Soil nutrient content and water level variation drive mangrove forest aboveground biomass in the lagoonal ecosystem of Aldabra Atoll

Abstract

Lagoonal mangrove ecosystems are vital for carbon capture, protection of coastlines and conservation of biodiversity. Yet, they are decreasing globally at a higher rate than other mangrove ecosystems. In addition to human drivers, local environmental factors influence the functioning of lagoonal mangrove ecosystems, but their importance and combined effects are relatively unknown. Here, we investigate the drivers of mangrove functioning, approximated by mangrove aboveground biomass (AGB), in a protected lagoonal mangrove ecosystem on Aldabra Atoll, Seychelles. Based on a survey of the mangrove forest structure in 54 plots, we estimated that the mean mangrove forest AGB was 82 ± 13 Mg ha⁻¹. The total AGB of the mangrove area (1720 ha) was nearly 140,600 Mg, equivalent to about 66,100 Mg of carbon stored in the standing biomass on Aldabra. To assess the direct and indirect effects of soil nutrient content, water level variation and soil salinity on mangrove AGB, we used a structural equation model. Our structural equation model explained 82 % of the variation in mangrove AGB. The soil nutrient content (concentration of essential macronutrients in the soil column) had the greatest influence on mangrove AGB variation. Additionally, high variation in water level (change in water depth covering a location) increased mangrove AGB by increasing nutrient content levels. Our results highlight the important contribution of Aldabra's lagoonal ecosystem to Seychelles' carbon storage and the role of hydroperiod as a regulator controlling the availability of crucial nutrients needed for the functioning of mangroves within lagoonal systems. We suggest conservation managers worldwide focus on a holistic ecosystem-level perspective for successful mangrove conservation, including the protection and maintenance of nutrient cycling and hydrological processes.

Full text

ZurichOpenRepositoryandArchiveUniversityofZurichUniversityLibraryStrickhofstrasse39CH-8057Zurichwww.zora.uzh.chYear:2022SoilnutrientcontentandwaterlevelvariationdrivemangroveforestabovegroundbiomassinthelagoonalecosystemofAldabraAtollConstance,Annabelle;Oehri,Jacqueline;Bunbury,Nancy;Wiesenberg,GuidoLB;Pennekamp,Frank;A’Bear,Luke;Fleischer-Dogley,Frauke;Schaepman-Strub,GabrielaAbstract:Lagoonalmangroveecosystemsarevitalforcarboncapture,protectionofcoastlinesandconservationofbiodiversity.Yet,theyaredecreasinggloballyatahigherratethanothermangroveecosystems.Inadditiontohumandrivers,localenvironmentalfactorsinuencethefunctioningoflagoonalmangroveecosystems,buttheirimportanceandcombinedeectsarerelativelyunknown.Here,weinvestigatethedriversofmangrovefunctioning,approximatedbymangroveabovegroundbiomass(AGB),inaprotectedlagoonalmangroveecosystemonAldabraAtoll,Seychelles.Basedonasurveyofthemangroveforeststructurein54plots,weestimatedthatthemeanmangroveforestAGBwas82±13Mgha−1.ThetotalAGBofthemangrovearea(1720ha)wasnearly140,600Mg,equivalenttoabout66,100MgofcarbonstoredinthestandingbiomassonAldabra.Toassessthedirectandindirecteectsofsoilnutrientcontent,waterlevelvariationandsoilsalinityonmangroveAGB,weusedastructuralequationmodel.Ourstructuralequationmodelexplained82%ofthevariationinmangroveAGB.Thesoilnutrientcontent(concentrationofessentialmacronutrientsinthesoilcolumn)hadthegreatestinuenceonmangroveAGBvariation.Additionally,highvariationinwaterlevel(changeinwaterdepthcoveringalocation)increasedmangroveAGBbyincreasingnutrientcontentlevels.OurresultshighlighttheimportantcontributionofAldabra’slagoonalecosystemtoSeychelles’carbonstorageandtheroleofhydroperiodasaregulatorcontrollingtheavailabilityofcrucialnutrientsneededforthefunctioningofmangroveswithinlagoonalsystems.Wesuggestconservationmanagersworldwidefocusonaholisticecosystem-levelperspectiveforsuccessfulmangroveconservation,includingtheprotectionandmaintenanceofnutrientcyclingandhydrologicalprocesses.DOI:https://doi.org/10.1016/j.ecolind.2022.109292PostedattheZurichOpenRepositoryandArchive,UniversityofZurichZORAURL:https://doi.org/10.5167/uzh-223719JournalArticlePublishedVersion
ThefollowingworkislicensedunderaCreativeCommons:Attribution4.0International(CCBY4.0)License.Originallypublishedat:Constance,Annabelle;Oehri,Jacqueline;Bunbury,Nancy;Wiesenberg,GuidoLB;Pennekamp,Frank;A’Bear,Luke;Fleischer-Dogley,Frauke;Schaepman-Strub,Gabriela(2022).Soilnutrientcontentand

waterlevelvariationdrivemangroveforestabovegroundbiomassinthelagoonalecosystemofAldabraAtoll.EcologicalIndicators,143:109292.DOI:https://doi.org/10.1016/j.ecolind.2022.1092922

Ecological Indicators 143 (2022) 109292
Available online 12 September 2022
1470-160X/© 2022 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
Original Articles
Soil nutrient content and water level variation drive mangrove forest
aboveground biomass in the lagoonal ecosystem of Aldabra Atoll
Annabelle Constance
a
,
b
,
*
, Jacqueline Oehri
a
, Nancy Bunbury
b
,
c
, Guido L.B. Wiesenberg
d
,
Frank Pennekamp
a
, Luke A’Bear
b
, Frauke Fleischer-Dogley
b
, Gabriela Schaepman-Strub
a
a
Department of Evolutionary Biology and Environmental Studies, University of Zurich, Zurich, Switzerland
b
Seychelles Islands Foundation, PO Box 853, Victoria, Mah´
e, Seychelles
c
Centre for Ecology and Conservation, University of Exeter, Cornwall Campus, Penryn TR10 9FE, UK
d
Department of Geography, University of Zurich, Zurich, Switzerland
ARTICLE INFO
Keywords:
Blue carbon
Coastal ecosystem
Conservation
Coralline
Islands
Structural equation model
ABSTRACT
Lagoonal mangrove ecosystems are vital for carbon capture, protection of coastlines and conservation of
biodiversity. Yet, they are decreasing globally at a higher rate than other mangrove ecosystems. In addition to
human drivers, local environmental factors inuence the functioning of lagoonal mangrove ecosystems, but their
importance and combined effects are relatively unknown. Here, we investigate the drivers of mangrove func-
tioning, approximated by mangrove aboveground biomass (AGB), in a protected lagoonal mangrove ecosystem
on Aldabra Atoll, Seychelles. Based on a survey of the mangrove forest structure in 54 plots, we estimated that
the mean mangrove forest AGB was 82 ±13 Mg ha
−1
. The total AGB of the mangrove area (1720 ha) was nearly
140,600 Mg, equivalent to about 66,100 Mg of carbon stored in the standing biomass on Aldabra. To assess the
direct and indirect effects of soil nutrient content, water level variation and soil salinity on mangrove AGB, we
used a structural equation model. Our structural equation model explained 82 % of the variation in mangrove
AGB. The soil nutrient content (concentration of essential macronutrients in the soil column) had the greatest
inuence on mangrove AGB variation. Additionally, high variation in water level (change in water depth
covering a location) increased mangrove AGB by increasing nutrient content levels. Our results highlight the
important contribution of Aldabra’s lagoonal ecosystem to Seychelles’ carbon storage and the role of hydro-
period as a regulator controlling the availability of crucial nutrients needed for the functioning of mangroves
within lagoonal systems. We suggest conservation managers worldwide focus on a holistic ecosystem-level
perspective for successful mangrove conservation, including the protection and maintenance of nutrient
cycling and hydrological processes.
1. Introduction
Mangrove ecosystems are hotspots for inorganic carbon capture
(Donato et al., 2011) and play a major role in regulating nutrient uxes,
trapping sediments, and protecting tropical coastlines (Duarte et al.,
2013, Carr et al., 2017). Mangroves provide a unique habitat for globally
threatened vertebrate populations (Luther and Greenberg, 2009) and
support high diversity and abundance of sh (Serafy et al., 2015), rep-
tiles, and marine megafauna in nearby reefs and seagrass meadows
(Sievers et al., 2019, Zhao et al., 2020). Despite their critical importance,
mangroves worldwide are in decline primarily because of human ac-
tivity, and are further threatened by the effects of climate change
(Woodroffe et al., 2016, Goldberg et al., 2020). Lagoonal mangroves
occupy 11 % of the total mangrove area and are being lost at a rate twice
that of other mangrove ecosystems (Worthington et al., 2020).
A lagoon ecosystem consists of a shallow coastal waterbody inter-
mittently separated from larger bodies of water by a natural barrier
(Kjerfve, 1994; Vadineanu, 2004). A key feature of lagoons is the
mangroves growing in this transitional environment between land and
open ocean (Vadineanu, 2004). Lagoonal mangroves on oceanic islands
usually occur within a carbonate setting (i.e., coralline islands) and are
exposed to high wave energy and lower sediment input than mangroves
in estuarine or deltaic environments (Ewel et al., 1998, Twilley et al.,
2018, Raw et al., 2019, Constance et al., 2021). In these harsh
* Corresponding author.
E-mail addresses: annabelle.constance@ieu.uzh.ch, constanceannabelle@hotmail.com (A. Constance).
Contents lists available at ScienceDirect
Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind
https://doi.org/10.1016/j.ecolind.2022.109292
Received 1 June 2022; Received in revised form 4 August 2022; Accepted 6 August 2022

