Estuaries and Coasts

, Volume 40, Issue 5, pp 1215–1226 | Cite as

Is Climate Change Shifting the Poleward Limit of Mangroves?

  • Sharyn M. Hickey
  • Stuart R. Phinn
  • Nik J. Callow
  • Kimberly P. Van Niel
  • Jeff E. Hansen
  • Carlos M. Duarte
Article

Abstract

Ecological (poleward) regime shifts are a predicted response to climate change and have been well documented in terrestrial and more recently ocean species. Coastal zones are amongst the most susceptible ecosystems to the impacts of climate change, yet studies particularly focused on mangroves are lacking. Recent studies have highlighted the critical ecosystem services mangroves provide, yet there is a lack of data on temporal global population response. This study tests the notion that mangroves are migrating poleward at their biogeographical limits across the globe in line with climate change. A coupled systematic approach utilising literature and land surface and air temperature data was used to determine and validate the global poleward extent of the mangrove population. Our findings indicate that whilst temperature (land and air) have both increased across the analysed time periods, the data we located showed that mangroves were not consistently extending their latitudinal range across the globe. Mangroves, unlike other marine and terrestrial taxa, do not appear to be experiencing a poleward range expansion despite warming occurring at the present distributional limits. Understanding failure for mangroves to realise the global expansion facilitated by climate warming may require a focus on local constraints, including local anthropogenic pressures and impacts, oceanographic, hydrological, and topographical conditions.

Keywords

Climate change Global change Mangroves Range shifts Temperature 

Introduction

Mangrove forests are highly productive and valuable ecosystems occupying a narrow intertidal fringe along tropical, subtropical, and warm temperate coasts. Most abundant and diverse through the tropical region, mangroves are integral to many ecosystem services, which are valued at approximately US$200,000/ha annually (Costanza et al. 2014). Climate change has led to the poleward migration of isotherms at rates averaging 27.5 km per decade across the globe (Burrows et al. 2011). Pursuantly, the poleward expansion of the leading edge of a wide diversity of terrestrial and marine species is reported, with average rates of 6.1 km decade−1 (Parmesan and Yohe 2003) and 72 km decade−1 (Poloczanska et al. 2013) respectively. Whilst mangroves display high levels of trait plasticity, particularly associated with their tolerance of salinity, and low temperatures (Lovelock et al. 2016), these variables are also known to be constraining factors to distribution (Krauss et al. 2008). Mangroves are influenced by both air and surface temperature regimes (Giri et al. 2011). As such, climate change should be providing opportunities for the poleward expansion of mangrove forests. However, there has been a minimal (Saintilan et al. 2014) systematic effort to test whether mangrove biogeographical limits are migrating poleward across the globe.

An understanding of the global trends in the mangrove population is largely unknown due to the scarcity of appropriate data, inconsistency of methods, variation in research effort (i.e. limited effort in areas that are remote), with limited species diversity or areas at the poleward limits of established mangroves communities, and poor temporal resolution of the available data sets and past studies. Remotely acquired satellite imagery provides an avenue to document land cover changes and extrapolate land surface temperature (LST) temporally across the globe. LST is commonly used in climatic and ecological studies as it provides an indicator of the heat exchange between the atmosphere and Earth’s surface (Tan et al. 2010).

Remote sensing and Geographic Information Systems (GIS) provide an avenue to systematically acquire and analyse global data that is cumbersome to achieve solely with in situ measurements. Advancement in technologies has led to the availability of various satellite data ranging from low to high resolution, including the freely available Landsat series, MODIS and more recently, Sentinel-2. Landsat data has previously been used in ecological monitoring studies, to map the global areal extent of mangroves (Giri et al. 2011) and to assess typhoon damage (Long et al. 2016). Whilst the Landsat series has a coarse resolution (30 m), it provides freely accessible temporal global cover not provided by other satellites. However, for mangrove forest assessment, the coarser resolution does provide limitations, such that small patches of mangrove forest are difficult to decipher. In fact Giri et al. (2011) note this difficulty in their thorough assessment of the global mangrove extent, in which they conclude that aerial photographs are better suited in determining small mangrove patches. With a lack of high-resolution data freely available at the global poleward sites, this study combines the Google Earth imagery timeline and Google street view to determine mangrove poleward range-edge locations. The high-resolution imagery provided in Google Earth has been utilised in previous ecological studies, including forest biomass estimates (Ploton et al. 2012) and land use monitoring (Jacobson et al. 2015).

Here, we test the hypothesis that warming has occurred at the locations of the range edge of mangroves, and that this warming is leading to a poleward migration of the biogeographical range of mangroves across the globe. We report findings on the biogeographic range limits of mangroves across regions in the northern and southern hemispheres and assess possible range shifts at these locations. This is a twofold process; firstly, we identify, through a thorough review of the literature, the reported distributional range edges for mangroves, and we then examine the current distributional limits using recent imagery available through Google Earth to validate the current range-edge sites of distribution. In doing so, we overcome the limitations of work to date that are minimal and unclear reporting of poleward sites globally. We then assess these sites in relation to decadal temporal changes in LST and air temperature (AT).

Methods

Poleward Mangrove Distribution

Current global locations (Fig. 1) that represent the latitudinal poleward limits of mangrove vegetation were established through a synthesis of the literature and available datasets (Fig. 2). Other site-specific factors reported to constrain mangrove development (including topography, hydrology and salinity, where data were available) were also noted at this stage for each of the range-edge locations (Table 1). Occurrences of mangroves were verified through visual checks in Google Earth and literature sources (Fig. 2; S1). This involved locating the site using current Google Earth imagery and visually checking its presence and potential extension in the historical imagery that was available, via the imagery timeline bar (Fig. 2). We defined evidence of a poleward shift in the leading edge as an established mangrove patch on the coastline poleward of the reported site. To distinguish between a latitudinal extension and areal growth at the existing site, the presence of a topographical feature (e.g. estuary or groyne) was required. A positive finding was then indicated in the imagery, by the presence of mangrove trees, topographically separated from the existing site, or as reported in the literature (Fig. 2). Where the mangrove site could not be located from imagery (S1), geographic location reported in literature was accepted. The resolution of the imagery ensured that only established plants would be visible, ensuring that isolated seedlings were not included. Dates of imagery varied between sites, with some sites having imagery for more dates than others (S1).
Fig. 1

Locations of the reported latitudinal biogeographic limits for mangroves for each region

Fig. 2

Model diagram of steps undertaken to determine poleward sites. Asterisk, site name and was changed for each location. Number sign, steps undertaken for visual checks in Google Earth as part depicted in inset. Images represent imagery from Google Earth for the Bunbury, Australia site. Refer to Appendix S1 for imagery and literature information

Table 1

Reported latitudinal range limits of naturally occurring mangrove distribution globally, including derived temperature. Avicennia spp. dominates occurrence at sites

