Mangrove
vegetation structure
dynamics and regeneration
Thesis Philosophiae Doctor Scientiarum
Farid Dahdouh-GuebaS
General Discussion
Published as :
Dahdouh-Guebas, F. & N. Koedam, 2002. A synthesis of existent and potential mangrove vegetation structure dynamics from Kenyan, Sri Lankan and Mauritanian case-studies. Meded. Zitt. K. Acad. overzeese Wet./Bull. Séanc. Acad. r. Sci. Outre-Mer 48(4): 487-511.
Research frameworks
for studies on vegetation structure dynamics
It
is known that, because of direct factors such as exploitation and clear
cutting (Kairo,
1995) and because of indirect factors such as siltation and
groundwater fluxes (Tack & Polk,
1999), mangrove forests are adversely
affected both quantitatively and qualitatively all around the globe (e.g.
Pernetta,
1993a,b,c,d,e;
Rützler & Feller,
1996).
Research groups are trying to quantify this decline from different
angles using remote sensing. However,
it is equally important to link this analysis to fieldwork that monitors the
qualitative changes as well. The
latter aims for example at the selective unsustainable utilization or
exploitation of certain mangrove tree species or at the patterns of succession, both of which can lead to a change in floristic composition
or
vegetation structure. Research
on changes in mangrove forests and on the regeneration
potential, including solutions to keep the latter at a level
allowing forest rejuvenation must necessarily be considered.
Only
recently the importance of mangroves has been acknowledged and efforts to
restore them arose. Understanding
mangrove vegetation structure dynamics
in
a particular area is a prerequisite to conservation
and
management directives, such as the establishment, protection and management of
re-afforestation plots in the framework of regeneration
projects
(e.g. Lee,
1996; Caloz &
Collet, 1997).
Dahdouh-Guebas et al.
(2000a)
emphasize that there is a need for a methodology that allows to express
reliable predictions about the state of mangroves using a relatively small
input from vegetation field work, and to decide whether a mangrove stand at a
certain location has the potential to successfully renew and rejuvenate or
whether anthropogenic
pressure renders human interference such as restoration imperative. A
monitoring
system
is needed to decide whether human interference is desirable, since artificial
restoration may be appreciated less than natural regeneration.
Field (1998b) stated that ‘natural regeneration of mangroves should
be the first choice of any rehabilitation programme, unless there is
irrefutable evidence that it will be unsuccessful’.
A clear understanding of the nature and dynamics of local
mangrove ecosystems will be the best guide to any restoration
programme (Field,
1996). The
first step is to collect information about the actual state of the mangrove
forest, emphasizing different vegetation layers, but also about past changes
in that particular vegetation. Where
such studies concentrate on the diversity of mangroves it is important to
assess on the appropriate spatial, taxonomic and temporal scale (Farnsworth,
1998). The second step is to
integrate such findings in the management and decision-making process.
It has been shown that remote sensing and GIS-based forestry studies can generate results that can be directly used in forest management planning (e.g. Holmgren et al., 1997; Holmgren & Turesson, 1998). Applicable findings (when focusing on vegetation layers of different age) can for instance include the prediction of future changes in the mangrove forests. In addition, combination of these data with local and global ecosystem data (biological, hydrological, physico-chemical, geographical,…), socio-geographical or –economic data, particularly in a GIS environment, allows to assess future changes under different scenarios (e.g. exploitation, conversion, natural catastrophes or sea level rise) and to adopt conservation strategies by interfering appropriately, if at all.
What encompasses 'dynamics' in the literature ?
A
verification of the term 'dynamics' in recent literature on marine science
or forestry, relevant to the study of general mangrove ecology, reveals that
this term is being used in an environmental, a faunal or a floral context.
