AN ANALYSIS OF MANGROVE FORESTS

ALONG THE GAMBIA RIVER ESTUARY:

IMPLICATIONS FOR THE MANAGEMENT OF ESTUARINE RESOURCES

Robert R. Twilley 1

Center for Environmental and Estuarine Studies^

Horn Point Laboratories

P.O. Box 775

Cambridge, Maryland 21613

Consultant Report to
The University of Michigan
Gambia River Basin Studies

River Resources Team

Great Lakes and Marine Waters Center

The University of Michigan
International Programs Report No. 6

1985

1 Present address: Department of Biology, University of Southwestern Louisiana,
Lafayette, Louisiana 70504

2 Contribution No. TS-34-85 from the Center for Environmental and Estuarine
Studies, University of Maryland

TABLE OF CONTENTS

Page

SUMMARY AND CONCLUSIONS v

INTRODUCTION 1

STUDY AREA 2

The Barrage 7

MANGROVE ANALYSIS 10

Mangrove Community Types • 10

Species Distribution and Zonation • 13

Forest Structure 16

Areal Distribution 19

Litter Productivity 23

Background • 23

The Gambia 26

Exchange of Detritus and Nutrients 30

Detritus 31

Nutrients 38

Allochthonous Organic Matter 46

Background 46

The Gambia 47

Mangroves and Fisheries 50

Introduction 50

The Gambia 52

MANGROVES AND ESTUARINE RESOURCES 54

Mangrove Stress 54

Disease • • 55

Drought . . . 56

Impact of Proposed Development 60

REFERENCES 66

APPENDICES

iii

SUMMARY AND CONCLUSIONS

Mangroves along the Gambia River estuary cover an area of 60,000 to 70,000
ha with approximately 12% of this total occurring above the proposed barrage
at Balingho.

Most of the mangroves in the meso- and oligohaline sections of the estuary
represent riverine forests inhabited by Rhizophora racemosa , R. harisonii ,
R. mangle , and Avicennia africana . Inland of these forests are basin
mangroves dominated by the latter two species. In the higher salinity areas
of the river, and up in the larger bo Ions, fringe and scrub mangrove forests
are more common.

The riverine and basin forests are dominated by trees >20 m, while forests
in higher salinity areas have less structure and are generally <7 m.
Based on forest structure statistics of the mangroves of The Gambia,
the following litter productivity rates (t dry mass •ha'^yr""^-) were
estimated for categories based on tree height: Rhizophora (>20 m) - 18.8;
Rhizophora (7-20 m) a 10.4; Avicennia (>7 m) » 11.6; Avicennia plus
Rhizophora (<7 m) * 4.0.

Ratios of litter export to litter productivity were estimated based on
results of other studies that have demonstrated the following relationship
between the magnitude of detritus export and tidal amplitude: riverine
forests a 0.94; fringe forest « 0.5; and basin forests a 0.2. Ratios used
for mangroves in The Gambia were 0.73 for Rhizophora >20 m and 7 - 20 m in
height, and 0.5 for Avicennia >7 m and Avicennia plus Rhizophora <7 m tall.

Personal observations of minimum leaf litter accumulation in the riverine
and basin mangrove forest of The Gambia substantiate the use of ratios above
0.5 for these forests.

Based on these estimates of litter productivity and export to litter
productivity ratios, detritus export was 289 gC'nT^yr" 1 . Total export for
mangroves along the Gambia River estuary was 181,040 tC'yr"" 1 with
40,080 tC'yr" 1 occurring above the barrage and 140,960 tC^yr" 1 below the
barrage.

Organic carbon inputs to the Gambia River estuary from phytoplankton net
productivity was estimated at 14,782 tC'yr" 1 based on 1Z *C02 uptake studies
in 1983.

Organic carbon inputs to the Gambia River estuary from river discharge at
Goulombo is equal to 15,960 tC-yr -1 .

Total organic carbon loading to the Gambia River estuary is 211,782
tC - yr -1 , of which about 85% was contributed by allochthonous inputs from
mangroves .

The nonconservative behavior of nitrate, phosphate, and total suspended
solids in bolons suggests that mangrove forests are a sink for these
constituents. Such trapping mechanisms are occurring where the
concentrations of these constituents in estuarine waters are high.
The food web associated with mangrove bolons are representative of other
mangrove ecosystems. There are four levels of the food chain with detritus
serving as a major energy source: detritivores , mixed trophic level, middle
carnivores, and higher carnivores.

Fish that were associated with the mangrove food chains in the bolons
constituted nearly 51.5% of the total catch of the artisanal fishery from

vi

Banjul to Kaur in 1983. In the upper river, from Kaur to Fatuto, nearly 30%
of the annual catch is represented by Tilapia , which is a detritivore found
in mangrove bolons. Collectively, these trophic levels of the mangrove food
chain represent 40% of the grand total catch of fish for the Gambia River
estuary.

Mangroves in The Gambia can be classified as existing in a very dry life
zone, and are thus very susceptible to slight changes in hydrology.
Therefore the present decline and death of mangrove trees along the Gambia
River estuary can be attributed to the decline in precipitation over the
last decade.

The proposed barrage at Balingho will increase the salinity of typically
meso- to oligohaline waters of the estuary to hypersalinity levels.
Rhizophora racemosa and R. harisonii are very susceptible to salinities
>l5°/°° and will not tolerate the change in conditions due to the barrage.
The proposed barrage at Balingho will also create a freshwater impoundment
upriver with a water level at + 1.7 m GD. This high level of stagnant water
will result in a loss of mangrove trees with heights >7 m in the area above
the barrage due to drowning of their root systems.

These losses of mangroves above and below the barrage will result in a loss
of nearly 147,920 tC^yr"" 1 of detritus to the estuary. This decline is
equivalent to 82% of the organic matter inputs to the estuary that are
presently supporting a detritus-based fishery.

Mangrove forests of The Gambia provide a vital natural resource to the
artisanal fishery of the Gambia River estuary.

vii

INTRODUCTION

The management of forested wetlands has received considerable attention in
the last 15 years because of increased human exploitation of these wetlands as a
forest resource, while at the same time natural resource managers have become
more cognizant of the importance of these plant communities to the ecology of
adjacent aquatic ecosystems. In the intertidal zone of the tropics, forested
wetlands or mangroves are a resource in the form of a variety of forestry
products and as an important component of estuarine ecosystems* The ex-
ploitation of mangroves for forestry products includes saw timber, building
material, fence posts, fuelwood, tanins, and charcoal. In some areas of the
world, such as Malaysia, these forests are managed for sustained yield to
provide these resources. Mangroves have also been acknowledged as contributing
to the productivity of coastal ecosystems by providing detritus to various
levels of estuarine food chains (Heald 1969, W. E. Odum 1971, Twilley 1982,
Twilley 1985). They are also important in the nutrient and sediment dynamics of
estuaries as well as contributing to erosion and flood control in coastal areas.

Because of these two important roles of mangrove ecosystems as forests and
estuarine resources, human manipulations of estuarine ecosystems that may alter
or destroy these forests should be carefully evaluated to understand their
potential impact on the entire estuarine basin. In the Gambia River estuary,
located in West Africa, a salt barrage is proposed in the oligohaline region
(Balingho), which will alter the hydrology and salt distribution of this
tropical estuary. This report is an analysis of the mangrove forests in this
estuary, and of the potential impact of this barrage to estuarine resources
associated with these forested wetlands. This report is supplemented with

observations made in the field during a trip from 29 June to 2 July 1984 along
the Gambia River estuary from Banjul to Georgetown (Appendix A).

STUDY AREA

The Gambia River estuary is located approximately 13 °N latitude along the
west coast of Africa (Fig. !)• The climate in this area is defined as Sub-
Guinean within the Tropical Sudanese zone (Richard-Molard 1949) and as tropical
semi-dry modified to tropical semi-wet nearer the sea. The climate is dominated
by alternating dry Harmattan air mass, which originates in the Sahara, and
southwesterly monsoon of humid oceanic air. As a result, there are distinct wet
and dry seasons with the former season occurring from June to October (Fig. 2).
Rainfall varies considerably from year to year in quantity (mean of about
1,140 mm) and distribution; the highest precipitation is usually in August. Due
to Sahel drought, which has lasted through the 1970s, rainfall in the basin has
followed a negative slope (m » -19.9 mnryr"" 1 ) for the last 23 yr of record (Arid
Lands Information Center 1981) (Fig. 3). Rainfall occurs in high intensity
storms with more than 50% of rain accumulated at Bombey, Senegal, falling at an
intensity >27 mm*hr~^, and intensities >100 mm*hr~* are not uncommon (Jones and
Wild 1975).

The difference in maximum and minimum monthly mean air temperatures during
the year are only about 3.8 °C for coastal locations, compared to 5.5 °C at Kaur
and 7.6°C at Basse which are more than 175 km from the mouth of the Gambia River
estuary. The mean annual average of maximum and minimum air temperatures also
increases inland from 25.1°C at Cape St. Mary and 25.5°C at Yundum, to 28.4°C at
Kaur and 28.1 °C at Basse (Fig. 2). These high annual air temperatures result in
evapotranspiration rates (based on Penman method) ranging from 4.8 to 15.0
nnrd"" 1 , with higher rates occurring in April and May during the dry season.

CT3

a
u
<

CO

cu

!3

CO

•H

s

a

H

CO

4J
CO

a

CD

>

e

cd

a

cu

o

C
O

CO
O
O

o

F M

M

N

M

M

N

FIG. 2. Climatic diagrams of air temperature and precipitation for Yundum (near
the mouth of the estuary) and Kaur (184 km upriver). Solid areas represent the
rainy season, vertical lines the period when rainfall exceeds evaporation, and
the dotted area the period when evaporation exceeds rainfall. The years of
record are in parentheses following the label.

1500

E
E

1000-

s

o

UJ

a.

500

/
/
/

1931-1960
□ 1968-1977

7\

/

/
/

/

PI

/

/
v

CASAMANCE GAMBIA SALOUM SENEGAL

FIG. 3. Change in precipitation along the west coast of Africa from Saloum to
Casamance by comparing annual totals from 1931 to 1960 to those from 1968 to
1977 (from Marius 1981).

The Gambia River originates in the moist evergreen forests of the Fouta
d'Jallon plateau in the Guinean hinterland and flows 1,200 km to the Atlantic
Ocean (Fig. 1). The total drainage area is 78,000 km 2 with moderate drainage
gradient in Guinea (slope of 1-4%), low gradient in Senegal (slope of - <1%),
and virtually no slope over the last 500 km from Gouloumbo, Senegal, to Banjul.
Collectively, the geology of the continental sub-basin consists of sedimentary
formations (64%), metamorphics (20%), and intrusives (16%); a small band of
Quaternary alluvium deposits border the Gambia River bed. Soil parent material
is fine textured with little or no sand since the combined silt and clay
fractions comprise more than 90% of the mineral material. These iron rich
deposits are partially the products of laterite-rich uplands (Dunsmore et al.
1976). The high clay content, consisting of kaolinite with some silica, renders
these soils resistant to infiltration.

Mean historical runoff is 231 mm or about 21% of precipitation. Annual
mean discharge since 1953 has varied from 90-460 nr* e s~^, with decreasing trend
since 1970 as a result of the Sahel drought, and peak discharge occurs in early
to mid September (Lesack et al. 1984). The transport of solutes by the Gambia
River are classified as a river with low runoff (<5 L*s~^*km~2) f warm (>15°C),
and low relief. Runoff loss of Na, K, SO4, N, and P from the Gambia River com-
pared to precipitation loading rates unto the watershed are low, possibly re-
flecting the scarcity of essential nutrients in the watershed for upland plant
communities (Lesack et al. 1984). The transport of particulate material also is
very low and fits in better with high runoff rivers (>15 L*s~ 1# km~ 2 ) than low
runoff rivers (Lesack et al. 1984). The export of particulate and total organic
carbon is extremely low compared to rivers around the world (Schlesinger and
Melack 1981, Meybeck 1982), with respective transport rates of 0.12 and 0.38
f knfZ-yr"" 1 (Lesack et al. 1984).

The Gambia River is tidal to Gouloumbo Bridge in Senegal, 526 km upstream
from Banjul, during the dry season. As a consequence of low river discharge
(<5 m^ # s~^), two sequential tides are able to penetrate the river at any one
time. During this period, salinity normally reaches Carrols Wharf which is 219
km from the mouth. This was observed in July, 1983 (Fig. 4), United Nations
Development Programme (1974) reported a salinity of 1.0° / 0O at Kuntaur which is
248 km from Banjul. In wet season there is a rapid increase in river discharge
(>1,000 m^*s"^-) and fresh water may be observed only 75 km from the mouth of the
river (Herklots 1979). During October 1983, a salinity of <l°/°° was measured
at 125 km upstream (Fig. 4). Thus the saltwater wedge may travel annually from
about 100 km to 220 km upstream from the mouth of the estuary.

Tides in the Gambia River estuary are semidaily with unequal magnitude.
The difference between mean high and low tide levels at Banjul is about 1.2 m
(Fig. 5), with slightly lower tidal amplitude during the wet season. Mean high
water during spring tides (Gambia Datum, GD) is +1.4 m at Banjul, compared to
+1.7 m at Balingho and Kaur (188 km upriver) (Coode and Partners 1977). Extreme
spring tides at Banjul are highest from August to October and may exceed +2.2 m
GD.

THE BARRAGE

A salt-water barrage has been proposed at 1 km south of Yelitenda
(Balingho) to provide a reservoir for rice irrigation, a barrier to tidal
seawater, and a road crossing. The site is located at latitude 13° 25 f N and
longitude 15° 35 f W and is 127 km from the mouth of the estuary. Salinity in
this area may fluctuate from freshwater to about 16°/°° (Fig. 4), thus the use
of the river for irrigation is strictly seasonal. Also, the extraction of fresh

OS
hi

>

a.

UJ

o

z

T3

C
CO

<U

O

4=)

a
o

3
^

CO

•H

a

c

CO

4J
CO
CD

u
>

a

CO

O

60

o

rH
c0

•H 00

C ON

CO
CO

O 4J

o

•H
4-1

3

o

o o

fO

cu —

(°%) A1INI1VS

U

5-i
CU

xi pa

■u m

CO 00

•H ON

O r-4

5-!

• (U

6

• 0)

a >

+ 1.5

2+1.0
a

<

CD

< +0.5
o

> 0.0

UJ

I

-0.5

• J ■ « t L t

MEAN HIGH TIDE
( BANJUL, GD)

o- — o © « _.