Site

Location

Species present

Extension-limiting factors

Continent

Coastline

Country

Latitude

Longitude

Nabq

Asia

NW

Egypt

28.21

34.42

AM

T (Galal 2007); FS (Galal 2007)

Geisum Island

Africa

NE (western Geisum Island coastline)

Egypt

27.67

33.70

AM

T (Galal 2007)

Iouik

Africa

NW

Mauritania

19.90

−16.31

AG

To (Dahdouh-Guebas et al. 2005); FS (Morrisey et al. 2010)

Gqunube

Africa

SE

South Africa

−32.93

28.03

AM

T (Morrisey et al. 2010)

Catumbelo

Africa

SW

Angola

−12.45

13.48

AG

O (Harris et al. 2013); FS (Morrisey et al. 2010)

St Augustine

North America

NE

USA

29.97

−81.33

RM; AG; LR

T (Cavanaugh et al. 2014)

Abreojos

North America

NW

Mexico

26.80

−113.70

RM

T (Saintilan et al. 2014)

Bahia de los Angeles

North America

NW (Gulf of California)

Mexico

29.08

−113.55

AG

T (Cartron et al. 2005)

Bermuda

North America

NW (Bermuda)

Bermuda

32.35

−64.71

RM; AG; CE

T (Morrisey et al. 2010)

Laguna

South America

SE

Brazil

−28.47

−48.79

LR; AS

T (Soares et al. 2012); O (Soares et al. 2012); FS (Morrisey et al. 2010)

Piura River

South America

SW

Peru

−5.51

−80.89

AG

To (Woodroffe and Grindrod 1991); O (Clusener and Breckle 1987; Woodroffe and Grindrod 1991); FS (Morrisey et al. 2010)

Fuding

Asia

NE (Continent)

China

27.32

120.22

KO

T (Daidu and Congxian 2006)

Kiire

Asia

NE (Japan)

Japan

31.52

130.33

KO

T (Wakushima et al. 1994)

Sharm Zubeir

Asia

NW (Saudi Arabia)

Saudi Arabia

27.44

35.60

AM

O (Khalil 2004); FS (Khalil 2004)

Corner Inlet

Australia

SE

Australia

−38.91

146.30

AM

T (Morrisey et al. 2010; Boon et al. 2011)

Ohiwa Harbour

Australia

SE (NZ)

New Zealand

−38.02

177.15

AM

T (Morrisey et al. 2010; Duke et al. 1998; De Lange 1994)

Bunbury

Australia

SW

Australia

−33.32

115.65

AM

T (Morrisey et al. 2010; Macnae 1963)

Kawhia Harbour

Australia

SW (NZ)

New Zealand

−38.10

174.80

AM

T (Duke et al. 1998; De Lange 1994)

Mangrove species: Avicennia marina (AM), Avicennia germinans (AG), Rhizophora mangle (RM), Conocarpus erectus (CE), Laguncularia racemose (LR), Avicennia schaueriana (AS), Kandelia obovate (KO). Limiting factors: temperature (T), topography (To), oceanography (O), and freshwater/salinity (FS)

Temperature

LST was extracted for each site from derived Modis Terra and Aqua satellite data using ESRI ArcGIS. Linear rates of temperature change were derived from linear regression of temperature versus time for each site. Temperature data included mean monthly night LST, and mean monthly day LST, between 2000 and 2015. To minimise boundary (sea surface) effects on LST, and misclassification of water as land (Ji et al. 2015), the third landward pixel from the site was selected. Mangroves are exposed to air and seawater in the intertidal zone, and depending on their location in this zone (e.g. near-spring high-tide mark or closer to low-tide mark), high tides may not always result in prolonged submersion of mangrove roots (e.g. mangroves higher in the intertidal zone or closer to the terrestrial edge), resulting in greater (duration) exposure to air (Knight et al. 2008), hence greater influence from air and land temperatures. For this reason, we have selected to investigate LST and AT at sites.

Global monthly averaged AT data were accessed through NCAR Community Climate System Model, version 3.0, using the “Twentieth Century or Historical Climate Simulations (1970–1999)” (https://gisclimatechange.ucar.edu/gis-data) and extracted for sites using ESRI ArcGIS. Linear regression of the LST and AT undertaken in SPSS V.22.

Results

Poleward Mangrove Distribution

We identified a total of 18 natural latitudinal range limits for mangroves across the globe (Table 1; Fig. 1), delineating the area where mangroves occur at present. We found that latitudinal limit of mangroves ranges widely across the globe, from latitudes of 5° to 39°, with a median value of 28.77° latitude (Table 1). The mean latitude 28.22 (±8.49) of the observed northern (n = 10) and southern hemisphere (n = 8) mangrove distribution limits are similar, 28.02°N (±3.41°) and −28.46°S (±12.65°), respectively, yet, there was a greater number of higher latitude sites in the southern hemisphere, with eight sites occurring at 30° of latitude or greater compared with only two sites where mangroves occurred within this latitudinal range in the northern hemisphere. The highest latitude for mangrove occurrence is in the southern hemisphere, at Corner Inlet, south-east Australia (−38.91°). The south-west coastlines of both America and Africa have relatively low-latitude poleward extents of mangroves (<15°), with the Piura river in the southern hemisphere being the lowest range limit in latitude (Table 1). In many instances, planted populations across the globe were reported to have been established successfully further poleward of the latitudinal range of the natural mangrove habitats (S1), though unassisted poleward extension has not been observed at any of these locations.

Temperature

Air temperature was reported in the literature as a constraint to the mangrove population in 12 of 18 sites and the exclusive constraint at 10 sites (Table 1). Examination of the interpolated mean annual air temperature at these sites indicates that, on average, the poleward limit of mangroves is set at locations with mean air temperatures (±SE) ranging from 9.97 °C (±2.75 °C) in the lower range at Corner Inlet and 28.66 °C (±8.12 °C) in the upper range at Geisum Island, across examined years (Table 2 (from NCAR-modelled mean monthly maximum air temperature)). Warming rates at individual sites involved considerable uncertainty (Table 2), largely due to the limited time-span of the observations relative to the slow warming rate. However, the trend toward warming at range edge is consistent across sites and, therefore, robust, even if estimates of warming rates at individual range-edge locations carry considerable uncertainty. The mangrove stands at range-edge sites located at higher latitudes tended to support cooler annual mean temperatures (Table 2). The limiting role of cold temperatures in setting the range-edge limits of mangrove stands is demonstrated by the low latitudinal limits in the distribution of mangroves in SW America and SW Africa, both imposed by cold water currents and upwelling systems (Table 1). Geomorphology, oceanographic and salinity and freshwater levels were also identified as influencing mangrove populations at their range limits (Table 1). The extent of these factors differed between the literature and across sites. Strong Saharan winds were reported as influencing the topography and success of mangrove seedlings at Sharm Zubier, Saudi Arabia, but this was not mentioned as a factor at the closely located Nabq, Egypt (Table 1). However, aridity and decrease in rainfall or groundwater fluctuations were mentioned as possibly limiting factors at both of these locations, as well as Iouik located in Mauritania (Table 1).
Table 2