In an environmental context it has been used to refer to nutrient
dynamics (Rice &
Tenore, 1981; Newell,
1984; Blair,
1988; Tam et
al., 1990; Chen &
Twilley, 1999), DOC dynamics (Velimirov,
1986),
sediment or detritus dynamics (Brakel,
1984; Flores-Verdugo et
al.,1987) and hydrodynamics (Wolanski,
1992; Kitheka et
al., 1995; Kitheka,
1997). In
a faunal context 'dynamics' has referred to behavioural clustering dynamics (Gherardi
& Vannini, 1992), community dynamics (Syms & Jones,
2000) and spatial
and temporal dynamics (Lugomela,
1995). In a vegetation context 'spatial and temporal dynamics' has
been used as well (Smith & Huston,
1989; Murali et al., 1998), next to litter dynamics (Brown,
1984; Twilley et
al., 1997), biomass dynamics (de Boer,
2000), canopy
dynamics (Herwitz et al.,
1998) and population dynamics (Fromard et
al., 1998; Jiménez &
Sauter, 1991; Clarke,
1995).
In a number of cases terms as 'mangrove forest dynamics' (Smith et al., 1991), 'vegetation dynamics' (Heil & Van Deursen, 1996; Dahdouh-Guebas et al., 2000a) or simply 'dynamics' (Putz & Chan, 1986) have been used, all of these intending more or less 'changes in stand structure and composition'. It is in the latter context that the present paper is written, more precisely : ‘changes in stand extent, structure and composition’. Although to scientists who are focusing on vegetation it is evident that these simple terms have the above meaning, to others these terms might seem less meaningful. Therefore we suggest to adopt the term 'vegetation structure dynamics' for 'changes in stand structure and composition', both spatial and temporal.
Data acquisition and analysis in studies on vegetation structure dynamics
In
the past two decades remote sensing
technology has been given a leading role in the acquisition of
data on vegetation (e.g. Gang &
Agatsiva, 1992; Cohen et al.,
1996;
Ramachandran et al., 1998;
Dahdouh-Guebas et al., 1999, 2000a,
2000c) and both 'reviews' and 'recent advances' are continuously reported in
order to emphasize and compare the potential of various remote sensing
technologies in the past and for the future (e.g. Rehder & Patterson,
1986; Tassan,
1987; Aschbacher et
al., 1995; Blasco et al.,
1998;
Holmgren & Turesson, 1998; Hyyppä et
al., 2000).
The
integration of data on vegetation structure
dynamics from
different moments in time has become almost entirely dependent on remote
sensing
(e.g.
Heil & Van Deursen, 1996; Murali et
al., 1998; Dahdouh-Guebas et al.,
2000a), which usually constitutes the only retrospective basis of comparison
to actual vegetation data (Dahdouh-Guebas et
al., 2000a). Assessment of
factors related to the mangrove on a large scale (global or regional
distribution, cartographic inventories, land-use
conversion, conservation) and investigation of the regional or
global extent of mangroves (e.g. Spalding et al.,
1997), largely
rely on satellite imagery. For
periods of time starting before the existence of space-borne sensors, aerial
photography will
often provide the essential and the only data on changes in vegetation.
Whereas aerial photography, in addition, has been of an unequalled
quality in the study of vegetation structure dynamics until present, the
launch of Ikonos, the first
commercial Very High Resolution (VHR) Earth Observation satellite in September
1999 by Space Imaging (US), probably marks the beginning of a new remote
sensing era providing both panchromatic and multi-spectral images with a 1 m
to 4 m resolution. This
type of resolution combined with the multi-spectral character of the imagery
(incl. near-infra-red) may provide alternatives to the as yet unsolutioned
inability of identification
of
mangroves on a species level (Verheyden et al.,
subm.; Dahdouh-Guebas et
al., in prep.a).
However, for the present research only aerial photographs were available and their applicability to the investigation of mangrove vegetation and the study of mangrove vegetation structure dynamics was positively evaluated (loc. cit.). However, a providing correct mangrove tree species list is essential (Jayatissa et al., subm.) and eventually fieldwork must be carried out.
Fieldwork
or ground truthing, which remains imperative in remote sensing
studies, has concentrated on the adult vegetation in many
case-studies (Spalding et al.,
1997), but great benefit arises when
combining these data with other vegetation layers (Murali et al., 1998; Dahdouh-Guebas et
al., 2000a, subm.a). Next to
overlays between map data originating from different moments in time in a GIS-environment (Geographical Information
System) and a quantification of changes that occurred in the past (e.g.