Ecological Indicators 143 (2022) 109292
2
environmental conditions, productivity and mangrove aboveground
biomass (AGB) tend to be lower than in other mangrove ecosystems
(Schaeffer-Novelli et al., 1990, Ewel et al., 1998, Worthington et al.,
2020). Despite their lower AGB, however, the contribution of lagoonal
mangroves to the total soil organic carbon stock of mangroves is rela-
tively higher than other mangrove ecosystem types (Twilley et al.,
2018).
Mangrove AGB – the aboveground standing dry mass of live matter
from trees – is an established indicator of the organic carbon stock in
mangrove ecosystems. Mangrove AGB is driven by multiple factors that
operate at distinct spatial scales (Twilley et al., 2018). At global and
regional scales, climatic factors (Hutchison et al., 2014) such as tem-
perature, precipitation and cyclone frequency (Simard et al., 2018)
control the functional pattern of mangroves. At the local scale, drivers of
mangrove biomass include soil salinity, nutrients, pH, (Ukpong, 1997;
Krauss et al. 2006, Reef et al., 2010) sultes, moisture (Xiong et al.,
2018), hydroperiod (Crase et al., 2013), and their interactions with local
geomorphology (Twilley et al., 2018). Of these environmental drivers,
the inuence of salinity on mangrove growth rates and development is
the most pervasive (Ukpong, 1997; Casta˜
neda-Moya et al., 2006). An
increase in soil salinity likely increases mangrove growth rates up to an
optimum level above which mangrove growth is reduced (Krauss and
Ball, 2012). Especially on coralline islands with limited nutrient input
from land, mangrove soils have low nutrient availability (Duarte et al.,
1998, Reef et al., 2010). Hence, nutrients are often limiting for
mangrove growth (Casta˜
neda-Moya et al., 2020). The direct and indirect
effects of hydroperiod (frequency, duration, and depth of water covering
a location by the action of tides, rivers, and precipitation) on mangrove
AGB are complex (Woodroffe et al., 2016). Hydroperiod, like salinity
and pH, is likely a regulator for mangroves, which controls growth but is
not consumed (Twilley and Day, 2012). Alternatively, hydroperiod can
dominate the concentration and availability of resources (e.g., nutrients)
and other regulators in mangrove environments (Twilley and Day,
2012). For instance, higher variation in water level (depth of water
covering a location) can maximize aboveground productivity by pro-
moting nutrient exchange, increased oxygen availability to the roots and
a reduced accumulation of toxic sultes (Twilley et al., 1986, Krauss
et al., 2006, Casta˜
neda-Moya et al., 2011). Few studies have investigated
how these local environmental drivers interact and their combined
importance in explaining mangrove AGB (Twilley et al., 1986, Ukpong,
1997; Casta˜
neda-Moya et al., 2006, Krauss et al., 2006, Casta˜
neda-Moya
et al., 2011).
The aim of this study is to quantify the mangrove ecosystem structure
and function in a protected lagoon environment and to understand the
combined inuence and the relative importance of local environmental
drivers on mangrove AGB in a lagoonal ecosystem. We used data
collected from 54 plots on Aldabra Atoll, a UNESCO World Heritage Site
with the largest extent (1720 ha) of mangroves in Seychelles, to address
the following questions: (1) what are the AGB and environmental con-
ditions in the mangroves on Aldabra?; and (2) what is the relative role of
soil salinity, nutrients, and hydroperiod in driving variation in
mangrove forest AGB? Specically, our hypotheses are that (Fig. A1): (i)
greater water level variation positively affects mangrove AGB; (ii) soil
salinity increases mangrove AGB. We assume this hypothesis to be true
up to hypersaline conditions (>40 Practical Salinity Scale); (iii) higher
soil nutrients increase mangrove AGB. We assume that mangrove
growth is often nitrogen or phosphorus limited; (iv) soil nutrient con-
tents and (v) soil salinity are inuenced by water level variation. We
assume that water level variation is an indicator for the redistribution,
suspension, and deposition of sediments, minerals, and nutrients
occurring in the shallow water column within the lagoon ecosystem
(Gamito et al., 2004).
The rst question will provide information on the contribution of
Aldabra to Seychelles’ climate change mitigation and adaptation efforts,
Fig. 1. he two sampling site locations on Picard and Grande Terre island of Aldabra Atoll Islands. The insets show the location of (1) surveyed plots according to
mangrove area coverage on Picard Island (2) Grande Terre Island (3) Aldabra (red dot) within the Indian Ocean. Digital Globe, 2011, captured the satellite imagery
shown in insets 1 and 2.
A. Constance et al.

