Disponible à / Available at permalink :

231  Download (0)

Texte intégral


- - -

- - -

Dépôt Institutionnel de l’Université libre de Bruxelles / Université libre de Bruxelles Institutional Repository

Thèse de doctorat/ PhD Thesis Citation APA:

Di Nitto, D. (2010). To go with the flow: a field and modelling approach of hydrochorous mangrove propagule dispersal (Unpublished doctoral dissertation). Université libre de Bruxelles, Faculté des Sciences – Sciences biologiques, Bruxelles.

Disponible à / Available at permalink : https://dipot.ulb.ac.be/dspace/bitstream/2013/210046/4/649c1f0e-ea90-47fd-a22f-4a64d9c444f8.txt

(English version below)

Cette thèse de doctorat a été numérisée par l’Université libre de Bruxelles. L’auteur qui s’opposerait à sa mise en ligne dans DI-fusion est invité à prendre contact avec l’Université (di-fusion@ulb.ac.be).

Dans le cas où une version électronique native de la thèse existe, l’Université ne peut garantir que la présente version numérisée soit identique à la version électronique native, ni qu’elle soit la version officielle définitive de la thèse.

DI-fusion, le Dépôt Institutionnel de l’Université libre de Bruxelles, recueille la production scientifique de l’Université, mise à disposition en libre accès autant que possible. Les œuvres accessibles dans DI-fusion sont protégées par la législation belge relative aux droits d'auteur et aux droits voisins. Toute personne peut, sans avoir à demander l’autorisation de l’auteur ou de l’ayant-droit, à des fins d’usage privé ou à des fins d’illustration de l’enseignement ou de recherche scientifique, dans la mesure justifiée par le but non lucratif poursuivi, lire, télécharger ou reproduire sur papier ou sur tout autre support, les articles ou des fragments d’autres œuvres, disponibles dans DI-fusion, pour autant que :

Le nom des auteurs, le titre et la référence bibliographique complète soient cités;

L’identifiant unique attribué aux métadonnées dans DI-fusion (permalink) soit indiqué;

Le contenu ne soit pas modifié.

L’œuvre ne peut être stockée dans une autre base de données dans le but d’y donner accès ; l’identifiant unique (permalink) indiqué ci-dessus doit toujours être utilisé pour donner accès à l’œuvre. Toute autre utilisation non mentionnée ci-dessus nécessite l’autorisation de l’auteur de l’œuvre ou de l’ayant droit.

--- English Version ---

This Ph.D. thesis has been digitized by Université libre de Bruxelles. The author who would disagree on its online availability in DI-fusion is invited to contact the University (di-fusion@ulb.ac.be).

If a native electronic version of the thesis exists, the University can guarantee neither that the present digitized version is identical to the native electronic version, nor that it is the definitive official version of the thesis.

DI-fusion is the Institutional Repository of Université libre de Bruxelles; it collects the research output of the University, available on open access as much as possible. The works included in DI-fusion are protected by the Belgian legislation relating to authors’ rights and neighbouring rights.

Any user may, without prior permission from the authors or copyright owners, for private usage or for educational or scientific research purposes, to the extent justified by the non-profit activity, read, download or reproduce on paper or on any other media, the articles or fragments of other works, available in DI-fusion, provided:

The authors, title and full bibliographic details are credited in any copy;

The unique identifier (permalink) for the original metadata page in DI-fusion is indicated;

The content is not changed in any way.

It is not permitted to store the work in another database in order to provide access to it; the unique identifier (permalink) indicated above must always be used to provide access to the work. Any other use not mentioned above requires the authors’ or copyright owners’ permission.


D 03758










Thesis submitted in Fulfilment of the Requirements for the degree of Doctor of Philosophy in Science of the Vrije Universiteit Brussel and Université Libre de Bruxelles






for the dâgpre# ' iversiteit SiifSSl!!






Prof. Dr. Nico Koedam

Laboratory of Plant Biology and Nature Management Vrije Universiteit Brussel - VUB


Prof Dr. Farid Dahdouh-Guebas

Complexité et Dynamique des Systèmes Tropicaux Département de Biologie des Organismes

Université Libre de Bruxelles - ULB Dr. James Gitundu KAIRO

Mangrove System Information Service Gazi Bay Mangrove Research Centre

Kenyan Marine and Fisheries Research Institute- KMFRI Prof. Dr. em. Hugo Decleir

Laboratory of Physical Geography Vrije Universiteit Brussel - VUB

This research project was funded by the Flemish Interuniversity Council (VLIR)

Photos by Diana Di Nitto, unless indicated otherwise Vrije

Universiteit Brussel



D ' E U R O R £





‘Science may set limits to knowledge, but should not set limits to imagination’

Bertrand Russel


I hâve been working on this thesis for several years, but I was not alone. I wish to thank many people who contributed to the successflil ending of this dissertation.


...Nico and Farid, for your guidance, innovative suggestions, patience and support in ail the years we hâve been working together. It was a mémorable time, both at APNA and working in the field abroad. I leamed a lot ffom both of you, for which I express my deep gratitude.

... Paul Erftemeijer, Jan van Beek and Firmijn Zijl, you hâve ail been amazingly helpful and friendly, making sure that I would reach my deadline in time. Your support has made the différence in the end.

...Kairo for the support you gave me during my stay in Gazi Bay and Abdul and Shamara for your ffiendship and your assistance in the field. I would also like to thank the local people in Gazi Bay, Galle and Pambala for making my expériences abroad unforgettable.

.. .my dear colleagues for the nice times we had working together, for all the tea breaks, the support and the encouraging words.

.. .my dear friends for being there for me, and always making sure that I could relax and enjoy myself when I needed a break.

.. .my dear family and especially Teun for your support during all this time, 1 would not hâve reached this point if it wasn’t for all of you.



Mangrove ecosystems thrive in (sub)tropical, intertidal areas where adaptations like vivipary and the hydrochorous dispersai of propagules become an absolute necessity. As propagule dispersai and early growth allow for the replenishment of existing stands and colonization of new habitats, many authors recognize the importance of these stages in structuring mangrove populations and communities. However, when it cornes to the actual propagule dispersai and recruitment mechanisms, there is an apparent lacuna in the current understanding of mangrove ecology. The period between the mature propagule falling ffom the parental mangrove tree and the early growth of the established seedling, under various possible circumstances, remains in the dark. In this study we focus on this particular period by investigating both the places where these propagules end up as the pathways their dispersai units follow. And we go one step further.

Mangrove forests are being destroyed worldwide at a threatening pace despite their tremendous asset to Coastal human communities and associated biological species. The effect of human- induced (cutting and mangrove conversion to aquaculture ponds) as well as indirectly and/or

‘naturally’ evolving disturbances (sea level rise) on propagule hydrochory occupies an important place in this study.

Dispersai of water-buoyant propagules of the family Rhizophoraceae and Acanthaceae (now including the Avicenniaceae) was studied in Gazi Bay (Kenya), Galle and the Pambala-Chilaw Lagoon Complex (Sri Lanka). Our study sites differ both in tidal régime and végétation structure, covering an interesting variety of ecological settings to examine propagule dispersai.

Field data and experiments ranging ffom micro/ mesotopographical measurements and successive propagule counts to hydrodynamic and propagule dispersai experiments were collected or executed in situ.

Two main methodological approaches were employed. Firstly, we addressed the question on mechanisms of propagule recruitment and early establishment by creating suitability maps using Geographical Information Systems (GIS). Secondly, the combined set-up of hydrodynamic modelling and ecological dispersai modelling was developed to simulate propagule dispersai pathways influenced by dispersai vectors (tidal flow, ffesh water discharge, wind), trapping agents (rétention by végétation or aerial root complexes) and seed characteristics (buoyancy, obligated dispersai period).