Linear regression analysis of temperature data for range-edge sites

Site

R2

F statistic

p value

Linear regression equation

Temperature

Rate of warming

Mean (°C)

SE

°C/decade

SE

From NCAR-modelled mean monthly maximum air temperature (NCAR maximum AT (1970–1999))

 Bermuda

0.002

0.678

0.411

y = 0.011x − 0.811

21.54

2.25

0.11

2.25

 Kiire

0.000

0.109

0.742

y = 0.012x − 3.381

20.77

6.05

0.12

6.06

 St Augustine

0.001

0.216

0.643

y = 0.012x − 0.948

23.73

4.39

0.12

4.40

 Bahia de los Angeles

0.001

0.206

0.650

y = 0.013x + 0.034

26.33

4.78

0.13

4.79

 Nabq

0.001

0.213

0.645

y = 0.025x − 20.410

28.24

8.71

0.25

8.72

 Geisum Island

0.001

0.192

0.661

y = 0.022x − 14.417

28.66

8.12

0.22

8.13

 Sharm Zubeir

0.001

0.210

0.647

y = 0.02x − 12.677

26.71

7.10

0.20

7.11

 Fuding

0.000

0.067

0.796

y = 0.012x − 3.841

19.98

7.62

0.12

7.63

 Abreojos

0.001

0.202

0.653

y = 0.01x + 6.233

26.06

3.64

0.10

3.65

 Iouik

0.003

0.964

0.327

y = 0.014x − 0.083

26.82

2.27

0.14

2.27

 Piura River

0.011

4.136

0.043

y = 0.009x + 7.216

25.56

0.75

0.09

0.75

 Catumbelo

0.003

1.129

0.289

y = 0.008x + 10.89

26.24

1.20

0.08

1.20

 Laguna

0.002

0.669

0.414

y = 0.011x − 1.059

21.65

2.30

0.11

2.30

 Gqunube

0.001

0.392

0.532

y = 0.008x + 3.657

20.46

2.22

0.08

2.22

 Bunbury

0.001

0.338

0.561

y = 0.015x − 9.031

20.69

4.23

0.15

4.23

 Ohiwa Harbour

0.000

0.169

0.681

y = 0.009x − 0.784

17.83

3.74

0.09

3.75

 Kawhia Harbour

0.002

0.539

0.463

y = 0.014x − 11.61

16.75

3.19

0.14

3.20

 Corner Inlet

0.000

0.159

0.690

y = 0.011x − 5.934

16.16

4.58

0.11

4.58

 Nxaxo

0.001

0.392

0.532

y = 0.008x + 3.657

20.46

2.22

0.08

2.22

 Kobonqaba

0.001

0.392

0.532

y = 0.008x + 3.657

20.46

2.22

0.08

2.22

 Kwelera

0.001

0.392

0.532

y = 0.008x + 3.657

20.46

2.22

0.08

2.22

From NCAR-modelled mean monthly minimum air temperature (NCAR minimum AT (1970–1999))

 Bermuda

0.001

0.527

0.468

y = 0.011x − 2.233

20.30

2.57

0.11

2.57

 Kiire

0.000

0.124

0.725

y = 0.014x − 12.482

15.52

6.56

1.40

6.57

 St Augustine

0.001

0.190

0.664

y = 0.012x − 5.054

18.18

4.41

0.12

4.42

 Bahia de los Angeles

0.001

0.414

0.520

y = 0.14x − 7.885

20.42

3.64

1.40

3.64

 Nabq

0.002

0.694

0.405

y = 0.35x − 55.691

13.62

6.88

3.50

6.89

 Geisum Island

0.002

0.632

0.427

y = 0.032x − 46.567

16.69

6.58

0.32

6.58

 Sharm Zubeir

0.001

0.428

0.513

y = 0.026x − 32.351

18.49

6.42

0.26

6.42

 Fuding

0.000

0.070

0.791

y = 0.013x − 12.816

13.31

8.14

0.13

8.15

 Abreojos

0.001

0.441

0.507

y = 0.011x − 1.904

20.34

2.77

0.11

2.77

 Iouik

0.003

0.901

0.343

y = 0.14x − 5.66

22.21

2.43

1.40

2.43

 Piura River

0.004

1.539

0.216

y = 0.009x + 5713

23.16

1.16

0.09

1.16

 Catumbelo

0.014

5.267

0.022

y = 0.13x − 3.516

21.60

0.91

1.30

0.91

 Laguna

0.002

0.852

0.357

y = 0.014x − 10.777

17.18

2.51

0.14

2.51

 Gqunube

0.001

0.451

0.502

y = 0.009x − 2.283

15.85

2.23

0.09

2.24

 Bunbury

0.003

0.986

0.321

y = 0.015x − 16.065

14.21

2.52

0.15

2.52

 Ohiwa Harbour

0.002

0.887

0.347

y = 0.16x − 20.414

11.06

2.76

1.60

2.77

 Kawhia Harbour

0.003

1.050

0.306

y = 0.016x − 19.648

13.071

2.64

0.16

2.64

 Corner Inlet

0.002

0.705

0.402

y = 0.014x − 17.701

9.97

2.73

0.14

2.73

 Nxaxo

0.001

0.451

0.502

y = 0.009x − 2.283

15.85

2.23

0.09

2.24

 Kobonqaba

0.001

0.451

0.502

y = 0.009x − 2.283

15.85

2.23

0.09

2.24

 Kwelera

0.001

0.451

0.502

y = 0.009x − 2.283

15.85

2.23

0.09

2.24

From MODIS-derived daytime LST (MODIS LST Day (2000–2015))

