Uhl et al 1982

7/23/2019 Uhl et al 1982

http://slidepdf.com/reader/full/uhl-et-al-1982 1/7

Successional Patterns Associated with Slash-and-Burn Agriculture in the Upper Rio NegroRegion of the Amazon BasinAuthor(s): Christopher Uhl, Howard Clark, Kathleen Clark and Pedro MaquirinoSource: Biotropica, Vol. 14, No. 4 (Dec., 1982), pp. 249-254Published by: The Association for Tropical Biology and ConservationStable URL: http://www.jstor.org/stable/2388082 .

Accessed: 19/06/2014 10:59

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .

http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of

content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms

of scholarship. For more information about JSTOR, please contact [email protected].

.

The Association for Tropical Biology and Conservation is collaborating with JSTOR to digitize, preserve and

extend access to Biotropica.

http://www.jstor.org

http://www.jstor.org/action/showPublisher?publisherCode=tropbio

http://www.jstor.org/stable/2388082?origin=JSTOR-pdf



http://www.jstor.org/stable/2388082?origin=JSTOR-pdf

http://www.jstor.org/action/showPublisher?publisherCode=tropbio

7/23/2019 Uhl et al 1982


Successional PatternsAssociated with

Slash-and-BurnAgriculturen the

Upper Rio

Negro Region of the Amazon Basin

Christopher Uhl,' Howard Clark, and Kathleen

Clark

Institute f Ecology, University f Georgia,

Athens, Georgia 30602, U.S.A.

and

Pedro

Maquirino

San Carlos de Rio

Negro, TerritorioFederal de

Amazonas, Venezuela

ABSTRACT

In a

study of plant successionduring and

following lash-and-burngriculture earthe town of San

Carlos

de Rio

Negro

in

southernVenezuela

we found

that: 1.

Weeds established ust as readily n plots with Manihot

esculenta (the principal

crop plant

of

the region) present s

in

plots with

M.

esculen:a

removed.

2.

Repeated farm-plotweedings caused

woody

plants to decline

in

numbers nd biomassand

herbaceousplants

to

increase. 3. Forbs and grasses dominated

mmediately

followingfarm

ite abandonment, ut by one year

these had

begun to senesce and

fast-growinguccessionalwoody species

(particularlyVismia spp.) were common.

Standing crop biomass at one year was 773 grams dry weight

m-2.

4.

Several

microhabitat ypeswere presenton abandoned farm

ites. Grasses and forbs showed no microhabitat reference,

whereas

successionalwoodyindividualshad theirbest establishment ear slash and underfruit rees.

RESUMEN

En un

estudio de

sucesion, realizado

durante

y

luego

del

abandono

de

una

parcela

de

agricultura

e corte

y

quema,

en

los

alrededoresde

la

problacion

de San

Carlos de Rio Negro (Venezuela), encontramos

ue:

1.

Las

malezas

se establecieron

con la misma

rapidezen

las

parcelasde

Manihot

esculenta

que

es

el

principal

cultivo en

la

region), que

en las

parcelas

donde

se habian

eliminado

las

plantas

de

M. esculenta.2. El

desmalezado

repetido

de

las

parcelas

cultivadas

produjo

una

disminucion

n el

nuimero

biomasa de plantas leiiosas, y un

aumento n

las

plantas

herbaceas.

.

Inmediatamente

espues

de

ser

abandonada

la

parcela

cultivada

fue dominada por plantas herbaceas,pero

al

anio siguiente

stas

ya

habian

comen-

zado

a

senescer, ya

eran

comunes

as

especies

eniosas

e

crecimiento

apido (particulamente

Vismia

spp.).

La

biomasa

de esta parcela era de 773 gramos de peso seco

in-2

al

primer anio.

4.

En

las

parcelas

abandonadas

despues

de ser culti-

vadas se

encontraban

muchos

tipos diferentes e microhabitats. as plantas herbaceas

no

mostraban

ninguna preferencia

por

determinado

microhabitat,mientras ue

los

individuos

de

especies leniosas

e

establecian

mejor

cerca de

los

troncos

cortados

y

debajo

de

arboles frutales.