MEAN LOW TIDE
(BANJUL,GD)

M J J
MONTHS

N

FIG. 5. Water levels in the Gambia River based on Gambia Datum including mean
high and low tides at Banjul, and predictions of levels within the impoundment
upriver from the barrage at different demands for irrigation based on areas (ha)
cultivated with rice (from Coode and Partners 1977, Johnson 1978).

water at Kuntaur (231 km upriver) at a rate of 1 nr*s~^ during the dry season
would move the saline mixing zone upriver by 1 km (Johnson 1978). The salt
barrage is intended to prevent the upriver excursion of salinity which would
eliminate irrigation in historically freshwater areas, and to provide a
continuous source of water for rice fields in areas that originally could only
farm rice during the rain season.

The barrage will be built such that maximum water level will be +1.70 m
(GD), which is the approximate spring tide level in this area (Coode and
Partners 1977). The original hydrologic budgets of water table fluctuations
were based on irrigation demands for 24,000 ha of rice which would drop the
water table to -1.02 m (GD) by the end of May (Coode and Partners 1977; Fig. 5).
Water level changes based on different acreages of rice cultivation (Fig. 5)
would range from a low of +1.2 m (GD) if no rice was irrigated to the extremely
low levels estimated by Coode and Partners (1977) (calculated by Johnson 1978).
In all of these scenarios, minimum water levels were observed at the end of May,
and recovery of levels back to +1.70 m (GD) occurred by the end of August.
The threshold acreage of rice at which minimum water levels would be below the
natural tidal levels during May would be about 8,000 ha.

MANGROVE ANALYSIS

MANGROVE COMMUNITY TYPES

Lugo and Snedaker (1974) developed a classification scheme relating forest
physiognomy and functional characteristics among mangroves with their hydrology
and geomorphology. Six types of mangroves were identified (Fig. 6) that group
environmental factors such as tidal hydrology, rainfall, soil texture, soil
salinity, and juxtaposition to estuarine waters with the structure,

10

OVERWASH MANGROVE ISLANDS
1- overwashed by daily tides. 2-
high rate of organic exports. 3-
dominated by red mangroves but all
species may be present. 4-south
Florida, south coast of Puerto Rico!
5-sensitive to ocean pollution.

FRINGE MANGROVE WETLANDS
1-line water ways. 2-high rate of
organic exports. 3-dominated by rec
mangrove. 4- throughout south Flor-
ida, Puerto Rico and Florida's east
and west coast. 5-sensitive to
ocean pollution.

SCRUB MANGROVE WETLANDS
1-on extreme environments. 2-low
organic exports. 3-usually red or |2m
black mangroves. 4-southeast Florida
south coast of Puerto Rico, high
latitudes on west coast of Florida.
5-sensitive to further stress.

^_I s*K>^ £fcr*

/TV*!

HAMMOCK MANGROVE WETLANDS
1-on land rises in south Florida.
2-low export of organic matter.
3-all mangrove species. 4-south
Florida everglades. 5-sensitive to
fire and drainage.

RIVERINE MANGROVE WETLANDS
1-along flowing waters. 2-high ex-
port of organic matter. 3-all man-
grove species, reds predominate.
4-south Florida^ north coast of
Puerto Rico. 5-sensitive to
alterations of water flow.

r*>

18m

BASIN MANGROVE WETLANDS
1-in depressions or areas of slow
water movement. 2-high seasonal ex-
port of organic matter. 3-black
mangroves predominate. 4- inland
locations in south Florida and
Puerto Rico. 5-sensitive to alter-
ation of sheet flow, sea water in-
put, and prolonged high water.

FIG. 6. The six mangrove community types from Lugo and Snedaker (1974) .

11

productivity, and nutrient cycling characteristics of the forests. Although
this classification system was based on surveys mostly from mangroves in the
Americas, it has proven applicable to mangrove forests around the world; such is
the case for West Africa.

The mangroves along the Gambia River estuary can be grouped into four of
the mangrove types based on the scheme by Lugo and Snedaker (1974). The domi-
nant type of mangrove along the Gambia River estuary is the riverine mangrove
forest which occurs along most of its shoreline, except for near the mouth of
the estuary and up in some adjacent bolons. These forests are inhabited by all
three Rhizophora species found in this area and are characteristically greater
than 10 m in height; they are most noticeable from Tendaba to Elephant Island.
Inland of these riverine forests, which are inundated daily by tides, are basin
mangrove wetlands which are dominated by Avicennia . These areas are
infrequently flooded by tides (spring tides only) and consequently have a
different soil texture and chemistry than the riverine forests.

At the mouth of the Gambia River estuary there are two types of mangroves.
Fringe mangrove wetlands, dominated by Rhizophora mangle , occur along the
shoreline of bays and lagoons that have fairly constant annual salinity. The
tidal amplitude in this area is about 1.2 m, which would normally support
forests with tree heights >10 m. Yet because of the high evapotranspiration
rates characteristic of the Gambian climate, forests at the mouth of the Gambia
River estuary have tree heights from 5-7 m. These fringe forests are very
similar to those along the bays of south Florida, USA. In this same area, there
are stands of scrub mangroves occurring more frequently along the more inland
shorelines of the bays. Slight increases in the topography, due to storm
deposits or more inland juxtaposition from the estuary, along a coastline with

12

such high evapotranspiration rates result in extreme soil salinity conditions.
These scrub mangroves, dominated by Avicennia with minor occurrence of
Rhizophora and Laguncularia , are usually <2 m tall.

SPECIES DISTRIBUTION AND ZONATION

Mangroves along the coast of Senegal and The Gambia are the most northern
mangroves of the Atlantic type on the west coast of Africa located between 12°
and 16°N latitude (Marius 1981). These mangroves extend over an area of 500,000
ha (most of the area is unvegetated tidal flats), half of which are in the
estuary of the Casamance River and remainder in estuaries of the Gambia and
Saloum rivers. In the Gambia River estuary, mangrove forests are comprised of
six species including Rhizophora racemosa , R. harrisonii , R. mangle , Avicennia
africana , Laguncularia racemosa , and Cono carpus erectus . Speciation of
mangroves along West Africa is very similar to mangroves along the Atlantic
coasts of the Americas, yet different to mangroves colonizing the east coast of
Africa (Chapman 1976).

The typical distribution or zonation of these mangrove species along the
shoreline of the Gambia River estuary is also similar to observations along the
Americas 1 Atlantic coastline. In the mesohaline and oligohaline regions of the
estuary, Rhizophora racemosa occupies a thin strip adjacent to the riverbank.
This area of the intertidal zone is inundated twice daily by tides and is
commonly associated with recent deposits of soft alluvium of high clay and silt
mineral content. Yet these forests typically obtain tree heights greater than
20 m on an elaborate prop root system. Just inland of this fringe of R.
racemosa is a stand of R. mangle which is also located in that portion of the
intertidal zone flushed daily by tides. Mixed among both of these is R.

13

harrisonii . At the level of mean high water there is usually a mixture of R.
mangle and Avicennia africana which is less frequently flooded by tides compared
to the pure Rhizophora stands. This group of mangroves, from the monospecific
strip of Rhizophora racemosa to the mixed stand of R. mangle and A. africana ,
constitutes the riverine mangrove forests.

Inland of the mixed 11. mangle and Avicennia africana are normally mono-
specific stands of Avicennia which are located above mean high water levels.
This vegetation band marks a distinct transition zone in vegetation and soil
characteristics between the lower and upper intertidal areas. Associated with
this zone is usually a sharp rise in soil salinity due to the infrequent nature
of tidal inundation and high evapo transpiration rates (Fig. 7). This area of
mangrove may be classified as basin mangrove wetland.

Soil salinity continues to increase inland of this mangrove zone which is
indicated by a gradual decrease in tree height and density of A. africana as
distance from the estuary inland increases (Fig. 7). The adjacent more inland
zone is usually void of any vegetation due to hypersalinity (pH may drop but
salinity is considered a key factor) and is referred to as bare tanne. Inland
of this zone may be tannes vegetated with marsh plants including Sesuvium
portulacastrum , Philoxerus vermicularis , Paspalum vaginatum , Eleocharis mutata ,
an( * Eleocharis carribea . This zone is referred to as a herbaceous tanne.
These tannes represent the most inland extent of the intertidal zone. The se-
quence of these tannes (bare vs. herbaceous) may vary depending on topography
which controls where precipitation and runoff may concentrate, resulting in
slightly lower soil salinity. In many areas of the upper intertidal zone of
The Gambia, the hydrology of this area has been modified by dikes built by
humans for the production of rice.

14

MHW
MLW

8
6

MANGROVES

TANNES

RIVERINE ' BASIN

HERBACEOUS

BARE

E

CO

E

<
CO

| iooo r
Moo

ui

9 600

QE

q

x400
o

I 120
§ 40

55

j<__M K. If

JL

50 100

DISTANCE INLAND (m)

150

FIG. 7. Vegetation sequence of mangroves in the Gambia estuary including soil
chemistry at 0.5 m depth at distances inland from the estuary.

15

This vegetation sequence varies considerably along the Gambia River
estuary, most noticeably by the exclusion of the more inland mangrove associes.
In many areas, only the fringe of the riverine mangroves is present, forming a
thin strip of mangroves along the estuary • This sequence is not applicable to
mangrove forests along the mouth of the estuary where monospecific stands of II.
mangle form fringe mangrove wetlands along the bays. Also in this area are
scrub mangroves dominated with Avicennia which have no apparent zona t ion
pattern.

FOREST STRUCTURE

The diversity in structure of mangrove forests along the Gambia River
estuary is demonstrated in Figure 8 by the change in tree heights along the
longitudinal axis of the estuary. These data are personal observations (tree
heights measured with clinometer) at sites along the estuary during July, 1984,
and indicate higher forest structure in the mesohaline region of thfe estuary.
Also apparent is that the higher structured forests are along the north bank of
this estuary. This was most noticeable at about 150 km upriver at Bai Tenda and
Elephant Island where tree heights on the north bank were 20-30 m compared to
only 10 m on the south bank. As discussed above, mangroves at the mouth of the
estuary were normally 5-7 m in the fringe forests and only 1 m in the scrub
mangrove wetlands.

This diversity in forest structure is also reflected in a forestry survey
°* Rhizophora mangrove forest (>7 m tree height) along the Gambia River estuary
(Forster 1983; Table 1). The complexity index (Holdridge et al. 1971), which
uses the variables of tree density, basal area, species number, and tree height
to determine forest structure of Rhizophora forests in this survey, ranged from

16

(ui) S1H9I3H 33U1

17

-^

>>

$-1

E

CO

.*

3

4-J

^■^

CO

<u

5-i

q:

CD

>

Ui

•H

>

Pd

a:

Cd
•H

a.

43

3

&

cd

a

Ul

0)

o

j_j

z

SO

p

c

o

CO

r-4
cd

Q

CO

<u

4-J

•H

CO

u

cd

T3

<u

u

CO

cd

cd

a

CO

4-J

JB

60

•H

0)

JS

CD

CD

(4

H

00

•

W

P*

TABLE 1. Structure of tall Rhizophora mangrove forest (>7 m) along the Gambia
River estuary (from Forster 1983).

Western Lower River North Bank and MacCarthy
Division Division Island Division

Area
(ha)

4,692

4,291

Density
(N/ha)

371

385

Basal Area
(m 2 /ha)

11.8

15.7

Mean Diam
(cm)

20

23

Mean Height
(«)

11

12

Total Volume
(m 3 /ha)

103.0

153.2

Damaged Trees*
(% of density)

32.0

23.0

Dead Trees
(% of volume)

21.0

19.0

Structural Index®

0.48

0.73

Complexity Index &

1.92

2.92

5,461

472

30.3

23

13

208.9

21.0

8.0

1.25
5.00

Total

14,980
416

18.5

24

13

183.5

25

22

1.00
4.00

*including some signs of decay in the crown

^Structural Index - (Mean Height x Basal Area x Density) ? 1,000

Complexity Index » (Mean Height x Basal Area x Density x Species No.) r 1,000

18

1.92 to 5.00 with the lower number associated with the Western Division of the
estuary (south bank near the mouth of estuary). The average index for the
entire mangrove area along the Gambia estuary was 4.00.

Structural indexes of the mangrove forests above Yelitenda were calculated
from Johnson (1978) using size class distributions of three categories of
mangrove forests based on tree heights: Rhizophora , >20 m; Rhizophora , 7-20 m;
anc * Avicennia , >7 m. Using the median diameter at breast height (dbh) for each
5-cm size class starting at 10-cm dbh, and the density of each respective size
class, the basal area in a 1 ha stand of each mangrove category was determined
(Table 2). Assumptions of tree height are given in Table 2 along with cal-
culations of complexity index (Holdridge L971). Complexity indexes ranged from
4.4 to 11.2, with the tall Rhizophora forests having the greater structure.
These values are within a range of 2.3 - 72.0 for riverine mangrove forests in
Florida, Puerto Rico, Mexico, and Costa Rica (Pool et al. 1977; Table 2).
The most significant difference among these riverine mangrove forests is
density, with forests along the Gambia estuary being less dense.

AREAL DISTRIBUTION

Estimates of aerial distribution of mangroves along the Gambia River
estuary in Table 3 range from 45,000 ha (Giglioli and Thornton 1965) to 71,343
(Checchi and Co. 1981; from Johnson 1978 and Abell 1980). The most accurate
estimates based on planimetry of aerial photos are by Rodriguez-Bejaram (per-
sonal communiation) , Forster (1983), and Dunsmore et al. (1976) which range from
60,000 to 67,000 ha. Mangroves have been classified based on stand height (m)
with 7 m the delineation between tall and short forests (Table 4). Checchi and
Company (1981) used estimates above the proposed barrage at Balingho by Johnson

19

TABLE 2. Structural indexes (based on trees with dbh >10 cm) of mangroves in
the Gambia River estuary compared with riverine mangrove forests in North and
Central America.

Basal

Tree
Height

Density

(no./O.l

ha

Area
(m 2 /0.1
ha)

Complexity*
Index

Reference

The Gambia (above Balingho)
Rhizophora: >20 m

39

2.86

11.2

Johnson 1978

Rhizophora:

7-20 m

56

1.38

4.4

••

Avicennia:

>7 m

34

3.31

6.7

««

Florida, USA

Ten Thousand

Island

60

2.17

2.3

Pool et al. 1977

Puerto Rico

Vacia Talega

98

1.71

7.5

tt

Mexico

Roblitos

91

2.41

3.5

..