 Bermuda

0.002

1.676

0.196

y = 0.043x − 63.73

21.88

3.744

0.43

3.74

 Kiire

0.000

0.220

0.639

y = −0.025x + 70.780

19.86

6.145

−0.25

6.15

 St Augustine

0.003

2.220

0.137

y = 0.059x − 92.612

25.52

4.49

0.59

4.49

 Bahia de los Angeles

0.005

3.484

0.062

y = 0.15x − 261.906

39.05

9.14

1.50

9.12

 Nabq

0.001

0.711

0.399

y = 0.063x − 86.592

40.40

8.52

0.63

8.52

 Geisum Island

0.002

1.422

0.234

y = 0.083x − 126.187

41.09

7.94

0.83

7.94

 Sharm Zubeir

0.001

0.586

0.444

y = 0.055x − 66.974

43.93

8.20

0.55

8.20

 Fuding

0.001

0.583

0.445

y = −0.039x + 100.642

22.13

5.82

−0.39

0.58

 Abreojos

0.019

13.084

0.000

y = 0.185x − 332.295

39.96

5.87

1.85

5.82

 Iouik

0.042

30.118

0.000

y = 0.221x − 405.615

38.42

4.67

2.21

4.58

 Piura River

0.001

0.489

0.485

y = 0.023x − 12.392

33.29

3.70

0.23

3.70

 Catumbelo

0.044

31.667

0.000

y = 0.208x − 383.086

35.41

4.3

2.08

4.21

 Laguna

0.001

0.715

0.398

y = 0.027x − 29.239

25.17

3.64

0.27

3.64

 Gqunube

0.000

0.015

0.902

y = −0.004 + 32.769

24.77

3.65

−0.04

3.66

 Bunbury

0.009

5.940

0.015

y = 0.197x − 369.554

25.72

9.21

1.97

9.18

 Ohiwa Harbour

0.004

2.977

0.085

y = 0.071x − 127.708

15.66

4.71

0.71

4.70

 Kawhia Harbour

0.014

9.537

0.002

y = 0.126x − 236.372

16.05

4.65

1.26

4.63

 Corner Inlet

0.001

0.950

0.330

y = 0.048x − 81.074

16.05

5.64

0.48

5.64

 Nxaxo

0.004

2.469

0.117

y = 0.044x − 64.534

24.58

3.21

0.44

3.21

 Kobonqaba

0.000

0.202

0.653

y = 0.014x − 2.932

24.48

3.45

0.14

3.45

 Kwelera

0.001

0.758

0.384

y = 0.027x − 30.136

24.94

3.58

0.27

3.58

From MODIS-derived night-time LST (MODIS LST Night (2000–2015))

 Bermuda

0.000

0.006

0.940

y = −0.003x + 23.966

18.21

4.290

−0.03

4.29

 Kiire

0.004

2.636

0.105

y = 0.096x − 179.489

12.41

6.697

0.96

6.69

 St Augustine

0.002

1.049

0.306

y = 0.051x − 87.426

15.71

5.697

0.51

5.70

 Bahia de los Angeles

0.002

1.698

0.193

y = 0.068x − 123.244

14.20

5.972

0.68

5.97

 Nabq

0.001

0.346

0.556

y = 0.036x − 50.270

22.02

6.946

0.36

6.95

 Geisum Island

0.000

0.339

0.561

y = 0.035x − 48.269

21.31

6.762

0.35

6.77

 Sharm Zubeir

0.001

0.480

0.489

y = 0.036x − 50.392

22.50

5.949

0.36

5.95

 Fuding

0.000

0.036

0.849

y = 0.011x − 10.481

12.05

6.683

0.11

6.69

 Abreojos

0.003

1.942

0.164

y = 0.058x − 101.739

13.76

4.693

0.58

4.69

 Iouik

0.006

4.111

0.043

y = 0.069x − 121.504

16.76

3.861

0.69

3.85

 Piura River

0.025

17.460

0.000

y = 0.149x − 283.793

15.90

4.106

1.49

4.06

 Catumbelo

0.013

8.968

0.003

y = 0.069x − 118.402

20.05

2.615

0.69

2.60

 Laguna

0.047

33.974

0.000

y = 0.184x − 355.194

14.88

3.677

1.84

3.59

 Gqunube

0.013

9.438

0.002

y = 0.077x − 140.284

13.84

2.856

0.77

2.84

 Bunbury

0.021

15.175

0.000

y = 0.125x − 240.002

11.06

3.684

1.25

3.65

 Ohiwa Harbour

0.019

13.346

0.000

y = 0.099x − 191.900

8.08

3.115

0.99

3.09

 Kawhia Harbour

0.069

51.013

0.000

y = 0.209x − 412.716

7.46

3.447

2.09

3.33

 Corner Inlet

0.011

7.692

0.006

y = 0.075x − 141.856

8.49

3.082

0.75

3.07

 Nxaxo

0.085

64.409

0.000

y = 0.198x − 383.989

13.57

2.929

1.98

2.80

 Kobonqaba

0.026

18.244

0.000

y = 0.102x − 190.076

14.31

2.741

1.02

2.71

 Kwelera

0.049

35.774

0.000

y = 0.150x − 288.074

13.62

2.925

1.50

2.85

Exploration of the mean monthly annual minimum and maximum air temperature showed that between 1970 and 1999, air temperature has increased consistently across sites (Table 2 (from NCAR-modelled mean monthly maximum air temperature and from NCAR-modelled mean monthly minimum air temperature)) by a rate of at least 0.08 °C/decade (Table 2 (from NCAR-modelled mean monthly minimum air temperature)), with a greater rate of increase evident in the minimum air temperature (averaging 0.6 °C/decade) than the maximum (averaging 0.1 °C/decade). Similarly within hemispheres, minimum air temperature has increased at a greater rate (averaging 0.88 °C/decade in the Northern Hemisphere and 0.36 °C/decade in the Southern Hemisphere); however, rates of increase for both minimum and maximum air temperatures (averaging 0.15 °C/decade in the Northern Hemisphere and 0.10 °C/decade in the Southern Hemisphere) are greatest in the northern hemisphere (Table 2 (from NCAR-modelled mean monthly maximum air temperature and from NCAR-modelled mean monthly minimum air temperature)).

LST has also exhibited an increase at the majority of sites, ranging from 0.1 to 2.2 °C/decade (Table 2 (from MODIS-derived daytime LST and from MODIS-derived night-time LST)). Daytime LST was reported as decreasing at three sites, whilst night-time LST decreased at the Bermuda. However, overall, daytime and night-time LST increased across sites with average rates of 0.75 and 0.90 °C/decade, respectively (Table 2 (from MODIS-derived daytime LST and from MODIS-derived night-time LST)). Rates were similar between hemispheres for daytime LST (averaging 0.80 °C/decade in the Northern Hemisphere and 0.71 °C/decade in the Southern Hemisphere) (Table 2 (from MODIS-derived daytime LST)); however, night-time LST increased at a greater rate in the Southern Hemisphere (averaging 0.46 °C/decade in the Northern Hemisphere and 1.31 °C/decade in the Southern Hemisphere) (Table 2 (from MODIS-derived night-time LST)).