Since propagule dispersai is not solely determined by species-specific propagule characteristics {e.g. buoyancy, longevity, etc.), we emphasize that propagule sorting by hydrochory has to be viewed within its ecological context. The significance of dense végétation obstructing long distance dispersai (LDD in its définition of this work), mainly in inner mangrove zones, supports our main finding that propagule dispersai is largely a short distance phenomenon.

‘Largely’ is here understood as quantitatively, not excluding epic colonization events of rare but important nature. Propagule rétention by végétation and wind as a dispersai vector, deserve a prominent rôle in studies on propagule dispersai.


conversion to aquaculture ponds, imposes limitations on propagule recruitment due to reduced propagule availability and a decrease in suitable stranding areas where the architecture of certain root complexes, like prop roots and pencil roots, fimction as propagule traps. These types of pressure hâve more severe conséquences on propagule dispersai than the effect of sea level rise on mangroves. Mangrove forests, which are not situated in an obviously vulnérable setting, can be résilient to a relative rise in sea level if a landward shift of végétation assemblages and successful early colonization is not obstructed by human-induced pressures.

Also, and this renders mangrove forests vulnérable in spite of their intrinsic resilience, when the

‘capital’ of forest is severely reduced or impoverished as happens extensively worldwide, the

‘interest’ on this capital, understood as propagule availability, delivery and trapping, will not allow them to efficiently cope with sea level rise, putting sustainability of mangrove ecosystem services and goods at risk.

In a larger framework of mangrove végétation dynamics, knowledge on propagule dispersai will benefit management strategies for the conservation of mangroves worldwide, besides its flindamental interest to fiilly fathom the ecology of this particular marine-terrestrial ecotone formation.




Mangrove-ecosystemen gedijen in (sub)tropische intergetijdegebieden waar aanpassingen zoals viviparie en de verspreiding van hydrochore propagulen een absolute noodzaak zijn. Wanneer propagulen zich verspreiden en uiteindelijk vestigen, zorgen zij ervoor dat bestaande mangrovebestanden worden aangevuld of nieuwe leefgebieden worden gekoloniseerd. Om deze reden erkennen vele auteurs het belang van deze levensfase in het structureren van mangrovepopulaties en -gemeenschappen.

Toch merken wij een duidelijke leemte in de huidige kennis van de mangrove-ecologie wanneer het gaat over de werkelijke propagulenverspreiding en vestigingsmechanismen. De kennis over de période tussen het vallen van de voldragen, rijpe propagulen en de groei van de gevestigde zaailing, dit onder verschillende omstandigheden, is nog steeds zeer beperkt. In deze studie richten we ons dan ook op deze bijzondere période door zowel het eindpunt als het afgelegde traject van propaguleverspreiding te onderzoeken. We wensen echter een stap verder te gaan.

Ondanks het buitengewoon groot ecologisch en economisch belang van mangroven, worden deze ecosystemen wereldwijd aangetast en bedreigd aan een angstwekkend tempo. De directe antropogene invloeden (zoals houtkap en de conversie van mangroven naar gamaalkwekerijen) alsook de indirecte en/of de natuurlijke verstoring (zeespiegelstijging) op propagule hydrochorie zijn van groot belang in dit onderzoek.

Het onderzoek naar de verspreiding van drijfbare propagulen van de familie Rhizophoraceae Acanthaceae (nu met inbegrip van de vroegere Avicenniaceae) vond plaats in Gazi Bay (Kenia), Galle en Pambala-Chilaw Lagoon Complex (Sri Lanka). Onze studiegebieden, die zowel in getijderegime als in vegetatiestructuur verschillen, omvatten een uiteenlopende variatie in de ecologische context die zich ertoe leent om de propaguleverspreiding te onderzoeken.

Veldgegevens en experimenten, gaande van micro/mesotopografische metingen en opeenvolgende propagulentellingen tôt hydrodynamische en verspreidingsexperimenten werden verzameld of uitgevoerd in situ.

Er zijn vnl. twee belangrijke methodologische benaderingen toegepast. In eerste instantie richtten we ons op de mechanismen betreffende propaguleverspreiding en initiële vestiging door geschiktheidkaarten aan te maken met behulp van Geografische Informatie Systemen (GIS).

Ten tweede werd er een gecombineerde set-up van hydrodynamische en ecologische verspreidingsmodellering ontwikkeld om het traject te simuleren dat propagulen volgen doorheen hun verspreiding in een mangrovebos. Bij dit type van modellering werden de volgende factoren opgenomen: verspreidingsvectoren (getijdenstromen, zoetwaterafvoer, wind), retentiefactor (door de vegetatie of het bovengrondse wortelcomplex) en de verspreidingseigenschappen van propagulen (drijfbaarheid, période van ‘verplichte’


Aangezien propaguleverspreiding niet uitsluitend wordt bepaald door de soortspecifieke kenmerken van propagules (bv. drijfvermogen, levensduur, enz.), benadrukken we het feit dat ruimtelijke uitsorteren van propagules d.m.v. hydrochorie moet benaderd worden binnen een ecologische context. Wanneer de begroeiing van mangrove vegetatie zeer dicht is, belemmert dit de mogelijkheid tôt lange afstand verspreiding (LAV zoals in de omschrijving van dit werk), voomamelijk in de middenzones van mangrovegebieden. Het voorgaande ondersteunt onze belangrijkste bevinding dat propagule verspreiding zich grotendeels over korte afstanden voordoet. ‘Grotendeels’ is hier opgevat als kwantitatief, zonder het zeldzame maar


mangrovevegetatie en wind in de roi van verspreidingsvector verdienen een belangrijkere positie in studies betreffende propaguleverspreiding dan nu het geval is.

Antropogene druk op mangrove-ecosystemen, in het bijzonder houtkap of de conversie van mangrove ten behoeve van gamaalkwekerijen, legt beperkingen op de vestiging van propagulen, meer bepaald door een verminderde propagule-beschikbaarheid en een afhame van de geschikte gebieden waar de architectuur van bepaalde wortelcomplexen, zoals steltwortels en pneumatoforen, fiingeren als een ‘fuik’ voor propagulen. Deze directe gevolgen hebben een grotere impact op propaguleverspreiding dan het effect van zeespiegelstijging op mangroven.

Mangrovewouden, die niet gelegen zijn in een duidelijk kwetsbare omgeving, kunnen

‘veerkrachtig’ omspringen met een relatieve zeespiegelstijging op voorwaarde dat vegetatie- assemblages de mogelijkheid hebben om naar de landzijde te migreren en als succesvolle kolonisatie aan deze zijde niet verhinderd wordt door artificiële en destructieve activiteiten.

Bovendien, en dit maakt mangrovewouden kwetsbaar ondanks hun intrinsieke weerbaarheid, als het ‘boskapitaal’ sterk wordt verlaagd of verpauperd, hetgeen wereldwijd geen uitzondering is, zal de ‘ecologische rente’ op dit kapitaal, zijnde de beschikbaarheid, aanvoer en retentie van propagulen, deze mangrovewouden niet in staat stellen om efficient op een stijging van de zeespiegel in te spelen. De negatieve gevolgen voor de diensten en goederen geleverd door mangroven zijn op deze manier niet te onderschatten.

In een mimer kader van vegetatiedynamiek van mangroven zal een grondige wetenschappelijke kennis over propaguleverspreiding niet enkel ten goede komen aan de flindamentele drijfveer en vraag om de écologie en de sturende processen van deze systemen te doorgronden, maar eveneens aan het behoud van mangroven wereldwijd.