Ecological Indicators 143 (2022) 109292
3
given the massive potential of blue carbon (carbon captured and stored
by coastal and marine ecosystems) as efcient carbon sinks (Taillardat
et al., 2018). The second question will improve understanding of re-
sources and regulators as indicators for a vital ecosystem function of
lagoonal mangroves closely tied to AGB, namely inorganic carbon cap-
ture. This study can ultimately inform optimal conservation manage-
ment worldwide of the vulnerability of mangrove ecosystems to changes
in hydrological processes at the coastline, salinity and other local
environmental factors.
2. Methods
2.1. Study design
2.1.1. Study site
Aldabra Atoll (9◦24′S, 46◦20′E) lies in the Western Indian Ocean and
has a land area of 15,500 ha and a mean elevation of 8 m above sea level
(Fig. 1; Stoddart et al., 1971). Aldabra has four rim islands, with other
smaller islands scattered within the lagoon area of 20,500 ha (Hamylton
et al., 2018). Nearly three quarters of the lagoon drains at low tides
through the channels separating the main islands (Hamylton et al.,
2018), as Aldabra is inuenced by a tidal range of 2–3 m.
Aldabra’s climate has pronounced seasonal variation. At least 75 %
of the yearly rainfall (mean 975 mm year
−1
) falls during the north-west
monsoon from November to April. May to October is the driest period,
with predominantly southeast trade winds. The rainfall variation drives
the seasonality in net primary productivity on the atoll for most vege-
tation except mangroves (Haverkamp et al., 2017).
Aldabra’s mangrove forests occur on the coastal areas inside the
atoll’s lagoon, with an estimated total extent of 1720 ha (Walton et al.,
2019). A recent study using remotely-sensed data found that the extent
of mangrove forests on the atoll has increased by 60 ha over the past 20
years (Constance et al., 2021), although there was considerable varia-
tion within this period. Seven species of mangroves have been recorded
on Aldabra: Avicennia marina; Bruguiera gymnorhiza; Ceriops tagal; Rhi-
zophora mucronata; Lumnitzera racemosa; Sonneratia alba; Xylocarpus
granatum (Macnae et al., 1971).
Aldabra is a Ramsar Wetland Site of International Importance and
hence an ideal site to study the drivers of mangrove AGB in a lagoonal
environment with its naturally occurring stands of mangroves that have
been strictly protected since 1976. The mangrove forests support several
globally threatened and/or endemic populations of terrestrial and ma-
rine species, including the largest breeding population of frigatebirds in
the Indian Ocean (ˇ
Súr et al., 2013). The lagoon and surrounding man-
groves are an essential nursery for the second largest population of en-
dangered green turtles in the Western Indian Ocean (Mortimer et al.,
2011).
2.1.2. Sampling design
From October 2019 to November 2020, we surveyed the mangrove
ecosystem structure within two mangrove sites on Aldabra (see Fig. 1).
To ensure that we surveyed a representative set of plots and to capture
the variability in mangrove growing conditions on the atoll, we used a
stratied sampling design based on mangrove area coverage from a land
cover map (Walton et al., 2019). Aldabra’s mangroves do not show the
typical species zonation described for other mangrove areas (Macnae
et al., 1971). This means our sampling strategy based on mangrove
coverage does not indicate different dominant species. For each site, we
established nine monitoring blocks (40 ×40 m) by selecting three blocks
for each of the three mangrove area coverage categories (low: 8–38 %;
medium: 39–69 %; high: 70–100 %). We used a random selection
method for the blocks. We overlaid grids of sizes 40 ×40 m on the land
cover map and then used a table of random numbers to select the center
coordinates of nine grids as the center coordinates for the selected block.
Within each block, we surveyed three plots of 5 ×5 m. In total, we
surveyed 54 plots within 18 blocks across two mangrove sites on Alda-
bra. The surveys covered 1 % and 0.2 % of the total area of mangroves on
Picard and Grande Terre, respectively.
2.2. Measurements of ecosystem structure and function
2.2.1. Mangrove aboveground biomass
In each plot, we identied the species of all adult trees (i.e., trees >
1.5 m or with reproductive traits) and measured their height and
diameter at breast height (DBH, using a measuring tape). For trees taller
than 8 m, we used an inclinometer (Suunto PM-5/66 PC, Vantaa,
Finland) to measure height to an accuracy of ±0.5 m. We used mea-
surement guidelines to determine DBH for trees with anomalies, such as
stilt roots, at breast height (Kauffman and Donato, 2012). We calculated
the AGB for trees from a general tropical allometric equation (Equation
(1), Chave et al., 2005), as no specic allometric equations exist for
Aldabra mangrove tree species. Several studies showed its applicability
to estimating mangrove biomass when no site or species-specic allo-
metric model is available (Fayolle et al., 2013, Njana et al., 2015).
AGB =0.0509 •
ρ
•D2•H(1)
where AGB is the aboveground biomass per tree in kg,
ρ
is the wood
density in g cm
−3
, D is DBH in cm, and H is height in m. We retrieved
species-specic wood densities from the world agroforestry wood den-
sity database (https://www.worldagroforestry.org/sea/Products/
AFDbases/.
WD/Index.htm; Table A1). We chose the lowest density value for
each species from the database because species’ wood densities and AGB
tend to be lower for mangrove forests near the sea than in deltaic and
inland riverine mangrove systems (Komiyama et al., 2008). None of the
densities were specic to Aldabra but were comparable to values used in
other eld studies on mangrove trees in the region (Njana et al., 2017).
This equation has a standard error of 12.5 % generated for mangrove
stands with a maximum DBH of 42 cm (Chave et al., 2005). To represent
mangrove AGB per ha, we summed the total mangrove AGB of trees
within a plot and divided by the area of the plot. We calculated the total
AGB of Aldabra by multiplying the mean AGB for all plots by the total
mangrove extent on the atoll (Kauffman and Donato, 2012, Walton
et al., 2019). We used a conversion factor of 0.47 to transform mangrove
AGB to carbon stock of live mangroves (Kauffman and Donato, 2012).
2.2.2. Water level variation
We recorded the hydroperiod in 12 of 18 blocks using autonomous
water level loggers (DCX-22 CTD, Keller, Switzerland) from October
2019 to November 2020. The loggers recorded water level (water height
measured from ground level to water surface), salinity, and tempera-
ture, every 10 min for up to one year. The accuracy of the logger was ±
0.005 mH
2
O (meters of water pressure unit) at a compensated temper-
ature range of 10–60 ◦C. From these data, we calculated the water level
variation; the standard deviation of daily water level. To represent water
level variation for each plot, we used the average value of the daily
water level standard deviation measured over more than a year. We
measured the aboveground water level, because our sediment sampling
was limited in depth by the karst structure of the underlying limestone.
2.2.3. Soil salinity
To assess the soil column depth, we rst inserted a at scraper knife
into the soil until it touched the underlying limestone bedrock at three
locations within each plot, and measured the vertical distance from the
surface down to the tip of the knife. At one of these measuring locations,
A. Constance et al.

Ecological Indicators 143 (2022) 109292
4
soil samples were taken with a hand shovel every 10 cm along the
vertical soil column to account for differences with depth and species-
specic resource use. From the extracted soil, we measured soil
salinity employing a soil-to-distilled water solution ratio of 1:2.5 (FAO,
2006) using a handheld conductivity meter (WTW Cond 3310, Xylem
Analytics Germany, Germany). The salinity readings were compensated
to a standard temperature of 25 ◦C and were a unitless measure of
salinity according to the Practical Salinity Scale 1978. This measure of
salinity is roughly equivalent to g of dissolved salts per 1 kg of water. We
recorded soil salinity in the eld at least four times per plot during the
wet and dry seasons. To represent soil salinity for each plot, we used the
average value calculated from four temporal measurements and up to
four measurements along the soil column. We intended to measure soil
bulk density using volumetric soil cylinders and alternative methods.
However, it was impossible to get accurate volumetric samples in
waterlogged conditions in the eld. Furthermore, transportation was
difcult, so that samples could not be transferred to the laboratory fa-
cilities without disturbance. We recognized the comparative issues that
might arise from this shortcoming in the eld. As a result, we stan-
dardized the soil sample volume collected with the shovel according to
the intended method (pipe volume and depth). In this study, we,
therefore, report soil nutrient content and not stock (Section 2.2.4).
2.2.4. Soil nutrient content
The soil samples extracted from the eld (Section 2.2.3) were
analyzed in the laboratory for their texture, elemental nutrient con-
centrations, and pH. The soil analyses were carried out on ne earth
fractions (<2 mm). Collected soil samples were sieved through a 2-mm
mesh for the measurement of soil texture after oven drying (40 ◦C) for
48 h. Soil textural class (proportion of various particle-size classes) and
the percent of clay were estimated by the eld texture determination
method (FAO, 2006). The remaining analyses were done on aliquots of
soil material (ca. 5 g) that were milled to <50
μ
m to create homogenous,
representative sub-samples of the eld sample, and have greater
reproducibility of results. Essential macroelemental concentrations
(mass-%) of magnesium, sulfur, phosphorus, calcium and potassium
were determined with an energy dispersive X-ray uorescence spec-
trometer (SPECTRO X-LAB 2000, SPECTRO Analytical Instruments,
Germany). The total nitrogen concentration was determined using
elemental analysis isotope ratio mass spectrometry (EA-IRMS; Thermo
Fisher Scientic Flash HT Plus elemental analyzer equipped with a
thermal conductivity detector and coupled via ConFlo IV to a Delta V
Plus Mass Spectrometer). Henceforth, nutritional elements are referred
to as nutrients for simplicity. Finally, we calculated a soil nutrient
content variable for each plot by multiplying the mean concentration for
each of the six essential nutrients by the mean soil depth (three mea-
surements of soil depth per plot, Section 2.2.3). We assume the variable
is a fair approximation of the soil nutrient content in the mangrove
conditions (see also Donato et al., 2011).
2.3. Statistical analysis
2.3.1. Mangrove ecosystem structure and function
We summarized the information collected on ecosystem structure
and function from the eld survey in three main ways. Firstly, to make
generalizations about the lagoonal mangrove ecosystem on Aldabra, we
did a tabular summary of plot-level measures of mangrove AGB, soil
nutrient concentrations, other soil properties and hydroperiod (Table 1).
Secondly, we mapped and compared site differences for environmental
drivers and mangrove AGB using unpaired t-tests on block-level esti-
mates (mean values of plots, Fig. 2). We used block-level estimates for
the t-tests to account for the systematic correlation of data collected at
the plot level, which would violate the independence assumption of t-
tests. We checked that the data residuals were approximately normally
distributed and tested the homogeneity of variances. We used an un-
equal variance t-test, when two groups of data had unequal variances
based on the results of Levene’s test. Thirdly, to assess the soil nutrient
content conditions across Aldabra’s mangrove ecosystem, we ran a
principal component analysis (PCA) of the soil nutrient content vari-
ables (Fig. 3). The PCA was useful for summarizing the covariation
among the essential macronutrients and understanding the variation
among plots in relation to each of the essential macronutrients,
considering the plots’ AGB and dominant species.
2.3.2. Drivers of mangrove aboveground biomass
We constructed structural equation models (SEM) from a priori hy-
potheses derived from ecological theory to mechanistically link
mangrove AGB and its drivers using plot level information. First, we
explored the variables of interest (Section 2.2) to be tested in the sta-
tistical model. To capture the main gradients in the soil content of all six
macronutrients (Section 2.2.3), we derived a composite variable (Grace
and Keeley, 2006). Specically, we tted mangrove AGB as a function of
the six soil macronutrients using a general linear model and summed
their predicted estimates per plot. The resulting soil nutrient content
variable represented the combined inuence of the six macronutrients
(Grace et al., 2010). The remaining variables (soil salinity, water level
variation and mangrove AGB) were tted into the model as per their
Table 1
Summary of Aldabra’s mangrove ecosystem structure and function measured
within 54 plots. Minimum (min), maximum (max), average (mean), standard
deviation (SD) and the number of samples (N) from which the estimates were
derived for water, soil, plot and tree variables are shown.
min max mean ± SD N (%)
Tree height (m) 0.8 12.4 3.9 ±2.3 769
Diameter at breast height (cm) 0.3 32.4 5.9 ±4.6 769
Tree aboveground biomass (kg) <0.1 243.4 14.4 ±30.0 769
Plot aboveground biomass (Mg/ha) 1.2 381.6 81.7 ±95.4 54
Plot aboveground carbon (Mg/ha) 0.5 179.4 38.4 ±44.8 54
# trees per species recorded
Avicennia marina 24 (3)
Bruguiera gymnorrhiza 29 (4)
Ceriops tagal 183 (24)
Lumnitzera racemosa 9 (1)
Sonneratia alba 1 (0)
Rhizophora mucronata 523 (68)
Water level (cm) 0 173 17.2 ±26.6 12
a
Water salinity
b
0 61.9 13.2 ±13.5 12
a
Water temperature (◦C) 20.4 42.0 27.5 ±2.9 12
a
Soil salinity
b
0 19.2 8.3 ±3.1 288
Soil pH 7.2 ±0.8 87
Mean soil depth (cm) 13.7 ±5.9 54
Soil macronutrient concentration (mass %)
Magnesium 1.9 ±0.6 88
Sulfur 1.7 ±1.0 88
Phosphorus 0.8 ±1.2 88
Potassium 0.3 ±0.2 88
Nitrogen 1.2 ±0.5 79
Calcium 15.3 ±10.6 88
Soil clay (mass %) 3.2 ±3.3 89
Soil texture (mass %)
Sand 47 (53)
Loamy sand 23 (26)
Sandy loam clay-poor 2 (2)
Sandy loam 1 (1)
Loam 2 (2)
Highly organic 14 (16)
a
N refers to the number of sampling locations across sites for hydrology
measures from November 2019 to November 2020. The combined total of water-
related recordings was 484,736.
b
The measurement of salinity was unitless (Practical Salinity Scale 1978) but
is roughly equivalent to g of dissolved salts per 1 kg of water.
A. Constance et al.

