Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=trsl20

Download by: [Uni San Francisco de Quito ] Date: 05 May 2016, At: 15:27

Remote Sensing Letters

ISSN: 2150-704X (Print) 2150-7058 (Online) Journal homepage: http://www.tandfonline.com/loi/trsl20

Hyperspectral remote sensing of salinity stress onred (Rhizophora mangle) and white (Lagunculariaracemosa) mangroves on Galapagos Islands

Conghe Song , Brian L. White & Benjamin W. Heumann

To cite this article: Conghe Song , Brian L. White & Benjamin W. Heumann (2011)Hyperspectral remote sensing of salinity stress on red (Rhizophora mangle) and white(Laguncularia racemosa) mangroves on Galapagos Islands, Remote Sensing Letters, 2:3,221-230, DOI: 10.1080/01431161.2010.514305

To link to this article: http://dx.doi.org/10.1080/01431161.2010.514305

Published online: 27 Oct 2010.

Submit your article to this journal

Article views: 264

View related articles

Citing articles: 5 View citing articles

Hyperspectral remote sensing of salinity stress on red(Rhizophora mangle) and white (Laguncularia racemosa)

mangroves on Galapagos Islands

CONGHE SONG*†, BRIAN L. WHITE‡ and BENJAMIN W. HEUMANN†

†Department of Geography, University of North Carolina, Chapel Hill, NC 27599, USA

‡Department of Marine Sciences, University of North Carolina, Chapel Hill,

NC 28599, USA

(Received 29 April 2010; in final form 1 August 2010)

Mangroves are an assemblage of salt-tolerant woody hydrophytes that are the

foundational species along tropical and subtropical coastlines, estuaries, lagoons

and rivers. Once covering 75% of the world’s tropical and subtropical coastlines,

mangrove forests have been declining rapidly throughout the world in the last few

decades. Monitoring and modelling mangrove forest growth are critical for their

conservation. Salinity is one of the primary limiting factors for mangrove forest

growth. In this letter, we report the potential of using hyperspectral remotely

sensed data collected in the field for monitoring the photosynthesis rate of red

(Rhizophora mangle) and white (Laguncularia racemosa) mangroves with regard to

salinity gradient. Using photochemical reflectance index (PRI) as proxy for photo-

synthetic rate, PRI for both species are strongly related to salinity gradient,

indicating the potential of monitoring mangrove forest growth on a regional

scale using hyperspectral remote sensing.

1. Introduction

Mangroves are an assemblage of salt-tolerant woody hydrophytes that are the founda-

tional species along tropical and subtropical coastlines, estuaries, lagoons and rivers

(Tomlinson 1986, Hogarth 2007). Mangrove forests have historically covered as much

as 75% of the world’s tropical and subtropical coastlines (Alongi 2002) but have suffered

dramatic declines in areas because of coastal development, non-renewable resourceexploitation (e.g. clear cutting, mining, aquaculture), pollution, high rates of sedimenta-

tion, sea level rise and alterations of hydrology (Saenger et al. 1983, Farnsworth and

Ellison 1997, Ellison and Farnsworth 2001, Duke et al. 2007). Alongi (2002) estimated

that as much as a third of the world’s mangrove forests have been lost in the past 50 years.

Mangrove forests sustain productive, biologically unique and economically important

ecosystems (Lugo and Snedaker 1974, Tomlinson 1986, Kathiresan and Bingham 2001,

Hogarth 2007). Mangroves provide valuable ecosystems goods and services including

wood, fibre, food, high primary productivity, linking terrestrial and marine systems,buffering coastline from erosion and supporting biodiversity (Alongi 2008). Mangrove

forest composition is often characterized by a strong zonation in community composition

based primarily on soil salinity related to tidal inundation (Tomlinson 1986), although

*Corresponding author. Email: song@email.unc.edu

Remote Sensing LettersISSN 2150-704X print/ISSN 2150-7058 online # 2011 Taylor & Francis

http://www.tandf.co.uk/journalsDOI: 10.1080/01431161.2010.514305

Remote Sensing Letters

Vol. 2, No. 3, September 2011, 221–230

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

other geomorphic, edaphic, climatic and biotic factors can create more complex patterns

(Onuf et al. 1977, Ewel et al. 1998, Lee 1998, Ellison 2002).