Isla La Palma
Rio de las Canas

145
103

5.59
5.61

41.3
27.7

Costa Rica
Mo in

Boca Barranca

118
66

9.53
3.18

72.0
6.0

Complexity Index - (Tree Height x Density x Basal Area x No. Species) rlOOO
(Holdridge et al. 1971)

20

TABLE 3. Estimates of mangrove distribution along the Gambia River
estuary.

Mangrove Area (ha) Reference

66,000 Brunt 1959

45,000 Giglioli and Thornton 1965

67,000 Dunsmore et al. 1976

71,343 Checchi and Co. 1981

66,900 Forster 1983

(1978) and estimates below the barrage by Abell (1980) to calculate mangrove
distribution by stand height for Rhizophora and Avicennia for the Gambia
estuary. Based on these estimates, nearly half the mangroves are >7 m (^36,000
ha) and 10% (6,125 ha) are >20 m. However, Forster (1983) found that about
52,000 ha or 78% of the total coverage was <7 m. He states that these estimates
are about double the estimates for short mangroves by Brunt (1959). Thus
Brunt's value of about 30,000 ha of mangrove <7 m was similar to those figures
presented by Checchi and Company (1981 )• These differences are extreme and
further analyses are needed to correct the confusion. This is a very important
statistic since the only estimates of litter productivity for the Gambia River
estuary are based on stand height.

Another important statistic concerning mangrove cover is the areal dis-
tribution for the two dominant genuses of mangroves in this estuary, Rhizophora
and Avicennia . Most of these forests are monospecific zones of each genus and
follow a zonation pattern of Rhizophora inhabiting the fringe mangrove zone
adjacent to estuarine waters, and Avicenni a occurring inland of the fringe zone.
This zonation pattern probably reflects some gradient in tidal hydrology (and

21

TABLE 4. Two estimates of the areal distribution of mangroves in the Gambia
River estuary. Classification system uses tree heights (m).

Classification

Region

<7 m

7-20 m

>20 m

Total

Reference

Estimate I

Below Barrage
Rhizophora
Aviceimia
Subtotal

11,880
21,590
33,470

18,330

8,310

26,640

2,520

2,520

32,730
29,900
62,630

Abell 1980

Above Barrage
Rhizophora
Avicennia
Subtotal

757
237
994

3,406

708

4,114

3,605

3,605

7,768

945

8,713

Johnson 1978

Total Estuary
Rhizophora
Avicennia
Total

12,637
21,827
34,464

21,736

9,018

30,754

6,125

6,125

40,498
30,845
71,343

Checchi and Company
1981

Estimate II

Western Division

16,800

4,

700*

21 , 500

Forster 1983

Lower River

Division

12,400

4,

300

16,700

North Bank

Division

22,600

5

,500

28,100

McCarthy Is.
Division

100

500

600

Total Estuary

51,900

15

,000

66,900

*Trees >7 m

22

thus other soil characteristics such as salinity, hydrogen sulfide, etc.), which
is significant to estimates of potential detritus export from these intertidal
forests. Thus estimates of areal distribution of each genus may enable a
prediction of detritus transport rates to the estuary. The only figures
available are from Checchi and Company (1981) which estimate that 57% of total
mangrove coverage, or 40,498 ha, are of the genus Rhizophora .

The impacted area of mangroves above the proposed barrage is 8,713 ha
according to Johnson (1978) and 7,930 ha from Rodriquez-Bejarano (personal
communication); and both estimates represent 12% of the total area according to
figures from Checchi and Company (1981), or 13% based on figures for total
mangrove cover by Forster (1983). The area above the barrage represents 59% of
mangrove trees >20 m, yet only 3% of the trees <7 m inhabit this area indicating
an unequal distribution of tree height along the salinity gradient.

Forster (1983) noted that their estimates for total mangrove cover were
similar to those by Brunt (1959), indicating that a large-scale loss of mangrove
forests may not be occurring. The two estimates do have large differences in
the distribution of mangroves according to the 7 m demarcation in tree height as
discussed above. Based on these differences, about half of the tall mangrove
trees (>7 m) have been eliminated since 1959 and replaced by mangrove forests
<7 m.

LITTER PRODUCTIVITY
Background

The purpose of studies on the production and decomposition of litter in
forests is to better understand the role of litter dynamics in such ecosystem
properties as nutrient cycling, succession, and stability (Bray and Gorham

23

1964). Litter dynamics have been extensively studied in terrestrial forests
(Jenny et al. 1949, Brown 1980), while less work has been done on the production
and turnover of litter in forested wetlands • Such information is particularly
needed for intertidal forested wetlands, such as mangroves, in order to
understand the coupling of these forests to the productivity of coastal
fisheries* Other interests in the litter production of mangroves include the
production of peat that occurs in many of these coastal swamps, and the
optimization of managed forests for exploitations such as charcoal, posts, chip
wood, and tanins.

Litter production values for mangrove forests worldwide range from 1.20
fha""^ # yr~^ for scrub mangroves in south Florida to 23.4 t •ha""-***yr~^ for a 20-
yr-old managed forest in Malaysia (Table 5). Grouping mangrove forests by the
scheme of Lugo and Snedaker (1974) used in Table 5 separates mangroves into
categories that represent forests with similar hydrography. There is a gradient
in the frequency in tidal inundation from scrub to basin to fringe mangrove
forests, and riverine mangroves are frequently inundated by freshnets from
rivers as well as tides. It has been suggested by Pool et al. (1975) that
litter production rates in mangroves are a function of water turnover within the
forest, and the rank of the means of litter production in Table 5 (riverine >
fringe > basin > scrub) supports this hypothesis.

An association between wetland hydrology and net primary production has
been observed for a variety of wetland communities including Spartina marshes
(DeLaune et al. 1979, Steever et al. 1976), Zizania aquatica (Whigham and
Simpson 1977), and freshwater forested wetlands (Connor and Day 1976, Brown
et al. 1979, Brown 1981, Brinson et al. 1981). Wharton and Brinson (1979)
suggested that the production of alluvial swamps was dependent on water

24

TABLE 5, Summary of total litter fall rates for different types of mangrove
forests.

Mangrove Type

Litter Fall
(f ha" 1# yr -1 )

Reference

Scrub Mangroves

Turkey Pt., SE Florida, USA
Turkey Pt., SE Florida, USA
Turkey Pt., SE Florida, USA

Mean + (SE)

2.71
1.20
1.68
1.86 (0.55)

Pool et al. 1975
Snedaker and Brown 1981

Basin Mangroves
Monospecific

Fort Myers, SW Florida, USA
Rookery Bay, SW Florida, USA
Rookery Bay, SW Florida, USA
Clam Bay, SW Florida, USA
S. Thailand
Mixed Forest

Fort Myers, SW Florida, USA
Rookery Bay, SW Florida, USA
Piiiones, Puerto Rico
Turkey Pt., SE Florida, USA

Mean + (SE)

Fringe and Overwash Mangroves

Ten Thousand Is., Florida, USA
Ten Thousand Is., Florida, USA
Turkey Pt., SE Florida, USA
Turkey Pt., SE Florida, USA
North River, Florida, USA
Cieba, Puerto Rico

Mean + (SE)

Riverine Mangroves

Ten Thousand Is., Florida, USA
Ten Thousand Is., Florida, USA
Chokolskee Bay, Florida, USA
Chokolskee Bay, Florida, USA
Gordon River, SW Florida, USA
Gordon River, SW Florida, USA
Vacia Talega, Puerto Rico
El Encanto River, Columbia
Sungai Merbok Estuary, Malaysia
Matang Mangrove Forest, Malaysia

Mean + (SE)

3.51

Twill ey et al. (in

press)

5.38

4.69

5.79

Heald et al. 1979

6.70

Christenson 1978

8.68

Twilley et al. (in

press)

7.51

9.70

Pool et al. 1975

7.50

Snedaker and Brown

1981

6.61

(0,

.70)

10.24

Snedaker and Brown

1981

9.81

10.82

7.71

Pool et al. 1975

8.76

Heald 1969

6.64

Pool et al. 1975

9.00

(0,

.72)

10.66

Snedaker and Brown

1981

11.73

11.75

Sell 1977

11.83

14.43

9.09

14.45

Pool et al. 1975

14.09

Hernandez and Mullen 1979

10.07

Ong et al. 1981

23.4

Ong et al. 1982

12.98

(1.

.01)

25

movement, not only as a source of silts and clays, but also a supply of
nutrients and aeration for optimal growth. Together with the trends observed
for different groups of mangroves, these results for a variety of wetlands
suggest that the chemical and kinetic properties of tides in intertidal
environments may ultimately be expressed in the magnitude of litter and/or net
primary production of the plant community.

The Gambia

There is no published information on the magnitude of litter productivity
of mangroves along the Gambia River estuary. In order to evaluate the potential
importance of these forests to estuarine resources, litter productivity is a
very important index. This report estimates litter productivity of mangroves in
the Gambia estuary using an association between forest structure and litter
productivity based on information for mangrove forests in United States, Puerto
Rico, and Costa Rica (Pool et al. 1975, Pool et al. 1977). An index of forest
structure for eight sites in these three provinces was based on tree height
(Ht), basal area (BA), and tree density (D) for 0.1 ha plots using the formula:
(Ht x BA x D) 1,000. This is similar to Holdridge et al.'s (1971) complexity
index, with the omission of the number of species present. Structural index (X)
for each site (Pool et al. 1977) was positively correlated with the litter
productivity estimates (Y) made at each site (Pool et al. 1975) (Fig. 9).
The equation [Y » 4.89 (X) + 5.05, r 2 « 0.85] results in a Y intercept of + 5.05
fha""^ # yr~*. Such a high intercept is due to the fact that structural
statistics were gathered only on trees with dbh >10 cm. Since all forestry
statistics for mangroves in The Gambia are based on this dbh, this relationship
was used.

26

>

O

=>
Q
O

ac
a.

<r

UJ

0.5 1.0 1.5 2.0

STRUCTURE INDEX (DBH > 10 cm)

FIG. 9. Relationship among structural indexes of mangrove forests and their
litter productivity estimates •

27

Using this relationship between forest structure and litter productivity,
estimates of the productivity of mangroves along the Gambia River estuary were
made based on previously discussed forest structure statistics (Table 6).
The structural indexes for the four categories of mangrove forests (based on
tree heights) used by Johnson (1978) are: Rhizophora trees >20 m s 2.81;
Rhizophora trees 7-20 m « 1.09; Avicennia trees >7 m * 1.34; and Avicennia plus
Rhizophora trees <7 m « no forestry statistics available. The resulting litter
productivity estimates varied from 10.4 to 18.8 t •ha~**yr""% which are within
the range for riverine mangroves listed in Table 5 (Twilley et al. in press).
The litter productivity value for the forests with tree heights <7 m was based
on a value for a mixed forest in South Florida that upon observation by the
author was similar to the structure of these forests in The Gambia (Fort Myers,
SW Florida, USA; See Table 5). This value is considered a conservative
estimate.

The areal production of litter for mangrove forests above and below
Balingho are calculated from areal distributions estimates of mangroves by
Johnson (1978) and Abell (1980), respectively. The total production was 583,670
fyr~l, of which 115,380 f yr"^ (20%) occurred in the impacted area above the
proposed barrage at Balingho (Table 6).

Estimates were also made of litter productivity of Rhizophora forests with
tree heights >7 m in The Gambia using forestry statistics by Forster (1983).
This category of forests produced 132,780 f yr" 1 of dry litter mass, most of
which occurred in the North Bank and MacCarthy Division of The Gambia (Table 7).

28

TABLE 6. Estimates of litter production (dry mass) for the mangroves above
and below Yelitenda based on information from Johnson (1978) and Abe 11 (1980).
Mangrove categories are based on species and tree heights (m) .

Category

Structural
Index

Production

Estimate

(fha^'yr" 1 )

Area
(ha)

Annual

Production

(f yr" 1 x 10 3 )

Above Yelitenda

Rhizophora: >20

m

2.81

18.8

3,605

67.77

Rhizophora: 7-20

m

1.09

10.4

3,406

35.42

Avicennia: >7

m

1.34

11.6

708

8.21

Avicennia plus
Rhizophora: <7

m

NA

4.0

994

3.98

Subtotal

9,713

115.38

Below Yelitenda

Rhizophora : >20 m

Rhizophora : 7-20 m

Avicennia : >7 m

Avicennia plus
Rhizophora : <7 m

Subtotal

Total

2.81

18.8

2,520

47.38

1.09

10.4

18,330

190.63

1.34

11.6

8,310

96.40

NA

4.0

33,470

133.88

62,630

468.29

71,343

583.67

29

TABLE 7. Estimates of litter production (dry mass) for Rhizophora mangroves
with tree heights >7 m based on data from Forster (1983).

Production Annual

Structural* Estimate Area Production
Category Index (fha^'yr" 1 ) (ha) (fyr"" 1 x 10 3 )

Western Division

0.48

7.4

4,692

34.72

Lower River Division

0.73

8.6

4,291

36.90

North Bank Division

1.25

11.2

5,461

61.16

MacCarthy Division

1.00

9.9

Total

14,444

132.78

*Structural Index =* (Basal Area x Tree Height x Density) r 1000; based on
0.1 ha plots

EXCHANGE OF DETRITUS AND NUTRIENTS

Estuarine ecosystems are coupled to intertidal wetlands by tides that
create a two-way flux of materials. Problems associated with the measurement of
these tidal fluxes have resulted in a controversy surrounding the hypothesis
that intertidal wetlands affect the productivity and nutrient cycles of
estuarine waters (Teal 1962, Odum and de la Cruz 1963, de la Cruz 1965, Day et
al. 1973, Haines 1977, Haines 1979, Nixon 1980, Odum 1980). Estimates of
organic carbon export from intertidal wetlands range from 45% (Teal 1962,
Axelrad et al. 1976, Roman 1981) to less than 1% (Heinle and Flemer 1976) of
their net production, while other intertidal wetlands may import organic carbon
(Woodwell et al. 1977). The present view is that the exchange of organic carbon
in intertidal wetlands may be site specific depending on the geomorphology and
tidal hydrology of the region (Mann 1975, Odum et al. 1979, Welsh et al. 1982).
Mangroves represent nearly 75% of the intertidal vegetation in the tropics and

30

the exchange of detritus and nutrients with these forests may strongly influence
the productivity of tropical coastal ecosystems.