Ecological Indicators 143 (2022) 109292
5
Fig. 2. Distribution of the mangrove aboveground biomass, mean daily water level variation, soil salinity, soil phosphorus and soil nitrogen content (soil con-
centration * soil depth of column) within and across surveyed mangrove sites on Picard (left) and Grande Terre (right) islands on Aldabra. The block names (e.g., AH,
KM) and their mean values for the corresponding measurements are displayed across rows. Source of the underlying land cover map: Walton et al., 2019.
A. Constance et al.

Ecological Indicators 143 (2022) 109292
6
plot-level values described in Section 2.2.
We modeled AGB using general linear mixed-effects models as our
study design was nested (three plots measured within each block). The
specied models (Table A3) were tted with the block as a random effect
(random intercept). We accounted for the spatial autocorrelation
(proximity of plots) that could inuence AGB variations by specifying an
exponential spatial correlation structure within the models based on the
plot coordinates. Model assumptions and residuals were assessed visu-
ally, and the goodness of t was determined using AIC and Fisher’s C
statistic. We used R (R Core Team, 2020) with the packages nlme
(Pinheiro et al., 2020) to perform linear mixed analysis, and piece-
wiseSEM (Lefcheck, 2015) for the SEM.
3. Results
3.1. Mangrove ecosystem structure and function
A total of 769 mangrove trees of six species were surveyed (Table 1).
Rhizophora mucronata was the dominant species (68 % of individuals).
The mangrove forest AGB (mean ±SE) was 81.7 ±13.0 Mg ha
−1
. The
total mangrove AGB, estimated at 140,598 Mg, is equivalent to 66,081
Mg carbon stored in the standing mangrove biomass on Aldabra. The
water level and salinity (mean ±SD) were 17.2 ±26.6 cm and 13.2 ±
13.5, respectively. The mean soil salinity calculated from 288 soil
samples was 8.3 ±3.1, while the mean soil depth varied slightly across
plots (13.7 ±5.9 cm). The highest concentration of soil essential
Fig. 3. The relationships between soil
nutrient content variables across plots.
Individual symbols represent plot ob-
servations. Observations are colored ac-
cording to their plot aboveground
biomass, and their shape shows the
dominant mangrove species (most
frequent) recorded at the respective
plot. The blue arrows depict the strength
(arrow length) and direction of soil
nutrient content gradients [nitrogen (N),
phosphorus (P), potassium (K), sulfur
(S), calcium (Ca), and magnesium (Mg)]
across plots. The nutrients S, K, Mg and
N contributed >90 % to the major
nutrient content gradient (rst principal
component PC1). Ca, P and N contrib-
uted nearly 90 % to the second major
nutrient content gradient (second prin-
cipal component PC2). A plot observa-
tion that is on the same (opposite) side
as a given soil content variable has a
high (low) value for this variable.
Fig. 4. Structural equation model showing drivers of aboveground mangrove biomass in the lagoonal environment of Aldabra (model tting parameters are Fisher’s
C =0.87, P =0.65, degrees of freedom =2). Solid and dashed arrows represent signicant (P < 0.05) and non-signicant paths, respectively. Numbers adjacent to
arrows are the standardized effect sizes of the relationship (Grace et al., 2012; unstandardized coefcients are given in Table A3). The coefcient of determination
(R
2
) value reports the total variation in a dependent variable (soil nutrient content, mangrove AGB) that is explained by the combined predictor variables (mangrove
tree image from ian.umces.edu/symbols/).
A. Constance et al.