Protecting and managing mangrove forests are important international conserva-

tion objectives (Macintosh and Ashton 2003). Achieving these objectives will require

the development of techniques for accurately mapping and monitoring the distribu-tion and health of mangrove forests in an era of accelerated climate change.

Mangrove forest monitoring is no easy task. Such forests are notoriously difficult to

access and survey on foot because of the flooded, soft sediment environments in which

they grow and the typically high densities of stems and prop roots. Remote sensing

provides an operational alternative to map and monitor mangrove forest on a timely

basis for a wide range of spatial coverage. Numerous studies have utilized remote

sensing to delineate, detect change, map species distribution and estimate structural

metrics such as leaf area index and above-ground biomass as reviewed by Walterset al. (2008). The remotely sensed data used include aerial photos, satellite images

(e.g. Landsat and Satellite Pour L’Observation de La Terre (SPOT)) as well as

airborne hyperspectral data (Green et al. 1998, Manson et al. 2001, Held et al. 2003,

Vaiphasa et al. 2007, Conchedda et al. 2008, Wang et al. 2008). Few studies explored

the potential of using hyperspectral remotely sensed data for monitoring or modelling

the growth in relation to the main environmental stressor, salinity.

Salinity is a major factor that influences mangrove forest growth. Although man-

groves have several mechanisms to tolerate saline conditions such as ion compart-mentation, osmoregulation, selective transport and uptake of ions, maintenance of a

balance between the supply of ions to the shoot and capacity to accommodate the salt

influx (Parida and Jha 2010), the costs of these mechanisms appear to come at the

expense of photosynthesis (Sobrado 2005). The objective of this study is to test the

suitability of using hyperspectral remote sensing in monitoring/modelling mangrove

forest growth on the Galapagos Islands. Specifically, we test the potential for detect-

ing the salinity-induced stress on photosynthesis for two common mangrove species,

the red and white mangroves, on Isla Isabela, Galapagos.

2. Methodologies

2.1. Theory

Photosynthesis is a biological process in which photosynthetically active radiation

(PAR) captured by leaf pigments is used to produce hydrocarbon in chlorophyll. Part

of the energy absorbed by chlorophyll is always lost as heat or fluorescence. The

xanthophyll cycle plays a key role in regulating the energy flow (Bilger and Bjorkman

1990). Under conditions of excess light, the absorbed PAR exceeds photosynthetic

capacity. The excessive PAR causes the deepoxidation of violaxanthin to zeaxanthin

through antheraxanthin. The process is reversed under limiting light condition

(Demmig-Adams 1990). Therefore, the xanthophyll cycle is directly related to photo-synthetic radiation-use efficiency. Gamon et al. (1990) identified that the reflectance

feature at 531 nm is associated with the conversion of violaxanthin to zeaxanthin. A

spectral index, the physiological reflectance index (PRI), was defined to quantify

canopy photosynthetic function as in the following:

PRI ¼ R531 � R570

R531 þ R570(1)

222 C. Song et al.

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

where R531 and R570 are reflectance at 531 and 570 nm, respectively. PRI was later

renamed as photochemical reflectance index by Gamon et al. (1992). Several subse-

quent studies found that PRI is highly correlated with photosynthetic radiation-use

efficiency (Penuelas et al. 1995, Gamon and Surfus 1999, Sims and Gamon 2002).

Nichol et al. (2006) examined the radiation-use efficiency with experimental man-grove forests and found that PRI is an effective indicator of photosynthetic activity.

In this study, we further test the feasibility of using PRI for monitoring mangrove

forest growth in field conditions. Because salinity is a major stress factor for man-

grove growth, we hypothesize that there be a statistically significant relationship

between PRI and salinity gradient. If this relationship exists, it would serve as

theoretical basis for development of new approaches to monitor/model mangrove

forest growth using remotely sensed data.

2.2. Study sites

The Galapagos Islands, located 1000 km off the coast of Ecuador, are a volcanicallyactive archipelago consisting of 13 large islands, 4 of which have human populations,

and 188 small islands and rocks. The Galapagos Islands were declared a national park

in 1959 (the park consists of 97% of land area), a UNESCO World Heritage Site in

1978 and a UNESCO Biosphere Reserve in 1987. The Galapagos Islands lie on the

western edge of the Atlantic–East Pacific mangrove complex. Mangrove forests

consist of three species common in this region: red (Rhizophora mangle), black

(Avicenna germinans) and white (Laguncularia racemosa) mangroves, as well as sev-

eral mangrove associates. In the Galapagos Islands, mangrove forests form dense butsmall patches in protected coves and lagoons along an otherwise barren or arid coast.