Verheyden,
1997), an overlay of a map with data from present-day vegetation
layers (e.g. as plots or transects) may provide insight into the
present and possibly future dynamics
of
the mangrove (Dahdouh-Guebas, et al.,
2000a, subm.a).
If
the vegetation layers with adult, young and juvenile trees are considered
there can be either an absence or a presence for each of these.
Table 1 summarises the possible combinations of vegetation layers and
defines the type of vegetation structure
dynamics that can form the basis for such combinations.
A
forest, or a species within a forest, without adult individuals has a
pioneering or colonising nature (colonisation dynamic type, hereafter referred
to as C-type or displaying C-dynamics). Examples
of species with a colonising nature are Avicennia and Sonneratia.
A less obvious example, but encountered on beaches away from mangroves
in both Kenya and Sri Lanka, is Bruguiera gymnorrhiza (pers. obs.).
From the case-study of Galle (Dahdouh-Guebas et al.,
2000a) it
is clear that in Sector 3 Rhizophora apiculata must have had a
colonising nature in the past.
A
forest with a presence of adult trees and an absence of either young or
juvenile ones is declining (degradation dynamic type or D-type / D-dynamics).
It is remarkable that this can be illustrated with the very same Sector
3 of Galle at present, since no young or juvenile trees were found during the
recent fieldwork missions (loc. cit.).
Another example to illustrate a D-type is the condition of the Parc
National du Banc d’Arguin in Mauritania, where adult Avicennia germinans
trees usually do not show young or juvenile trees in their understory (Dahdouh-Guebas & Koedam, in
press). A forest with adult trees and without either young or
juvenile ones may be threatened with decline as well, unless there is a
transient lack of younger specimens or an accelerated growth.
The latter can be very acute in forest areas where Rhizophora
mucronata dominates the canopy, Ceriops tagal dominates the young
understory and a mix of both species is dominating the juvenile understory.
When canopy
A
forest with adult, young and juvenile trees is generally rejuvenating
(rejuvenation dynamic type or J-type / J-dynamics).
However, it may also be declining depending on the similarity in
distribution of adult, young and juvenile trees.
Whereas the term ‘decline’, as used above, refers to a decline of
age structure on a particular place and will be referred to as ‘vertical
decline’, this term can also be used with respect to the area coverage of a
forest or species, hereafter called ‘horizontal decline’.
If this horizontal decline is purely surface bound we refer to
‘quantitative horizontal decline’, which is not considered at this stage. When all vegetation layers are represented in the field we
will obviously also refer to J-dynamics, but if there are significant shifts
in species composition
from
‘mangrove species’ towards ‘non-mangrove species’ we will refer to
‘qualitative horizontal decline’. This
should be taken sensu lato and applies in case of both shifts from
strict or major mangrove components towards the minor mangrove components and
shifts from mangroves species in general towards mangrove associates
or
non-mangrove species.
Table
2 is showing how the data from the past on the spatially static or dynamic
nature of a forest can be combined with distribution data
from the present from all vegetation layers in order to evaluate the status of
the mangrove as being spatially static (i.e. without spatial changes
over time) or spatially dynamic (i.e. with spatial changes over time).
It must be highlighted however that a spatially static forest does not
imply a static nature of all processes. As
a matter of fact there is a steady-state condition underlying the spatially
static or spatially dynamic nature of a forest.
A spatially static forest such as Sector 1 in the mangrove of Galle (Dahdouh-Guebas et al.,
2000a) supports rejuvenation and other
processes in its understory.
A
spatially static forest with a similar distribution of adult, young and
juvenile trees, for instance, is obviously rejuvenating : the younger trees
develop close to the adult ones but the vegetation patches themselves do not
displace. In case of a strong
dissimilarity between the above distributions the spatially static forest
might be declining and possibly requiring human interference, whereas in case
of a spatially dynamic forest dissimilar distributions might be perfectly
normal (Tab. 2).
This type of analysis, the results of which can be shown using a clear and highly qualitative graphical design, can be supported by a parallel statistical analysis that is based on the same type of data and generates more quantitative and testable results (Figs 1 and 2). Detrended correspondence analysis (DCA), canonical correspondence analysis (CCA) and non-linear multi-dimensional scaling (NMDS) are particularly adapted tools for this type of research, which in addition also allow to include environmental data that may help in the explanation of the observed vegetation structure (Cannicci et al., 2000; Dahdouh-Guebas et al. subm.a, subm.b).