Detritus

Background * Organic carbon export estimates for riverine and fringe man-
groves represent rates for intertidal wetlands regularly inundated by tides.
In an Australian riverine mangrove forest, particulate organic matter export was
420 gC # m" 2, yr -1 f which was about 94% of litter fall (Boto and Bunt 1981).
Particulate detritus export from fringe mangroves in south Florida was estimated
at 186 gC*m"" 2# yr"' 1 f or only 42% of litter fall (Heald 1969). The tides in the
riverine forest were 3 m/tide, for an annual tidal amplitude of about 2,190 m;
compared to 0.5 m/tide in south Florida or 355 m*yr""^ in the fringe forest.
Both of these export estimates are for leaf litter and particulate carbon export
only, and do not include export of DOC to adjacent estuarine systems.

Total organic carbon export from infrequently flooded basin mangroves was
64 gC'nT^yr" 1 (Twilley 1985). The annual tidal amplitude inside the basin
forests was only 12 m, and as a result only 22% of the litter fall inputs to the
forest floor were exported (Twilley 1982). In forested wetlands with no tidal
influence, TOG export was only about 6% of the litter production rate (Day et
al. 1977, Mulholland 1981). These comparisons suggest that tides influence what
proportion of litter produced in mangroves is exported, and that the magnitude
of this organic carbon export from mangroves is directly related to the
cumulative tidal amplitude within the forests (Fig. 10).

Nearly 75% of the TOC exported from the basin mangrove forests was
dissolved organic carbon. DOC also dominated total organic carbon flux from
intertidal marshes (Axelrad et al. 1976, Happ et al. 1977, Pendleton 1979, Roman

31

400-

Al

1

E

300

o

9

m

K

200

(E

O

Q.

X

LU

100

10 100 1,000 10,000

CUMULATIVE TIDAL AMPLITUDE, m/yr

FIG. 10. Annual net export of organic carbon from mangrove forests in relation
to the cumulative tidal amplitude per year in the forest. Export values (left
to right) are for basin (Twilley 1985), fringe (Heald 1969), and riverine (Boto
and Bunt 1981) mangroves.

32

1981) j, and more than 93% of the total organic carbon export from bottomland
hardwood swamps (Mulholland and Kuenzler 1979). Mulholland and Kuenzler (1979)
attributed higher DOC export from forested wetlands compared to uplands to
higher water residence time in wetlands together with leaching of organic carbon
from the litter* Leaching of DOC from mangrove leaf litter, particularly from
black mangrove leaves, accounted for the initial high loss of leaf dry mass
during; decomposition (Twilley 1982), thus providing a source of DOC to the
surface water of these forests. The effect of water residence time on DOC
concentrations coincides with the observation in basin mangrove forests that
these concentrations decreased during August and September when tidal frequency
increased (Twilley 1985).

Rainfall also increases the export of DOC from basin mangroves, similar to
results for salt marshes (Harris et al. 1980, Roman 1981). The inputs of
organic carbon from the forest canopy to the surface water via stemflow and
throughfall increased the DOC concentrations in the surface water. This mech-
anism, and probably others, account for DOC being the major form of organic
carbon export from most intertidal wetlands, which has only been substantiated
in recent studies. During the infancy of the "outwelling" concept of detritus
flux from intertidal wetlands only particulate detritus was measured, thus
grossly underestimating the potential contribution of organic carbon from these
systems to estuarine waters.

Rates of organic carbon export from mangroves are dependent on the volume
of tidal water inundating the forest each month, which may be influenced most by
tidal frequency (Twilley et al. in press). As a result, export rates are
usually seasonal in response to the seasonal rise in mean sea level (msl), with
greatest quantities of export occurring from August to October and lowest values

33

from December to May for mangroves in South Florida* This supports the proposal
by Kjerfve et al. (1978) that an increase in msl during the summer may influence
material exchange in tidal creeks as a result of increased tidal inundation of
wetlands along the southeastern coast of the USA.

The residence time of litter on the forest floor of mangroves is controlled
by the relative influence of leaf decomposition and export. A comparison of the
litter production, storage, and export characteristics of riverine (Boto and
Bunt 1981), fringe (Heald 1969, Pool et al. 1975), and basin mangroves (Twilley
et al. in press) summarizes the effect of these two factors on litter dynamics
among mangrove forests (Fig. 11). The annual estimates of cumulative tidal
amplitude in these three forest types is 2,190 m, 355 m, and 15 m, respectively.
As discussed by Twilley (1985), the export of organic carbon from mangroves is
associated with increased tidal influence, as may be litter production. Yet the
two are not proportional, since litter production in the riverine forest is 1.7
times that of the basin forest, while the difference in litter export is 7.3
times higher in the riverine forest. As tidal influence within mangrove forests
increases, leaf export dominates organic carbon export, and thus the ratio of
organic carbon export to litter production increases. Such ratios are 0.94 in
riverine forests (Boto and Bunt 1981) where leaf litter on the forest floor is
negligible. Consequently, leaf decomposition rates for Rhizophora leaves are
not proportional to residence times of litter in either fringe or riverine
forests since most of the leaf litter is exported.

For basin mangrove forests, organic carbon export is only about 20% of the
carbon input to the forest floor via litter fall (Twilley 1985), and litter
standing crop ranges from 100 to 500 g dry mass # m~2. Although the seasonal
standing crop values of litter decreased as tidal influence increased, this

34

LU

±:q. 10

LL.

uj

s

e

CO

ro

CO CO
(1)

CQ M-l

CO o
O

rH U
CD

C U

CO tH

•-N CD
O 4J

u

60 *H

d J*

d <u

CO J->

4J CO

CO «H
O

JH O

CD CQ

•U CO

4J CO

U
CD

o

d o

€
o
J-l

CD

u

CO

0) 0) CO
>> CO 60

CD

4-> i-i

CO o

CD U

U CO

o

d d

•H ^
CO CO
CO CO

CO 4J M-4

w d ^

o m

W «H N

CD W ON

MOH

CO 3

JH T3 •

O O iH

4J *j CO

co a

4J
* 4J CD

/"n co

H (U H

rH a o
CO O

^ TJ CU

d

l-i CO *

cd a\

4J 3 vO

CO

4J

•H T3

CO rH

O CO

4J G* CD

3 SB
d. O >«■•

c a

•H CD CD
TJ 60

d d

O W-H
,£> C M

CO TJ

a 3 *

iH ^

a a rH

•H COO

CO rH

60 CO
U CD
O CO
CO
<4-4 O
O rH

C
3

>» U 3 <*h

CO
6

a

3
CO

O

CD

4J O
O 4-1

o

d ^

CO
CD

U «H
O M
CX CD
X >
CD *H

O
CD

a

CO

o

d
o
a

CO

>, s

a co
o

O CD
00 «H

d S

•H
T3 CD

d £

CO 4J
CO H

o

d

CO co
u

iH rH
rH 3
CO CO
(4-i CD
W

CD Tj

4J CD

4J JS •
•H 60rH
rH «H |
CD W

CD > >>
*C •

H cOCN
CD i

m a

• CO •

4J O

CO d 60
CD O

U 3

O T3 H

CD

CO CO

cd co 3
> -Q o

O rH

U rH l*H

2P

CO co d

a <u «o

u

<^ OCN
CO Ci< I

cq a

CD >> •
*-■ CO CJ
O.CQ* 60

35

response can be explained by the effects of hydrology on leaf decomposition
(Twilley et al. in press). Export of leaves from basin mangroves was negligible
during the summer (Twilley 1985) when minimum litter values were observed*

The residence times of leaf litter in basin mangroves vary from 0,13 to
0.51 yr (Table 8). Leaf litter residence times for monospecific basin forests
at Rookery Bay are similar to values for riverine forests, which indicates that
under some conditions the relative influence of export and decomposition on
litter dynamics may be equal. However, as indicated in Figure 11, the short
residence time of leaf litter in riverine systems represents a loss from the
system, while leaf fragmentation in basin forests results in the formation of
peat deposits which are characteristic of these inland forests (Davis 1943).
The organic carbon content of alluvial mangrove soils are only 4-5% (dry mass)
(Hesse 1961, Giglioli and Thornton 1965), compared to about 20% for basin
mangrove forests (Coultas 1978, Twilley 1982).

The Gambia . Estimates of detritus export from mangroves along the Gambia
River estuary were based on litter productivity values calculated above, and
on assumptions of the percentage of this litter that is exported annually.
As previously discussed, these export : production ratios for litter are strongly
influenced by tidal hydrology, and range from 0.20 for basin mangrove forests to
0.94 for riverine mangroves. Based on these considerations, and a tidal ampli-
tude of 1.2 m in the riverine forests of the Gambia estuary, an E:P ratio of
0.75 was applied to forests in the inter tidal zone flushed daily by tides
(Rhizophora >20 m and Rhizophora 7-20 m) . Avicennia forests and mixed forests
°f Rhizophora and Avicennia inhabit more inland areas of the intertidal zone and
thus E:P ratios of 0.5 were applied to these sites. Based on these E:P ratios

36

00 •

c <~»

o a

iH w
CO

CO
CO 4J

o) x:

> 60

O -H
U CI)

oo,n
cs

CO cu
6 cu

at

O
U T3

c

O CO

JQ CU

U -H

CO O

o cu
a

CO CO

co
3

T3 O

c

CO T3

CU

CO CO

CO CO

CO J2
B
cu

>s U

U CO
T3

CO
CO CU

CO -H

u

u o
^ 00

o cu
a- 4J

X CO

cu a

CO cu

3 >

4J O

•H U

U (30

4J C

CU «0

-a s

CO u
CU CO

jj 3

CO 4J

CO

•H CU

4-)

CO J-l

W CU

>

•H
• ft*
00

CO

<: co

o
a

w

CO

o

* -4

o

,fir-4

u I

CO U

ro

O

CO

i-4

CO

CO

X

X

r-»

>>l

M

H

P

>S

•K

P*

•• CO

co
o

iH X
CO

O I

CO I

CU U

<q •

r-4

M I

CU CO

P4 .C

O

CM

on

no
o>

f~* o

00

o

o
sr

o

co

nO

oo

NO

en

CM

CM

CO

vO

as

o

o

00

CO
00

wo

•

o

•

vO
CM

OS
ON

o

co
oo

o
m

o
m

co

ON

CM

WO

O

CM

00

o

on

MD

vO

o
m

m

N£>

•

CO
ON
CM

co

u
o

00

cu
u
CO

r**

CM

i-i

00

00

r^

>d-

CM

ON

CO

•

•

•

•

•

r*»

m

00

CO

m

NO

CO

i-4

00

co

co

vO

o

ON

o

NO

a\

oo
oo

on

CM

co oo

CO VO

r-4 <r

NO

CO

oo

00
00

B
o

CM
/\

CO

C ••

<D ^ CO

U U

•H O

<u a

JH O

N

CU «H

§ s

«5

O

NO

o

00

-a-

vO

o

•

•

•

•

00

o

r-4

<*

o

CM

I

V

co

3

••

a

CO

CO

••

p4

u

CO

co

O

o

•H

•H

X

X

c

cs

a

a

c

a

o

o

cu

cu

N

N

o

o

•H

•H

•H

•H

^3

x:

>

>

CCS

&

<*

<s

CO

o
u

C73

CO
CU

CU

O
•H

CU
PQ

O
CM

O
CM

I

V

co

3

••

••

a

CO

CO

CO

• •

u

14

U

CO

CO

O

o

o

•H

•H

£.

x:

J3

c

C

a

a

a

a

C

o

o

o

CU

CU

N

N

N

o

o

•H

•H

•H

•H

•H

jC

42

•^

>

>

peS

Pd

cd

<s

<J

*o

3
CO

o

CU

00

c

•H

U
O
O
O

CO

CO

a

3
O

00

CU

%

u

00

c

CO

a

cu

CO

cu

x:

o

o

3
O

a

u
cu

o
a

XJ

CU ^N

CO

M-4 CO

O CU

u
o a

•H

CO «H
0^ >-^

CM

00
ON

cu

co

CO

co

6

u

e

CO

u

00
CU

a

oo

M-4

o

c
o

■H
iJ
CO

V4

4J

c

cu

o

fi
o
a

3
O

,o

u

CO

a

oo

c

•H

s

3

CO
CO

<

■K

37

and litter productivity estimates for each category of forests, the export of
detritus from mangroves of the Gambia River estuary was calculated for areas
above and below Balingho (Table 8). Litter dry mass is usually about 48% carbon
(Twilley 1982); from this assumption estimates of organic carbon export from
these riverine mangroves were made. Total export was 181,040 tC-yr"" 1 , of which
40,080 t* yr" 1 (22%) occurred from above Balingho. Based on a total area of
mangroves along the Gambia River estuary (Johnson 1978, Abell 1980; equals
71,343 ha), detritus export is estimated at 289 g C-nT^yr"" 1 (per m 2 of
mangrove) .

Cumulative tidal amplitude in the mangroves of the Gambia estuary is
876 nryr*" 1 (1.2 m amplitude x 2'day"" 1 x 365 d'yr"" 1 ). Using the relationship
between tidal hydrology and detritus export in Figure 10, the magnitude for
export from mangroves of the Gambia estuary should range from 200 to 400
gC*m~ 2 *yr~*-. The estimate above based on litter production and export to pro-
duction ratios falls within this range.

Annual patterns in precipitation and levels of spring high tides (Fig. 12)
indicate that detritus export would be maximum from July to October.

Nutrients

Introduction . Intertidal forested wetlands may function as either a source
or a sink of nutrients to estuarine waters. Nixon (1980) argued that the
paradigm in ecology that wetlands are a nutrient sink contradicts the "out-
welling" concept of marshes proposed by Odum (1980) . Yet in his review, Nixon
(1980) summarized that the "outwelling" of detritus from marshes may be about
100 gC nf^yr* 1 , less than previously estimated, while both the net flux of
nitrogen and phosphorus may be from coastal waters into the marsh. Thus in-

38

4J

U

o
u

4

(0S6I-988I '39VU3AV J*fr9)

(inrNva'ww) nvdNivy

o

00

in

o

10

nnrisiva 4 w) 3011 oNiads nvsw

u
o
£4

e

O

CO

0)

SO

c

•H
V-i

CO
O

GO

co

T3

T3

C
co

a
o

>>

CU

M

V4

CO

a.