SLASH-AND-BURN

AGRICULTURE is thepredominant

farmingmethod n roughly 0 percent f

the

ex-

ploitable oils of the world nd supports

ver 250

million

eople Hauck 1974).

In

the

Amazon

Basin

an estimated

6,000

km2

f non-floodedorest ere

cleared y

slash-and-burngriculturistsrom 966to

1975, making his the mostcommon orm f land

use

in this

region

Meyers 1979).

This

paper

de-

scribes

uccession

uring

he

period

f

farm-plotse

and for the first

year following bandonmentn

theupperRio

Negro region.

THE

STUDY

SITE

The

studywas conductedn a 0.25 ha

tierra irme

(i.e., non-flooded)

lot ocated km east of thevil-

lage

of San

Carlosde

Rio Negro, lose to the con-

fluence f the Guainia

and Casiquiarerivers

10

56' N, 670 03' W, 119

m

elevation).

A

species-rich

evergreenorest

with a canopyheight f 25-35

m

was initially

resent n thisplot.This

forest as cut

and burned

n late 1976.The plot was

thenfarmed

for hreeyears sing traditional

racticesUhl and

Murphy

981)

and

then

bandoned

o natural uc-

cession. 30

x

50

m

plot

n

the center f

the farm

plot was designated

s the study lot.

The climate f the

region s relatively

seasonal

with mean nnual emperature

f 26?C

and

mean

annual ainfall f 3500mm.The months rom cto-

ber

through

arch

re less

rainy,

ut

monthly

ain-

fall

tillgenerally

xceeds 00 mm.

The

terrain

onsists f rolling

hills,

with

only

the

higher

ierra irme reas

being

used

for

farming.

Soils on thehigher reas

ontain n A horizon om-

posed mostly

f

fine sand that overlies

kaolinite

clay

nd

abundant erric

oncretionsHerrera 979).

These soils

are

very ow

in

nutrientsecauseof

in-

BIOTROPICA 14(4): 249-254

1982 249

IContribution

f

the InternationalAmazon

Project coordin-

ated by the Ecology Center of

the InstitutoVenezolano de

Investigaciones

ientificas IVIC),

Caracas,

Venezuela. This

work was funded

by the U.S. National Science

Foundation

in

cooperationwith IVIC.

This content downloaded from 185.2.32.185 on Thu, 19 Jun 2014 10:59:10 AMAll use subject to JSTOR Terms and Conditions





7/23/2019 Uhl et al 1982


tensive

eaching nder

humid

ropical

onditionsor

millions

f

years

nd

the

absence

of

unweathered

parentmaterial

hich

ould serve s a fresh

ource

of nutrients

Fittkau

t al.

1975).

METHODS

PRE-ABANDONMENTUCCESSION.-Farm

plotsnthe

San

Carlos

region re

planted

rimarilyo

Manihot

esulenta

rantz.

known

ocallys

yuca),

with

pine-

apple

plants

nd

fruitrees

resentt low densities.

We

studied he

ffect f

yuca n weed

establishment

by

comparing

he

number

f

weeds nder

uca

lants

with

the

number

resent

n

plots

whereyuca

had

been

artificially

emoved.

ll

weeds were

removed

from

oth

reatments

t

the

beginning

fthe

xperi-

ment.

reatments

ere

paired,

x 2

m in

size,

nd

replicated

ight imes. 2

m

buffertripwas

cleared

around he

non-yucalots.

San Carlos farm lots are usuallyweeded once

in

the

first ear

ndtwo

times year

hereafterntil

plot

abandonment.ll

weeds are pulled

out of the

soil

except or

prouts

r

an

occasionalarge

ndivid-

ual which

must

e cut.The effect f

repeated

arm-

plot

weedingson

density,

iomass, nd

life-form

composition f

recolonizingegetation

as studied

by

pulling

all

non-crop lants

from

27

randomly

located

1

x

1.5

m

permanent

lots t

10, 16, 21,

26,

and

31

months.

While

thismimicked

ocal

practices,

it

resulted

n an

underestimatef

root

tanding rop

because

portion

f

the roots

was left

n

the

soil.