Types of vegetation structure dynamics in mangroves
Basically,
the vegetation structure
of
mangroves can be typified as zoned on one hand (e.g. in the Kenyan
sites), with a vegetation ‘zone’ defined as a long band-like patch of
vegetation, or as non-zoned on the other hand, which then displays a mosaic
pattern
of ‘vegetation patches’, the latter defined as a polygon with no
determined shape or area (e.g. Galle, Sri Lanka).
In some cases however, zonation
may
be very irregular or restricted to a particular part of the tidal gradient,
and be termed as a ‘partial’ or ‘semi-zonation’ (e.g. Pambala,
Sri Lanka). Both ‘zones’ and
‘patches’ would have a certain, often monospecific floristic composition. However,
there are also a number of recurrent mangrove assemblages
, such as the ones listed by Macnae
(1968). This author points out
that Walter & Steiner (1937) named the zones that they observed in
East-Africa after the dominant tree in the assemblage, a way of identifying
zones or patches that is still much in use today (cf. e.g. Gallin et
al., 1989).
Whereas
the zonation
-issue and particularly the causes of
its formation have been much debated in the history of mangrove research,
little has been said about vegetation structure
dynamics
, let alone terming some of the types.
Dahdouh-Guebas et al.
(2000a)
introduced the term ‘moving mosaic
’ (Figs 1 and 3) for the type of
vegetation structure dynamic that displays relatively large vegetation patches
to apparently ‘move’ from one area in a mangrove forest to another area
(disappearance and appearance), or put alternatively : for the type of
vegetation structure dynamic that displays a certain area of a forest that
changes in species composition
over
time, and may even interact with terrestrial vegetations such as sedges and
coconut
plantations.
A vegetation structure dynamic displaying vegetation patches to extend
or to grow, rather than to ‘move around’, can similarly be termed a
‘growing mosaic’.
Dahdouh-Guebas
et al. (2000a) suggest
that a moving mosaic
vegetation structure dynamic may be typical for mangroves that
are characterised by an irregular topography instead of the frequently
encountered intertidal slope. In
areas where mangroves are clearly zoned, changes in vegetation structure often
follow a rather pronounced intertidal slope (Dahdouh-Guebas et al.,
subm.a). The vegetation structure
dynamics that occur under these circumstances can be typified as ‘shifting
zones’ (Fig. 3), if the zones are displaced
entirely, or as ‘growing zones’ (Figs 2 and 3) if the zones are
becoming larger (positive growth, e.g. seaward grey patches in Gazi,
Fig. 2) or smaller (negative growth, e.g. landward grey patches in
Gazi, Fig. 2). Some clear
examples of the latter types of transgressive (sometimes introgressive)
vegetation structure dynamics can be found as responses of mangroves to
selective cutting by people (Dahdouh-Guebas et al.,
2000b, subm.a;
Kairo et al., in prep.), to sea-level change or to altered tidal
hydrodynamics (Woodroffe,
1990, 1995, 1999;
Wilton
& Saintilan, 1999), and to natural
events (Stevens & Montague,
1999; Wilton
& Saintilan, 1999; Nguyen et al.,
2000). In the latter two cases,
however, the ‘shifting zone’ concept applies to the entire mangrove
ecosystem rather than to vegetation assemblages specifically.
Vegetation
structure
dynamics
of
mangroves is also associated to succession
, particularly in a situation in which a
naked or denuded habitat is colonised and further develops. We recognise three categories : floristic accretion
, floristic invasion
and
floristic dominance/extinction (Fig. 4). ‘Floristic
accretion’ occurs when a first pioneering species is in part responsible for
the development
of
new adjacent zones, mostly located more landward.