Les écosystèmes de mangroves (forêts de palétuviers) se développent dans les zones intertidales (sub)tropiques où des adaptations telles que la viviparie et la dispersion hydrochorique des propagules est une nécessité absolue. Puisque la dispersion et le développement initial permet le réapprovisionnement de la forêt et la colonisation de nouveaux habitats, plusieurs auteurs reconnaissent l’importance de ces stades dans la structuration des populations et des communautés végétales de mangroves. Cependant, il existe une lacune apparente dans la compréhension actuelle de l’écologie des mangroves qui se situe au niveau de la distribution même des propagules et au niveau de leurs mécanismes de recrutement. La période entre la chute d’une propagule mature de l’arbre de mangrove parentale et le développement initial de la plantule, dans les différentes conditions possibles, reste dans l’obscurité. Dans cette étude on se focalise particulièrement sur cette période en investiguant les localités précises où ces propagules surgissent comme sur les trajets que ces unités de dispersion auront suivi. Et nous allons au-delà.

Malgré leur bénéfices surprenants à l’égard des communautés humaines et des espèces biologiques associées, les forêts de mangroves sont détruites partout au monde à un taux menacent. L’effet des perturbations anthropiques (déboisement et conversion de mangroves en basins d’aquaculture) comme ceux des perturbations indirectes et/ou naturelles (augmentation du niveau de la mer) sur l’hydrochorie des propagules occupent une position centrale dans cette étude.

La dispersion des propagules flottantes de la famille des Rhizophoraceae et des Acanthaceae (maintenant incluant les Avicenniaceae) a été étudiée à Gazi Bay (Kenya), Galle et le Complexe de Pambala-Chilaw Lagoon (Sri Lanka). Nos sites d’études different en régime de la marée comme en structure de la végétation, couvrant une variété de conditions écologiques intéressante afin d’étudier la dispersion des propagules. LFne gamme de données de terrain et d’expériences, des mesures micro/mesotopographiques et des comptages successifs de propagules jusqu’aux expériences d’hydrodynamique et de dispersion de propagules, a été collectionnée ou exécuté in situ.

Deux méthodologies principales ont été employées. Premièrement, on a adressé la question sur les mécanismes de recrutement de propagules et sur l’établissement initial par la synthèse de cartes de convenance dans une plateforme de Système d’ Information Géographique (SIG).

Deuxièmement, la combinaison de la modélisation hydrodynamique et la modélisation écologique de la dispersion à été réalisée afin de simuler les trajets de dispersion des propagules mais aussi des vecteurs de dispersion (le flux de la marée, le versement d’eau douce, le vent), des agents d’accrochage (rétention par la végétation ou par les racines aériennes) et des caractéristiques des semences (flottabilité, période de dispersion obligatoire) qui agissent dessus.

Puisque la dispersion de propagules n’est pas seulement déterminée par les caractères spécifiques à l’espèce {e.g. flottabilité, longévité, etc.), on accentue que le triage de propagules par l’hydrochorie doit être vue dans son contexte écologique.


abrévié dans ce travail), surtout à l’intérieur de la mangrove, soutient notre résultat principale que la dispersion de propagules est essentiellement un phénomène à courte distance.

‘Essentiellement’ est entendu ici comme quantitativement, ne pas excluant les événements de colonisation épique qui ont une nature rare mais importante. La rétention de propagules par la végétation et le vent en tant qu’agent d’accrochage méritent un rôle éminent dans les études de dispersion de propagules.

La pression anthropogénique sur les écosystèmes de mangroves, plus particulièrement le déboisement et la conversion en basins d’aquaculture, impose des limitations sur le recrutement des propagules dues à la disponibilité réduite de propagules et à une réduite de zones d’accrochage qui conviennenL où l’architecture de certains complexes de racines, comme les grandes racines échasses et les pneumatophores du type crayon, fonctionnent comme des pièges de propagules. Ces types de pressions ont des conséquences plus sévères sur la dispersion de propagules que les effets de l’augmentation du niveau de la mer. Les forêts de mangroves, qui ne se trouvent pas dans une situation manifestement vulnérable, peuvent être résilientes à une augmentation du niveau de la mer relative à condition que le déplacement des assemblages végétaux vers l’intérieur du territoire et la colonisation initiale ne soit pas obstrué par des pressions anthropogéniques. En plus, et ceci rend la forêt de mangrove vulnérable malgré sa résilience intrinsèque, si le ‘capital’ forestier est réduit ou appauvri comme c’est continuellement le cas partout dans le monde, la ‘rémunération de capital’, entendu comme la disponibilité, l’apport et l’accrochage des propagules, ne la permettra pas de faire face à l’augmentation du niveau de la mer, ce qui mettra à risque la durabilité des biens et des services écosystémiques de la mangrove.

Dans un cadre de la dynamique des mangroves plus large, la compréhension de la dispersion de propagules bénéficiera les stratégies de gestion pour la préservation globale des mangroves, ceci à coté de l’intérêt fondamental de noyauter l’écologie des cette formation écotone marine- terrestre particulière.




2D/3D 2 Dimensional - 3 Dimensional ppt part per trillion ADI Altemating Direction Implicit PSU Practical Salinity Unit

AHT Ail High Tides r" square of the corrélation coefficient

AR4 Fourth Assessment Report RADP Randomly Amplified Polymorphie DNA

BE Bâte Ela RMSE Root Mean Squared Error

Bias« normalized bias SA Sensitivity Analysis

DC Dutch Channel SDD Short Distance Dispersai

DGPS Difï'erential Global Positioning System

SET Surface Elévation Tables

DNA DeoxyriboNucleic Acid SHT Spring High Tides

DTM Digital Terrain Model SKI Static Kinematic Program

DS Dry season SLR Sea Level Rise

EADAP Africa Database and Atlas Project Sw Seaward

e.g. exempli gratta TAR Third Assessment Report

EHT Equinoctial High Tides TIN Triangulated Irregular Network

EMS Environmental Monitoring System TSH Tidal Sorting Hypothesis

et al.= et alu UNEP United Nations Environment Program

GIS Geographical Information System UNESCO United Nations Educational,

Scientific and Cultural Organization GLOSS Global Sea Level Observing


UTM Universal Transverse Mercator

HAD Height above Datum WGS World Geodetic System

lOC Intergovemmental Océanographie Commission

WS Wet season

IPCC Intergovemmental Panel on Climate Change

KPA Kenya Ports Authority LDD Long Distance Dispersai

Lw Landward

ME Marambettiya Ela MHT Medium High Tides

MH WN Mean High Water at Neap tide MHWS Mean High Water at Spring tide ML WN Mean Low Water at Neap tide MLWS Mean Low Water at Spring tide na data not available

NHT Normal High Tides ObsP observation point

ODP Obligated Dispersai Period

PE Pol Ela

pers. obs. Personal observation(s)


Tbesis outline 1

Chapter 1 7

General Introduction and Objectives

Cbapter2 25

Digital terrain modelHng to investigate the effects of sea level rise on mangrove propagule establishment

Cbapter 3 47

Biotic and abiotic factors driving propagule dispersai in a microtidal mangrove forest

Cbapter 4 77

Mangroves facing cbmate change’ response to projected scénarios of sea level rise

Cbapter 5 105

Modelling mangrove propagule dispersai in Chilaw Lagoon (Sri Lanka):

sensitivity analysis and implications for shrimp farm réhabilitation

Cbapter 6 135

Mangrove propagule rétention and dispersai in a macrotidal environment (Gazi Bay, Kenya): modelling the effects of trapping by végétation and sea level rise

Cbapter 7 163

General conclusions and recommendations

Appendiœs 177

References 209

Curriculum vitae 228


Thesis outline



For every researcher it is a challenge to sail into unknown waters. In the field of mangrove ecology, views on the hydrochorous dispersai of mangrove propagules are scarce and more importantly rarely revised. There is in fact a research gap involving the period between the mature propagule falling from the parental tree and the early growth of the established seedling.