Ecological Indicators 143 (2022) 109292
7
macronutrients recorded from 88 soil samples (mean ±SD) was 15.3 ±
10.6 % for calcium.
Mangrove aboveground biomass was similar on Picard and Grande
Terre (P >0.05, t =1.52, df =11). Spatially, the AGB on Picard tended
to decrease with increasing distance from the lagoon, while on Grande
Terre, there was no clear trend in AGB with distance to the primary
branching of the lagoon (Fig. 2). The daily variation in water level was
highest on Picard (P < 0.05, t =2.69, df =6). The most water-exposed
block on Picard experienced on average 37.2 cm SD in daily water level
and the lowest value recorded was for Aldabra was on Grande Terre (7.4
cm). Changes in the water level tracked the changes in diurnal tide in all
blocks on the atoll (Fig. A3). Soil salinity was marginally higher on
Grande Terre than on Picard (P =0.05, t =-1.72, df =16) and the inland
blocks on Grande Terre tended to have the highest mean salinity
(9.1–11.2). The mean nitrogen content of blocks on Grande Terre was on
average nearly twice as high as on Picard (P < 0.05, t =-2.25, df =11).
The highest nitrogen content was measured in blocks further inland
from the lagoon on Grande Terre. Phosphorus content was similar be-
tween Grande Terre and Picard (P >0.05, W =54), although values
peaked in the more water-exposed blocks on Picard.
Plots with higher values of phosphorus, magnesium, calcium, and
sulfur had on average higher mangrove AGB (Fig. 3). The correlations
among soil nutrient content variables (depicted by corresponding angles
in Fig. 3) varied: potassium and sulfur, as well as calcium and phos-
phorus, were strongly positively correlated. In contrast, calcium and
nitrogen were negatively correlated. Ceriops tagal largely dominated
lower-nutrient plots (opposite side of soil content variables), whereas
Rhizhophora mucronata was more widespread in higher-nutrient plots.
Plots dominated by Avicennia marina tended to have higher nitrogen
content. The proportion of variance explained by each principal
component is provided in Table A2 and a biplot including the rst three
principal components (PC3 explains 15 % of the variation) is shown in
Fig. A2.
3.2. Drivers of mangrove aboveground biomass
We investigated the combined inuence and the relative importance
of local environmental drivers on mangrove AGB in a lagoonal
ecosystem with structural equation models. Our SEM were supported for
the lagoonal mangrove ecosystem since the covariate structure implied
by the SEM did not differ from the covariance structure in the observed
data (Fisher’s C, df =2, P >0.05, Fig. 4). Soil nutrient content, water
level variation, and soil salinity explained 82 % of the variation in
mangrove AGB. Soil nutrient content was the most inuential driver of
mangrove AGB with a standardized coefcient of 0.80 (Table A3)—a
change of 1 SD in the total soil nutrient content is estimated to result in a
0.80 SD change in mangrove AGB. Water level variation had a similar
positive effect on mangrove AGB (0.58; standardized coefcient)
through its positive effect on soil nutrient content (see Fig. 4). Specif-
ically, higher variation in the water level promoted higher soil nutrient
content, which positively inuenced the AGB. The total soil nutrient
content variation explained by the daily water level variation was 52 %.
We did not observe an indirect effect of water level variation on
mangrove AGB through soil salinity (path removed as it did not improve
model tting (see Table A3 for the full model). Soil salinity and water
level had weak direct effects on mangrove AGB (P >0.05; Table A3).
4. Discussion
Our study found that mangrove forest aboveground biomass (AGB)
on Aldabra Atoll is 81.7 ±13.0 Mg ha
−1
(mean ±SE). This gure is
comparable to a global mean estimate of 73.5 Mg ha
−1
for lagoonal
mangrove ecosystems, the most threatened type of mangrove globally
(Worthington et al., 2020), and sits in the middle range for mangrove
AGB reported in the Western Indian Ocean (Alavaisha and Mangora,
2016, Benson et al., 2017, Njana et al., 2017). The total mangrove AGB
of 140,598 Mg, equivalent to 66,081 Mg carbon stored in the standing
biomass on Aldabra, is the highest mangrove AGB carbon stock contri-
bution to the Seychelles and one of the highest reported for a coralline
island (Simard et al., 2018). We therefore suggest that the discussion of
the key role of mangrove ecosystems in carbon storage (Donato et al.,
2011) is broadened to include lagoonal mangrove ecosystems and
coralline islands.
Soil nutrient content was the most important driver of mangrove
AGB in our study, supporting hypothesis iii. Our analysis suggests that
higher contents of phosphorus, magnesium, calcium and sulfur on
Aldabra are correlated with higher mangrove AGB (see Fig. 3). Phos-
phorus is essential for plant metabolism, whereas magnesium is an
alkaline element taken up in ionic form by plants (Schulze et al., 2019).
The magnesium ion is involved in chlorophyll-related functions within
the plant and is comparatively more abundant in seawater than the
other macronutrients (Ukpong, 1997; Schulze et al., 2019). However,
the most important antagonistic (negatively correlated) elements to the
uptake of Mg
2+
are Ca
2+
and K
+
(Schulze et al., 2019). A complex range
of interacting effects between different essential nutrients and the
calcium-rich soils on Aldabra is likely occurring and supports the use of a
composite variable to represent the total effect of nutrient content on
mangrove AGB.
Hydroperiod, specically a high variation in daily water level,
positively inuenced soil nutrient content on Aldabra, and therefore
mangrove AGB. This nding supports hypothesis iv that in a lagoonal
system, a high water level variation likely redistributes and suspends
nutrients and other sediments from the sediment to the surface layer in
the water column. Furthermore, lower water level variation probably
reduces the aeration of soil layers and can have adverse effects on
bacteria that facilitate nutrient uptake by mangroves (Twilley et al.,
1986, Krauss et al., 2006, Casta˜
neda-Moya et al., 2011, Krauss et al.,
2014). Water level variation is therefore a critical regulator contributing
to the nutrient content and the extent to which mangrove growth is
resource-limited within a lagoon ecosystem (Krauss et al. 2006, Reef
et al., 2010). Therefore, monitoring, maintaining and understanding
natural or human-mediated changes to water level is crucial for man-
aging this ecosystem.
Our study did not provide enough evidence to support hypotheses ii
and v. Soil salinity did not have an effect on AGB either directly or
indirectly through water level variation, despite soil salinity being long
linked to mangrove physiology, growth rates and structure (Ukpong,
1997; Casta˜
neda-Moya et al., 2006, Crase et al., 2013). Soil salinity
recorded in the majority of places on Aldabra (mean ±SD =8.3 ±3.1)
is outside the hypersaline range for mangrove growing conditions (Chen
and Ye, 2014, Ball and Pidsley, 1995) and thus provides little explana-
tory power compared to soil nutrient content, which is a more integrated
measure of soil depth and nutrient concentration. However, plot-level
salinity was only available for an average of four time points across
the wet and dry seasons, which might have limited our power to detect
an effect of salinity. In addition, we lack information on the effects of
freshwater sources, including groundwater, on the soil salinity observed
in our study site, which likely reduces the physiological salinity stress for
mangroves on Aldabra. Based on our analyses, salinity effects on
mangrove function might be less important in areas with moderate salt
content, although more evidence is needed on the factors regulating
salinity levels directly.
We observed clustering patterns of surveyed plots in relation to the
nutrient content gradients on Aldabra (see Fig. 3). Rhizhophora mucro-
nata dominated in plots (27/54) with relatively low to relatively high
nutrient content. This explains the species’ dominance in lagoonal
mangrove forests on Aldabra (68 % of all mangrove trees recorded).
Ceriops tagal thrived in areas of relatively low nutrient content while
Avicennia marina dominated in soils with high nitrogen content. Ac-
cording to experimental studies, Avicennia marina has a greater capacity
to assimilate excess nitrogen for growth than other mangrove species,
including Ceriops tagal (Yates et al., 2001; Naidoo, 2009). The
A. Constance et al.