The harvestedlants n eachplot weregrouped s

forbs,

rasses, uccessional

oody

tems,

orest

rees,

and

vines.

Plants n

each

groupwere

divided nto

root, tem, nd

eaf

fractions,

ven-driednd

weighed.

The

effect f

farming

n

the

soil

seed

bank

was

studied y

comparing

he

number

nd

species om-

position

f

seeds

germinating

n

soil

samples

aken

from

he

farm ite

at

the timeof

abandonment

nd

from n

adjacent

ndisturbedorest

lot. Each

soil

samplewas

20

x

20

x

5 cm

deep. All

samples

were

spread

n

trays nd set

1.7 m

above

the ground n a

sunny ocation

on

large

tables.

Contamination

y

wind-

or

animal-borneeeds was

checked

hrough

surveysffive dditionalrays,ontainingoilwhich

had

beenoven-dried

t

250? C for

ive

hours, laced

among

he

ample rays

nder

tudy. rayswere

ur-

veyed

for new

germinations

wo

times week

for

two

months,

fter

which

ime the soil

in

each

tray

was

turned

ver nd

mixed.

urveys

ontinued ntil

no

further

erminations

erefound.

POST-ABANDONMENT

SUCCESSION.-Slash-and-burn

farm

ites

n

the

San

Carlos

rea

areactively

sedfor

two to threeyears.

armersmaycontinue o visit

their

plots for several dditional

years to harvest

fruits r an occasional atchof

mature uca.How-

ever,

ecause ittle r no weeding

ccurs, uch sites

wereregardeds

abandoned or hepurposes f this

study.

We recorded

he

height

nd

identity

f all in-

dividuals

:

5 cm tall n the

permanent

x 1.5m

plots

at 4, 10, and 16 months

ollowing he final

farm-site eeding.

Biomass f all plants

2

m tallwas estimated

one

year fter bandonmenty

harvestingll plants

presentincluding

ineapple lants) in 25 plotsof

1 M2.

Harvested lantswere

separated y life form

and

plant part, ven-dried,nd

weighed.

The

bio-

mass

(in grams) of

all

plants >

2 m

tall

in the

study

plot was estimated y the

regression f

di-

ameter2X height

both in cm) on stem weight

[y

=

(0.49x) -29.5;

n=

111 and

leaf weight y-

(0.17x) +76.4; n

111.

Diametermeasurements

weremade t

10

cm

height.Valuesofr2were

> 0.90

in

both ases.Rootbiomasswas estimated

y eparat-

ing all rootsfrom

oil to a depth f 50 cm in five

0.25 m2randomly

ocated lots.

At the

time f farm-site

bandonment,here

re

severaldistinct

microhabitatsresent.

Much of the

area

contains are

soil, but large,partially

ecom-

posed

ogs

still over he oil n some

places.Clumps

of

vegetatively

eproducingineapple lants

nd so-

latedfruit

rees 4-10

m

tall)

are

also

present.

inal-

ly,

soil

is

mounded nd

planted

o

yuca

n

patches

where t appears hat final ropis possible. he

effect

f

these

microhabitatypes n

the establish-

mentof successional

egetation ollowing arm-site

abandonment

as tested

n

a

25

x

25 plot

ocated

n

a

recentlybandoned

arm ite

ocated bout

150

m

from

he

main

tudy

ite.The

plot

was cleared f all

vegetation

nd slash

xcept

for

five ashew

rees.A

3

x

3

m

plot

was

formed round ach

cashew

ree.

The

remaining

our

reatmentsere

andomlylaced

in

3

x

3

m

plots

with m

border

trips

etween

lots.

The

bare

soil

plots

were

simply

marked

ff.

Pine-

apple plants

were

transplanted

nto their

ssigned

plots. oil was moundedn theyucaplots ndyuca

stem tocksnsertednto hemounds.

arge ogs

were

piled

in the

slash

plots.

All

herbaceous

nd

woody

plants

hathad

become

stablished

n

each

of

these

plots

were

ounted nd dentified

fter ne

year.

ix-

teen oil cores

6

cm diameter

5

cm

deep)

were

taken

rom ach

of

the

bare

oil, lash,

nd

fruit ree

plots

t

the

conclusion f the

experiment.

he cores

from

ach

plot

were

pooled

nd

set

n a

sunny pot

to

allow

germination

f

soil seeds s

described

bove.