Warming and Poleward Expansion at Mangrove Range-Edge Locations

Despite significant warming at the range limit in both air and land surface temperatures, recent poleward biogeographic shifts of mangrove populations have only been reported for 1 of the 18 range-edge locations, located on the south-east African coast (Table 1). Within half a century, the mangrove distribution of South Africa transgressed to 0.3°S and occupied three southern estuaries where mangroves had previously been reported to be absent (Table 1; Fig. 3). The rate (0.04/year) of transgression was greatest at the first reported latitudinal extension between 1962 and 1969 when mangroves expanded from Kobonaqaba to Kwelera (Fig. 3), although the range (0.005°/year) continued to expand again in 1982 (Kwelera to Gqunube) but at a much slower rate (Fig. 3). The distances between the estuaries were greatest at the first recorded mangrove extension event (Kobonaqaba to Kwelera, Fig. 3). However, this comparison is dependent on the timing of the surveys and it is possible that the actual year that mangroves appeared at each new location predated the time of the surveys. No further range expansion was reported after 1982, and we were unable to locate any mangrove stand south of this location by scrutinising satellite images. Hence, the resulting poleward extension between 1962 and to date is 1.66 km per decade, but this represents a maximum, with the rate declining to 0.64 km per decade if indeed no further extension has occurred to date. LST is similar across the South African sites, with only small differences in temperature means apparent (daytime LST, 24.45–24.94 °C; night-time LST, 13.57–14.31 °C). LST does not seem to follow a linear latitudinal trend here (Table 2 (from MODIS-derived daytime LST and MODIS-derived night-time LST)).
Fig. 3

Location of mangrove temporal latitudinal biogeographic locations on the south-east African coast. The extent of mangroves was reported to be in Nxaxo and Kobonqaba in 1962. By 1969, it was reported at Kwelera, and in 1982, it was reported at the current poleward site, Gqunube

Discussion

Despite increases in AT/LST at most sites, we did not find evidence of poleward migration of mangroves since 1982. Additionally, we found evidence that mangrove communities are extant in geographic regions where the average air temperature are below the commonly held minimum for survival and reproduction (16 °C air and (associated 24 °C water minimum) (Gilman et al. 2008)).

Our synthesis indicates that the latitude at which the poleward limit of mangrove distribution is located varies significantly across the globe. Likewise, the temperature regimes at these poleward extents are highly variable, with day and night LST ranging from 5 to 55 °C in 2015. Air temperature means across the studied period ranges from 9.97 to 28.6 °C across sites. In fact, the poleward limit at Corner Inlet occurs where the average minimum temperature is well below that threshold, recording an annual minimum average temperature of 9.6 °C in 2015 (BOM 2014).

The results presented indicate that despite an increase in temperature across the time periods, the data we located indicated that whilst mangroves may have increased in area at high latitudinal sites (Cavanaugh et al. 2014; Saintilan et al. 2014), they are not consistently extending their latitudinal range across the globe. Decreasing rates in temperature are likely to be a result of pixel size and mixing, such that minimal land area at peninsula and island locations containing water interference, or mountainous topography adjacent to estuaries, with possible snow/ice interference.

However, if mangrove poleward limits were limited solely by temperature, these would be expected to be migrating poleward at the rates of migration of isotherms, of 30.6 km per decade in the Northern Hemisphere and 13 km per decade in the Southern Hemisphere, at the latitudinal bands where mangrove poleward range limits are found (Burrows et al. 2011). In addition, mangroves at their poleward limits would have been expected to experience changes in phenology (Burrows et al. 2011), with advanced flowering and seed release, for which we have been unable to find any (supporting data in the) literature for locations at the poleward limit. Indeed, literature directly researching temporal changes in flowering and seed germination in mangroves at high latitudes is limited, rendering the assessment of changes in mangrove phenology and reproduction at their poleward limits cumbersome. Hence, despite evidence for climate change at the range edge of mangroves, the analysis presented here leads us to reject the hypothesis that mangrove range limits are expanding poleward with increased (average) temperatures and isotherm migration.

Recent literature has reported that an increased CO2 atmosphere, as associated with climate change, would likely benefit mangroves, increasing their areal distribution (Gilman et al. 2008). Further, findings by Reef et al. (2016) report that increased atmospheric CO2 increased the optimum temperature for photosynthesis in seedlings by 4 °C. The absence of poleward mangrove migration may be a result of climatic conditions (with other mentioned conditions) not being suitable further toward the poles as opposed to temperature limitations here.

In addition to air temperature, the poleward distributional limit of mangroves is believed to be set by oceanographic features preventing dispersal, topography, orography and salinity, such as at Corner Inlet where, at the southern land mass, there is no further land for mangroves to transgress. Propagules must be able to access suitable intertidal habitat for colonisation to occur, as such favourable ocean conditions are required to transport the floating propagules whilst viable to coastlines that are geomorphically suitable for establishment, such as those with open estuary entrances and intertidal depositional environments (Saintilan and Rogers 2009; Saintilan et al. 2014).

Lack of evidence for poleward mangrove expansion across the globe despite increasing temperatures indicates that mangrove populations could be restricted by other factors, or that mangroves are responding to climate change in a more complex (landward encroachment and partial range filling) way than the simple expectation of a gradual poleward expansion (Saintilan et al. 2014) in line with increasing temperatures. The fact that stable mangrove stands have been created by deliberately planting seedlings at locations at higher poleward latitudes than current limits in several locations provide evidence that temperature regimes suitable for mangrove growth occur poleward of their current limits (S1), indicating that temperature is not the sole limiting factor to mangrove distribution.

The Iouik site (Mauritania) like many of these high latitude sites, has a small mangrove population (n = 11 trees), which limits the potential output of propagules, reducing the probability of range extension. Further, whilst tree height was averaged at 1.32 m at this site, trees 30 cm tall were recorded as “flowering profusely” (Dahdouh-guebas and Koedam 2001), suggesting they were adult trees despite their small size. Indeed, stunted mature mangrove trees <0.5 m in height have also been found at locations near the range limit in SW Australia (Duarte, personal observation, Fig. 4). Mangroves located at high latitudinal sites typically show physiological and morphological adaptation, often leading to their presence being dismissed as an outlier and thus not reported as the poleward limit, rendering the establishment of temporal poleward timelines difficult.
Fig. 4

Dwarf mangrove tree. Shark Bay, Western Australia. Photograph: Carlos M. Duarte

Due to the resolution of imagery used in this review, stands composed of trees of this height and number may not be detected. Whereas, careful field assessments may reveal the presence of seedlings that could not be detected from spatial methods. However, isolated seedlings may not yet signal an established population and may represent transient events. Indeed, identifying a location as the poleward limit requires the presence of a stable stand derived from natural colonisation. The dynamic nature of the poleward limit means mangrove populations may experience transient shifts in either poleward or equatorial directions in association with extreme warm or cold years and/or years of profuse reproduction as reported at St Augustine, Florida associated with an expansion in mangrove area after decline caused by an extreme frost event (Cavanaugh and Al 2014; Cavanaugh et al. 2014; Giri and Long 2014) and in El Nino years in New Zealand (Lovelock et al. 2010). Indeed, extreme cold or warm events can cause retreat or progression, respectively, of the poleward leading edge of mangrove biogeographical ranges (McMillan 1971; Cavanaugh et al. 2014) and result in oscillations that lead to no-clear poleward progression of the leading range edge. However, the broad-scale approach deployed in this study, which concentrates on distribution as opposed to areal changes, ensures that we are reporting on established migration across all poleward sites, and not small-scale changes, that is, site-specific changes to areal cover, which may represent transient recruitment events. This ensures that only mangrove poleward distributional range extensions and related global temperature trends are reported here.