Ecological Indicators 143 (2022) 109292
8
dominance of these species in relation to nutrient content likely reects
their ecological niches in nature; information that can inform manage-
ment decisions, such as selecting species for restoration projects, or
understanding the likely effects of higher-nutrient environments on a
species. In this context, Aldabra serves as a crucial example of how
conservation can support the function of lagoonal ecosystems and
enable them to adapt to (or even benet from) future disturbances.
We observed spatial differences in nutrient content within and across
surveyed mangrove sites that suggest additional factors are contributing
to soil nutrient content on Aldabra. For example, we recorded higher
nitrogen content on Grande Terre in blocks several hundred meters
inland with the lowest water level variation (see Fig. 2). Because the vast
majority of terrestrial organic matter in these areas is of chelonian origin
(Falc´
on and Hansen, 2018), we suggest that the Aldabra giant tortoise
plays a role in the nutrient cycle within the mangroves (Macnae et al.,
1971, Grubb, 1971). Furthermore, Picard hosts one of Aldabra’s four
breeding frigatebird colonies (ˇ
Súr et al., 2013) and thousands of red-
footed boobies nest in its mangroves. These seabirds likely contribute
nutrient-rich guano to the mangrove forest where they nest (Wu et al.,
2018; Reef et al., 2010; De La Pe˜
na-Lastra, 2021). Grazing by dugongs
and green turtles provides greater nutrient availability to seagrass eco-
systems even up to one year later (Aragones et al., 2006). We expect
these processes to be relevant for the mangrove-lagoon ecosystem on
Aldabra with its large green turtle population (Mortimer et al., 2011)
and small population of dugongs. Loss of this macrofauna can greatly
degrade the amount and location of nutrients available to the mangrove
ecosystem, underscoring the conservation relevance of these threatened
populations and their habitat.
One of the strengths of the structural equation modeling used in our
study is that additional paths can be added as more data become
available (Grace et al., 2012). For Aldabra, we anticipate that the
addition of data quantifying freshwater sources, soil properties
(including moisture, sulte concentrations, rates of organic matter
decomposition) and biotic interactions (e.g., macrofauna abundance)
could strengthen the models predicting mangrove AGB. The SEM pre-
sented in our study can be used in the conservation management of
lagoonal systems, for example, by guiding prioritization of the most
appropriate locations for restoration where hydroperiod regulators can
redistribute nutrient supply that support mangrove forest functioning.
In the absence of direct human stressors, we show that soil nutrient
content and water level variation directly and indirectly drive mangrove
AGB in a lagoonal ecosystem. Water level variation can be used as an
indicator of the effects of hydroperiod on mangrove ecosystem func-
tioning. Further, mangrove species on Aldabra appeared to have
different niches in relation to soil nutrient content. The spatial distri-
bution of soil nutrients across the atoll suggests a direct contribution of
macrofauna—potentially from breeding seabird populations and the
nearby seagrass meadows, reefs and terrestrial landscapes—to
mangrove ecosystem function. These ndings underscore the relative
importance of nutrient cycles and hydroperiod in mangrove ecosystem
functioning at local scales. Integrating these processes in conservation
management demands a holistic ecosystem-level perspective on
mangrove conservation that can deliver climate change mitigation and
adaptation objectives while also enhancing global biodiversity.
CRediT authorship contribution statement
Annabelle Constance: Conceptualization, Methodology, Formal
analysis, Writing – original draft, Writing – review & editing, Project
administration. Jacqueline Oehri: Methodology, Formal analysis,
Writing – review & editing. Nancy Bunbury: Conceptualization, Project
administration, Supervision, Writing – review & editing. Guido L.B.
Wiesenberg: Methodology, Validation, Supervision, Writing – review &
editing. Frank Pennekamp: Methodology, Formal analysis, Writing –
review & editing. Luke A’Bear: Methodology, Validation. Frauke
Fleischer-Dogley: Supervision, Project administration. Gabriela
Schaepman-Strub: Conceptualization, Methodology, Supervision,
Formal analysis, Writing – review & editing, Project administration.
Declaration of Competing Interest
The authors declare that they have no known competing nancial
interests or personal relationships that could have appeared to inuence
the work reported in this paper.
Data availability
Data will be made available on request.
Acknowledgements
A special thanks go to the Seychelles Islands Foundation (SIF) and
staff for their invaluable support. In particular, we would like to thank
Cheryl Sanchez, Jude Brice and Julio Agricole for the constructive
feedback and support for the eldwork planning and data collection, as
well as the SIF staff who helped immensely to collect eld data: Mikael
Esparon, Germano Soru, Maria Bielsa, Marvin Roseline, Jessica Mou-
mou, Joel Bonne, Jake Letori, Matt Waller, Bruno Mels, Martin van
Rooyen, Brian Souyana and Mersiah Rose. Our sincere thanks go to
Ronny Rose and Christina Quanz for handling the immense logistics of
the eldwork to the remote eld site. Special thanks to Jeannine Sure-
mann, Aline Hobi and Esmaeil Taghizadeh for their great help pro-
cessing and analyzing the soil samples, and Ilja van Meerveld for advice
on the water level loggers. We would like to thank the University
Research Priority Program Global Change and Biodiversity of the Uni-
versity of Zurich (URPP-GCB) and the Western Indian Ocean Marine
Science Association for funding towards equipment used in the eld
research. The URPP-GCB further supported the contributions of Anna-
belle Constance, Frank Pennekamp and Gabriela Schaepman-Strub.
Many thanks to Elena Plekhanova and Raleigh Grysko for helping
with the writing. Finally, we would like to thank the reviewers and the
editor for their careful reading of our manuscript and their insightful
comments that greatly improved the quality of the manuscript.
Appendix A
Table A1
Wood density values used to estimate mangrove AGB for species sur-
veyed on Aldabra. Source: World Agroforestry Center (www.worldag
roforestry.org/sea/Products/AFDbases/WD/Index.htm).
Species Wood density (g cm
¡3
)
Avicennia marina 0.79
Bruguiera gymnorrhiza 0.63
Ceriops tagal 0.87
Lumnitzera racemosa 0.75
Rhizophora mucronata 0.94
Sonneratia alba 0.62
Xylocarpus granatum 0.59
Table A2
The standard deviation, proportion of variance explained by each principal
component, and the cumulative proportion of variance explained for the prin-
cipal component analysis on soil nutrient content conditions across Aldabra’s
mangrove ecosystem.
PC1 PC2 PC3 PC4 PC5 PC6
Standard deviation 1.78 1.11 0.92 0.52 0.43 0.32
Proportion of Variance 0.55 0.21 0.15 0.05 0.03 0.02
Cumulative Proportion 0.55 0.76 0.91 0.95 0.98 1
A. Constance et al.

Ecological Indicators 143 (2022) 109292
9
Table A3
Regression coefcients for the structural equation model for mangrove above ground biomass (AGB) in the lagoonal ecosystem of Aldabra presented in Fig. 4.
Standardized coefcients (scaled by standard deviations) of signicant values (P < 0.05) are bolded. The path coefcients and their corresponding standard error,
standard estimate and test statistic (critical value) are provided.
Goodness of t
AIC Fisher’s C Degrees of freedom P value
24.868 0.868 2 0.648
Coefcients
Response Predictor Estimate Standard Error Degrees of freedom Critical value P value Std.estimate
Mangrove AGB Soil salinity 0.0489 0.6475 17 0.0755 0.9407 0.0069
Mangrove AGB Soil nutrient content 0.8736 0.1299 17 6.7256 0 0.7976
Mangrove AGB Water level variation 16.9857 14.7679 9 1.1502 0.2797 0.1551
Soil nutrient content Water level variation 73.176 15.6446 9 4.6673 0.0012 0.7301
Table A4
Structural equation model for initial hypothesized relationships between variables that inuence mangrove aboveground biomass on Aldabra Atoll.
Goodness of t
AIC Fisher’s C Degrees of freedom P value
34.868 0.868 2 0.648
Coefcients
Response Predictor Estimate Standard Error Degrees of freedom Critical value P value Standard estimate
Mangrove AGB Soil salinity 0.0489 0.6475 17 0.0755 0.9407 0.0069
Mangrove AGB Soil nutrient content 0.8736 0.1299 17 6.7256 0 0.7976
Mangrove AGB Water level variation 16.9857 14.7679 9 1.1502 0.2797 0.1551
Soil nutrient content Water level variation 730.176 15.6446 9 4.6673 0.0012 0.7301
Mean soil salinity Water level variation −7.2102 4.5621 9 −1.5804 0.1485 −0.4659
Table A5
Structural equation model without variables that had no signicant effect on mangrove aboveground biomass on Aldabra Atoll.
Goodness of t
AIC Fisher’s C Degrees of freedom P value
22.000 0 0 1
Coefcients
Response Predictor Estimate Standard Error Degrees of freedom Critical value P value Standard estimate
Mangrove AGB Soil nutrient content 0.8756 0.1265 18 6.9192 0 0.7994
Mangrove AGB Water level variation 16.4905 13.2920 9 1.2406 0.2461 0.1505
Soil nutrient content Water level variation 730.176 15.6446 9 4.6673 0.0012 0.7301
Fig. A1. Conceptual diagram of direct and indirect effects of soil nutrient content, salinity and water level variation on mangrove aboveground biomass in a lagoonal
environment. Arrows represent unidirectional relationships among variables. An indirect effect of a driver variable on mangrove aboveground biomass is read when
arrows connecting the driver variable to mangrove aboveground biomass is connected through a second variable (mangrove tree image from ian.umces.
edu/symbols/).
A. Constance et al.