250

Uhl,

Clark,

Clark, and

Maquirino






7/23/2019 Uhl et al 1982


RESULTS AND DISCUSSION

PRE-ABANDONMENT SUCCESSION.-The actual pres-

ence of yuca plantshad no apparent

ffect n the

ability fweedsto

reestablish

fter

arm-plot

eed-

ing, lthough he

yuca canopy

did

reduce ight

nd

soil-surface

emperaturetable 1). More than80

percent f

all

establishments

n

both

reatment

ypes

were the forbspecies,

Borreria latifolia

(Aubl.)

Schum. nd

Eupatorium erasifoliumSch.

Bip.)

Baker.Grasses omprisedn additional 0 percent r

moreof

all

establishments

n

both treatments

ith

Paspalum ecumbens w. the most ommon pecies.

The process f succession

egan

nd was

curtailed

through eeding

ive

imes

efore

he farm itewas

finally bandoned.

hese repeatedweedings ad no

clear effect n the

number

f

reestablishinglants

(table 2).

Plant

density

as low at the

first,econd,

and

fourthweedings

nd

relatively igh after

he

third nd fifth

weedings.Weedings hree

nd

five

occurred oward he end of the wet season,while

weedings wo and four ccurred t

the

end

of the

relativelyry eason.

Uhl et al. (1981) have shown

that reduced ainfall t San Carlos

owers he rate

of seed germinationnd seedling

stablishmentn

disturbedites.

The repeatedweedings

id

appear

to affect he

life-formompositionf the reestablishingegeta-

tion table 2).

Forest-treeensityt the first eed-

ingwas0.65individualslot-' all of sprout rigin),

butby

he

final

weeding ensity

ad declined o

0.11.

Hence, the

repeatedweedingshad apparently

x-

hausted

he

sprouting

eserves f manycut forest

trees. uccessional

oody pecies omprised 4 per-

cent i.e., 12.1 individuals lot-')

of the total tems

present t thefirstweeding, ut

by the fifth eed-

ing they

lso

had

declined n numbersccountingor

only

2

percent

f the total ndividuals resent. n

contrast,he density f forbs nd

grasseswas low at

thefirst

wo

weedings, uthigh hereafteri.e.,more

than25 individuals lot-'

forweedings , 4,

and

5).

Weed biomass measurements aralleled density

values.

Havel

(1960)

in

New Guinea,

Snedaker

(1970)

in

Guatemala,

nd

Kellman

1970)

in

the

Philippines eported

imilar rends.

High forb

nd

grass stablishment

n the

aban-

doned

arm

lot

mayhaveresultedrom

n

abundance

of favorable ermination

urfaces here. n

the San

Carlos

region,

oot mat

(a

surface

mat of forest

treeroots

5-30

cm

thick),

charred

wood,

and

bare

soil surfaces re the main seed-bed

ypespresent

following

orest

utting

nd

burning.

he

bare

soil-

surface ype ncreases

with timebecausethe root

mat decomposes nd the charredwood is washed

away

n

heavy

ains.

or

example,

1

percent

f

the

farm-siteurface rea was bare

soil one year after

forest learing, ut by

the

time

of abandonment9

percent

'f

his

ite

had

exposed

oil. n

separate

ur-

face olonizationxperimentsUhl

et al. 1981), forb

and

grass pecies

howed

ood establishment

n these

bare-soil urfaces

s

compared

o

successional

oody

species.

The results

f

the

seed-bank

urvey table 3)

TABLE

1. A

summaryof differences

etween

plots

with

yuca present nd adjacent plots where

yuca was

removed.

Yuca present Yuca removed

Number

of

species present 9

7

Number

of

woody

establishmentsa 0.7 + 0.8b 0.0

Number

of

herbaceous

establishments 15.6? 8.53 12.3?+

9.6

Total numberof

establishments 16.3

?

8.42

12.3

?