Our study sites are located near Puerto Villamil on Isla Isabela, which is the largest

island of the Galapagos archipelago (figure 1). We selected two sites to study the

salinity stress on mangrove growth with hyperspectral remote sensing along a pro-

tected bay. Site 1 is dominated by red mangroves along an estuary with deep sediment

deposits and fed by a landward freshwater spring. The freshwater input creates a

strong salinity gradient, particularly at low tide. We identified 10 spots (points to the

left in the top panel in figure 1) along the salinity gradient and sampled leaf reflectanceand salinity simultaneously. Site 2 is located to the east of Site 1 and is dominated by

white mangroves. This site lacks freshwater input and lies on top of a lava bedrock

ledge with shallow sediment deposits and tidal pools.

2.3. Hyperspectral reflectance measurements

We used the FieldSpec HandHeld device manufactured by ASD Inc. (at Boulder, CO,

USA) to take the hyperspectral reflectance measurements. The spectral range spans

325–1075 nm. The spectral resolution is 3.5 nm with a sampling interval of 1.6 nm. The

processing software resamples the reflectance with 1.0 nm resolution. The instrument

was calibrated by the manufacturer just before taking to the field. To control for noisein the field, we equipped the FieldSpec HandHeld with a plant probe and a leaf clip,

accessories provided by the manufacturer. The plant probe comes with a halogen bulb

(2901 K) to provide a constant source of light. The plant probe and the HandHeld

device is connected with a fibre optic wire. The leaf clip is specially made to simplify the

spectral measurement of leaves with the plant probe. Together they can gently enclose a

circular leaf area with 1 cm in diameter. Because of the relatively low temperature from

the halogen bulb, we found that reflectance signals below 400 nm are noisy.

Hyperspectral remote sensing of red and white mangroves 223

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

Because of the fact that there are significant variations in the spectral signature for

different leaves on the same tree depending on its age and physical condition, we

selected the mature healthy leaves for the hyperspectral measurements. Leaves were

carefully selected such that all were fully grown with no mechanical damage. We

found that the second leaf from the top of a shoot best met these requirements. The

location of the spectral measurements can also influence the reflectance signal,particularly whether the main leaf vein in the middle is enclosed. Although measure-

ments avoiding the main leaf vein are ideal, it is impossible to avoid the main leaf vein

in the spectral measurement with the leaf clip because of the relatively small leaf size of

white mangrove. To facilitate comparison between red and white mangroves, we

measured the spectral signal at the tip of the leaves across the main vein for both

species. To reduce random error, each spectral measurement is an average of five

samples.

Figure 1. The study sites for hyperspectral measurements for red (left) and white (right)mangroves. The red mangroves are located along an estuary with upland freshwater input,whereas the white mangrove site does not have freshwater input. The study site is located 3 kmwest of the town of Puerta Villamil (A) on Isabela Island (C), located to the east of FenerandinaIsland (B) and to the west of Santa Cruz (D) and San Cristobal (E) Islands.

224 C. Song et al.

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

Measurements were carried out on land because of the difficulty and risk of work-

ing with valuable electronic equipment in the intertidal swamp. Gamon et al. (1990)

found that leaf reflectance at 531 nm can change within minutes after switching from

dark to full light in field conditions without detachment. Because of the difficulty in

access, detachment of leaves from the tree is a common practice in hyperspectralmeasurements of mangrove leaves (Sims and Gamon 2002, Vaiphasa et al. 2007,

Wang and Sousa 2009). In our experiment, the light intensity was constant for all

leaves as a result of using the plant probe, and the use of a leaf clip maximally reduces

noise from other sources. We assume that similar change in PRI among all leaves will

occur within minutes after detachment, thus not significantly altering the relation-

ships between PRI and salinity.