This is the case for pioneering mangrove species such as Avicennia
or Sonneratia. ‘Floristic
invasion’ occurs when an established zone is invaded by another species that
develops within the original zone and forces the original species to retreat
(Fig. 2). This may be the process
underlying the double zonation
often
observed in Avicennia
marina
(e.g. Dahdouh-Guebas et al.,
subm.d). Finally, in a particular
vegetation structure comprising different assemblages
with
a dominant species, one may develop to become the dominant assemblage at the
expense of other species or assemblages (Fig. 1). We term this case ‘floristic dominance
’ with respect to the dominating
species and ‘floristic extinction’ with respect to the retreating and
disappearing ones. In some case
floristic invasion and floristic dominance may be difficult to distinguish, or
an interaction between both may exist (cf. Fig. 1).
The ease with which these processes can be distinguished in part also
depend on the regularity with which imagery can be obtained.
Whereas
the vegetation structure
dynamics
at
lower latitudes take place against the background of the multispecific nature
of the mangrove stands (incl. the ‘behaviour’ of forest patches
with
different compositions with respect to one another), at the highest latitudes
where mangroves occur it is somewhat different. In the Parc National du Banc d’Arguin (PNBA), at the
northern biogeographical
limit
of mangroves along the West-African coast, Avicennia
germinans is
the sole mangrove tree species that constitutes the mangrove ecosystem.
‘Vegetation structure dynamics’ as defined above (i.e. changes in stand extent, structure and composition) must be interpreted in its context. Basically, the ‘extent’ has still the very same meaning in the PNBA, but the scale we are considering at these higher latitudes is different and in many cases we are considering fragmented small populations on a large area rather than continuous fringes. Contrary to mangroves at lower latitudes, the ‘structure’ does not include zonation issues, mosaics or other vegetation patches on a substantial area, because we are dealing with a monospecific mangrove. For the same reason there is little point in describing a ‘composition’, unless all the non-mangrove beach and sebkha vegetation is included. Therefore ‘vegetation structure’ is limited to the extent and fragmentation of the few mangrove populations left and to their physiognomy. Whereas the latter has not been an issue in Kenya or Sri Lanka (probably because there are other vegetation features that are more conspicuous), in Mauritania the different mangrove physiognomies were the most remarkable features of the vegetation structure and comprised four different types : high tree formations, wide tree formations, ‘shrub’ formations and ‘sebkha’ formations, which were obviously no phases in a vegetation development. The sole possible case would be for the sebkha formation to evolve into a shrub formation. Whether this is actually the case requires a long-term monitoring , preferably using aerial photography if useful, and is subject to future research.
Mangrove regeneration and its constraints as an integrated application
Investigations
on the status of mangroves in Kenya revealed that three types of forest states
can be recognised : mangrove in a virtually pristine condition (Kiunga and
Lamu, North Kenyan coast; Kairo et al.
1999), mangrove that is
anthropogenically adversely influenced (MidaCreek
and other creeks between Mombasa and Malindi, central Kenyan coast) and
mangrove that is anthropogenically degraded (GaziBay
and other creeks between Mombasa and Vanga, South Kenyan coast) (Kairo, in
prep.). In South-West Sri Lanka
the occurrence of mangrove forests in a highly fragmented way, is mainly due
to man as well (De Silva &
Balasubramaniam, 1984-85). Studies based on sequential aerial photography
in
both countries have shown that the dynamics
in
vegetation structure
in
sites disturbed by man probably requires human interference to rehabilitate
the mangrove (Dahdouh-Guebas et al.,
2000a, subm.a).
A prediction following from combination with investigations on the
distribution
of
young and juvenile trees confirm this (Dahdouh-Guebas et al.,
subm.b).
The above study therefore leads to a suggestion of both forest areas
and tree species that should be considered in artificial regeneration
.
However,
both areas and species are exposed to a number of threats. Certain mangrove areas are subject to high propagule
predation rates
(Smith et al.,
1989; McKee,
1995a; McGuinness,
1997b; Dahdouh-Guebas et
al., 1997, 1998;
Dahdouh-Guebas, subm.).
This biotic factor affects the choice of the site in mangrove
restoration
. Understanding
such constraints to mangrove regeneration
obviously contributes to an improvement and a development
at
the level of artificial plantations and silviculture (Gong &
Ong, 1995).
Dahdouh-Guebas et al.
(1999b), Ballerini et al.
(2000),
Cannicci et al. (2000) and
Dahdouh-Guebas et al. (subm.c) provide a first step in the
understanding of crabs’ feeding behaviours by analysing the diets of crabs
and their zonation in the forest with respect to mangrove trees.