Ecological Indicators 143 (2022) 109292
10
Fig. A2. Soil nutrient content variables plotted with the three most descriptive principal components. Spheres represent plot observations and are colored according
to the dominant mangrove species recorded at the respective plot. The red arrows depict the strength (arrow length) and direction of soil nutrient content gradients
[nitrogen (N), phosphorus (P), potassium (K), sulfur (S), calcium (Ca), and magnesium (Mg)] across plots. The nutrients S, K, Mg and N contributed >90 % to the
major nutrient content gradient (rst principal component PC1). Ca, P and N contributed nearly 90 % to the second major nutrient content gradient (second principal
component PC2). CA and P contributed 90 % to the third principal component.
Fig. A3. Example of the daily trends of water level for three surveyed blocks (JM, MH and QL) on Aldabra. See Fig. 2 for location of loggers of JM, MH and QL.
A. Constance et al.

Ecological Indicators 143 (2022) 109292
11
References
Alavaisha, E., Mangora, M., 2016. Carbon Stocks in the Small Estuarine Mangroves of
Geza and Mtimbwani, Tanga, Tanzania. Int. J. Forest. Res. 2016.
Aragones, L.V., Lawler, I.R., Foley, W.J., Marsh, H., 2006. Dugong grazing and turtle
cropping: grazing optimization in tropical seagrass systems? Oecologia 149 (4),
635–647.
Ball, M.C., Pidsley, S.M., 1995. Growth Responses to Salinity in Relation to Distribution
of Two Mangrove Species, Sonneratia alba and S. lanceolata, in Northern Australia.
Funct. Ecol. 9, 77–85.
Benson, L., Glass, L., Jones, T., Ravaoarinorotsihoarana, L., Rakotomahazo, C., 2017.
Mangrove Carbon Stocks and Ecosystem Cover Dynamics in Southwest Madagascar
and the Implications for Local Management. Forests 8, 190.
Carr, M.H., Robinson, S.P., Wahle, C., Davis, G., Kroll, S., Murray, S., Schumacker, E.J.,
Williams, M., 2017. The central importance of ecological spatial connectivity to
effective coastal marine protected areas and to meeting the challenges of climate
change in the marine environment. Aquat. Conserv. Mar. Freshwater Ecosyst. 27,
6–29.
Casta˜
neda-Moya, E., Rivera-Monroy, V.H., Twilley, R.R., 2006. Mangrove zonation in the
dry life zone of the Gulf of Fonseca, Honduras. Estuaries Coasts 29, 751–764.
Casta˜
neda-Moya, E., Twilley, R.R., Rivera-Monroy, V.H., Marx, B., Coronado, C., Ewe, S.,
2011. Patterns of Root Dynamics in Mangrove Forests Along Environmental
Gradients in the Florida Coastal Everglades, USA. Ecosystems 14, 1178–1195.
Casta˜
neda-Moya, E., Rivera-Monroy, V.H., Chambers, R.M., Zhao, X., Lamb-Wotton, L.,
Gorsky, A., Gaiser, E.E., Troxler, T.G., Kominoski, J.S., Hiatt, M., 2020. Hurricanes
fertilize mangrove forests in the Gulf of Mexico (Florida Everglades, USA). Proceed.
Natl. Acad. Sci. U.S.A. 117, 4831–4841.
Chave, J., Andalo, C., Brown, S., Cairns, M.A., Chambers, J.Q., Eamus, D., Folster, H.,
Fromard, F., Higuchi, N., Kira, T., Lescure, J.P., Nelson, B.W., Ogawa, H., Puig, H.,
Riera, B., Yamakura, T., 2005. Tree allometry and improved estimation of carbon
stocks and balance in tropical forests. Oecologia 145, 87–99.
Chen, Y., Ye, Y., 2014. Effects of salinity and nutrient addition on mangrove Excoecaria
agallocha. PLoS ONE 9, e93337.
Constance, A., Haverkamp, P.J., Bunbury, N., Schaepman-Strub, G., 2021. Extent change
of protected mangrove forest and its relation to wave power exposure on Aldabra
Atoll. Global Ecol. Conserv. 27.
Crase, B., Liedloff, A., Vesk, P.A., Burgman, M.A., Wintle, B.A., 2013. Hydroperiod is the
main driver of the spatial pattern of dominance in mangrove communities. Glob.
Ecol. Biogeogr. 22, 806–817.
De la Pe˜
na-Lastra, S., 2021. Seabird droppings: Effects on a global and local level. Sci.
Total Environ. 754, 142148.
Donato, D.C., Kauffman, J.B., Murdiyarso, D., Kurnianto, S., Stidham, M., Kanninen, M.,
2011. Mangroves among the most carbon-rich forests in the tropics. Nat. Geosci. 4,
293–297.
Duarte, C., Geertz-Hansen, O., Thampanya, U., Terrados, J., Fortes, M., Kamp-Nielsen, L.,
Borum, J., Boromthanarath, S., 1998. Relationship between sediment conditions and
mangrove Rhizophora apiculata seedling growth and nutrient status. Mar. Ecol.
Prog. Ser. 175, 277–283.
Duarte, C.M., Losada, I.J., Hendriks, I.E., Mazarrasa, I., Marb`
a, N., 2013. The role of
coastal plant communities for climate change mitigation and adaptation. Nat. Clim.
Change 3, 961–968.
Ewel, K.C., Twilley, R.R., Ong, J.E., 1998. Different Kinds of Mangrove Forests Provide
Different Goods and Services. Global Ecol. Biogeograp. Lett. 7, 83–94.
Falc´
on, W., Hansen, D.M., 2018. Island rewilding with giant tortoises in an era of climate
change. Philosoph. Trans. Roy. Soc. B: Biol. Sci. 373, 20170442.
FOOD AND AGRICULTURE ORGANIZATION OF THE UNITED NATIONS 2006.
Guidelines for soil description. 4th ed. Rome, Italy.
Fayolle, A., Doucet, J.-L., Gillet, J.-F., Bourland, N., Lejeune, P., 2013. Tree allometry in
Central Africa: Testing the validity of pantropical multi-species allometric equations
for estimating biomass and carbon stocks. For. Ecol. Manage. 305, 29–37.
Gamito, S., Gilabert, J., Diego, C.N.M., P´
erez-Ruzafa, A., 2004. Effects of Changing
Environmental Conditions on Lagoon Ecology. In: G¨
onenç, E., Wolin, J.P. (Eds.),
Coastal Lagoons: Ecosystem Processes and Modeling for Sustainable Use and
Development. CRC Press, Boca Raton. Florida.
Goldberg, L., Lagomasino, D., Thomas, N., Fatoyinbo, T., 2020. Global declines in
human-driven mangrove loss. Glob. Change Biol. 26, 5844–5855.
Grace, J.B., Anderson, T.M., Olff, H., Scheiner, S.M., 2010. On the specication of
structural equation models for ecological systems. Ecol. Monogr. 80, 67–87.
Grace, J.B, Keeley, J.E, 2006. A structural equation model analysis of postre plant
diversity in California shrublands. Ecol. Appl. 16 (2), 503–514.
Grace, J.B., Schoolmaster, D.R., Guntenspergen, G.R., Little, A.M., Mitchell, B.R.,
Miller, K.M., Schweiger, E.W., 2012. Guidelines for a graph-theoretic
implementation of structural equation modeling. Ecosphere 3.
Grubb, P., 1971. The Growth, Ecology and Population Structure of Giant Tortoises on
Aldabra. Philosoph. Trans. Roy. Soc. London Ser. B, Biol. Sci. 260, 327–372.
Hamylton, S., Hagan, A., Bunbury, N., Fleischer-Dogley, F., Spencer, T., 2018. Mapping
the lagoon at aldabra atoll, western Indian Ocean. Atoll Res. Bull. 2018, 45–59.
Haverkamp, P.J., Shekeine, J., de Jong, R., Schaepman, M., Turnbull, L.A., Baxter, R.,
Hansen, D., Bunbury, N., Fleischer-Dogley, F., Schaepman-Strub, G., 2017. Giant
tortoise habitats under increasing drought conditions on Aldabra Atoll—Ecological
indicators to monitor rainfall anomalies and related vegetation activity. Ecol. Ind.
80, 354–362.
Hutchison, J., Manica, A., Swetnam, R., Balmford, A., Spalding, M., 2014. Predicting