3

u

y-i

CO

<u

CO

J-»

cs

CU

^

>

<u

•H

4J

(^

4J

CO

CO

a-H

£>

tH

n

CO

CO

3

a

c

cu

<J

•c

4-i

•

a

CN

•H

«— 1

/■N

•

CO

O

00

M

ON

fa

1— »

39

tertidal wetlands may be able to function both as a source of detritus and a
sink of nutrients in estuarine ecosystems. However, as discussed above for
detritus flow, the function of a system as sink or source of nutrients is
probably dependent on site-specific characteristics such as regional hydrology
and geomorphology (Odum et al. 1979).

The net flux of nutrients among intertidal wetlands and coastal waters may
depend on the transformations of nutrients among inorganic and organic forms,
and the fate of these forms of nutrients in the marsh. Two major mechanisms
associated with the loss of nutrients within the marsh are burial of organic and
inorganic materials, and gaseous exchange with the atmosphere.

Burial is associated with the accumulation of organic matter from plant
material that remains in the marsh following losses due to decomposition and
export. Recently it has been demonstrated that the belowground biomass of
intertidal wetlands may be more productive than aboveground structures, and thus
represents a large potential source of organic matter for the accumulation of
peat. Nutrients assimilated by the biomass during the growing season are lost
from the system via burial of this plant detritus deep into the sediments.

Sediments entering the marsh from adjacent estuarine waters may also be
transporting nutrients adsorbed to their surfaces, and these particles are
deposited and buried in the peat of intertidal wetlands. These adsorption
mechanisms are also associated with organic matter within the marsh, and have
been shown to be important in the accumulation of metals in marsh systems.
Thus nutrient sinks via burial may result from the absorption and adsorption
of nutrients entering intertidal wetlands from either coastal waters and/or
upland runoff.

40

For carbon, sulfur, and nitrogen, the net exchange of nutrients with the
atmosphere may be paramount to understanding the source or sink characteristic
of an intertidal wetland. Since nitrogen is most often indicated as potentially
limiting primary production of estuarine ecosystems, we will emphasize its
atmospheric cycle in this discussion. Large pools of organic material and
bacterial biomass in the top decimeter of the marsh surface result in a high
demand for electron acceptors. Denitrif ication is a collection of processes by
which nitrate (NO3) is utilized as an electron acceptor and nitrogen is
converted to either N2O or N2. The latter form of nitrogen gas may be the
dominant product of denitrif ication in marine and estuarine intertidal wetlands.
Since N2 is highly insoluble and so little of it is recycled by nitrogen
fixation, denitrification represents a potential loss of nitrogen from the
system. Denitrification rates are dependent on the supply of nitrate, thus
nitrification rates (NH4 ■> NO3) and/or the input of NO3 from estuarine waters
and/or upland runoff, strongly influence the magnitude of this nitrogen sink in
intertidal wetlands.

The magnitude of nutrient loss via burial and atmospheric exchange relative
to the recycling of nutrients within the wetland, along with the net exchange of
nutrients with estuarine waters at the wetland boundary, will determine the
function of these systems as a nutrient source or sink in estuarine ecosystems.

The Gambia . Sources and sinks of nutrients in the Gambia River estuary
were evaluated using mixing diagrams of dissolved nutrients during October and
November 1983 (Berry et al. 1985). Concentrations of nitrogen, phosphorus, and
total suspended solid were determined on samples collected along longitudinal
transects of the estuary as well as along the Bai Tenda bolon during cruises

41

aboard the R/V Laurentian . Mixing diagrams are plots of salinity along the "X"
axis and the nutrient concentrations along the "Y" axis. A straight line
between the two end members is called the theoretical dilution line and
represents concentrations based on dilution along the salinity gradient.
Variation from this theoretical line represents concentrations that result from
processes that cause either a sink (below the theoretical line) or source (about
the theoretical line) of that nutrient along the estuarine axis (see Liss 1976
for discussion).

Dilution plots of nitrate along the Gambia River estuary indicate a source
of this nutrient occurs in the oligohaline regions of the estuary during
October. This source of nitrate is due to nitrification in the water column as
indicated by a decrease in ammonium concentrations and a sag in the percentage
of saturated dissolved oxygen in this upper region of the estuary. Nitrate
concentrations reached nearly 300 yg'L""^ near Elephant Island about 200 km from
the estuary mouth (Fig. 13). By November, the source of nitrate in low salinity
waters was diminished with peak concentrations of only 225 yg'L"* (Fig. 13).
This dilution plot also indicates that a process resulting in a nitrate sink may
be occurring further down the estuary. Phosphate also is being produced along
the estuary in October and November, particularly in the mesohaline areas where
concentrations may reach 25 yg # L~l. Dilution plots of total suspended solids
indicate that the estuary retains a major portion of the sediment load within
the system. Sediment deposition was most noticeable in the oligohaline areas
where concentrations decreased from 100 to about 25 mg'L"" 1 .

Mixing diagrams were also used to evaluate the net flow of nutrient ex-
change in the mangrove bolons (Fig. 14). Changes in nutrient concentration at a
station during a tide may only reflect the movement of nutrients in a parcel of

42

OCTOBER

NOVEMBER

300

200-

9
3

UJ

5 1001

Z

£ 100 -

CO

Q

8 75

Q
UJ

a

z
u

Q.
CO

3
CO

o

O

•

300-

200

100 -

50-

25"

" X

•

1 , L_

100

~ 20 -

;>n -

10 -■

■ M

1

30
SALINITY (%o)

10

20

30

FIG. 13. Mixing diagrams for nitrate, total suspended solids, and phosphate
along the Gambia River estuary during October and November, 1983.

43

OCTOBER

150

"" ^^ #

100

-J

9
9

ui 50

CC

2

1

1

NOVEMBER

_ 10-

9
3

UJ

<
X
Q.
CO
O
X
0.

1000

J_

200-

100-

8
6

41-

2 "

1500 2000 3500 4000

CONDUCTIVITY (umhos/cm)

A

A

1

1

4500

FIG. 14. Mixing diagrams for nitrate, ammonium, and phosphate along the Bai
Tenda bolon in the Gambia River estuary during October and November, 1983.

44

water and not any net change due to processes within the bolon. Evapotrans-
piration causes higher salinities (conductivities) in the headwaters of a
mangrove bolon, thus salt can be used as a tracer of water exchanging with
mangroves during a tide* Using conductivity and nutrient concentrations of
water at five stations in the Bai Tenda bolon during ebb tides, dilution plots
were made to evaluate the fate of these nutrients in the bolons. Negative
slopes were observed for nitrate and phosphate indicating that concentrations of
these nutrients are much higher in the estuary than in the bolon headwaters.
This was particularly evident for nitrate that varied from 210 yg # L~^ at the
mouth of a bolon to about 10 yg*L""^ in the mangroves. For both these nutrients,
dilution plots indicate that concentration may be only attributed to dilution
during October; yet during November, a sink for both nutrients was observed
along the bolon. A loss of suspended solids was also observed within the bolons
with deposition occurring in the headwaters. In contrast to these parameters,
diagrams of ammonium indicated that this reduced form of nitrogen was being
produced within the mangroves, probably a result of high mineralization rates
associated with these organic soils.

These dilution plots suggest that mangroves along the Gambia River serve as
a nutrient sink for nitrate and phosphate, and trap total suspended material
from estuarine waters. These three parameters have peak concentrations along
the mesohaline and oligohaline regions of the estuary where mangrove forests are
most prominent. Nitrate may be utilized in the sediments of these plant
communities as an electron acceptor (denitrif ication) and phosphate may be
adsorbed by sediments or organic material. Mangroves have been recognized as
contributing to the stabilization of shorelines by accumulating riverborne
sediments. However, these functions are only suggested by the data analysis

45

presented, and the role of these forested ecosystems in nitrogen, phosphorus,
and sediment budgets of the Gambia River ecosystem awaits more complete
investigation.

ALLOCHTHONOUS ORGANIC MATTER
Background

There are three major sources of organic matter utilized by secondary con-
sumers in estuaries: autochthonous production from phytoplankton and benthic
autotrophs (algae and seagrasses), allochthonous detritus from upland watersheds
(riverine input), and finally allochthonous detritus from intertidal wetlands.
Collectively, these pools of organic matter support the high consumption of
energy characteristic of estuarine ecosystems. Many economically important
fisheries, such as shrimp, use the estuary for a nursery consuming large quan-
tities of food during their most active stages of growth. Thus, the sources of
organic matter in estuaries must be sustained in order to maintain these high
rates of secondary productivity.

Much controversy has existed over the relative importance of these three
sources of organic matter in estuarine ecosystems. Salt marshes were estab-
lished as the major source of detritus sustaining secondary production in
estuarine waters based on their high net primary production and detritus export
rates (Teal 1962). Turner (1977) has also demonstrated the relationship between
fishery yields in some coastal waters (kg of shrimp harvested) to the amount of
marsh land in the area. However, Haines (1977, 1979) has argued that detritus
from phytoplankton and riverine sources may be more important than marsh
detritus in estuaries. She found that the del 13 C ratio of detritus in
estuaries reflected the C-3 autotrophic pathway of phytoplankton and forested

46

wetlands rather than the C-4 mechanism associated with saltmarsh net production.
Other investigators using del ^C data have observed a diversity of ratios
implicating all three sources of nutrition for secondary consumers depending on
which source of detritus was most available (Macko and Zieman 1983). Sources of
detritus can vary considerably among estuaries and are related to the
geomorphological, hydrological, and physical characteristics of the system (Mann
1975, Odum et al. 1979, Welsh et al. 1982).

The importance of a source of detritus in an estuary may thus be related to
its contribution to the organic matter pool within the system. In south
Florida, Macko and Zieman (1983) concluded from del *-*C ratios that the major
source of nutrition for penaeid shrimp in Rookery Bay was mangrove detritus.
Based on a budget of autochthonous and allochthonous inputs of organic matter to
this estuary, Twilley (1982) also concluded that mangrove detritus was a major
source of energy for secondary productivity in this system. Since detritivores
are basically nonselective in acquiring their food, compared to herbivores,
their dependence on a specific source of detritus may be related to its
availability. Thus a budget of the relative sources of organic matter from
autochthonous net production, river discharge, and mangrove export to the pool
of organic matter in the Gambia estuary may reflect the importance of these
sources of detritus to the fisheries of this system.

The Gambia

The annual net production of phytoplankton of the Gambia estuary was
estimated from data collected on four cruises on the R/V Laurentian (Healey et
al. 1985). The estuary was divided into three segments (Lower - Banjul to
Mootah Point; Middle - Mootah Point to Kauntaur; Upper - Kauntaur to Goulombo)

47

and net production was measured at four times during the year within each
segment using ^C uptake in light and dark BOD bottles incubated at various
depths at each station. Net production was converted to a daily per m^ rate by
integrating production with depth and assuming a production period of 9 h*d~ .
Mean net productivity ranged from 0.7 to 152.9 mg G •m~2*<i~l for the estuary,
and these minimum and maximum rates occurred in the mid-estuary stations
(Table 9). There was no consistent seasonality with minimum and maximum rates,
except that lower rates in the mid and upper estuary stations were associated
with higher sediment load during high river discharge. Annual net production
was 20,45, 25.54, and 10.55 gC*m~2*y r ~l a t the lower, middle, and upper estuary
stations of the Gambia River (Table 9).

Based on these rates, and the areal dimensions of each segment, net primary
productivity of the Gambia estuary was estimated at 14,782 tC'yr"" 1 (Table 9).
Nearly 57% of this total occurred from Banjul to Mootah Point. Riverine input
of organic carbon to the Gambia River estuary at Goulombo is estimated at 15,960
tC*yr~^ (Lesack et al. 1984). Allochthonous inputs of organic carbon from
mangrove wetlands along the Gambia estuary were estimated at 181,040 tC'yr" 1
(Table 10). Thus, the total amount of organic carbon available for secondary
production within the Gambia River estuary annually is about 211,782 tC
(Table 10). Of this total, about 85.5% is contributed by the mangrove forests,
while phytoplank tonic and riverine sources account for about 7% each. There are
no estimates for benthic net production from algae and/or seagrasses, yet this
is probably minimal due to the high turbidity of this system.

Estimates for phytoplankton and riverine inputs are based on actual organic
carbon measurements. Inputs of detritus from mangroves are based on estimates
of litter fall and export: litter production ratios (these export ratios are

48

TABLE 9. Seasonal and annual estimates of net primary productivity ( 14 C
uptake) of phytoplankton in the Gambia River estuary. (Lower - Banjul to
Mootah Point; Middle ■ Mootah Point to Kauntaur; Upper » Kauntaur to Goulombo).

Description

Lower

Middle

Upper

Sum

August
October
December
March

Seasonal Rates (mg C'nf^d"" 1 )

52.2 121.5 62.3
78.0 0,7 12.7
64.0 2.5 13.5

30.3 152.9 26.5

Per m 2

(gC-m^-yr"" 1 ) 20.45

Per segment
(tC-yr" 1 )

8,405

Annual Rates

25.54 10.55

5,849

528

21.42

14,782

TABLE 10. Estimates of annual inputs of total organic carbon to the Gambia
River estuary (Banjul to Goulombo) from net primary primary production of
phytoplankton, riverine discharge, and mangrove export.

Source

T0C Inputs
(tC/yr)

Reference

Phytoplankton
Gambia River
Mangroves

14,782

15,960

181,040

Healey et al. 1985
Lesack et al. 1984
This study

Total

211,782

49

based on tidal amplitude). The estimates of detritus export from mangroves
along the Gambia River estuary are within the range for mangroves under similar
tidal amplitude. Even if these estimates are off by 50% (90,520 tOyr" 1 ),
mangrove detritus would still represent nearly 75% of the total organic carbon
pool available for secondary productivity.

MANGROVES AND FISHERIES
Introduction

One of the most important assets of estuaries as a natural resource to
humans is their function as a nursery for a diverse group of fisheries.
Numerous species of fish and crustaceans utilize estuaries during the juvenile
stages of their development following spawning in either coastal or freshwater
systems. Estuaries are productive nurseries because they provide an adequate
supply of food, have a diversity of environmental conditions (including salintiy
and temperature), and maintain habitat protection from predation. Best
management practices for estuaries should include optimization for all three of
these characteristics, yet scientific understanding is lacking in establishing
key linkages between these parameters and secondary productivity of estuarine
ecosystems.