9.6

Midday light

at soil

surface

on

overcast ay (lux) 7,500?+9,500 22,750?+ 9,700

Midday temperature

t

soil

surface n sunnyday

(0C) 40.5?0.9

46.8?4.4

Soil

percentagemoisture

after three days

without ain 6.4?0.93

6.9?+1.16

aOnly plants

?

5

cm tall

were

considered s

established.

bNumber

of

plants presentper

2

x

2

m plot

?

95 % con-

fidence nterval.

TABLE 2. The mean number

of plants and grams dry-weight

iomass

in 27 1 x 1.5 m permanent

lots at each of

five

weedings.a

Forbs

and grasses

Woody plants

Total individuals

Weeding

Density Biomass

Density

Biomass

Density Biomass

1 6.1?

2.7b

7.1?

5.4 12.7?6.0

16.6+6.5

18.8? 7.4 23.7?

9.1

2

9.7? 5.6 55.2?24.9

2.2?0.9

8.5?4.7

11.9?

5.7

63.7?24.9

3

72.6?30.8 41.7?16.3

3.4.2.3

76.0?30.8

45.1?16.7

4

28.2+16.2

24.2?12.5

2.2?1.2

5.0+4.1 30.4?16.2

29.2?13.0

5

53.1?22.6 27.5?16.7

1.8?0.8

3.2?2.5

54.9?22.7 30.7?16.6

aIndividuals

<

3 cm

tall not

considered

n

density stimates.

"95 % confidence

nterval.

Slash-and-Burn Agriculture

n

Amazon

Basin 251






7/23/2019 Uhl et al 1982


also help to explain he

observed hift rom

woody

toherbaceous

ife-formominance

uring he

period

of

farm-plotse. The estimated

umber

f germin-

able

seeds

m-2

was 581 in the

farm ite at the time

of

abandonmentnd177 in

the djacent,ndisturbed

forest ite.

Another orest

ite, several

kilometers

away,was

also ampled t this ime

nd

had

a germin-

able seedbankof 752 seeds

m-2

(Uhl et al. 1981)

suggestinghat

eed-bank ize varies

with ocation.

Successionalwoodyspecies

dominatedhe germin-

able seed bank in

both the forest

ites;however,

most f

the dentified

erminationsn the farm

ite

were

forbs r grasses. orb

and grass eeds

ncrease

in

abundance n

farm-siteeed banks

becausetheir

life ycles re

shorterhan he nterval

etweetweed-

ings. n contrast,

uccessional oody

pecies reweed-

ed out before

heycan produce eed

locally,

nd

after everalweedings he

seed bankthatgives rise

to

thisgroup

s

greatly

educed.

POST-ABANDONMENT

SUCCESSION.-Four

months

after he last

farm-ploteeding,

rasses nd forbs

accountedor

3

percent

f the

ndividuals

?

5

cm

tall

n

the27

permanent X 1.5

m

study lots table

4).

The

most bundant

orb pecies

t this imewere

Eupatoriumerasifolium

nd Phyllanthusp. with .2

and

4.1 individuals

lot-',respectively.he most

common rass species n

the plotswere

Paspalum

decumbens nd

Panicum pilosum Sw.

Total plant

density eclined t the

10-

and

16-month urveys

because

many rasses

nd

forbs

ad died.

Mostdeath

resulted ither

rom hading y taller lants

r

from

senescence ollowing

ruiting. ll other ife forms

either ncreased

r remained hesame

between he

four-

nd

16-month

urveys. orest rees

had

very

low

densitieshroughouthe

study eriod nd

were

onlypresent s sprouts.

At

16 months,

uccessional oodyplantsdomi-

nated he site

withthe tallestndividuals

-4 m

in

height.

f

199

individuals

2

m tall

present

n

the

30 X 50

m

study lot

t

16 months,

1

percent

ere

Vismia

auriformis

Lam.) Choisy, 19 percent

Vismia

apurensis

eich,

and

9 percentwereCecropia

ficifolia

nethlage.

n additional

ight uccessional

woody pecies

had

from ne

to seven ndividuals

-

2

m

tall

present. hese

successional oody

pecies

presumably

riginated rom seeds

dispersed nto

the

site from ndividuals

inging he margin f

the

clearing.The margins f farmsites are seldom

farmed

because these areas

burnpoorlyand are

quickly vergrown ith uccessional

egetation.We

TABLE 3.