Global Patterns in Mangrove Forest Biomass. Conservat. Lett. 7, 233–240.
Kauffman, J.B., Donato, D.C., 2012. Protocols for the measurement, monitoring and
reporting of structure, biomass and carbon stocks in mangrove forests. Center for
International Forestry Research.
Kjerfve, B. 1994. Chapter 1 Coastal Lagoons. In: Kjerfve, B. (ed.) Elsevier Oceanography
Series. Elsevier.
Komiyama, A., Ong, J.E., Poungparn, S., 2008. Allometry, biomass, and productivity of
mangrove forests: A review. Aquat. Bot. 89, 128–137.
Krauss, K.W., Ball, M.C., 2012. On the halophytic nature of mangroves. Trees 27, 7–11.
Krauss, K.W., Doyle, T.W., Twilley, R.R., Rivera-Monroy, V.H., Sullivan, J.K., 2006.
Evaluating the relative contributions of hydroperiod and soil fertility on growth of
south Florida mangroves. Hydrobiologia 569, 311–324.
Krauss, K.W., McKee, K.L., Lovelock, C.E., Cahoon, D.R., Saintilan, N., Reef, R., Chen, L.,
2014. How mangrove forests adjust to rising sea level. New Phytol. 202, 19–34.
Lefcheck, J.S., 2015. PIECEWISESEM: Piecewise structural equation modelling in R for
ecology, evolution, and systematics. Methods Ecol. Evol. 7, 573–579.
Luther, D.A., Greenberg, R., 2009. Mangroves: A Global Perspective on the Evolution and
Conservation of Their Terrestrial Vertebrates. Bioscience 59 (7), 602–612.
Macnae, W., Westoll, T.S., Stoddart, D.R., 1971. Mangroves on Aldabra. Philosoph.
Trans. Roy. Soc. London B, Biol. Sci. 260, 237–247.
Mortimer, J., von Brandis, R., Liljevik, A., Chapman, R., Collie, A., 2011. Fall and Rise of
Nesting Green Turtles (Chelonia mydas) at Aldabra Atoll, Seychelles: Positive
Response to Four Decades of Protection (1968–2008). Chelonian Conserv. Biol. 10,
165–176.
Naidoo, G., 2009. Differential effects of nitrogen and phosphorus enrichment on growth
of dwarf Avicennia marina mangroves. Aquat. Bot. 90, 184–190.
Njana, M.A., Bollandsås, O.M., Eid, T., Zahabu, E., Malimbwi, R.E., 2015. Above- and
belowground tree biomass models for three mangrove species in Tanzania: a
nonlinear mixed effects modelling approach. Ann. Forest Sci. 73, 353–369.
Njana, M.A., Zahabu, E., Malimbwi, R.E., 2017. Carbon stocks and productivity of
mangrove forests in Tanzania. Southern Forests: J. Forest Sci. 80, 217–232.
Pinheiro, J., Bates, D., Debroy, S., Sarkar, D. & Team, R. C. 2020. nlme: Linear and
Nonlinear Mixed Effects Models. R package version 3.1-149.
Raw, J.L., Godbold, J.A., Van Niekerk, L., Adams, J.B., 2019. Drivers of mangrove
distribution at the high-energy, wave-dominated, southern African range limit.
Estuar. Coast. Shelf Sci. 226.
Reef, R., Feller, I.C., Lovelock, C.E., 2010. Nutrition of mangroves. Tree Physiol. 30,
1148–1160.
Schaeffer-Novelli, Y., Cintr´
on-Molero, G., Adaime, R.R., de Camargo, T.M., 1990.
Variability of Mangrove Ecosystems along the Brazilian Coast. Estuaries 13,
204–218.
Schulze, E.-D., Beck, E., Buchmann, N., Clemens, S., Müller-Hohenstein, K., Scherer-
Lorenzen, M. (Eds.), 2019. Plant Ecology. Springer Berlin Heidelberg, Berlin,
Heidelberg.
Serafy, J.E., Shideler, G.S., Araujo, R.J., Nagelkerken, I., 2015. Mangroves Enhance Reef
Fish Abundance at the Caribbean Regional Scale. PLoS ONE 10, e0142022.
Sievers, M., Brown, C.J., Tulloch, V.J.D., Pearson, R.M., Haig, J.A., Turschwell, M.P.,
Connolly, R.M., 2019. The Role of Vegetated Coastal Wetlands for Marine
Megafauna Conservation. Trends Ecol. Evol. 34, 807–817.
Simard, M., Fatoyinbo, L., Smetanka, C., Rivera-Monroy, V.H., Casta˜
neda-Moya, E.,
Thomas, N., van der Stocken, T., 2018. Mangrove canopy height globally related to
precipitation, temperature and cyclone frequency. Nat. Geosci. 12, 40–45.
Stoddart, D.R., Taylor, J.D., Fosberg, F.R., Farrow, G.E., 1971. Geomorphology of
Aldabra Atoll. Philosoph. Trans. Roy. Soc. London B, Biol. Sci. 260, 31–66.
ˇ
Súr, M., Bunbury, N., van de Crommenacker, J., 2013. Frigatebirds on Aldabra Atoll:
population census, recommended monitoring protocol and sustainable tourism
guidelines. Bird Conser. Int. 23, 214–220.
Taillardat, P., Friess, D.A., Lupascu, M., 2018. Mangrove blue carbon strategies for
climate change mitigation are most effective at the national scale. Biol. Lett. 14.
Twilley, R.R., Lugo, A.E., Patterson-Zucca, C., 1986. Litter Production and Turnover in
Basin Mangrove Forests in Southwest Florida. Ecology 67 (3), 670–683. https://doi.
org/10.2307/1937691.
Twilley, R.R., Rovai, A.S., Riul, P., 2018. Coastal morphology explains global blue carbon
distributions. Front. Ecol. Environ. 16, 503–508.
Ukpong, I.E., 1997. Vegetation and its relation to soil nutrient and salinity in the Calabar
mangrove swamp, Nigeria. Mangroves and Salt Marshes 1, 211–218.
Twilley, R.R. & Day Jr., J.W. 2012. Mangrove Wetlands. In: J.W. Day Jr., J., W.M, Kemp,
A, Y´
a˜
nez-Arancibia, B.C. CRUMP (ed.) Estuarine Ecology.
Vadineanu, A. 2004. Identication of the lagoon ecosystems. In: G¨
onenç, I. E. & Wolin,
J. P. (eds.) Coastal Lagoons: Ecosystem Processes and Modeling for Sustainable Use and
Development. Boca Raton, Florida: CRC Press.
Walton, R., Baxter, R., Bunbury, N., Hansen, D., Fleischer-Dogley, F., Greenwood, S.,
Schaepman-Strub, G., 2019. In the land of giants: habitat use and selection of the
Aldabra giant tortoise on Aldabra Atoll. Biodivers. Conserv. 28, 3183–3198.
Woodroffe, C.D., Rogers, K., McKee, K.L., Lovelock, C.E., Mendelssohn, I.A., Saintilan, N.,
2016. Mangrove Sedimentation and Response to Relative Sea-Level Rise. Ann. Rev.
Mar. Sci. 8, 243–266.
Worthington, T.A., Ermgassen, Z.U., Friess, D.A., Krauss, K.W., Lovelock, C.E.,
Thorley, J., Tingey, R., Woodroffe, C.D., Bunting, P., Cormier, N., Lagomasino, D.,
Lucas, R., Murray, N.J., Sutherland, W.J., Spalding, M., 2020. A global biophysical
typology of mangroves and its relevance for ecosystem structure and deforestation.
Sci. Rep. 10, 14652.
Wu, L., Liu, X., Fang, Y., Hou, S., Xu, L., Wang, X., Fu, P., 2018. Nitrogen cycling in the
soil-plant system along a series of coral islands affected by seabirds in the South
China Sea. Sci. Total Environ. 627.
A. Constance et al.

Ecological Indicators 143 (2022) 109292
12
Xiong, Y., Liao, B., Proftt, E., Guan, W., Sun, Y., Wang, F., Liu, X., 2018. Soil carbon
storage in mangroves is primarily controlled by soil properties: A study at Dongzhai
Bay, China. Sci. Total Environ. 619–620, 1226–1235.
Yates, E.J., Ashwath, N., Midmore, D.J., 2001. Responses to nitrogen, phosphorus,
potassium and sodium chloride by three mangrove species in pot culture. Trees 16,
120–125.
Zhao, Q., Stephenson, F., Lundquist, C., Kaschner, K., Jayathilake, D., Costello, M.J.,
2020. Where Marine Protected Areas would best represent 30% of ocean
biodiversity. Biol. Conserv. 244.
A. Constance et al.