A key linkage that has received some scientific investigation is between
the types of food webs associated with the input of organic matter to an
estuary. In a profile of mangrove communities of South Florida, Odum et al.
(1982) summarized a plethora of studies documenting the diverse fauna associated
with these productive ecosystems. They concluded that a continuum of mangrove
communities from freshwater to marine environments serves as a nursery for
numerous species of estuarine-dependent fishes that are spawned offshore

50

including striped mullet, gray snapper, sheepshead, spotted sea trout, red drum,
and silver perch. The relative importance of mangroves may vary along this
continuum depending on the ability of other estuarine primary producers such as
seagrasses and phytoplankton communities to inhabit the area. In the
Fakaha tehee Stand of riverine mangroves of South Florida, Carter et al. (1973)
ranked the mangrove-fringed bays as the most important nursery grounds among the
habitats available for fisheries. Odum et al. (1979) established that within
these types of riverine mangrove systems, leaf material and associated
microflora were the foundation of a detritus food chain that supported a mixed
trophic level of herbivores, carnivores, and omnivores. This mixed trophic
level included amphipods, shrimp, polychaetes, crabs, molluscs, and a few fish,
which supported middle and higher level carnivores , many of which were
economically important fisheries. Thus a linkage among mangrove productivity,
detritus food webs, and estuarine food chains was documented.

A more quantitative linkage between the productivities of intertidal
wetlands and fisheries would establish the magnitude of the impact of wetland
destruction to estuarine productivity. Turner (1977) calculated a relationship
betweetn shrimp landings (kg'yr" 1 ) to the areal coverage of intertidal vegetation
for northeastern Gulf of Mexico and on a more regional scale in Louisiana.
Macnae (1974) was less successful in establishing a similar type of model
between the harvests of prawn shrimp and mangrove coverage because of insuffi-
cient statistics, yet for the few examples listed there was a positive trend
between these two variables. Although detailed quantitative models of detritus
food chains are still lacking, surveys and studies such as that by Odum et al.
(1979) establish that for estuaries to continue their function as nurseries,
they must provide a large source of food to support the recruitment of juveniles
into these areas.

The Gambia

Food webs associated with mangrove bolons along the Gambia River estuary
are described in a list of invertebrates (Appendix B) and fish (Appendix C)
sampled in these intertidal ecosystems. Detritus-consuming invertebrates are
dominated by snails ( Tympantonus fuscatus var. fuscatus) and crabs ( Sesarma
huzardi) that inhabit the muds of bolons. Both penaeid ( Penaeus duorarum ) and
prawn shrimp ( Nematopalaeron hastafus) consume mangrove detritus suspended in
the water column. These detritivores, plus a group of fish including three
species of Tilapia ( Tilapia sp., T\ occidentalis , and T. heudeloti) and two
others ( Liza falcipinnis and Pellonula vorax ), constitute the basic linkage
between mangrove litter production and complex food webs in the Gambia River
estuary. Through this group of organisms energy from these productive forests
may be transfered to a hierarchy of consumers that collectively represent the
secondary productivity of this estuary.

A description of a food web in the Gambia estuary was developed from in-
formation on the gut contents of fish inhabiting the Bai Tenda bolon (Table 11).
From this information the fish were arranged into four groups representing a
hierarchy of trophic levels as used by Odum et al. (1979) for the mangrove
system in south Florida, USA: detrivores, mixed trophic level, middle
carnivores, and higher carnivores. The detritivores were listed above and
include fish with only detritus material within their guts. The mixed trophic
level was dominated by Ghrysycthys nigrodigitalis and £. furcatus , and their
stomachs also included detritus along with representatives of other trophic
levels (e.g., crabs, shrimp, fish). The guts of middle consumers had only
representatives of the middle hierarchy of the food web such as molluscs, crabs,
and shrimp, and were dominated by Polydactylus quadrifilis and Bostrychus

52

x:

>s

a
c
3

3

cr
3

3

x:

•3

s

CO

u
3

3

4-»
CO

3

V4

3

>

a

cd
O

3

o

f-4

o

ua

CO

TS

G

CD

H

•H

CO

«

3

J2

u

3

•

i-4

CO

u

T3

3

0)

00

r-4

a

u

a

•H

CO

3

CO

J3

CO 3

*3
<4-l 3
O 3
O
00 *W
3
i-l r-4

4-> CO
CO i-»
«H U
>J 3
4-J

CO

• a

*-4 (4-1

o

CO
u
O
H

J3
00

CO

u

o

CO

a

V4

JC
CO

CO

d
en

3

V4

co co

r-4 «D

3 3

O CO

CO r-4

CQ 04

>

* 3
•3 «H
3 CO
3 u
co O

3

Q

r-4 f-» <r cm

I I I I I N«^H I (S

I I I I I HH I N I

CH f-4

I I I

W^ N I I

I I I I ^111

I I I I I

^ I I I I

on l

I I

I l I I

I I I I I

^ I I I I

m r-t t

<^^ h h h r-r^<N«-ii-*

I I I I I m i^ cm r-< r-4

r-4 cm I ll I m I

I I

I I -4

111. — I

I I I I I

I I I I I

I ]

I I

-I CM -4 J I I | I I I

1 I I I I

I I I I

I I I I I

I I I I I

CM

m

m

r-4

i***

en

00 ^

CM

co

1

1

1

1

1

1

i

1

1

1

r-» 1-4

r*

CO

3

CO

4J

•H

CO

CO

CQ

3

u

co

a

•H

CO

•H

•H

3

cu

i-4

3

3

00

CO

(4

t-»

•H

4-1

3

3

CO

«f»

3

CO

3

o

U-«

CO

3

3

3

•H

►J

•^

■u

3

M

u

•H

3

r-4

4-»

iH

r-4

E!

o

CQ

4->

CJ

o

U

CO

3

3

3

3

CO

«H

M

a

CO

M

CO

CO

T3

•H

3

CO

O

«H

00

3

CO

4-1

U

a

00

(4

00

CO

1

a

CO

CQ

O

s

3

a

3

«H

«H

3

9

o

•H

3

3

a

3

3

I** 4

a

3

3

>

3

3

r-«

3

U-l

o

eg

o

CO

31

cu

CO J-i

o

«*-4

3

3

3

3

T3

^

CU

CJ

rH

CO

co

>

3

f-»

3 m

>

X2

<4-4

3

CO

U

CO «H

•H

O

T3

H»

CO

CO

<U

>N

X*

H4

J3

CQ

3

-w 3

M

o

Xi

w • a

2 a-H

o

>

3

92

>>

>>

.e

3

2

O

3

00

CO

3

3

3

■3

3

a

0)

04

Jfi

jfi

CO

4J

CO

S

3

rH

CU

H 3

•H

•H

O

U

CQ

o to o

o

CO

J3

Q

4->

-u

3

-3

^

>^

>>

3

al 3

-«S

a

a

3

3

> <-»

i-4

n

jfi

X

i-4

O

CQ

O

A

4J

CU

4=

u

o

o

3

i-4

3

M CO

co

CO

3

CO

a

a

3

•H

CO

U

o

CQ

cu

o

u

M

3

>>

>>

H «H

<4-4

•H

3

«H

>>

t4

a

3

T-*

a

Q

3

^

>,

cd

XL

J=

4->

00

O

M Ou

Puj

O

a

«

CO

CO

i-4

O

4J

u

TJ

« H

u

i

a

O

3

3

O

g.3

CO

CO

r-4

CO

H

>>

>>

u

J2

4J

o

o

>!

OJ-H

u

•H

•H

CO

O

M

N

r-4

r-4

1-4

u

u

3

u

CU

o

cu

r-4

OJJS

CO

o

a

a

O.

Ui

•3

Cjd «H

•H

•H

cu

•H

h-4

-3

j3

o

>>

CO

M

i-4

O

•H O

o

H4

cu

3

3

■u

>^

Q H

»-3

H

04

H

>:

CJ

CJ

fe

04

04

X ^»

(U

w

CO

«

33

as

93

93

CO

93

53

africanus (Table 11). The higher carnivores had only fish in their guts and
included Hemichromis fasciatus and Hepsetus odoe .

The potential importance of this mangrove food web to the local economy may
be evaluated using catch statistics of the artisanal river fishery for July 1982
- June 1983 (R. Moll, personal communication). Catches of Kujeli ( Polydactylus
quadratilis ), Catfish ( Chrysichthys nigrodigitalis and £. furcatus ), and Jotor
(Fonticulus elongatus ) constituted nearly 51.5% of the total catch of the arti-
sanal fishery from Banjul to Kaur. Diets of the latter two groups represent
mixed trophic levels with significant amounts of detritus, while Kajeli are
middle carnivores feeding on shrimp and crab. In the upper river, from Kaur to
Fatuto, nearly 30% of the annual catch is represented by Tilapia , which is a
detritivore found in mangrove bolons. The three species of Talapia found in the
bolon all had only detritus in their guts (Table 11). Collectively, these
trophic levels of the mangrove food chain represent 40% of the grand total catch
of fish for the Gambia River estuary.

MANGROVES AND ESTUARINE RESOURCES

MANGROVE STRESS

An increase in the mortality of mangroves along the Gambia River estuary
has been noticed over the past several years. Dead mangrove trees are par-
ticularly evident along the more inland areas of the intertidal zone in bolons
that extend several km from the main body of the estuary, e.g., Bintang bolon.
According to Johnson (1978), patches of dead trees could not be detected in
aerial photographs taken in January and February 1972, and that most of the
impact was associated with trees <7 m in height. An understanding of what

stress or stressors may be causing this increased mortality is needed in order

54

to discuss the impact of further changes in the ecology of the Gambia River
basin* Two scenerios discussed below have been proposed as the cause of
increased tree mortality of the mangrove forests - disease and drought.

Disease

Teas and McEwan (1982) cited a gall disease, simlar to one caused by a
fungus [ Cylindrocaroen didymum (Hartig) Wallenw] in Florida, as responsible for
mass mortality of Rhizophora sp. in the Gambia River estuary. Trees >20 m tall
were most susceptible and galls were widespread throughout the mangroves along
this estuary. Based on this evidence, Checchi and Company (1981) predicted that
within 3-5 years there would be a mass loss of Rhizophora along the river,
suggesting that present harvesting should be undertaken immediately to utilize
what is presently left of this natural resource.

Jimenez et al. (1984) suggested that galls are common among mangrove
forests as a normal occurrence in forest development. By comparing several
forests, they summarized that abnormal tree mortality by gall disease would
require infestation of more than 25% of stems per ha. Since older trees are
more susceptible to the fungus, infestation may represent a higher percentage of
total basal area of the forest. Even in cases where there were abnormal
infestations of the gall in mangrove forests, it was suspected that some other
stress had increased the susceptibility of the forest to this disease.

In The Gambia, dead trees comprise less than 15% of the total mangrove area
according to a recent report by Forster (1983). This percent is well within the
natural range for tree mortality among mangrove forests (Jiminez et al. 1984),
indicating that there is not a mass mortality phenomenon occurring in mangrove
forests along the Gambia River.

55

Drought

Intertidal wetland forests do not fall into the classification of world
life zones or plant formations by L.R. Holdridge (1967; Fig. 15) * These life
zones are based on available precipitation, evapotranspiration, and mean
temperature (biotemperature). Thus, the structure of plant communities can
generally be described in relation to the net availability of water
(precipitation - evapotransporation), as well as to temperature. Yet intertidal
plant communities have the advantage of obtaining water from several sources
other than just precipitation; these include groundwater flow, river flooding,
and/or tidal inundation. In The Gambia, where water loss by evapotranspiration
is high relative to precipitation, the high tidal amplitude and riverine
discharge of the Gambia River basin subsidize the high demand for water from
high annual mean temperatures (26 °C). Thus these forests are able to generate
structure uncharacteristic of the very dry life zone in which they exist.
The ability of mangrove forest structure to be independent of those character-
istics for a particular life zone was also evident for forests in Florida,
Puerto Rico, Mexico, and Costa Rica (Pool et al. 1977). In this survey by Pool
et al. (1977), mangrove forests with the second and fourth highest complexity
index occurred in tropical dry forest life zones; these forests had more
structure than those mangrove forests occurring in moist tropical life zones.

Yet these life zone characteristics do emphasize a key point concerning the
ecology of mangrove forests in arid environments - that they are very
susceptible to slight changes in hydrology. For mangrove forests in arid life
zones, small shifts in precipitation patterns become very noticeable in mangrove
forest structure (Cintron et al. 1978). Cyclic patterns in mangrove forest
succession and structure have been documented for arid environments of Puerto

56

I I ' | ' ! M'l'l'l'l'lllllllllllllllllllllll

I ] i I . I ilililihl.lllll!

u

cu

•

cu

/•■s

O0r-

T3

r*»

•H

C^

U

fH

T3

N^

rH

O

•

33

rH

CO

!*

Xi

U

cu

CO

(3

rH

O

o

•H

o

4J

PL4

CO

E

►»

&

o

*M

T3

cu

4J

jo

c

•H

CO

M

rH

a

a

CO

CD

u

T3

o

CO

CO

CO

cu

a

CO

o

cu

N

>

o

CU

n

t+-l

00

•H

c

rH

CO

6

T3

iH

M

M

O

O

M-l

S

CO

M-l

cu

O

C

o

a

N

o

•H

CO

4J

CU

ctJ

u

a

CO

•H

a

*H

•H

•H

TD

CO

d

CO

•H

co

iH

X

CJ

O

X>

cu

x:

V4

-U

CO

rH

M

a

o

&o

M-i

c

CO

4-)

CO

O

U

cu

to

ft

CO

•H

cu

P

X!

H

fe

57

Rico in relation to cyclic patterns in rainfall (Cintron et al. 1978) * Similar
observations have been made for mangroves in south Florida (Davis and Hilsenbeck
1974). In all of these cases, a reduction in rainfall resulted in increased
soil salinity which resulted in an increase in tree mortality. This increased
mortality resulted in vegetation shifts from forests to "tannes" or "salinas"
(areas within the intertidal zone void of vegetation) that expanded at the
expense of the area previously colonized by mangroves*

The cyclic nature of these shifts in mangrove forest structure observed in
Puerto Rico results from changes in soil salinity stress which influences tree
mortality and areal coverage. Increased rainfall (sometimes in the form of a
hurricane) stimulates regeneration of forests within unvegetated "tannes" with
pioneer mangrove species such as Rhizophora . The mechanisms associated with
this flip/ flop control of hydrology on mangrove forest structure is demonstrated
in the model of mangrove ecosystem in Figure 16. The tank which depicts soil
water (diagrams use energy language characters as in H. T. Odum 1971) has inputs
and losses of water as discussed above. Since transpiration is so high, the
turnover time of the tank is low and thus changes in inputs quickly result in
increases in salinity which translate to stress on the mangrove forest. Cintron
et al. (1978) suggested that management of mangroves in dry forest life zones
should account for this cyclic phenomenon by maintaining buffer zones for the
expansion of mangroves during "wet" years. Also, tree mortality should be
accepted as a natural process in these forests, and thus short-term observations
should be placed in the perspective of the cyclic nature of these systems.