2.4. Salinity measurements

Salinity was measured in situ at each of the mangrove sites using a Hydrolab Quanta

CTD (Hach Environmental, Loveland, CO, USA) with specific conductivity, tem-

perature and depth sensors. Salinity provided by the instrument is in specific con-

ductivity at 25�C in milliSiemens per centimetre (mS cm-1). Seawater salinity is,52 mS cm-1. Transects were made at each site during different days and at different

times during the tidal cycle, to investigate both spatial gradients and tidal cycle

variability. At Site 1, the spring-fed estuary, near-bottom and near-surface water

salinity was measured during both high and low tides, along the seaward increasing

salinity gradient. Salinity throughout the estuary was the highest during low tide, and

these values were used for analysis with PRI.

Salinity was also measured at Site 2, the white mangrove site along the lava rocky

shore. At low tide, the white mangroves are above the sea level among tide pools in thelava bedrock. We measured the salinity of the silt close to the root for the individuals

from which we took leaf samples. To measure the salinity of silt, we mixed 1 part of silt

with 5 parts bottled freshwater in volume (Henschke and Herrmann 2007). We also

measured the tide pool water close to the root of the sampled individuals.

3. Results and discussions

The relationship between PRI and salinity gradient is shown in figure 2 for red

mangroves at low tide. The relationship is statistically significant (r ¼ 0.69;p ¼ 0.03). Salinity is negatively correlated with PRI, indicating that salinity is a stress

factor to red mangrove growth. Given that multiple other environmental factors, such

as nutrients and wave impacts, may influence mangrove growth, it is clear that salinity

is a major environmental factor controlling red mangrove growth.

The white mangroves found on the lava rocky shore are found in an intertidal region

with high variability in tidal inundation, wave forcing and salinity stress. This site is

more energetic and temporally variable than the red mangrove site (figure 1). Thus, the

relationships between PRI and physical factors are expected to be more complex. Therelationship between PRI and silt salinity is negatively correlated (figure 3(a)), but

the relationship is weaker (r ¼ 0.53; p ¼ 0.14) than that for the red mangrove. Several

factors may contribute to the weaker relationship between PRI and salinity for white

mangroves. First, there is a smaller salinity dynamic range in the samples, which may

mask the effect of salinity stress. Second, the silts sampled are not uniform as they

contain varying proportions of very fine roots and sand. These non-silt components

contribute to the salinity noise weakening its relationship with PRI. Third, the sampling

Hyperspectral remote sensing of red and white mangroves 225

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

itself in the field may contribute to salinity noise as the volume of the silt is measured

with a measurement cup and visual evaluation. Finally, as mentioned, this site is more

energetic and likely to be influenced by other physical factors like wave stress.

Nonetheless, given these uncertainty factors, the relationship between salinity and

PRI is reasonably strong for white mangroves.The relationship between pool water salinity and PRI is reversed (figure 3(b)). The

pool water salinity is strongly influenced by evaporation and therefore by the time

elapsed since the last seawater inundation. Thus, salinity can be highly variable and is a

strong function of phase in the spring–neap tidal cycle as well as wave conditions. The

pool size and the time of evaporation determine the pool water salinity along the tidal

range. In general, the higher the pool is in the intertidal range and the further from the

seafront, the longer the evaporation time since the last high tide. Thus, the salinity

gradient of pool water on shore is an indirect measurement of height in the intertidalregion and distance to the seafront. Thus, we hypothesize that figure 3(b) primarily

reflects the impact of waves on the growth of white mangrove, rather than salinity.

Comparing figures 2 and 3(a) suggests that red mangroves are more sensitive to

salinity than white mangroves according to the magnitude of the slope of the regres-

sion line. This finding is consistent with salinity experiment as reported in the

literature (Chen and Twilley 1998). Mangrove forests have several mechanisms to

cope with salinity, leading to local accumulation of salt in the root zone. The fact that

red mangrove is generally distributed at the seafront is testimony to its lower toleranceof salinity as the frequent wash of tidal water helps dilute the local salt accumulation.

The freshwater input from upland further reduces the salinity stress, thus red man-

groves generally grow well in estuaries. The capability of identifying the effect of

salinity gradient on mangrove forest growth with hyperspectral remotely sensed

signals makes it possible to monitor mangrove forest growth on a regional scale.