This is astonishing, because mangrove species are thought to be dépendent on the hydrochorous process for colonization of new habitats and their extensive biogeographical species range suggest the efficiency of this strategy. The quest for environmental information, being a diverse combination of biotic and abiotic factors, is necessary to fill this scientific gap and to unveil in which site-specific conditions propagules are recruited or which pathways they follow towards their final point of recruitment and early establishment. To challenge the poorly tested views on mangrove dispersai ecology nowadays, we need to go beyond the highly valued, yet standard procedures of field-based experiments. At this stage, it is timely to additionally explore the efifect of intertwined variables on possible propagule behaviour by constructing models that actually simulate processes of hydrochory. The pioneer character and the uncertainties of computer models in their approach to simulate real propagule dispersai, should not limit us to carefully scrutinize apparent dispersal-environment relationships and to bend current views on propagule dispersai where necessary. In addition, our research also intends to identify those drivers which require doser scrutiny in future research.

This research was carried out in three study sites intended to represent varions ecological settings for mangrove formations: Gazi Bay in Kenya, Galle and the Pambala-Chilaw Lagoon Complex in Sri Lanka. The focus is on water-buoyant propagules of the family Rhizophoraceae and Acanthaceae (genus Avicennia formerly better known under Avicenniaceae).

The présent PhD dissertation is composed of 7 chapters. Starting with Chapter 1 -General introduction and objectives - we outline the sparse but exisüng views on propagule hydrochory.

Furthermore, we address the different environmental factors included in this study and pinpoint the weak links in the current knowledge of their rôle in propagule dispersai processes. Aller a comparison between the site-specific characteristics of the study areas, we close this chapter by defining our main objectives. CHAPTER 2 and Chapter 3 address meclianisms of propagule recruitment or early establishment respectively in a macrotidal area (Gazi Bay, Kenya) and microtidal area (Galle, Sri Lanka). The effect of microtopography, top soil texture and above- ground root complex on propagule recruitment was investigated by field experiments and afterwards translated into GIS-based suitability maps for stranding or self-planting of propagules. The effect of dégradation (tree cutting), sea level rise and microtopography-altering burrowing activities of the mangrove mud lobster Thalassina anomala, is treated. Given the intertidal position inhérent to mangrove ecosystems, a more in-depth GIS analysis on the effect of different sea level rise scénarios on mangrove resilience is envisaged in Chapter 4. Apart Ifom simulating possible landward shifts of mangrove végétation assemblages in Gazi Bay (Kenya), we call for an adaptive management strategy conceming conservation on a régional scale.



Thesis outline

In Chapter 5 and Chapter 6 we model propagule dispersai pathways by implementing the important drivers afïecting the fate of a fallen propagule being dispersed throughout a mangrove area. We apply this methodology to an implication of shrimp pond restoration in the Pambala- Chilaw Lagoon Complex (Sri Lanka) and in view of sea level rise in Gazi Bay (Kenya).

Chapter 7 complétés this thesis with a general conclusion explaining the entire research framework and we express renewed views on propagule dispersai and its applicability both at a local and global scale.

Although this PhD thesis ends here, our work is not yet expected to be finished. We elaborate on recommendations as we still need to know more in order to fully unravel the processes of hydrochory and the effect of human-induced or naturally evolving disturbances on propagule distribution. Therefore we aiso set a research agenda to deepen our insight in mangrove dispersai ecology and its relation to climate change through sea level rise. Given the ecological and economical importance of mangrove services and goods, knowledge on propagule dispersai will benefit management strategies for the conservation of mangroves worldwide.






ntroduction and







1.1. Long distance dispersai versus short distance dispersai

Most mangrove species are typically dispersed by water-buoyant propagules (Tomlinson, 1986). These dispersai units bave the opportunity to take advantage of estuarine, Coastal and océan currents in order to replenish existing stands and to colonize new suitable habitats.

However, propagules do not disperse uniformly as their dispersai ability is expected to vary with each taxon, dépendent on several factors including the viability and buoyancy of propagules, the rate and direction of surface currents, the water conditions and the availability of suitable habitats (Duke et ai, 1998; Sousa et al, 2007).

Both local establishment and the potential for long distance dispersai (LDD'), hâve important implications for mangrove ecology and évolution (Sousa et ai, 2003b). On the one hand, local dispersai limits the exchange between adjacent sites and may consequently lead to significant genetic variation which was observed for Avicennia germinam (Dodd et ai, 2002; Dodd and Rafii, 2002; Ceron-Souza et al, 2005) and A. marina (Maguire et al, 2000). At a local scale, in the zonated Gazi mangrove System (Gazi Bay, Kenya), it was observed that a genetic différentiation may occur between landward and seaward A. marina zones (Dahdouh-Guebas et al, 2004a). In the latter study the causes were not investigated, but the very local dispersai of mangrove propagules and trapping as we hâve observed in the dense intermediate Gazi mangrove may offer an explanation for this species.

On the other hand, biogeographic patterns of mangrove distribution, species diversity and évolution, hâve been largely shaped by variations in LDD patterns amongst species and the land barriers they encounter (Ricklefs and Latham, 1993; Duke, 1995; Duke et al, 1998; Ellison et al, 1999; Gain et al, 2000; Duke et al, 2002; Alongi, 2009).

Mangrove dispersai over long distance is a much quoted example of water dispersai (Guppy, 1906; Ridley, 1930; Van der Pijl, 1969; Ricklefs and Latham, 1993), however this may not be correct for the modal or average distances their propagules disperse within a mature mangrove forest {e.g. Yamashiro, 1961; McGuinness, 1997; Sousa et al, 2007). It is however so that rare, epic colonisation of new, remote habitats can hâve pronounced conséquences when establishment is effective, e.g. the natural dispersai of Rhizophora mangle to many islands of the Hawaiian archipelago over hundreds of kilométrés in about a decade (Sauer, 1988) or the more surprising recolonisation of the Krakatoa in 1883 by Xylocarpus sp. (being amongst the first) (Whittaker et al, 1989).

Harper (1977) already highlighted the general inadequaey of quantitative information about plant dispersai and the tendency to focus on rare colonising events rather than the spatial and temporal patterns of dispersai that might influence population processes. This leads us to the duality of propagule dispersai efïiciency. Propagules must be optimally fit to rejuvenate and colonise at a local scale, which requires effective and fast establishment in the soil and root development. In contrast, effective long distance dispersai nécessitâtes a prolonged period of longevity and preferably a less pronounced embryonic development. The viviparous propagule, after a lack of dormancy while still attached to the parental tree, should subsequently hait its

' LDD in this context is used in the sense of propagule movement over océans



development for the time of dispersai, if precocious and inappropriate development is to be avoided.

1.2. Supplyside theory

In the 1980s, an apparent shift from a curiosity in the rôle of post-recruitment biotic interactions to a renewed interest in the influence of a variable propagule supply, gave rise to the term

‘suppiy-side ecology’(Sousa et al., 2007). Supply-side ecology implies that propagule availability in ecological Systems linked by dispersai can hâve a great or greater influence on community structure than post-recruitment biotic interactions, (Lewin, 1986; Hughes et al., 2000; Sousa et al., 2007).