The

number

of germinable

eeds

in soil

samples

taken from the

farm plot at the time

of aban-

donment and from

an adjacent

matureforest

plot.

Farmedsite

Forest ite

Number of

samples

15

13

Mean

number f

germinations er

sample 23.3

12.88a

7.08?2.29

Number of

germinations

-2

581

177

Percentage f total

germinations

growing to

an

identifiable

tage 74

69

Forest rees,

ercentage f total

identified

erminations 0 1

Successional

rees,

ercentage

of total

identified

erminations 8

95

Forbs and grasses,

percentage

of total

identified

erminations 92 1

'95% confidence nterval.

TABLE

4.

Changes

n

density

nd

species

number n

27

permanent

x

1.5

m

plots during

the

first

6

months

f

succes-

sion

following

he

final farm-plotweeding.

Months following

final weeding

4

10 16

Mean number

of individuals plot-1:

forbs

7.8

+

3.9b

7.8?+ 4.2 3.1 ?+3.2

grasses

19.8?10.6

6.3 ?3.3

1.7+2.3

successional

woodye 1.8+ 0.8

3.1? 1.0

3.2?+1.0

primary orestwoody 0.04?+0.08 0.04?+0.08 0.07?+ 0.15

Total

29.4+13.4

17.2?6.0 8.1?5.4

Number of species,

all

plots:

forbsa

5

6

4

grasses

8 6 6

successionalwoodyc 6

10

9

primary

orestwoody

1 1

3

Total

20

23

22

:Includes

herbaceous reepers.

h95

% confidence

nterval.

'Includes woody vines.

252 Uhl, Clark, Clark, and

Maquirino






7/23/2019 Uhl et al 1982


have

observed ther

newly

abandoned

farm

plots

in

the environs

f SanCarlos

nd

found hem

o

be

similar

to the

studyplot

in

vegetation

ife-form

composition

nd

structural

evelopment.

Standing

rop

biomass n the

abandoned

arm

siteone

year fter

he final

yuca

harvest

i.e., 16

months

fter he

final

weeding)was

773 grams

ry

weight

m-2,

a value imilaro otheriteratureeports

for

bandoned

eotropical

arm

ites

(table 5).

Of

this

total,

30 grams

were

aboveground

uccessional

species

biomass nd

278

gramswere

biomass

from

crop

plants till

present n

the site.

Leaf

biomass f

the

successional

egetation

as 234

g

m-2

and

stem

biomass

was

196

g

m-2

while oots ccounted

or

nly

about

10

percent

f the

total

non-cropiomass. he

low

root

biomassvalue

may

be

an

artifact f

the

small

ample

ize. n

excavations

f

entire oot

ys-

tems

of several

ommon

uccessionalree

species,

root

biomass

has

generally

een

15-20

percent f

total iomassUhl,pers. bs.).

The

woody

speciespresent n the

abandoned

farm ite

ppeared o

have

greater

ensities

earde-

cayingogsand

under

ruit

rees.

Grasses nd

forbs,

by

contrast,

ere

ble

to

establish

n

open,bare

soil

areas

s

well s

near

ogs

nd

under

ruit

rees. hese

establishment

referencesere

demonstratedy the

results f

the

microhabitat

xperiment

table 6).

The

cashew

and

slash

treatmentsad

significantly

more

woody stablishments

p

<

0.01,DuncanMul-

tiple

Range

Test) than did

the

other

treatments,

while

no

significant

ifferencesn

the

numbers

f

forb

and

grass colonizers

were

detectable

mong

treatments.ecausepineapple ndyucagrewpoorly,

these

treatments

ere

probably

imilar o

the bare

soil

plots n terms

f

ight nd

soil

temperature.

TABLE

5.

The

standingrop

biomass

f

a

farm

ite

one

year

fter

bandonmentt

San

Carlosde

Rio

Negro

nd

at sites

lsewhere.