Salinity is one of the key environmental factors that distinguish upland forests from

mangrove forests. The multiple taxa of mangroves indicate multiple independent

evolutionary adaptations to saline conditions from terrestrial forests (Duke et al.2002). Although mangrove forests are among the most productive ecosystems in the

world (Clough et al. 1997), salinity is an environmental stressor, not a facilitator (Lin

and Sternberg 1993, Ball 2002, Saenger 2002). The gradient of salinity along intertidal

coastal areas contributes to the zonation of the mangrove forests. Currently, there are

very limited approaches available in modelling and monitoring mangrove forest

y = −0.0092x + 0.1135r = 0.6925 p = 0.0288

−0.06

−0.03

0.00

0.03

0.06

0.09

6.00 8.00 10.00 12.00 14.00 16.00

Low tide salinity (mS cm−1)

PR

I

Figure 2. Salinity stress on PRI for red mangroves.

226 C. Song et al.

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

growth (Chen and Twilley 1998, Berger and Hildenbrandt 2000, Berger et al. 2008).

Our finding points to a new direction in modelling and monitoring mangrove forest

growth using hyperspectral imagery. In general, hyperspectral remotely sensed data

are very limited, but the situation will significantly improve given the fact that the

United States is planning to launch the Hyperspectral Infrared Imager in the coming

years. Realistic modelling of mangrove forest growth in the Galapagos Islands willsignificantly aid the conservation effort to protect the endemic species that reside in

the mangrove forests.

4. Conclusions

This study found that red mangrove forests show a statistically significant salinity

stress effect on PRI, an indicator of radiation-use efficiency in plant photosynthesis.White mangrove forests have a reasonably strong relationship between salinity and

PRI, but not as strong as the red mangroves. Results from this study point to a new

direction to monitor/model mangrove forest growth on a regional scale using hyper-

spectral remote sensing. Future research is needed (1) to examine the effects of

multiple variables, including salinity, wave-action and soil nutrients, on photosynth-

esis and (2) to scale spectral relationships for satellite-based remote sensing

applications.

(a)

y = −0.0082x + 0.0832 r = 0.5274 p = 0.1362

−0.04

−0.02

0.00

0.02

0.04

0.06

6.00 7.00 8.00 9.00 10.00

Soil salinity (mS cm−1)

PR

I

(b)

Pool water salinity (mS cm−1)

y = 0.0042x − 0.2211r = 0.7098 p = 0.0485

−0.01

0.00

0.01

0.02

0.03

0.04

52.00 54.00 56.00 58.00 60.00 62.00

PR

I

Figure 3. Salinity stress on PRI for white mangroves: (a) silt salinity; (b) tide pool watersalinity.

Hyperspectral remote sensing of red and white mangroves 227

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

Acknowledgement

The authors thank Dr. Sarah Lee, Kristin Blank, Evan Raczkowski, Stacey Frisk and

Royce Brown for their field assistance, and the Galapagos National Park for provid-

ing transportation and a field guide. Funding for this pilot study was provided by the

Center of Galapagos Studies at UNC, Chapel Hill, NC, USA.

References

ALONGI, D.M., 2002, Present state and future of the world’s mangrove forests. Environmental

Conservation, 29, pp. 331–349.

ALONGI, D.M., 2008, Mangrove forests: resilience, protection from tsunamis, and responses to

global climate change. Estuarine Coastal and Shelf Science, 76, pp. 1–13.

BALL, M.C., 2002, Interactive effects of salinity and irradiance on growth: implications for

mangrove forest structure along salinity gradients. Trees-Structure and Function, 16,

pp. 126–139.

BERGER, U. and HILDENBRANDT, H., 2000, A new approach to spatially explicit modeling of

forest dynamics: spacing, ageing and neighbourhood competition of mangrove trees.

Ecological Modelling, 132, pp. 287–302.

BERGER, U., RIVERA-MONROY, V.H., DOYLE, T.W., DAHDOUH-GUEBAS, F., DUKE, N.C.,

FONTALVO-HERAZO, M.L., HILDENBRANDT, H., KOEDAM, N., MEHLIG, U., PIOU, C. and

TWILLEY, R.R., 2008, Advances and limitations of individual-based models to analyze

and predict dynamics of mangrove forests: a review. Aquatic Botany, 89, pp. 260–274.