These observations by Cintron et al. (1978) are very applicable to the
mangrove forests of The Gambia which are also located in an arid life zone.
Marius (1981) has documented the change in mangroves along the Saloum, the

58

8

u

o

60

CO

o

00

as

rO

u

u

CO

e

cu

4-1
CO

>>

CO

o
o

cu

cu
>
o
u

00

e

B

o

00

u

CU

c
cu

o

CU /-\

O r-*

• 3

^4 O

ta 33

59

Gambia, and Casamance estuaries in the last decade due to the increased drought
conditions in this region. Data collected by Marius (1981) from 1971 to 1978
for a mangrove forest in Senegal (Fig. 17) indicate that vegetation shifts have
resulted in a decline of mangrove forest cover due to the expansion of inland
"tannes." Intertidal areas vegetated by Rhizophora and Avicennia during 1971
were areas of dead trees in 1974, and by 1978 these areas were bare land.
Soil water properties of these areas also changed t salinity of areas once
vegetated by mangroves increased from 60 ms - cm"^ in 1971 to 150 ms'cm"-'- by 1978,
and pH of soil water in these areas declined from about 7 to <5. The levels of
both of these properties (salinity and pH) shifted to values characteristic of
vegetated and unvegetated "tannes," and are the result of a decrease in rainfall
in the region (Marius 1981, see Fig. 3 of this report). Decreases in pre-
cipitation inputs to the soil water compartment result in changes in many soil
water properties such as salinity, pH, hydrogen sulfide content, and
concentration of other potentially toxic constituents; thus the exact mechanism
causing the decline in vegetation is not known. However, this phenomenon is
identical to those described by Cintron et al. (1978) for other mangroves in
arid environments and demonstrates the sensitive nature of these forests to
changes in hydrology.

IMPACT OF PROPOSED DEVELOPMENT

Justification for the proposed salt barrage at Balingho on the Gambia River
estuary is based on controlling water levels and salt distribution in the river
in order to increase agricultural production in the region. The restriction of
salt water at Balingho will result in an increase in salinities in an area of
the estuary that was formerly mesohaline. The lack of any freshwater flow to

60

E

^

P^^

KN.W

R^?

:^

kw^v^v

^sss

UJ

s

UJ
003

Ui •»
CD «o

CO

00

CU

.X3

00

O r-4

CU W
CO

C CO

cd 3

U «H
4J ILi

2P e
a o
o ?-»

0EA0
TREES , TANNE

UJ

s^S^N^

kVvVV\

or

<

KV

.2

>
CO

yj

IN^

^XSSsN

CO

kXVsS

0)

■•I

NNX^

L^V^

z

^^ ■

Kxs>sx:

kS>^v>

z
z

Ui

o

co

UJ

cr
o

fc^<S^

LVSN

£\S

<

ac

O

z

&

i

co

UJ
<T
O
U.

- 1 "

kS>sSNSS

— l. -i ■ '

• •

U 4J

a) cd

§-°

a a
a °

O T3

t-4 CU
U CO

03 cd
cu

00 00

cu rv.

> ON

T3

e o

CO 4-1
5-1 r-l

cu r*.

Cd r-4

5 s

r-4 O
•H U

O M-4

CO

rH
«4-J 03

O 60

cu

>> c

4J CU
•H W
C

•h e

cd

CO

T3

c

cd

PS

a
c cd

5-4

cd

3

CO

CU

CO

M

o

o
o

o

IO

(uioog *o) h<* (uio/sui) A1INHVS 30N3n03S N0J1VX393A

CO o
CU
00 CU

e c
cd o

XI N
U
iH

cd

• T3

f-» «U
S-i

• cu

M C

61

modulate the salinity of this region will drastically change the character of
this region of the estuary. Estuaries need fresh water in order to be
estuaries. The impoundment of fresh water by the proposed dam at Balingho will
result in the formation of a marine ecosystem downstream.

The encroachment of higher salinity water into this mesohaline region of
the Gambia estuary and the cessation of a seasonal freshet during the rain
season will have deleterious effects on mangrove ecosystems. As explained
above, mangrove forests that exist in arid environments are very susceptible to
changes in hydrology. Many mangroves in dry forest life zones are able to
develop into highly structured forests, but only with the subsidy of some
periodic "washout" by fresh water; such is the case for the mangroves in the
Balingho region of the Gambia River estuary. Even under present conditions, the
effects of increased stress as a result of a decline in precipitation can be
observed in mangrove forests in this area. The encroachment of "tannes" into
areas previously colonized by mangroves is evident in the Bintang bolon and has
been documented for estuaries in Senegal (Marius 1981). The "whispey"
appearance of the tips of tree limbs in the canopy of Rhizophora >20 m tall is
particularly common at Yelitenda. This aberrant formation in mangrove trees
results from the lack of water transpiring to the canopy due to increased
salinity of water in the soil (S. Snedaker, personal communication). Increased
salinity and a decline in riverine freshwater flow will only exacerbate these
two phenomenon. The tall (>7 m) Rhizophora and Avicennia trees will die and the
total area of mangrove cover will decline soon after the barrage is built. This
impact below the barrage includes 29,160 ha of mangroves (trees >7 m) which
represents an estimated 334.41 x 10 3 fyr"" 1 of dry mass from litter fall (from
Table 6). Based on estimates of detritus export from these forests, that is

62

equivalent to a loss of 109.9 f yr"^ x 1CH of organic carbon from the estuary
below the proposed barrage at Balingho.

Water levels above the barrage will be maintained at +1.70 m GD and salt
water penetration will be prohibited thus forming an expansive freshwater im-
poundment in this region of the Gambia River. Evidence from studies on the
effects of impoundments on mangroves varies as to which species are more
susceptible to constant flooding; yet personal observation of impoundments in
South Florida, USA, suggests that responses are specific to the magnitude and
time interval of the change. Rhizophora can exist under flooded conditions only
if the water level does not entirely cover the lenticels on their prop roots.
Avicennia have root modifications called pneumatophores for transporting air to
roots in an anaerobic environment. These pneumatophores are able to adapt to
increasing water levels by elongation; for instance pneumatophores of Avicennia
trees in the Mosquito Lagoon impoundment in Florida, USA, has pneumatophores
that are 3 feet long. If the water level change occurs before pneumatophores
can adapt, then these "root breathing" organs will be inundated and the trees
will die.

As previously discussed (see Fig. 5) J( water levels will be maintained at
+1.7 m GD from July to December, and depending on the acreage of rice culti-
vation, the water levels will drop from -1-1. 2 m (0 ha of cultivation) to -1.0 m
GD (24,000 ha of cultivation) during the dry season. For 6 months of the year
water levels will be above high tide level (at Banjul) and only at rice culti-
vation of 8,000 ha will water levels drop to normal low tide levels. Since the
duration of these water level changes will be much longer than the normal diel
frequency of tides, it is anticipated that: the root system of existing mangrove
forests will not be able to adapt to this impoundment. Mangrove forests with

63

much less structure may survive and the areas abandoned by the death of the
larger trees will most likely be colonized by less productive marsh species
rather than mangroves since the impoundment will be fresh water. A conservative
impact of mangrove loss along the Gambia River above the barrage includes trees
>7 m which represents 7,719 ha of forests (from Table 4). This is equivalent to
about 111.4 x 10^ t dry mass /yr of litter in the watershed, which is equal to a
loss of about 39.12 x 10^ tC # yr~^- of allochthonous organic carbon to the
estuary. Since tidal amplitude will be nonexistant in the impoundment, which is
the key mechanism for the transport of organic carbon from wetlands, this
estimate of the decline in allochthonous organic carbon import to the river
above Balingho is a gross underestimate.

Together, the loss of allochthonous organic carbon to the Gambia River
estuary below and above Balingho as a result of the proposed barrage equals
147.92 x 10 3 tC'yr"" 1 . This loss is equivalent to 82% of the present estimate of
total detritus input from mangroves to the estuary. Since inputs of organic
carbon from phytoplankton and riverine discharge are so low, this decline in
allochthonous detritus from mangroves represents nearly 70% of the total inputs
of organic matter to this system. This loss of organic matter to an estuary
that presently supports a diverse detritus food chain could have a negative
impact on the fisheries of the Gambia estuary. The productivity of mangroves is
coupled to the estuary and this energy is transferred along a food web that
produces a source of protein for the people of The Gambia. Thus the loss of
this mangrove productivity in the form of decreased detritus flow into the
Gambia estuary as a consequence of the barrage should be evaluated against the
gain in agricultural production that may be provided in the region. As sug-
gested by this study, the mangroves of The Gambia may be a vital national

64

resource to the artisanal fisheries of this estuary. With better utilization of
this fishery by the Gambian government, these forests could contribute more
significantly to the local economy as a natural resource, rather than as
forestry products or freshwater impoundments as presently proposed.

65

REFERENCES

Abell, T.M. 1980. Letter to R.J. McEwan, 28 May 1980. Overseas Develop-
ment Administration, Land Resources Development Center. Surbiton,
Surrey . England .

Arid Lands Information Center. 1981. Draft environmental profile on
The Gambia. University of Arizona. Tucson, Arizona. National
Park Service contract No. CX-000 1-0-0003. 85 pp.

Axelrad, D.M. , K.A. Moore, and M.E. Bender. 1976. Nitrogen, phosphorus,
and carbon flux in Chesapeake Bay marshes. Virginia Water Research
Center. Bull. 79, Virginia Polytechnical Institute, Blacks burg, Virginia.

Berry, T.D. , R.A. Moll, and G.L. Krausse. 1985. Physical and chemical envi-
ronment of the Gambia River, West Africa, 1983-1984. International Pro-
grams Report No. 9, Great Lakes and Marine Waters Center, The University of
Michigan, Ann Arbor, Michigan.

Boto, K.G. , and J.S. Bunt. 1981. Tidal export of particulate organic

matter from a Northern Australian mangrove system. Estuarine, Coastal
and Shelf Science 13:247-255.

Bray, J.R. , and E. Gorham. 1964. Litter production in forests of the world.
Advances in Ecological Research 2:101-157.

Brinson, M.M. , A.E. Lugo, and S. Brown. 1981. Primary productivity,
decomposition, and consumer activity in freshwater wetlands.
Annual Review of Ecology and Systematics 12:123-161.

Brown, S. 1980. Rates of organic matter accumulation and litter production in
tropical forest ecosystems, pp. 118-139. In The role of tropical for-
ests on the world carbon cycle. (Ed. S. Brown, A.E. Lugo, and B. Liegel),
Symposium held at Institute of Tropical Forestry, Rio Piedras, Puerto
Rico. U.S. Department of Energy, CONF-800350.

Brown, S. 1981. A comparison of the structure, primary productivity, and
transpiration of cypress ecosystems in Florida. Ecological Monographs
51:403-427.

Brown, S. , M.M. Brinson, and A.E. Lugo. 1979. Structure and function of

riparian wetlands, pp. 17-31. In Strategies for protection and manage-
ment of floodplain wetlands and other riparian ecosystems. Symposium Pro-
ceedings, U.S. Department of Agriculture. GTR-W0-12.

Brunt, M. 1959. The Gambia land use and vegetation survey. Proc. 3rd
Int. Afr. Soils Conf. Dalaba 1:389-394.

66

Carter, M.R. , L.A. Burns, T.R. Cavainder, K.R. Dugger, P.L. Fore, D.E. Hicks,
ILL. Revells, and A.W. Schmidt. 1973. Ecosystem analysis of the
Big Cypress Swamp and estuaries. U.S. Environ. Prot. Agency, Region 4,
Atlanta, Ga. EPA 904/9-74-002.

Chapman, V.J. 1976. Mangrove Vegetation. J. Cramer, Germany.

Checchi and Company. 1981. Mangrove feasibility study. Gambia Forestry
Project No. 635-0205. Final Report. Washington, D.C.

Christenson, B. 1978. Biomass and primary production of Rhizophora

apiculata BL. in a mangrove in southern Thailand. Aquatic Botany 4:43-52.

Cintron, G. , A.E. Lugo, D.J. Pool, and G. Morris. 1978. Mangroves of arid

environments in Puerto Rico and adjacent islands. Biotropica 10:110-121.

Connor, W.H. , and J.W. Day, Jr. 1976. Productivity and composition of a
bald cypress-water tupelo site and bottomland hardwood site in a
Louisiana swamp. American Journal of Botany 63:1354-1364.

Coode and Partners. 1977. The Gambia barrage study. Final Report. 20 Sta-
tion Road, South Norwood, London, UK.

Coultas, C.L. 1978. The soils of the intertidal zone of Rookery Bay,
Florida. Soil Science Society of America 42:111-115.

Davis, G.E. , and C.E. Hilsenbeck. 1974. The effects of watershed manage-
ment on the Shark River Slough-Whitewater Bay estuary of Everglades
National Park, Florida. Draft Report, Everglades National Park.
Homestead, Florida. 16 pp.

Davis, J.H. , Jr. 1943. The natural features of southern Florida. Geological

Survey Bulletin No. 25. Department of Conservation, Tallahassee, Florida.

Day, J.W. , Jr., W.G. Smith, P.R. Wagner, and W.C. Stowe. 1973. Community
structure and carbon budget of a salt marsh and shallow bay estuarine
system in Louisiana. Center for Wetland Resources, Louisiana State
University, Baton Rouge. Public. No. LSU-SG-72-04.

Day, J.W., Jr., T.J. Butler, and W.H. Connor. 1977. Production and nutri-
ent export studies in a cypress swamp and lake system in Louisiana,
pp. 255-269. In M. Wiley (ed.), Estuarine Processes, Vol. 2.
Academic Press, N.Y.

de la Cruz, A. A. 1965. A study of particulate organic detritus in a Georgia
salt marsh estuarine ecosystem. Ph.D. thesis. Univ. of Georgia,
Athens, Georgia.

DeLaune, R.D. , R.J. Buresh, and W.H. Patrick, Jr. 1979. Relationship of

soil properties to standing crop biomass of Spartina alterniflora in a
Louisiana marsh. Estuarine and Coastal Marine Science 8:477-487.

67

Dunsmore, J., A.R. Blan, G. Lowe, D. Moffatt, I. Anderson, and J. Williams.
1976. The agricultural development of the Gambia: An agricultural
environmental and socio-economic analysis. LRD 22.