Mass

drywt, g m-2

Aboveground

biomass:

Grasses

127

Forbs

9

Successional rees

272

Primary orest rees 0

Vines

22

Cashew trees

39

Pineappleplants

239

Belowground

biomass:

All

groups combined

65

San

Carlos,

Total

773

Guatemala,

Total

aboveground

(Snedaker

1970)

836

Costa

Rica,

Total

aboveground

(Ewel

1971)

1067

TABLE

6.

The

mean numberof grass and forb and woody

colonizers

n

five treatmentsfter ne year. Only

individuals more than 10 cm tall were consid-

ered as established.

Mean number f colonizers/plot

Treatment Grasses and forbs Woody plants

Bare Soil 184+ 65.7 0.0

Yuca 130+ 62.3 0.2?0.6

Pineapple 137? 130.3 0.6?1.1

Cashew

146+108.0

10.2?8.0

Slash

185+115.5

10.2?4.2

'95 % confidence nterval.

Treatment ifferences

n

woodyplant establish-

ment ouldoccur

because

f differential

ispersal

o

certain

microhabitats.ndeed, soil-germinableeed

checks t the conclusion f the experimentevealed

that

oils under he cashew reeshad an estimated

932 germinable oody eeds

m-2

(all successional

species) as compared o 126 in the slash nd 74 in

thebare

soil

plots.Fruit rees, uch s cashew,may

supply ood s well as perch itesforbirds.Differ-

entialgerminationf seeds once in a microhabitat

and/or

differential

ortalityf seedlings ollowing

germination ay also be importantn explaining

higher stablishmentn fruit reeand slash micro-

habitats.

lash

presence esultsn lower oilevapora-

tion and lower

oil-surface

emperaturesUhl

et

al.

1981), therebyrovidingavorableoil-moistureon-

ditions

or erminationnd seedling rotectionrom

midday

eat.

The presence f fruit

rees

lso protects

seedlingsrom irect eating,ndsteadyitternputs

from

hese rees

may

ikewise

avor stablishment.

CONCLUSIONS

The

early

uccession

ollowing lash-and-burngri-

culture

t San

Carlos

s similar

n

life-form

ompo-

sition nd

structureo other iteratureeports rom

both

New

and

Old

World

tropical

arm

ites.

ym-

ington 1933) reportedhat bandoned

arm

lots

n

Malaysia

were dominated

y grasses (particularly

Imperata ylindrica), orbs, nd successional rees

and

shrubs.

ellman

1970)

found hat

early

uc-

cessional ites in the Philippineswere coveredby

stands

f

herbs, ut after ne year, oftwood

rees

began

to

overtop

he

herb

ayer

nd soon

formed

complete

over.

Clayton (1958) reported

hat

a

thicket f close-growingrees nd shrubs s formed

soon

after arm-sitebandonmentn Nigeria. n the

neotropics,

udowski

1961) and Ewel (1971) de-

scribed

early-fallowegetation

n

Costa

Rica as

tanglesdominated y a mixture f grasses, erbs,

vines,

nd

woody ioneers;

nedaker

1970) report-

Slash-and-Burn

griculture

n Amazon Basin 253






7/23/2019 Uhl et al 1982


ed that

ne-year-oldallowsn

Guatemalawere

dom-

inated y woody

hrubswhich

ormed more-or-less

uniform

anopy

t 2

m;

and

Blum

(1968) reported

a

similar egetation

tructurefter ne

year

n

Pan-

ama.

The time

required o

regrow forest n a

de-

forested

ite t San

Carlos

depends

n how the

forest

is removed nd on the manner f land use follow-

ingremoval. or

example, hen

heforests cut

for

timber, ecoverys

rapid

because manyforest

ree

species

have the ability

o sprout rom

ut stumps

(Uhl

et al.

1981).

If

sites re

cut and

burned

nd

then

immediatelybandoned,

uccession

proceeds

more

lowly ecause

burning educes

prouting o-

tential nd

reduces he

woodyspecies

seed bank

(Uhl

et

al.

1981).