BILGER, W. and BJORKMAN, O., 1990, Role of the xanthophyll cycle in photoprotection eluci-

dated by measurements of light-induced absorbance changes, fluorescence and photo-

synthesis in leaves of Hedera canariensis. Photosynthesis Research, 25, pp. 173–185.

CHEN, R. and TWILLEY, R.R., 1998, A gap dynamic model of mangrove forest development

along gradients of soil salinity and nutrient resources. Journal of Ecology, 86, pp. 37–51.

CLOUGH, B.F., DIXON, P. and DALHAUS, O., 1997, Allometric relationships for estimating biomass

in multi-stemmed mangrove trees. Australian Journal of Botany, 45, pp. 1023–1031.

CONCHEDDA, G., DURIEUX, L. and MAYAUX, P., 2008, An object-based method for mapping and

change analysis in mangrove ecosystems. ISPRS Journal of Photogrammetry and

Remote Sensing, 63, pp. 578–589.

DEMMIG-ADAMS, B., 1990, Carotenoids and photoprotection in plants: a role for the xantho-

phyll zeaxanthin. Biochimica et Biophysica Acta, 1020, pp. 1–24.

DUKE, N.C., LO, Y.Y. and SUN, M., 2002, Global distribution and genetic discontinuities of

mangroves – emerging patterns in the evolution of Rhizophora. Trees-Structure and

Function, 16, pp. 65–79.

DUKE, N.C., MEYNECKE, J.O., DITTMANN, S., ELLISON, A.M., ANGER, K., BERGER,

U., CANNICCI, S., DIELE, K., EWEL., K.C., FIELD, C.D., KOEDAM, N., LEE, S.Y.,

MARCHAND, C., NORDHAUS, I. and DAHDOUH-GUEBAS, F., 2007, A world without

mangroves. Science, 317, pp. 41–42.

ELLISON, A.M., 2002, Macroecology of mangroves: large-scale patterns and processes in tropi-

cal coastal forests. Trees-Structure and Function, 16, pp. 181–194.

ELLISON, A.M. and FARNSWORTH, E.J., 2001, Mangrove communities. In Marine Community

Ecology, M.D. Bertness, S.D. Gaines and M.E. Hay (Eds) (Sunderland, UK: Sinauer

Associates).

EWEL, K.C., ZHENG, S.F., PINZON, Z.S. and BOURGEOIS, J.A., 1998, Environmental effects of

canopy gap formation in high-rainfall mangrove forests. Biotropica, 30, pp. 510–518.

FARNSWORTH, E.J. and ELLISON, A.M., 1997, The global conservation status of mangroves.

Ambio, 26, pp. 328–334.

GAMON, J.A. and SURFUS, J.S., 1999, Assessing leaf pigment content and activity with a

reflectometer. New Phytologist, 143, pp. 105–117.

228 C. Song et al.

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

GAMON, J.A., FIELD, C.B., BILGER, W., BJORKMAN, O., FREDEEN, A.L. and PENUELAS, J., 1990,

Remote sensing of the xanthophyll cycle and chlorophyll fluorescence in sunflower

leaves and canopies. Oecologia, 85, pp. 1–7.

GAMON, J.A., PENUELAS, J. and FIELD, C.B., 1992, A narrow-waveband spectral index that

tracks diurnal changes in photosynthetic efficiency. Remote Sensing of Environment, 41,

pp. 35–44.

GREEN, E.P., MUMBY, P.J., EDWARDS, A.J., CLARK, C.D. and ELLIS, A.C., 1998, The assessment

of mangrove areas using high resolution multispectral airborne imagery. Journal of

Coastal Research, 14, pp. 433–443.

HELD, A., TICEHURST, C., LYMBURNER, L. and WILLIAMS, N., 2003, High resolution mapping of

tropical mangrove ecosystems using hyperspectral and radar remote sensing.

International Journal of Remote Sensing, 24, pp. 2739–2759.

HENSCHKE, C. and HERRMANN, T., 2007, Testing for soil and water salinity. Fact Sheet No.

66/00. Government of South Australia Primary Industries and Resources.

HOGARTH, P.J., 2007, The Biology of Mangroves and Seagrasses (Oxford, UK: Oxford

University Press).

KATHIRESAN, K. and BINGHAM, B.L., 2001, Biology of mangroves and mangrove ecosystems.