Experimental designs to analyse the phenomenon of propagule predation were set up by e.g. Smith et al. (1989), Osborne & Smith (1990), McKee (1995a), McGuinness (1997b), Dahdouh-Guebas et al. (1997, 1998), Steele et al. (1999), Dahdouh-Guebas (subm.) and Allen et al. (in prep.). The results found in the present study are summarized in table 1.
Table 1. A synthesis of the findings on propagule predation in Kenya
and Sri Lanka (Dahdouh-Guebas et al.,
1997, 1998; Dahdouh-Guebas,
subm.).
|
Kenya |
Sri Lanka |
|
differential
predation among forest zones : more
predation in landward and Rhizophora dominated zones |
differential predation among forest patches
: more predation in Excoecaria dominated
patches |
|
no
differential predation among mangrove propagules : all
species are predated |
differential
predation among mangrove propagules :
Avicennia
predated more than Buguiera, which in turn is predated more than Rhizophora |
|
differential
predation among mangrove crabs : more
predation by Neosarmatium spp. and |
differential
predation among mangrove crabs : more
predation by Neosarmatium spp. and Chiromanthes spp. |
Also
propagule predation by other animals seems
to affect tree species to different degrees, for instance propagules from Rhizophora
apiculata Bl.
seem to be much less appreciated by the snail Terebralia palustris L.
than those of Bruguiera
gymnorrhiza
(L.) Lam (Dahdouh-Guebas,
subm.). Predation
or parasitism by insects was observed in B. gymnorrhiza, but not in
other species (loc. cit.).
In
the PNBA in Mauritania, the two main crab species observed were Uca tangeri
Eydoux and Callinectes marginatus A. Milne-Edwards, which are an algal
feeder and an animal predator respectively.
Propagule predation was not observed at all in the PNBA (Dahdouh-Guebas
& Koedam, in press).
It
has also been shown how propagule predation
and
vegetation structure
dynamics
may
be inter-linked through hydrology. The
model that Dahdouh-Guebas
(subm.) introduced for this link, is inspired on the phenomenon of propagule predation
but includes elements from vegetation structure dynamics as well.
It is briefly repeated here.
The model starts from an adult tree or forest with mature propagules. Before the propagules fall from the tree, predation or parasitism by insects may occur. When a propagule falls there are two possible situations with respect to the water level : low water and high water. The theory states that when the water level is low (predominantly during the dry season), propagules that fall on the mangrove soil plant themselves (planting strategy of Van Speybroeck, 1992) or have more possibilities to strand (stranding strategy of Van Speybroeck, 1992), exposing them more to propagule predators, which in turn are very mobile and predate considerably. The predation of planted propagules (vertical) is initially lower than from stranded (horizontal) propagules (Dahdouh-Guebas et al., 1997). Some of the propagules are not predated, initially survive and establish. When the water level is high (predominantly during the wet season), the forest is often permanently flooded for a period, and the fallen propagules drift away through the water. They are much less likely to be affected by propagule predators, which at that time are stuck on the mangrove roots. When the water level lowers, predation might still take place, but is less. Dahdouh-Guebas et al. (1997) indeed found that mature propagules are predated less than freshly gathered ones. Therefore, more propagules establish after a period of dispersal, during which such a mature stage can be reached. The propagules that establish are exposed to either favorable or unfavorable environmental conditions. These are influenced by environmental factors such as tidal inundation, microtopography, soil texture, aridity and salinity among others, some of which may be inter-linked. Propagules developing under unfavourable conditions will die, whereas those developing under favourable conditions will survive, further develop and grow into an adult tree. As a matter of fact, the post-establishment situation is less simple than introduced here, because between the stage of propagule and that of an adult tree there are many other events. One point in the model is that the zonation, which is partially present in Pambala, must be the result of differences in salinity along the tidal gradient during the dry season, but that the dispersion of propagules to lead to any zonation in the first place, is controlled by the wet season. Once the water level lowers again and propagules can establish, propagule predators further control this establishment.
Propagule predation may
thus be one link in the chain of events leading to a particular vegetation
structure.