In the field of mangrove ecology, the Tidal Sorting Hypothesis (TSH) by Rabinowitz (1978b) was one of the first théories involving supply-side dynamics. The TSH suggests that the interacting effects of water depth and propagule size may explain the differential distribution of mangrove species with tidal élévation (i.e. zonation) (see also section 3.1).

This widely cited hypothesis has been challenged by a number of researchers and lead to the conclusion that the delivery of diaspores to ail sections of the intertidal zone is not solely steered by tidal action but compétitive interactions and variables (e.g. prédation, forest structure) also play a significant rôle (e.g. Osbome and Smith, 1990; Smith, 1992; Sousa et al., 2007). These critical re-evaluations may hâve exposed the weaknesses of the TSH in explaining the vertical distribution as such, yet a rather dogmatic or unsupported view on mangrove distribution being solely determined by physiological tolérance was punctured, giving rise to several studies on the dispersai properties of mangrove propagules including buoyancy, period of obligate dispersai, longevity and period of establishment (e.g. Clarke, 1993b; Clarke et ai, 2001; Drexler, 2001; Allen and Krauss, 2006b; Saifullah et al., 2007) and the importance of the dispersai agent ‘flowing water’ was acknowledged by several authors (e.g. Clarke, 1993b; McKee, 1995b; Stieglitz and Ridd, 2001). Furthermore, Van Speybroeck (1992) elaborated on the proposed dispersai strategies (self-planting versus stranding) of mangrove trees of the family of the Rhizophoraceae.

1.3. Scientific gap in mangrove ecology

Despite the emergence of the studies mentioned in the previous section, there is still an apparent scientific gap with respect to the actual dispersai and recruitment mechanisms and the rôle of abiotic factors (e.g. hydrodynamics, wind) on propagule dispersai. Propagule dispersai and early growth are both critical stages in a plant’s life cycle, however most research focus has been on the effect of biotic factors on propagule growth or survival (e.g. Palihawadene and Pinto, 1989; Terradosa et al, 1997; Matthijs et al, 1999; Delgado et al, 2001; Kitaya et al, 2002; Krauss and Allen, 2003b; Krauss and Allen, 2003a; Cannicci et al, 2008). Citing Duke et al. (1998), ‘if a species is présent, the environment must be suitable for it, but the opposite does not apply’, we emphasize the importance of the period between the fall of a mature propagule and early growth of the seedling. As mentioned above, several studies hâve investigated the dispersai properties of mangrove propagules and recognized the interest of dispersai vectors (tidal flow, freshwater discharge) in studies on propagule dispersai.

Surprisingly enough, there are only a few publications on végétation as a trapping agent in the rétention of propagules (Mazda et al, 1999a; Stieglitz and Ridd, 2001; Sousa et al, 2007;

Gurnell et al, 2008). Given the complexity of propagule dispersai mechanisms, it is impérative




to exceed the highly valued, yet isolated studies on propagule dispersai and to Itnk the different drivers of propagule dispersai in studies on propagule distribution within the ecological settings of a mangrove forest. This introduction gives an overview of these significant drivers (dispersai vectors (see section 2.), trapping agent (see section 2.3) and propagule dispersai characteristics (see section 3.) to better comprehend their rôle with respect to propagule recruitment and dispersai.

2. P

hysical processes in mangroveecosystems

The importance of tidal flooding in mangroves was first demonstrated by Watson’s classification of inondation classes (1928). His suggestion that different species of mangroves occur among distinctive zones, each with a characteristic hydroperiod, gave rise to studies on the relevance of physical processes {e.g. tidal régime, sédimentation) structuring mangrove forests {e.g. Chapman, 1944; Lugo and Snedaker, 1974; Carlson étal, 1983; Bunteto/., 1985).

Tidally driven currents are the most important type of water movement in mangrove ecosystems, while sea waves and groundwater flow can also play a significant rôle. Tidal motion of sea water does not only form the physical environment that supports mangrove forests; these tidally reversing flows also ensure material exchange {e.g. nutrients, dissolved oxygen, mangrove litter) between the mangrove areas and the open sea (Woodroffe, 1985a, b;

Wolanski et al, 1990; Bouillon et al, 2000).

2.1. Classification of mangrove landforms

Mangrove landforms hâve been classified in

varions ways by different researchers {e.g. Lugo and Snedaker, 1974; Thom, 1982; Citron and Novelli, 1984). To discuss the hydraulics of mangroves, we follow the classification by Cintrôn and Novelli (1984) based on topographie features: riverine forest, fringe forest basin forest (Figure 1.1)

Riverine forest (R-type); Floodplains along river drainage channels or tidal creeks, which are inundated by most high tides and exposed during low tides. Dissipation of wave energy along the tidal creeks protects these types of mangroves to direct exposure to sea waves.

Fringe forest (F-type): This landform can be described as mangrove areas along shorelines facing the open sea and therefore directly exposed to both tidal and wave action.

Basin type (B-type): dépréssions with mangrove areas that are inundated by spring high tides during wet seasons, but rarely flooded by high tides during dry seasons (John and Lawson, 1990).

Figure 1.1: Classification of mangrove forest types based on topographie features (after Cintrôn and Novelli, 1984): (a) Riverine forest type; (b) Fringe forest type; (c) Basin forest type



2.2. Behaviour of water in mangrove areas

2.2.1. Surface water hydrodynamics

When mangroves with a common, gentle slope gradient from land to sea are inundated by tidal action, this usually leads to widely flooded areas (Wolanski et al., 1992). Around mean tidal level, the inundation of a mangrove area starts. When high tides are reached, the water level in the forest coïncides with the tidal level in the creek or open sea. During ebb tides, the water level within the mangrove area deviates from that in the creek due to considérable flow résistance by the complex above-ground root complexes like prop roots and pneumatophores (Mazda et al., 1997a; Mazda et al., 1999b) (see 2.3.). The inundation cycle is quite similar for F-type and B-type forests, though F-type forests are flirther affected by wave action while B- type forests can become ponded with stagnant water during long dry seasons (Mazda et al.,


Within R-type forests, mangroves are flooded by brackish waters consisting of both ffeshwater discharged from the catchment area and offshore sait water entering through the tidal creeks. A common feature of many mangrove areas with tidal creeks is a tidal flow asymmetry in which the peak current velocity is higher at ebb tide than at flood tide. Tidal asymmetry appears to flush out coarse sédiment from the creek, ensuring the creek depth and material exchange between the mangrove area and open sea (Wolanski et al., 1980; Wolanski, 2006). This can also be expected to drag away mangrove-derived material including propagules.

2.2.2. Ground water hydrodynamics

The importance of groundwater flow in mangrove areas is often overshadowed by a much higher rate of water flux by surface water movement. However, the signifîcant rôle of groundwater flow in determining soil properties and maintaining mangrove ecosystems has been widely investigated (Ridd and Sam, 1996; Hughes et al, 1998; Susilo and Ridd, 2005).

The behaviour of groundwater varies among the different types of mangroves. Findings by Mazda and Ikeda (2006) indicate that groundwater levels near a creek within an R-type forest descended up to 15cm after tidal inundation had ceased. In contrast, only minor changes were observed in an F-type forest. The descent speed of groundwater is also slower than that of surface water; however the presence of numerous animal burrows and sédiment layers increases permeability (Héron and Ridd, 2003; Susilo and Ridd, 2005).

