growth Land Growth
Camelias II 0-5 4.0 4.4 4.5 4.8
5-20 4.6 4.5 5.0 5.0 -
20-40 4.6 4.7 5.0 4.9
Average of 0-5 5.6 6.2 5.8 5.8 6.4 5.8
First Four 5-20 5.6 5.5 5.6 5.7 5.7 5.8
Fields 20-40 5.7 5.8 5.8 6.0 6.2 6.1