Forster, H. 1983. Evaluation of the national forest inventory of the

Gambia. German Agency for Technical Cooperation, Ltd. Report No. 10.

Gambia Ports Authority. 1983. Tide tables for Banjul 1983. Banjul,
The Gambia.

Giglioli, M.E.C., and I. Thornton. 1965. The mangrove swamps of Keneba,

lower Gambia River basin. I. Description notes on the climate, mangrove
swamps and the physical composition of their soils. Journal of Applied
Ecology 2:81-103.

Haines, E.B. 1977. The origin of detritus in Georgia salt marsh estuaries.
Oikos 29:254-260.

Haines, E.B. 1979. Interactions between Georgia salt marshes and coastal
waters: A changing paradigm, pp. 35-46. In R.J. Livingston (ed.),
Ecological Processes in Coastal and Marine Systems. Marine Science
Volume 10, Plenum Press, N.Y.

Happ, G. , J.G. Gosselink, and J.W. Day, Jr. 1977. The seasonal distribution
of organic carbon in a Louisiana estuary. Estuarine, Coastal and Marine
Science 5:695-705.

Harris, R.C., B.W. Ribelin, and C. Dreyer. 1980. Sources and variability
of suspended particulates and organic carbon in a salt marsh estuary,
pp. 371-384. In P. Hamilton and K.B. Macdonald (eds.), Estuarine
and Wetland Processes with Emphasis on Modeling. Marine Science
Volume 11, Plenum Press, N.Y.

Heald , E.J. 1969. The production of organic detritus in a south Florida
estuary. Ph.D. dissertation, University of Miami, Coral Gables.

Heald, E.J., M.A. Roessler, and G.L. Beardsley. 1979. Litter production
in a southwest Florida black mangrove community, pp. 24-32. In Pro-
ceedings of the Florida Anti-Mosquito Association 50th Meeting,
April 22-25, 1979.

Healey, M. J. , R.A. Moll, and D. Page. 1985. Plankton assemblages of the
Gambia River, West Africa. International Programs Report No. 10,
Great Lakes and Marine Waters Center, The University of Michigan,
Ann Arbor, Michigan.

Heinle, D.R. , and D.A. Flemer. 1976. Flows of materials between poorly
flooded tidal marshes and an estuary. Marine Biology 35:359-373.

Herklots, P.G. 1979. The Gambia second phase rural development programme.
Annex 4. Irrigation developoment . Consultancy Report, UK Overseas
Development Administration.

68

Hernandez, A,, and K. Mullen. 1979. Productividad primaria neta en

un manglar del pacifico Colombiano. Memorias del Simposium sobre
Pacifico Colombiano. Universidad del Valle. Cali-Colombia. S.A.

Hesse, P.R. 1961. Some differences between the soils of Rhizophora and
Avicennia mangrove swamps in Sierra Leone. Plant Soil 14:335-346.

Holdridge, L.R. 1967. Life Zone Ecology. Tropical Science Center,
San Jose, Costa Rica. 206 pp.

Holdridge, L.R. , W.C. Grenke, W.H. Hatheway, T. Liang, and J. A. Tosi, Jr.
1971. Forest Environments in Tropical Life Zones. Pergamon Press,
New York.

Jenny, H. S.P. Gessel, and F.T. Bingham. 1949. Comparative study of de-
composition rates of organic matter in temperate and tropical regions.
Soil Science 68:419-432.

Jimenez, J. A., A,E. Lugo, and G. Cintron. 1984. Tree mortality in man-
grove forests. Unpublished manuscript. Institute of Tropical
Forestry, Rio Piedras, Puerto Rico.

Johnson, M.S. 1978. Inventory of mangroves above the proposed Gambia

River barrage at Yelitenda, The Gambia. Land Resources Development
Center, Project Report 54. Surbiton, Surrey, England.

Jones, M.J., and A. Wild. 1975. Soils of the West African Savanna.
Commonwealth Agricultural Bureaux. 246 pp.

Kjerfve, B., J.E. Greer, and R.L. Crout. 1978. Low frequency response
of estuarine sea level to non-local forcing, pp. 497-513. In M.L.
Wiley (ed.), Estuarine Interactions. Academic Press, New York.

Lesack, L.F., R.E. Hecky, and J.M. Melack. 1984. Transport of carbon,
nitrogen, phosphorus, and major solutes in the Gambia River, West
Africa. Limnology and Oceanography 29:816-830.

Liss, P.S. 1976. Conservative and non-conservative behavior of dissolved
constituents during estuarine mixing, pp. 93-130. In J.D. Burton
and P.S. Liss (eds.), Estuarine Chemistry. Academic Press, New York.

Lugo, A.E., and S.C. Snedaker. 1974. The ecology of mangroves. Annual
Review of Ecology and Systematics 5:39-64.

Macko, S.A. , and J. Zieman. 1983. Stable isotope composition and amino
acid analysis of estuarine plant litter undergoing decomposition.
Estuarine Research Federation Biennual Meeting, Virginia Beach.

Macnae, W. 1974. Mangrove forests and fisheries. FAO/UNDP Indian
Ocean Programme. IOFC/DEV/7434.

69

Mann, K.H. 1975. Relationship between morphometry and biological

functioning in three coastal inlets of Nova Scotia, pp. 634-644.
In L.E. Cronin (ed.), Estuarine Research. Volume 1. Academic
Press, New York.

Marius, C. 1981. Acid sulphate soils of the mangrove area of Senegal and
Gambia. In H. Dost and N. van Breeman (eds), Proceedings of the
Bangkok Symposium on Acid-Sulphate Soils. I. L.R.I, publication 31,
The Netherlands.

Meybeck, M. 1982. Carbon, nitrogen, and phosphorus transport by world
rivers. American Journal of Science 282:401-450.

Mulholland , P.J. 1981. Organic carbon flow in a swamp-stream ecosystem.
Ecological Monographs 51:307-322.

Mulholland, P.J. , and E.J. Kuenzler. 1979. Organic carbon export from
upland and forested wetland watersheds. Limnology and Oceanography
24:960-966.

Nixon, S.W. 1980. Between coastal marshes and coastal waters — a review
of twenty years of speculation and research on the role of salt marshes
in estuarine productivity and water chemistry, pp. 437-525. In P. Ham-
ilton and K.B. MacDonald (eds.), Estuarine and Wetland Processes with
Emphasis on Modeling. Plenum Press, N.Y.

Odum, E.P. 1980. The status of three ecosystem-level hypotheses regarding
salt marsh estuaries: tidal subsidy, outwelling, and detritus-based
food chains, pp. 485-495. In V.S. Kennedy (ed.), Estuarine Perspectives.
Academic Press, Inc. N.Y.

Odum, E.P., and A. A. de la Cruz. 1963. Detritus as a major component of
ecosystems. A.I.B.S. Bulletin No. 13:39-40.

Odum, H.T. 1971. Environment, Power, and Society. John Wiley and Sons, N.Y.

Odum, W.E. 1971. Pathways of energy flow in a south Florida estuary.
Sea Grant Bulletin No. 7, Univ. of Miami Sea Grant Program.

Odum, W.E., J.S. Fisher, and J.C Pickral. 1979. Factors controlling the
flux of particulate organic carbon from estuarine wetlands, pp. 69-80.
In R.J. Livingston (ed.), Ecological Processes in Coastal and Marine
Systems. Marine Science Volume 10, Plenum Press, N.Y.

Odum, W.E., C.C. Mclvor, and T.J. Smith, III. 1982. The ecology of the
mangroves of south Florida: A community profile. Fish and Wildlife
Service/Office of Biological Services, Washington, D.C. FWS/0BS-81/24.

Ong, J.E., W.K. Gong, and C.H. Wong. 1981. Ecological monitoring of the
Sungai Merbok estuarine mangrove ecosystem. School of Biological
Sciences, Universiti Sains Malaysia, Penang, Malaysia.

70

Ong, J.E., W.K. Gong, and C.H. Wong. 1982. Productivity and nutrient

status of litter in a managed mangrove forest. Symposium on mangrove
forest ecosystem productivity, BIOTE.OP-UNESCO, Bogor, Indonesia.

Pendleton, E. 1979. Tidal import and export of organic detritus and organ-
isms in a North Carolina salt marsh creek system. Ph.D. dissertation,
North Carolina State University, Raleigh.

Pool, D.J., A.E. Lugo, and S.C. Snedaker. 1975. Litter production in
mangrove forests of southern Florida and Puerto Rico, pp. 213-237.
In Proceedings of the International Symposium on Biology and Management
of Mangroves (Ed. G.E. Walsh, S.C. Snedaker, and H.J. Teas). Institute
of Food and Agricultural Sciences, University of Florida, Gainesville.

Pool, D.J., S.C. Snedaker, and A.E. Lugo. 1977. Structure of mangrove forests
in Florida, Puerto Rico, Mexico, and Costa Rica. Biotropica 9:195-212.

Richard -Molard, J. 1949. Afrique Occidentale Francaise. Paris, from Giglioli,
M.E.C., and I. Thornton. 1965. The mangrove swamps of Keneba, lower
Gambia River basin. I. Descriptive notes on the climate, the mangrove
swamps, and the physical composition of their soil. J. of Applied Ecology
2:81-103.

Roman, C.T. 1981. Detrital exchange processes of a Delaware salt marsh.
Ph.D. dissertation. University of Delaware, Lewes, Delaware.

Schlesinger, W.H., and J.M. Melack. 1981. Transport of organic carbon in
world's rivers. Tellus 33:172-187.

Sell, M.G., Jr. 1977. Modelling the response of mangrove ecosystems to

herbicide spraying, hurricanes, nutrient enrichment and economic devel-
opment. Ph.D. thesis. University of Florida, Gainesville, Florida.

Snedaker, S.C, and M.S. Brown. 1981. Water quality and mangrove

ecosystem dynamics. Final Report to United States Environmental
Protection Agency EPA-600/4-81-022. Gulf Breeze, Florida.

Steever, E.Z., R.S. Warren, and W.A. Niering. 1976. Tidal energy subsidy

and standing crop production of Spartina alterniflora . Estuarine Coastal
and Marine Science 4:473-478.

Teal, J.M. 1962. Energy flow in the salt marsh ecosystem of Georgia.
Ecology 43:614-624.

Teas, H.J. , and R.J. McEwan. 1982. An epidemic dieback gall disease of

Rhizophora mangroves in the Gambia, West Africa. Plant Disease 66:522-523.

Turner, R.E. 1977. Intertidal vegetation and commercial yields of
penaeid shrimp. Transactions of the American Fisheries Society
106(5): 411-416.

71

Twilley, R.R. 1982. Litter dynamics and organic carbon exchange in

black mangrove ( Avicennia germinans ) basin forests in a southwest

Florida estuary* Ph.D. dissertation. University of Florida, Gainesville.

Twilley, R.R. 1985. The exchange of organic carbon in basin mangrove

forests in a southwest Florida estuary. Estuarine, Coastal and Shelf
Science 20: 543-557.

Twilley, R.R. , A.E. Lugo, and C. Patterson-Zucca. Production, standing
crop, and decomposition of litter in basin mangrove forests in
Southwest Florida. Ecology: in press.

United Nations Development Programme. 1974. Hydrological and topographical
studies on The Gambia River Basin. Final report, 5. Reading, England:
Howard Humphreys and Sons, for UNDP.

Welsh, B.L. , R.B. Whitlatch, and W.F. Bohlen. 1982. Relationship between
physical characteristics and organic carbon sources as a basis for
comparing estuaries in southern New England, pp. 53-67. In V.S. Kennedy
(ed.), Estuarine Comparisons. Academic Press, N,Y.

Wharton, C.H, , and M.M. Brinson. 1979. Characteristics of southeastern

river systems. Strategies for protection and management of floodplain
wetlands and other riparian ecosystems, pp. 32-40. In Symposium Proceed-
ings, U.S. Department of Agriculture, Washington, D.C. GTR-WO-12.

Whigham, D., and R.L. Simpson. 1977. Growth, mortality and biomass partition-
ing in freshwater tidal wetland populations of wild rice ( Zizania
aquatica var. aquatica ). Bulletin Torrey Botanical Club 104:347-351.

Woodwell, G.M., D.E. Whitney, C.A.S. Hall, and R.A. Houghton. 1977. The Flax
Pond ecosystem study: exchanges of carbon in water between a salt marsh
and Long Island Sound. Limnology and Oceanography 22:833-838.

72

APPENDIX A. List of sites visited along the Gambia River estuary during
June, 1984.

Site Distance Upriver (km)

Oyster Bolon

Lamin Island 5

Bintang 60

Duntu Malang Bolon 100

Tend aba Camp 1°°

Yelitenda 129

Bai Tenda 145

Elephant Island 150

Jessadi Wharf Town 192

Kudung Tenda 235

Pasari Island 235

Georgetown 291

73

APPENDIX B. List of invertebrates collected in mangrove bolons along the
Gambia River estuary.

Mollusca

Tymp ant onus fuscatus var. fuscatus
Neritina adansoniana

Crustacea

Penaeid Shrimps

Parapenaeopsis atlantica
Penaeus duo r arum

Prawn Shrimp

Nematopalaemon hastatus

Caridean Shrimp

Paleomonetes sp.

Mud Crabs

Sesarma huzardi
Sesarma elegans
Panopeus africanus

Blue Crabs

Callinectus pallidus
Callinectus amnicola
Callinectus marginatus

74

APPENDIX C. List of fish species caught in the Bai Tenda bolon along
the Gambia River estuary during 1983.

Aplocheilichthy s normani
Bostrychus africanus
Chrysichthys furcatus
Chrysychthys nigrodigitatus
Cynoglossus senegalensis
Elops lacerta
Elops senegalensis
Fonticulus elongatus
Hemichromis bimaculatus
Hemichromis fasciatus
Hepsetus odoe
Hydro cynus brevis
Hyporhamphus picarti
Ilisha africana
Liza falcipinni s
Pellonula vorax
Plectorhynchus macrolepis
Polydactylus quadrif ilis
Pomadasys jubellini
Porogobius schlegeli
Psettias sebae
Pseudotolithus senegalensis
Pythonichthys m arcrurns
Schilbe mystus
Strongylura senegalensis
Synodontis gambiensis
Tilapia brevimanus
Tilapia hendeloti
Tilapia occidentalis
Tilapia rheophila
Tilapia sp.
Trachinotus f al catus

75