Frost has been determined to influence mangrove decline and growth in mangrove canopy (Wang et al. 2011; Cavanaugh et al. 2014). It is well documented that mangroves at their latitudinal limit are susceptible to low temperatures, and as such, vulnerable to frost damage (Ellis et al. 2006). Frost affects nutrient absorbance from leaf senescence in mangroves, a process critical for nutrient absorption as mangrove soils tend to be low in nutrient concentrations (Ellis et al. 2006; Wang et al. 2011). With stressors such as low temperatures, and frost, mangroves are unable to undertake nutrient resorption from leaves due to mass leaf fall and leaf damage (Ellis et al. 2006; Wang et al. 2011). However, in the Northern Hemisphere, frost has been reported to be declining in intensity and duration, beneficial to mangrove areal expansion as demonstrated at St Augustine, USA (Cavanaugh et al. 2014). However, recent studies have demonstrated that at least in southern Australia, frost events have increased in duration (Crimp et al. 2016), suggesting the influence may be localised and not globally similar. As such, we have not individually analysed temporal frost records in this study, though acknowledging this as a potential constraint to poleward expansion that requires greater site-specific research. However, by including monthly LST, extreme hot or cold events would be included in the mean monthly value.

Poleward sites located near populated areas provide easier access for reporting. However, many may have been anthropogenically affected, by reclamation, or exploitation for resources such as wood. In fact, Saintilan et al. (2014) also noted the difficulty in determining changes in mangrove populations in China and Taiwan due to extensive clearing, such that apparent poleward extension, when reported, may represent recovery of prior habitat lost to human activities rather than range expansion responses to climate change. In fact, many of the sites suitable as potential habitats for range expansion have been subjected to anthropogenic modification to varying grades, though tending to have received some degree of rehabilitation.

Out of the 18 range-edge locations present worldwide, only one presented evidence for poleward expansion, with the mangroves at the south-east African range edge expanding poleward at 0.64 km per decade, much slower than the average rate of poleward expansion with climate change reported for either terrestrial (6.1 km per decade (Parmesan and Yohe 2003)) or marine (72 km per decade (Poloczanska et al. 2013)) organisms. Whilst no further expansion has occurred since 1982 despite warming of air and land surface temperatures on this coastline (Table 2; Fig. 3), Steinke and Ward (2003) demonstrated that seedlings from northern South African estuaries could propagate along much of the southern South African coastline, reiterating that oceanographic and topographic conditions also have a fundamental role in restricting the poleward expansion of mangroves through constraining propagule dispersion.

Whilst this study has concentrated on mangroves at a global scale, individual species have their own range, and as evident from this review, Avicennia sp. is distributed at most of the poleward sites and is representative of the mangrove latitudinal range (Spalding et al. 2010; Quisthoudt et al. 2012). Hence, our conclusion that the latitudinal limit of mangroves is not extending refers to that of species conforming to the poleward range limit, dominated by Avicenna sp., and does not preclude that some mangrove species may be extending their individual range but not the overall mangrove range poleward. Individual genera and species may have individual latitudinal ranges that may be temperature related (Quisthoudt et al. 2012). Consequently, mangrove poleward range extension may be occurring for species with lower latitudinal ranges. This could result in changes with species composition at sites that could affect ecosystem services, including the fauna mangrove forests support. Modelling of mangrove species dynamics with climatic changes in South African estuaries predicts species-specific latitudinal range shifts by 2050 (Quisthoudt et al. 2013). Species-specific changes to biogeographic limits due to climate change have also been predicted for other vegetation, including seaweeds (Bartsch et al. 2012; Quisthoudt et al. 2013). The development of climate envelopes for mangrove species would provide insight into the possible response of mangrove populations to climate change, especially for those populations distributed at the latitudinal range-edge limits.

This report provides a comprehensive account of modern poleward mangrove populations but is limited temporally by past studies and the availability and resolution of data and imagery. This study has been able to build on the critical works (Saintilan et al. 2014), with the GIS datasets prepared by Giri et al. (2011) and Spalding et al. (2010), by investigating range expansion of sites (as opposed to areal extension at site) and by measuring LST and AT at poleward sites to establish temperature climates here. The comprehensive historical reviews that precede available imagery of the South African mangrove population enabled the poleward expansion to be mapped here, highlighting the significance of temporal scale to ecological studies. The remotely acquired satellite and modelled data provided access to comparable data at all sites and across multiple time points which was necessary to fill the gaps of current and past research. Utilising a broad-scale approach as this study has enables large-scale patterns to be detected, ensuring any ecological findings are in fact globally influenced by climate change, and not solely site-specific environmental factors, which whilst important do not reflect global ecological changes. Further, the implementation of a broad-scale approach allows for significant sites to be detected for future management and further targeted research.

Based on the data sets used in this work, mangroves, unlike other marine and terrestrial taxa, may not be experiencing a poleward range expansion despite warming occurring at the present distributional limits. Whilst such data has limitations as discussed earlier, we believe that the use of processed satellite data enables researchers, to undertake global monitoring of mangrove range shifts that otherwise would be difficult to carry out on such a spatial scale. As such, we suggest that such a review of mangrove distribution be undertaken periodically to monitor any advancement in poleward range, in line with changes in climatic variables. Hence, enhanced efforts to observe mangrove dynamics at the 18 stands identified here as delineating their global extent are required, together with the development of climatic envelopes for the species involved, to understand their response to global change. In line with our findings, this would include monitoring a range of climatic factors such as aridity, extreme temperatures and environmental factors such as salinity. However, as discussed, such data is difficult to obtain at high resolution on a global scale, and as such, further studies may benefit from site-specific in situ monitoring where available. Building on the findings of this research, determining potential mangrove range-edge sites based on climatic, environmental and oceanographic modelling would be beneficial in determining areas of significance, such as for protection from development. In a context of global mangrove decline, disturbance at their range limits may prevent mangroves from realising their potential range with climate change. Hence, understanding failure for mangroves to realise the global expansion facilitated by climate change may require a focus on local constraints, including local anthropogenic pressures and impacts, oceanographic, hydrological and topographical conditions.

Notes

Acknowledgements

This work was supported by the CSIRO Flagship Marine and Coastal Carbon Biogeochemical Cluster (Coastal Carbon Cluster) with funding from the CSIRO Flagship Collaboration Fund.