Advances in Marine Biology, 40, pp. 81–521.

LEE, S.Y., 1998, Ecological role of grapsid crabs in mangrove ecosystems: a review. Marine and

Freshwater Research, 49, pp. 335–343.

LIN, G.H. and STERNBERG, L.D.L., 1993, Effects of salinity fluctuation on photosynthetic gas-

exchange and plant-growth of the red mangrove (Rhizophora mangle). Journal of

Experimental Botany, 44, pp. 9–16.

LUGO, A.E. and SNEDAKER, S.C., 1974, The ecology of mangroves. Annual Review of Ecology

and Systematics, 5, pp. 39–64.

MACINTOSH, D.J. and ASHTON, E.C., 2003, Draft code of conduct for the sustainable management

of mangrove ecosystems. World Bank, International Society for Mangrove Ecosystems,

and Centre for Tropical Ecosystems Research, Aarhus University, Denmark.

MANSON, F.J., LONERAGAN, N.R., MCLEOD, I.M. and KENYON, R.A., 2001, Assessing techni-

ques for estimating the extent of mangroves: topographic maps, aerial photographs and

Landsat TM images. Marine and Freshwater Research, 52, pp. 787–792.

NICHOL, C.J., RASCHER, U., MATSUBARA, S. and OSMOND, B., 2006, Assessing photosynthetic

efficiency in an experimental mangrove canopy using remote sensing and chlorophyll

fluorescence. Trees-Structure and Function, 20, pp. 9–15.

ONUF, C.P., TEAL, J.M. and VALIELA, I., 1977, Interactions of nutrients, plant-growth and

herbivory in a mangrove ecosystem. Ecology, 58, pp. 514–526.

PARIDA, A.K. and JHA, B., 2010, Salt tolerance mechanisms in mangroves: a review. Trees-

Structure and Function, 24, pp. 199–217.

PENUELAS, J., FILELLA, I. and GAMON, J., 1995, Assessment of photosynthetic radiation-use

efficiency with spectral reflectance. New Phytologist, 131, pp. 291–296.

SAENGER, P., 2002, Mangrove Ecology, Silviculture and Conservation (Dordrecht, The

Netherlands: Kluwer Academic Publisher).

SAENGER, P., HEGERL, E.J. and DAVIE, J.D.S., 1983, Global status of mangrove ecosystems. The

Environmentalist, 3 (Suppl. 3), pp. 1–88.

SIMS, D.A. and GAMON, J.A., 2002, Relationships between leaf pigment content and spectral

reflectance across a wide range of species and developmental stages. Remote Sensing of

Environment, 81, pp. 337–354.

SOBRADO, M.A., 2005, Leaf characteristics and gas exchange of the mangrove Laguncularia

racemosa as affected by salinity. Photosynthetica, 43, pp. 217–221.

TOMLINSON, P.B., 1986, The Botany of Mangroves (Cambridge, UK: Cambridge University

Press).

Hyperspectral remote sensing of red and white mangroves 229

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016

VAIPHASA, C., SKIDMORE, A.K., DE BOER, W.F. and VAIPHASA, T., 2007, A hyperspectral band

selector for plant species discrimination. ISPRS Journal of Photogrammetry and

Remote Sensing, 62, pp. 225–235.

WALTERS, B.B., RONNBACK, P., KOVACS, J.M., CRONA, B., HUSSAIN, S.A., BADOLA, R.,

PRIMAVERA, J.H., BARBIER, E. and DAHDOUH-GUEBAS, F., 2008, Ethnobiology, socio-

economics and management of mangrove forests: a review. Aquatic Botany, 89,

pp. 220–236.

WANG, L. and SOUSA, W., 2009, Distinguishing mangrove species with laboratory measure-

ments of hyperspectral leaf reflectance. International Journal of Remote Sensing, 30,

pp. 1267–1281.

WANG, L., SILVAN-CARDENAS, J.L. and SOUSA, W.P., 2008, Neural network classification of

mangrove species from multi-seasonal Ikonos imagery. Photogrammetric Engineering

and Remote Sensing, 74, pp. 921–927.

230 C. Song et al.

Dow

nloa

ded

by [

Uni

San

Fra

ncis

co d

e Q

uito

] a

t 15:

27 0

5 M

ay 2

016