2.2.3. Atmospheric processes

Although the mangrove canopy layer protects and shelters the soil surface undemeath, mangrove forests are still subjected to atmospheric éléments like e.g. sunlight, wind, rain, and évaporation (Snedaker and Snedaker, 1984; Kjerfve, 1990). In tropical mangrove régions, a high variation in rainfall throughout the year can hâve a signifîcant influence on the flooding patterns within these areas. A high freshwater discharge in the wet season brings changes in both water properties and hydrodynamics, including density stratification and vertical water circulation in tidal creeks (Ong et al, 1991). Dry seasons with high évaporation rates signify high values of salinity in groundwater and surface water layers (Hollins and Ridd, 1997). In addition, with respect to propagule dispersai, Sousa et al (2007) elaborated on the effect of




freshwater sheet flow during the rainy season that may render propagule dispersion more directional.

Finally, the atmospheric element ‘wind’ can generate surface waves in open water bodies and wide channels, however the adult mangrove canopy reduces the influence of wind and subsequently the création of surface waves (Mazda et al, 1990a).

2.3. Physical characteristics of mangrove végétation

Living on the edge of land and sea, mangroves hâve morphologically adapted their above- ground root System in order to allow growth in the loose, hypoxie soils of high energy Coastal environment (Hogarth, 1999). Various forms of aerial roots hâve developed with different functions (Figure 1.2), varying ffom gas-exchange, water uptake and anchorage for stabilization (Tomlinson, 1986).

Figure 1.2: Illustration of different above-ground root structures: A) pendi roots (Avicennia marina), B) knee roots (Ceriops tagal), C) peg roots (Sonneratia aiba) and D) prop roots (Rhizonhora mucronata). Ail oictures were taken in Gazi Bav, Kenya.

As shown in Figure 1.2, the density of these root structures can be very high. Végétation in general, and for mangroves in particular the complex above-ground root architecture of some mangrove species, offers résistance to the water flow and causes a decrease in the bed shear stress, subsequently leading to enhanced déposition inside and behind the végétation zone (Krauss et ai, 2003; Temmerman et al, 2005). Root Systems, which in mangroves concentrate near the substrate, influence the particle size distribution of the bottom sédiment in marginal en central parts of mangrove areas (Sato, 1978; Furukawa and Wolanski, 1996b)



Wave réduction by mangrove végétation in view of Coastal protection, bas been studied by e.g.

Sato (1978), Furukawa et al. (1997), Mazda et al. (1997a) and discussed by Feagin et al.

(2009). Although results point to the fonction of mangrove végétation as land ‘protector’, the quantitative mechanisms of wave réduction are not well understood due to the différences between species and the complex vertical configuration of mangrove trees. Furthermore, the latter studies focused on short-period waves such as sea waves, therefore we can not draw the line towards the protection of mangroves against tsunami waves. Several studies on the unique rôle of mangrove forests in the protection of coastlines, suggest that mangroves can absorb much of the energy of tsunamis and considerably limit the damage along these affected coastlines {e.g. Hamzah et ai, 1999; Dahdouh-Guebas et al, 2005b), yet more research on the drag force on mangrove trees is needed (Feagin et al 2009).

Given the complexity of mangrove root structures, it is unimaginable that these roots do not affect the hydrochorous dispersai of propagules throughout a mangrove. As the relationship between végétation density and the drag force is presumed to be dépendent on the mangrove species, the details of this relationship still hâve to be examined (Mazda et al, 1999a).

The modelling of propagule dispersai based on hydrodynamic flows and trapping by végétation, has never been done. However, the distribution of prawn larvae in a mangrove area in Malaysia has been modelled by means of a 2D diffusion model which leads to the conclusion that distribution is due to principally tidal currents and latéral trapping in mangrove-ffinged tidal creeks (Chong et al, 1996).

2.4, Links between mangrove topography and dynamic processes

Mangrove topography and the dense végétation inhérent to mangrove areas steer and hinder water currents forced by tides and waves (Wolanski et al, 1980; Mazda et al, 1997a; Mazda et al, 1997b). In turn, meandering in creeks and érosion of coastlines can occur when topography is modified through sédiment transportation by these water currents (Mazda et al, 1995;

Furukawa et al, 1997; Aucan and Ridd, 2000; Brinkman et al, 2005). Sédiment déposition and resuspension by currents is linked to turbulence generated by the interaction with the végétation (Furukawa and Wolanski, 1996b; Furukawa et al, 1997).

The interaction between water currents and the two ‘topographicaf features, being the wide areas of mangrove flooded daily and the above-ground root structures, is necessary to maintain a natural and healthy environment. Mangroves create their own topographical environment {e.g. bio-geomorphology) through a sériés of feedback processes in order to ensure their survival in these harsh intertidal conditions. A good example is the tidal asymmetry mentioned before. Larger ebb currents export bottom sédiments of the creek to the open sea and assure the depth of the creek. As this process continues, the size of the mangrove area will be reduced or végétation density will increase, which in tum decreases the ebb tidal currents leading to silt accumulation in the creeks and the start of an ‘isolation’ of the mangrove area from the open sea (Mazda et al, 1995; Aucan and Ridd, 2000; Wolanski, 2006).




3. P

ropagule dispersaicharacteristics

Most mangrove species share two characteristic propagule traits: dispersai by water and vivipary (Tomlinson, 1986). The dispersai capacity of propagules is influenced by many factors, including buoyancy, propagule size, shape and weight, salinity, tidal action, longevity, root growth characteristics and prédation on propagules, numbers allowing predator saturation (Rabinowitz, 1978a; Steinke, 1986; Smith, 1987; Smith, 1992; Clarke, 1993b; Clarke and Myerscough, 1993; Dahdouh-Guebas et al, 1997; Dahdouh-Guebas et ai, 1998; Clarke et al, 2001; Drexler, 2001; Clarke and Kerrigan, 2002; Sousa et al, 2003a; Allen and Krauss, 2006b;

Sousa et al, 2007). The focus within this study is on the dispersai of water-buoyant propagules of the familles Rhizophoraceae and Acanthaceae (now including the Avicenniaceae (APG, 2009)) (Figure 1.3). However, a review on species-specifïc dispersai characteristics is provided in the following chapters. In this part of the introduction we give an outline of the different current views on propagule dispersai.

Figure 1.3: Illustration of the different species included in this study on propagule dispersai. A) Rhizophora apiculata, B) Rhizophora mucronata, C) Ceriops tagai, D) Avicennia marina and E) Bruguiera gymnorrhiza. The shape of propagules of Avicennia oifidnaiis, which is not shown, are



3.1. Tidal Sorting Hypothesis

As mentioned before in section 1.2, the Tidal Sorting Hypothesis (TSH) by Rabinowitz (1978a) was one of the first hypothèses on propagule dispersai. Rabinowitz (1978a) started by defining 4 parameters that were essential to quantily the dispersai capacity of a propagule:

Longevity: the period during which a propagule remains viable Floating perioà. the period that a propagule remains floating

Period required for establishment: the period a propagule needs to take root in the soil and not be displaced by normal tide action

Obligated dispersai period: time taken for viable, floating propagules to start developing roots

By investigating these parameters through several experiments, Rabinowitz (1978a) discovered two contrasting patterns of dispersai which we now know as the TSH (Figure 1.4). On the one hand, mangrove species with smaller propagules (like Laguncularia racemosa and Avicennia germinans) are carried farther inland by flood tides than larger ones (like Rhizophora mangle), stranding and establishing in greater numbers at higher upper tidal élévations. On the other hand, larger propagules hâve a greater access to the soil surface in lower intertidal areas on the seaward side where water depth is greater. This type of ecological sorting at early life-history stages may lead to the habitat différentiation or ‘zonation’ commonly observed in mature mangrove stands (Rabinowitz, 1978a; Smith, 1987; McKee, 1995a; McGuinness, 1997; Ellison et al., 2000).