In the PNBA, the lack of success of Avicennia germinans North of the very last tree (which still produces numerous propagules) could be of a climatic nature (frost frequency) as reported for this species along the North-American West coast (Stevens, 1999).
Based
on the study of vegetation structure dynamics, propagule predation, etc… the
question as to which forest area needs rehabilitation can actually be answered
by providing a map-based regeneration scheme.
Within the issue of the mangrove tree species that can be used for
regeneration the results of studies on genetic differentiation can be
integrated (e.g. Abeysinghe,
1999), answering questions such as “To
which degree do different or fragmented populations differentiate on a genetic
level ?” and “How desirable is it then to collect propagules from a well
rejuvenating, but distant population ?”.
In mangrove regeneration studies, a most recent and new focus lies in the monitoring of re-afforested plots as a measure of the success of rehabilitation and of the degree of restoration of the original ecological functions or an alternative stable condition (Bosire, 1999; McKee & Faulkner, 2000; Bosire et al., subm.a, subm.b). As a matter of fact rehabilitation plots aim at a restoration of the natural habitat, which can be assessed by monitoring the secondary succession into the often monospecific artificial mangrove stands, as well as the macrofaunal recruitment in it.
A
first door to future research was opened with the revisit of the Point-Centred
Quarter Method or PCQM (Dahdouh-Guebas & Koedam,
subm.).
In order to understand mangrove vegetation structure
dynamics
, monitoring
and
fieldwork is required, the latter of which necessitates the application of
certain methods to acquire the data needed.
The PCQM is a widely used method to derive forest parameters such as
density
and
basal area
. However,
for forest characteristics measured at a particular time to be interpreted
within the framework of mangrove vegetation structure dynamics, it is
essential that they represent as closely as possible the actual field
conditions, or give a good estimation of the errors involved.
Dahdouh-Guebas & Koedam
(subm.) highlight a number of problems
associated to the PCQ-Method, make a first attempt at tackling these issues
and open the door to future empirical research that can be checked with actual
field observations. The investigation of more sites will lead to the
establishment of ‘PCQM accuracy correction factors’ (sensu
Dahdouh-Guebas & Koedam, subm.). An
in-depth comparison between different PCQM ‘rules’ (e.g. selection
of closest stem versus selection of central stem) and their effects on
the accuracy of derived forest structure parameters is also still
unchallenged.
The
abundance of mangrove juveniles on the tree and on the mangrove floor have
never been investigated in an integrative manner.
Phenology studies have always concentrated on characteristics of
leaves, flowers and fruits when these were still attached to the parental
tree, and terminated when these structures detached.
This approach leaves the question as to which propagules establish
where unanswered. In the study of
vegetation structure, its establishment and its dynamics, the answer to this
question may be very valuable.
Similarly,
the assessment of the natural propagule predation as opposed to the
experimental propagule predation may provide more insight into the
relationship between propagule predation and vegetation structure.
So far, propagule predation studies concentrated on experimentally
planted propagules, leading to figures of 100% predation.
However, in natural situations it is very unlikely that propagule
predators will jeopardise the survival of the plants that form their habitat
at a large scale. Mast seeding
with predator saturation may be an escape mechanism.
The
regular marking of cohorts of mature propagules, the monitoring of their
dispersal and of their fate on the mangrove floor, integrates and fills the
latter two gaps in scientific research on mangrove vegetation structure and
propagule predation.
With respect to remote sensing, the study of vegetation structure dynamics using the new very high resolution space-borne imagery (e.g. Ikonos satellite) must be purchased, but the commercial nature of these data and the extremely high prices associated to this imagery is at present a block to the use of these data for many academic research groups in both industrialised and developing countries.
References
See Bibliography
Published as :
Dahdouh-Guebas, F. & N. Koedam, 2002. A synthesis of existent and potential mangrove vegetation structure dynamics from Kenyan, Sri Lankan and Mauritanian case-studies. Meded. Zitt. K. Acad. overzeese Wet./Bull. Séanc. Acad. r. Sci. Outre-Mer 48(4): 487-511.
| Farid Dahdouh-Guebas' Mangrove Biocomplexity Homepage |
This page is maintained by Farid Dahdouh-Guebas, fdahdouh@vub.ac.be