Supplementary material

12237_2017_211_MOESM1_ESM.docx (17 kb)
ESM 1(DOCX 17 kb)

References

  1. Bartsch, I., Wiencke, C. & Laepple, T. 2012. Global seaweed biogeography under a changing climate: the prospected effects of temperature. Seaweed Biology: Novel Insights into Ecophysiology, Ecology and Utilization, pp. 47–66.Google Scholar
  2. BOM 2014. Bureau of Meteorology. Monthly mean minimum temperature: Corner Inlet (Yanakie). Available: http://www.bom.gov.au/jsp/ncc/cdio/weatherData/av?p_nccObsCode=38&p_display_type=dataFile&p_startYear=&p_c=&p_stn_num=085301. Accessed 22 Oct 2014
  3. Boon, P.I., T. Allen, J. Brook, et al. 2011. Victorian Saltmarsh Study: mangroves and coastal saltmarsh of Victoria: distribution, condition, threats and management. Melbourne: Institute for Sustainability and Innovation, Victoria University.Google Scholar
  4. Burrows, Michael T, David S Schoeman, Lauren B Buckley, Pippa Moore, Elvira S Poloczanska, Keith M Brander, Chris Brown, et al. 2011. The pace of shifting climate in marine and terrestrial ecosystems. Science 334(6056): 652–55. doi:10.1126/science.1210288.
  5. Cartron, J.-L.E., Ceballos, G., Felger, R.S. (eds) 2005. Biodiversity, ecosystems, and conservation in Northern Mexico. Oxford University Press.Google Scholar
  6. Cavanaugh, K.C., and E. Al. 2014. Reply to Giri and Long: freeze-mediated expansion of mangroves does not depend on whether expansion is emergence or reemergence. Proceedings of the National Academy of Sciences 111: E1449–E1449.CrossRefGoogle Scholar
  7. Cavanaugh, K.C., J.R. Kellner, A.J. Forde, D.S. Gruner, J.D. Parker, W. Rodriguez, and I.C. Feller. 2014. Poleward expansion of mangroves is a threshold response to decreased frequency of extreme cold events. Proceedings of the National Academy of Sciences of the United States of America 111: 723–727.CrossRefGoogle Scholar
  8. Clusener, M., and S.W. Breckle. 1987. Reasons for the limitation of mangrove along the west coast of northern Peru. Vegetation: 173–177.Google Scholar
  9. Costanza, R., R. de Groot, P. Sutton, S. van der Ploeg, S.J. Anderson, I. Kubiszewski, S. Farber, and R.K. Turner. 2014. Changes in the global value of ecosystem services. Global Environmental Change 26: 152–158.CrossRefGoogle Scholar
  10. Crimp, Steven Jeffery, Bangyou Zheng, Nirav Khimashia, David Lyon Gobbett, Scott Chapman, Mark Howden, and Neville Nicholls. 2016. Recent changes in southern Australian frost occurrence: implications for wheat production risk. Crop and Pasture Science 67(8):801.Google Scholar
  11. Dahdouh-Guebas, F., and N. Koedam. 2001. Are the northernmost mangroves of West Africa viable?—a case study in Banc d’ Arguin National Park. Mauritania. Hydrobiologia: 241–253.Google Scholar
  12. Dahdouh-Guebas, F., L.P. Jayatissa, D. Di Nitto, J.O. Bosire, D. Lo Seen, and N. Koedam. 2005. How effective were mangroves as a defence against the recent tsunami? Current Biology: CB 15: R443–R447.CrossRefGoogle Scholar
  13. Daidu, F., and L. Congxian. 2006. Complexities of China’ s Coast in Response to Climate Change. 2: 54–58.Google Scholar
  14. De Lange, W.P. & de lange, P.J. (1994) An appraisal of factors controlling the latitudinal distribution of mangrove (Avicannia marina var. resinifera) in New Zealand. Journal of Coastal Research 10, 539–548.Google Scholar
  15. Duke, N.C., J.A.H. Benzie, J.A. Goodall, and E.R. Ballment. 1998. Genetic structure and evolution of species in the mangrove genus Avicennia (Avicenniaceae ) in the Indo-West Pacific. Evolution 52: 1612–1626.CrossRefGoogle Scholar
  16. Ellis, William L., Justin W. Bowles, Amy A. Erickson, Nate Stafford, Susan S. Bell, and Melanie Thomas. 2006. Alteration of the chemical composition of mangrove (Laguncularia racemosa) leaf litter fall by freeze damage. Estuarine, Coastal and Shelf Science 68(1–2):363–371.Google Scholar
  17. Galal, Nasser. 2007. National Action Plan for the Conservation of Mangroves in Egypt. Ministry of Agriculture and Land Reclamation, Egyptian Government, Egypt. p 469.Google Scholar
  18. Gilman, E.L., J. Ellison, N.C. Duke, and C. Field. 2008. Threats to mangroves from climate change and adaptation options: a review. Aquatic Botany 89: 237–250.CrossRefGoogle Scholar
  19. Giri, C., E. Ochieng, L.L. Tieszen, Z. Zhu, A. Singh, T. Loveland, J. Masek, and N. Duke. 2011. Status and distribution of mangrove forests of the world using earth observation satellite data. Global Ecology and Biogeography 20: 154–159.CrossRefGoogle Scholar
  20. Giri, C.P., and J. Long. 2014. Mangrove reemergence in the northernmost range limit of eastern Florida. Proceedings of the National Academy of Sciences of the United States of America 111: E1447–E1448.CrossRefGoogle Scholar
  21. Harris, L., S. Holness, R. Nel, A.T. Lombard, and D. Schoeman. 2013. Intertidal habitat composition and regional-scale shoreline morphology along the Benguela coast. Journal of Coastal Conservation 17: 143–154.CrossRefGoogle Scholar
  22. Jacobson, Andrew, Jasjeet Dhanota, Jessie Godfrey, Hannah Jacobson, Zoe Rossman, Andrew Stanish, Hannah Walker, and Jason Riggio. 2015. A novel approach to mapping land conversion using Google Earth with an application to East Africa. Environmental Modelling & Software 72:1–9.Google Scholar
  23. Ji, Luyan, Peng Gong, Xiurui Geng, and Yongchao Zhao. 2015. Improving the accuracy of the water surface cover type in the 30 m FROM-GLC product. Remote Sensing 7(10): 13507–13527. doi:10.3390/rs71013507.CrossRefGoogle Scholar
  24. Khalil, A.S.M. 2004. Status of mangroves in the Red Sea and Gulf of Aden. PERSGA Technical Series No. 11.Google Scholar
  25. Knight, J.M., P.E.R. Dale, R.J.K. Dunn, G.J. Broadbent, and C.J. Lemckert. 2008. Patterns of tidal flooding within a mangrove forest: Coombabah Lake, Southeast Queensland, Australia. Estuarine, Coastal and Shelf Science 76(3): 580–593. doi:10.1016/j.ecss.2007.07.044.CrossRefGoogle Scholar