Throughout the years, several researchers hâve tested the TSH leading to a substantial amount of criticism based on contradicting experimental evidence. Sousa et al. (2007) grouped this wide-ranging criticism into three groups.

• Physiologists indicated that Rabinowitz’s reciprocal transplant experiments of propagules did not run long enough to confirm that différences in tolérance of edaphic conditions can explain zonation. They argued that propagules could still hâve been surviving on embryonic reserves buffering them ffom environmental stress. While this argument has merit, there are however a growing number of studies that show the same results (e.g. Smith, 1987; Jimenez and Sauter, 1991; Delgado et al, 2001).

• Several authors (Smith, 1987; Smith, 1992; Clarke et al, 2001; Delgado et al, 2001;

Sousa et al, 2007) do not align with the idea that seaward stands should be dominated by species that produce larger propagules while land inward smaller propagules prevail.

A suggestion was made that tidal action may deliver propagules to all sections of the intertidal area (Smith, 1992).

• When only looking at tidal action in areas with a high tidal range, the dispersai of propagules will be mainly directed towards landward areas as indicated by Rabinowitz (1978a). However, Sousa et al (2007) reassessed the TSH by conducting experiments in the same study area as Rabinowitz and found a highly directional movement of propagules towards more seaward areas. This is probably due to sheet flow of ffeshwater runoffin seasons with high rainfall.




Tidal Sorting Hypothesis



Outgoing tide


Stranded positions

as tides recedes, followed by seedling establishment

Figure 1.4: Illustration of the Tidal Sorting Hypothesis (TSH) by Rabinowitz (1978a).

Propagules are sorted with the incominq tide according to propagule size, whereafter Although the TSH may not by sufFicient to explain adult distribution patterns, it was one of the first hypothèses to raise the issue of propagule dispersai processes. The experiments on propagule dispersai are not incorrect, yet the several critics indicated that different factors (incl.

prédation, sheet flow) play a significant rôle in propagule dispersai, not only tidal sorting.

3.2. To go with the flow,.... or not

When a propagule falls from the parental mangrove tree, there are two possibilities or dispersai strategies that can occur. As been suggested for the family of the Rhizophoraceae, propagules use either the stranding or the self-planting strategy (e.g. Van Speybroeck, 1992). The self- planting strategy States that when torpedo-shaped propagules fall ffom the adult tree, they can self-plant in the muddy substrate. The stranding strategy implies that propagules are dispersed away from the parental tree and strand in areas further afield. When self-planted propagules are not firmly fixed within the soil, they can be displaced by tidal action. The ‘perfect’ fall of a propagule into a self-planted position could also be hindered by the well-developed canopy, younger tree layers and a complex root structure. Self-planting will not allow for long distance dispersai (LDD) by océan currents, yet, local establishment is highly promoted. Unless stranded propagules hâve the opportunity to uproot themselves (Tomlinson and Cox, 2000), e.g.

in microtidal areas where flooding area is limited, a tidal transportation of these propagules will



take place to areas further afield. LDD will become possible and eolonisation of suitable habitats further away can oceur.

Various authors bave been inclined to support either the self-planting (e.g. La Rue and Muzik, 1951) or the stranding strategy (e.g. Egler, 1948; Rabinowitz, 1978a; Rabinowitz, 1978b).

However, Van Speybroeek (1992) reported that self-planting appears to be the major meehanism of propagule dispersai in relatively undisturbed mangrove forests, while stranding proves to be dominant in eolonizing over-exploited and cleared mangrove forests. In addition, Cannicci et al. (2008) link propagule dispersai strategies to tidal régime and consequently propagule prédation. They State that when water level is low (or during dry seasons with little tidal influence) propagules can self-plant or strand of which stranded propagules are known to be predated most (Dahdouh-Guebas et al., 1997; Cannicci et al, 2008). However, the main dispersion drivers are high tides or land run-off and seasonal rivers during wet seasons in mangrove forests with a microtidal régime. When a forest is flooded permanently for a substantial period, falling propagules will drift away, making the propagules less vulnérable for prédation until stranding can occur.

When stranding propagules hâve the opportunity to go with the tidal flow, the dispersai distanee will be dépendent on both morphological features of the propagules and extemal environmental factors, like e.g. tidal action, location of propagule release, wind, sheet flow and rétention by végétation, débris. The suecess of this drift in eventual establishment then dépends on their longevity, the period required for establishment and obligated dispersai (time taken for viable, floating propagules to develop latéral roots).

On the one hand, examples of short distance dispersai (SDD) hâve been given by e.g.

Yamashiro (1961), Clarke and Myerscough (1991), McGuinness (1997) and Sousa et al.

(2007). These propagule dispersai experiments showed distances of e.g. maximum 8m for Ceriops tagal (observation period: 70 days) (McGuinness, 1997) and maximum 50m for Kandelia candel (observation period: 30 days) (Yamashiro, 1961). Espeeially species of Rhizophora are suggested to disperse over short distances (McKee, 1995a).

On the other hand, LDD of propagules can occur leading to propagules dispersai over thousands of kilométrés (Guppy, 1906; Sauer, 1988; Whittaker et al, 1989; Rieklefs and Latham, 1993).

The place of propagule release can play a significant rôle in dispersai distances travelled by propagules, as e.g. Clarke (1993b) indicated that Avicennia marina propagules which were released along tidal creeks could disperse over 500m during one single flood tide.

4. H



induced pressureson mangroves

Although mangroves and assoeiated eeosystems can be considered as one of the biologically most productive and socio-economieally most important eeosystems in the (sub)tropies, a global destruetion is threatening mangroves through numerous forms of human pressure. Apart ffom mangroves as a souree of wood, the main eause of this mangrove loss ean be attributed to an increased human pressure on Coastal eeosystems in general and the compétition for the eonversion of mangrove territoiy as a funetion of expansion of agriculture, aquaculture, inlfastructure and tourism (e.g. Dahdouh-Guebas et al., 2000b; Kairo et al., 2002; FAO, 2003;

Dahdouh-Guebas et al., 2005b; Bergquist, 2007; Duke et al., 2007; FAO, 2007; Bosire et al., 2008b; Walters et al., 2008).

Clear-felling, more than natural disturbance (Ferwerda et al., 2007), ean impose severe problems on propagule establishment because of the removal of aerial root complexes. Prop




roots and pencil roots, function as propagule and sédiment traps (Ellison, 1998; Bosire et al., 2003; Krauss et al., 2003; Dahdouh-Guebas et al., 2007). They can dissipate current energy and in this way trap sédiment in their structures (Mazda et al., 1997b; Phuoc and Massel, 2006) causing a modification in the grain size distribution of the substrate (Bosire, 1999).

Furthermore, (unregulated) land cover changes in upstream catchments can cause extensive siltation. This can lead to burial of propagules and juvéniles and eventually to an alteration of the topography on a local scale hence impacting propagule dispersai, establishment, survival and growth (Thampanya et ai, 2002; Mohamed, 2008b).

Mazda et al. (1999a) demonstrated by numerical experiments that material dispersion (which could be e.g. fine sédiments, nutrients, prawn larvae, mangrove propagules) in a riverine mangrove forest (R-type, see section 2.1) dépends on végétation density. When artificial thinning of a mangrove forest occurs, natural balance is disturbed through both a réduction in végétation density accompanied by a decrease in propagule supply and a change in propagule dispersai patterns. Furthermore, several studies hâve shown that fHnge type mangrove forests (F-type, see section 2.1) that hâve been partially deforested, yet also R-type forest with no direct contact to open sea, are extremely susceptible to severe érosion by sea waves {e.g. Furukawa et al., 1997; Mazda et al., 1997a; Massel et al., 1999; Mazda et al., 2006).