If sites re

cut

and

burned

nd

and

thenfarmed

or several

yearsbefore

bandon-

ment

this

study), ecoverys

even slower. he re-

peated

weedings

uring arming

sually xhaust

he

sproutingeserves f thosestems hatsurvive he

burn.

orbs,

rasses,

nd

successional

rees

ominate

on

newly

bandoned arm

ites

becausethey

have

many

eeds

present

n

the

seed bank

and/or

ecause

they have

easily

dispersed

eeds. Forest

trees,

by

contrast, ave

few, f

any, seeds in

farm-site

eed

banks

nd often

have

large,

orly

dispersed

eeds.

Therefore,

rimary orest

mustbe

closeto

thefarm

clearing or forestrees o colonize. n caseswhere

clearings re

very

arge i.e.,

seed

sourcesmany

km

away) and

land

use

prolongedi.e.,

sprouting

o-

tentialost),

it seems

afe o

predict hat

he return

to

primary

orestwill

takemany

hundreds

r per-

haps even

thousandsf

years.

ACKNOWLEDGEMENTS

Thanks re

extended o Getulio

Gomez and

Paulo Livino

for

many ours f

help n the

field;

rs.

Carl

Jordan, ruce

Haines,

Martin

Kellman,

nd an

anonymous

eviewer or

their

ritical

eading

f

the

manuscript;

nd

Saundra

Green

forhelp n preparationf themanuscript.

LITERATURE CITED

BLUM, K. E. 1968. Contributions oward an understanding f vegetational evelopment n the Pacific

owlands of Panama.

Ph.D.

diss. Florida State Univ., Tallahassee, Florida, U.S.A.

BUDOWSKI,G. 1961. Studies on forest uccession n Costa Rica and Panama. Ph.D. diss. Yale

University, onnecticut,

U.S.A.

CLAYTON,W. D.

1958.

Secondaryvegetation

nd the transition o Savanna near Ibadan, Nigeria. J. Ecol. 46:

217-238.

EWEL,J. 1971. Experiments n arresting uccessionwith cutting nd herbicides n five tropical environments. h.D. diss.

Univ.

North

Carolina, Chapel Hill,

North

Carolina, U.S.A.

FITTKAU,

E.

J.,W. JUNK,

H,

KLINGE, AND H.

SIOLI.

1975. Substrate nd vegetation n the Amazon region. Berichteder

Internaticnalen ymposien der InternationalenVereinigungfur Vegetationskunde Rinteln,

1969).

HAUCK, F. W. 1974. Shifting ultivation nd soil conservationn Africa.F.A.O. Soils Bull. 24: 1-4.

HAVEL, J. H. 1960. The deflection f secondary uccession n cleared mid-mountain ain forest by plantation tendings.

In, Symposium n the impact of man in the humid tropics.Goroka TPNG. Commonw. Gov. Print.

Office,Can-

berra,

Australia.

Pp. 339-343.

HERRERA, . 1979. Nutrient istribution nd cycling n an Amazonian caatinga forest n Spodosols in

southernVenezuela.

Ph.D.

diss. Univ. Reading, Reading, England. 244 pp.

KELLMAN, M. C. 1970. Secondaryplant succession n tropical montane Mindanao. Publication BG/2, Aust.

Nat.

Univ.,

Canberra,Australia.

MEYERS, N. 1979. The sinking rk. Pergamon Press, Oxford. 308 pp.

SNEDAKER, . 1970. Ecological studies on tropicalmoist forest uccession n eastern

owland

Guatemala.

Ph.D. diss.

Univ.

Florida, Gainesville, Florida, U.S.A. 131 pp.

SYMINGTON, . F. 1933. The study f secondary rowth n rain forest ites n Malaya. Malay. Forest

2:

107-117.

UHL, C., K. CLARK,H. CLARK,AND P. MURPHY. 1981. Earlyplant succession

fter

orest utting

nd

burning

n the

Upper

Rio Negro region of the Amazon Basin. J. Ecol. 69: 631-649.

, AND P.

MURPHY. 1981. A comparison f productivitiesnd energy alues between lash

and

burn

agriculture

nd

secondary uccession

n

theUpper Rio Negro region of the Amazon

Basin.

Agro-Ecosystems

:

63-83.

254

Uhl, Clark, Clark,

and Maquirino






Uhl et al 1982

Documents