  26. Krauss, K.W., C.E. Lovelock, K.L. McKee, L. López-Hoffman, S.M.L. Ewe, and W.P. Sousa. 2008. Environmental drivers in mangrove establishment and early development: a review. Aquatic Botany 89: 105–127.CrossRefGoogle Scholar
  27. Long, Jordan, Chandra Giri, Jurgenne Primavera, and Mandar Trivedi. 2016. Damage and recovery assessment of the Philippines’ mangroves following Super Typhoon Haiyan. Marine Pollution Bulletin 109(2):734–743.Google Scholar
  28. Lovelock, C.E., B.K. Sorrell, N. Hancock, Q. Hua, and A. Swales. 2010. Mangrove forest and soil development on a rapidly accreting shore in New Zealand. Ecosystems 13: 437–451.CrossRefGoogle Scholar
  29. Lovelock, C.E., K.W. Krauss, M.J. Osland, R. Reef, and M.C. Ball. 2016. The physiology of mangrove trees with changing climate. In Tropical tree physiology: adaptations and responses in a changing environment, ed. G. Goldstein and S.L. Santiago, 149–179. Cham: Springer International Publishing.CrossRefGoogle Scholar
  30. Macnae, W. 1963. Mangrove swamps in South Africa. Journal of Ecology 51: 1–25.CrossRefGoogle Scholar
  31. McMillan, Calvin. 1971. Environmental factors affecting seedling establishment of the Black Mangrove on the Central Texas Coast. Ecology: Ecological Society of America 52(5): 927–930. doi:10.2307/1936046.CrossRefGoogle Scholar
  32. Morrisey, D.J., A. Swales, S. Dittmann, M.A. Morrison, C.E. Lovelock, and C.M. Beard. 2010. The ecology and management of temperate mangroves. Oceanogr. Mar. Biol. Ann. Rev. 48: 43–160.Google Scholar
  33. Parmesan, C., and G. Yohe. 2003. A globally coherent fingerprint of climate change impacts across natural systems. Nature 421: 37–42.CrossRefGoogle Scholar
  34. Ploton, Pierre, Raphaël Pélissier, Christophe Proisy, Théo Flavenot, Nicolas Barbier, S. N. Rai, and Pierre Couteron. 2012. Assessing aboveground tropical forest biomass using Google Earth canopy images. Ecological Applications 22(3):993–1003.Google Scholar
  35. Poloczanska, E.S., C.J. Brown, W.J. Sydeman, W. Kiessling, D.S. Schoeman, P.J. Moore, K. Brander, J.F. Bruno, L.B. Buckley, M.T. Burrows, C.M. Duarte, B.S. Halpern, J. Holding, C.V. Kappel, M.I. O’Connor, J.M. Pandolfi, C. Parmesan, F. Schwing, S.A. Thompson, and A.J. Richardson. 2013. Global imprint of climate change on marine life. Nature Climate Change 3: 919–925.CrossRefGoogle Scholar
  36. Quisthoudt, K., J. Adams, A. Rajkaran, F. Dahdouh-Guebas, N. Koedam, and C.F. Randin. 2013. Disentangling the effects of global climate and regional land-use change on the current and future distribution of mangroves in South Africa. Biodiversity and Conservation 22: 1369–1390.CrossRefGoogle Scholar
  37. Quisthoudt, K., N. Schmitz, C.F. Randin, F. Dahdouh-Guebas, E.M.R. Robert, and N. Koedam. 2012. Temperature variation among mangrove latitudinal range limits worldwide. Trees 26: 1919–1931.CrossRefGoogle Scholar
  38. Reef, R, Slot, M, Motro, U et al. 2016. The effects of CO2 and nutrient fertilisation on the growth and temperature response of the mangrove Avicennia germinans. Photosynthesis Research 2.Google Scholar
  39. Saintilan, N. & Rogers, K. 2009. Coastal saltmarsh vulnerability to climate change in SE Australia. NSW Coastal ConferenceGoogle Scholar
  40. Saintilan, N., N.C. Wilson, K. Rogers, A. Rajkaran, and K.W. Krauss. 2014. Mangrove expansion and salt marsh decline at mangrove poleward limits. Global Change Biology 20: 147–157.CrossRefGoogle Scholar
  41. Soares, M.L.G., G.C.D. Estrada, V. Fernandez, and M.M.P. Tognella. 2012. Southern limit of the Western South Atlantic mangroves: assessment of the potential effects of global warming from a biogeographical perspective. Estuarine, Coastal and Shelf Science 101: 44–53.CrossRefGoogle Scholar
  42. Spalding, M, Kainuma, M, Collins, L. 2010. World Atlas of Mangroves. A collaborative project of ITTO, ISME.Google Scholar
  43. Steinke, T.D., and C.J. Ward. 2003. Use of plastic drift cards as indicators of possible dispersal of propagules of the mangrove Avicennia marina by ocean currents. African Journal of Marine Science 25 (May 2013): 169–176. doi:10.2989/18142320309504007.CrossRefGoogle Scholar
  44. Tan, K.C., H.S. Lim, M.Z. MatJafri, and K. Abdullah. 2010. Landsat data to evaluate urban expansion and determine land use/land cover changes in Penang Island, Malaysia. Environmental Earth Sciences 60: 1509–1521.CrossRefGoogle Scholar
  45. Wakushima, S., S. Kuraishi, and N. Sakurai. 1994. Soil salinity and pH in Japanese mangrove forests and growth of cultivated mangrove plants in different soil conditions. Journal of Plant Research 41: 39–46.CrossRefGoogle Scholar
  46. Wang, Wenqing, Siyang You, Yunbo Wang, Li Huang, and Mao Wang. 2011. Influence of frost on nutrient resorption during leaf senescence in a mangrove at its latitudinal limit of distribution. Plant and Soil 342(1–2):105–115.Google Scholar
  47. Woodroffe, C.D. & Grindrod, J. 1991. Mangrove biogeography: and sea-level change environmental the role of Quaternary. 18, 479–492.Google Scholar

Copyright information

© Coastal and Estuarine Research Federation 2017

Authors and Affiliations

  • Sharyn M. Hickey
    • 1
    • 2
  • Stuart R. Phinn
    • 3
  • Nik J. Callow
    • 1
  • Kimberly P. Van Niel
    • 1
    • 2
  • Jeff E. Hansen
    • 1
    • 2
  • Carlos M. Duarte
    • 2
    • 4
  1. 1.The School of Earth and Environmental SciencesUniversity of Western AustraliaCrawleyAustralia
  2. 2.The UWA Oceans InstituteUniversity of Western AustraliaCrawleyAustralia
  3. 3.School of Geography, Planning and Environmental ManagementThe University of QueenslandBrisbaneAustralia
  4. 4.Red Sea Research Center (RSRC)King Abdullah University of Science and TechnologyThuwalKingdom of Saudi Arabia

Personalised recommendations