Many efforts hâve been done to plant mangroves in extensive, degraded areas along mangrove coastlines ail over the world {e.g. Stevenson et al., 1999; Lewis III, 2005; Bosire et al., 2008b).

However, to promote and preserve effective plantings and their surrounding areas, it is very important to quantitatively understand the key physical and hydraulic processes in such a setting, as well as fundaments of mangrove autoecology, more specifically the patterns of their reproduction, propagule distribution and establishment, (Lewis III and Marshall, 1998;

Stevenson et al., 1999; Lewis III, 2005; Bosire et al., 2008b).

It must be possible for mangroves and human populations to live side by side in the future years to corne. As stated by Alongi (2009), the competing demands of humans and mangroves are manageable if proper management plans are based on relevant scientific information. Picturing the larger framework of mangrove végétation structure dynamics and approaches that consider and integrate the relationships between organisms, the physical environment, and humans, can offer crucial insight into the functional rôle of mangrove forests in the Coastal zone, including estimâtes of maximum sustainable yield of fish and wood. As mangroves are not only threatened by human-induced disturbances, naturally evolving disturbances in view of global climate change, especially sea level rise, also need spécial scrutiny to ensure that human needs do not conflict with the sustainability and conservation of mangrove ecosystems.

5. M








Living at the edge of land and sea, sea level rise (SLR) is believed to be the greatest climate change challenge that mangroves will face (Field, 1995; McLeod and Rodney, 2006). Climate change components like a rise in température, CO2 levels and précipitation will most likely hâve an influence on mangrove productivity, phenology and distribution {e.g. Woodroffe and Grindrod, 1991; Bail and Munns, 1992; Pemetta, 1993; UNEP, 1994; Field, 1995; Snedaker, 1995; Ellison and Famsworth, 1997; Erftemeijer and Hamerlynck, 2005), yet the effects of a SLR are presumed to be more significant (Gilman et al., 2008) and immédiate.

Throughout the Quatemary, mangroves hâve shown high resilience to disruptions from large sea level fluctuations over historié time scales (Woodroffe, 1990; Field, 1995). However, the adaptive capacity of mangroves and other Coastal wetlands to sea level rise (usually by



landward migration) is now severely limited in many localities by increasing human activities (Gilman et ai, 2006a; Gilman et al, 2008).

Tidal range and sédiment supply are two critical indicators of mangrove response to SLR (Gilman et al, 2008). Mangroves can expand or adapt if the rate of sédiment accretion is sufficient to keep up with SLR. For example, in western Jamaica, mangrove communities were able to sustain themselves because their rate of sédimentation exceeded the rate of the mid- Holocene SLR of ca. 3.8 mm/yr (Hendry and Digerfeldt, 1989). Rates of SLR between 9 and 12 cm over 100 years are suggested to stress mangroves, while faster rates could seriously threaten mangrove ecosystems (Ellison and Stoddart, 1991). On the other hand, a sédiment supply that is too excessive (e.g. resulting from poor agricultural practices, increasingly combined with extreme weather events) can bury aboveground root structures and ultimately lead to death (Ellison and Stoddart, 1991 ; Thampanya et al, 2002; Mohamed, 2008a).

As mangroves can adjust by expanding landward or laterally into areas of higher élévation, or even by growing upward in place (McLeod and Rodney, 2006), this ability to migrate is also determined by local conditions, such as infrastructure {e.g. roads, agricultural fields, dikes, urbanization, seawalls, ad shipping channels) and topography {e.g. steep slopes). The response of mangrove to SLR is strongly dépendent on the physiographic setting in which they occur.

Some ecologists for instance believe that mangrove communities are more likely to survive the effects of sea level rise in riverine, macrotidal, sediment-rich environments (Semeniuk, 1994;

Woodroffe, 1995) than in microtidal, sediment-starved environments like on many small islands {e.g. in the Caribbean) (Parkinson et al, 1994a).

Additionally, adaptive capacity will vary among species as some mangrove species appear to be more robust and résilient to SLR than others (Ellison and Stoddart, 1991). On the one hand, individual species hâve varying tolérances of the period, frequency, and depth of inundation, and on the other hand different végétation zones hâve different rates of change in sédimentation élévation (Krauss et al, 2003; Rogers et al, 2005; McKee et al, 2007). When landward areas become accessible during SLR. dispersai and early growth can hâve a significant influence on the future structure and dynamics of multispecies mangrove assemblages (Clarke et al, 2001;

Sousa et al, 2007). Species-specific compétition in landward migration areas may allow some species to outcompete others and to become more dominant within the newly formed species composition (Lovelock and Ellison, 2007).

Mangrove health and composition could be affected by projected increases in the frequency of high water events (Church et al, 2001; Church et al, 2004) due to changes in salinity, recruitment, inundation and wetland sédiment budget (Gilman et al, 2006a; Gilman et al, 2007). When storm surges flood a mangrove area that is also subjected to SLR, severe mangrove destruction can occur. This fatal combination was observed in mangrove forests in Southern Florida where acres of black mangroves {Avicemia germinans) suflfocated because flooding caused by hurricane Andrew was prevented from subsiding by sea level rise (Swiadek, 1997). A similar case of flooding, associated with heavy rains of the 1998 El Nino, resulted in mortality of mangrove forests in the Rufiji Delta in Tanzania (Erftemeijer and Hamerlynck, 2005). Also the El-Nino rains in 1997 caused siltation and a subsequently massive die-off of adult and young trees within a small Rhizophora mucronata stand in Gazi Bay (Kenya) (Dahdouh-Guebas et al, 2004b).

In this study we model propagule dispersai with a focus on the effect of sea level rise (SLR).

This is done by implementing different projected SLR scénarios: minimum (+9cm), relative (+20cm), average (+48cm) and maximum (+88cm), as projected by the Intergovemmental




Panel of Climate Change (IPCC) (IPCC, 2001). We based our analyses on SLR scénarios of the IPCC Third Assessment Report (TAR) of 2001 and not on those of the Fourth Assessment Report (AR4) in 2007, which respectively forecast a range from 9cm - 88cm by 2100 and a range from 18cm-59cm by 2090-2099. Due to a lack of published literature, AR4 models do not include uncertainties in climate-carbon cycle feedback nor do they include the full effects of changes in ice sheet flow. The AR4 projections however include a contribution due to increased ice flow from Groenland and Antarctica at the rates observed for 1993-2003, but these flow rates could increase or decrease in the future. The AR4 could hâve similar ranges to those of TAR if uncertainties were treated in the same way.

Figure 1.5: Decision tree to aid résilient site sélection for mangroves adapted from McLeod and Rodney (2006)

In studies on the effect of SLR, it is important not only to look at the responses of mangrove ecosystems to SLR, but also to investigate the possibilities of a mangrove area to enhance or maintain intrinsic resilience in the event of a rising sea level. To be successful in a changing World, where global climate change in conjunction with an increasing population puts high demands on Coastal wetlands (IPCC, 2001), conservation strategies must strive to reach the complementary key goals of maintaining biodiversity, promoting ecosystem values and enhancing resilience. In order to achieve these goals, a practical decision tree was developed by McLeod and Rodney (2006) which can aid managers to select protected areas based on resilience criteria (Figure 1.5). This decision tree can be applied once candidate sites of high biodiversity hâve been selected using biological criteria for factors indicating strong recovery potential, e.g. regarding peat building, landward migration, sédiment distribution and propagule dispersai. Natural resilience of mangroves to climate change gives hopes for their long-term survival. This should motivate scientists and managers globally to develop forward-looking strategies and to respond with innovative solutions




Sujets connexes :