Abstract
Background
Many mangrove ecosystems are periodically exposed to high velocity winds and surge from tropical cyclones, and often recover with time and continue to provide numerous societal benefits in the wake of storm events.
Scope
This review focuses on the drivers and disturbance mechanisms (visible and functional) that tropical cyclones of various intensities have on mangrove ecosystem properties around the world, as well as the potential ecosystem services role offered by mangroves along storm-ravaged coastlines. When viewed together, studies describe repeatable types of impact and a variety of responses of mangroves that make them ecologically resilient to high velocity winds, and which have served to advance the notion that mangroves are disturbance-adapted ecosystems.
Conclusions
Studies have documented massive tree mortality and forest structural shifts as well as high variability of spatial effects associated with proximity and direction of the tropical cyclone trajectory that influence biogeochemical processes, recovery of individual trees, and forest regeneration and succession. Mangroves provide coastal protection through surge and wind suppression during tropical cyclones, and yet are able to overcome wind effects and often recover unless some alternative environmental stress is at play (e.g. hydrological alteration or sedimentation). Structural elements of mangroves are influenced by the legacies imposed by past tropical cyclone injury, which affect their current appearance, and presumably their function, at any point in time. However, much is yet to be discovered about the importance of the effects of tropical cyclones on these fascinating botanical ecosystems, including the role of storm-based sediment subsidies, and much more effort will be needed to predict future recovery patterns as the frequency and intensity of tropical cyclones potentially change.
Keywords: Mangrove wetlands, hurricane, regeneration, storm visible effects, surge, tropical storm, typhoon, recovery
INTRODUCTION
Mangrove ecosystems occur throughout many tropical, sub-tropical and warm temperate regions of the world, and represent a unique botanical community comprised of a number of distinctive growth forms (Tomlinson, 2016). Mangroves have even expanded their distribution within warm temperate locations globally, facilitated by a prolonged post-Pleistocene recolonization period (Sherrod and McMillan, 1985; Woodroffe and Grindrod, 1991; Kennedy et al., 2016), accentuated by higher minimum winter temperatures at progressively higher latitudes over the late Holocene and Anthropocene (Osland et al., 2013; Saintilan et al., 2014). Taxonomically, mangroves are comprised of approx. 70 species or putative hybrids (Duke et al., 1998), and include tree, shrub, fern and palm life forms, collectively creating the potential for structural adaptability and resilience to major disturbances. In a recent review, 45 % of reported disturbances to mangrove area in the global literature were attributed to tropical cyclones (Sippo et al., 2018), ranking tropical cyclones as perhaps the top non-anthropogenic disturbance they must endure.
Tropical cyclone effects on mangrove ecosystems, while disruptive, are often not long lived, as some tree species can re-sprout quickly (e.g. Avicennia), while others rely on advance regeneration strategies (e.g. seedlings present prior to storm events) or post-storm seeding. Mangroves are described as disturbance-adapted communities (Lugo et al., 1981) and are often considered ecologically stable (Alongi, 2008), suggesting that they are able to take on structural disruptions from periodic storms and recover. Where tropical cyclone disturbance frequency is more intense (e.g. parts of the Caribbean region), mangrove structural complexity is reduced to lower statured canopies with few canopy emergents (Lugo and Snedaker, 1974). Where cyclone disturbance is rare (e.g. some Pacific islands, Panama), structural biomass and complexity can increase over time (Allen et al., 2001; Simard et al., 2019).
However, when the complexity of a mangrove is compromised by environmental modifications, especially those influencing tidal hydrological boundaries and characteristics, resistance to tropical cyclone disturbance is also compromised, making individual cyclone events potentially more impactful (Vogt et al., 2012; Lewis et al., 2016). Initial system properties matter tremendously, such that tropical cyclones with similar meteorological metrics affect the same mangrove forest differently based on conditions imposed by past disturbance (Peters et al., 2011). The legacy of past tropical cyclone disturbance and recovery in mangroves often defines the current ecosystem state, and this disturbance legacy is part of the ecology of mangrove forests in many locations.
Science programmes and policies have targeted mangrove ecosystems over the last few decades as structural protection afforded to coastal communities from wind, surge and tsunamis have become more evident (Danielsen et al., 2005; Dahdouh-Guebas et al., 2005; Das and Vincent, 2009). Along with protection for humans living in disturbance-prone environments, additional incentives for mangrove conservation related to carbon storage and flux have emerged as mangroves have also been identified as important components of the global carbon cycle (Bouillon et al., 2008; Donato et al., 2011). Because of these values attributed to mangroves, protection, rehabilitation and restoration activities have helped to reduce global mangrove area losses on an annual basis from the 1970s to the present (Friess et al., 2019). While this is seemingly good news, reports are nuanced by very high annual mangrove losses in specific regions (e.g. Indonesia, Malaysia and Myanmar; Feller et al., 2017), and an uncertain future for mangroves with climate change (Ward et al., 2016). In some regions, climate change is expected to increase the frequency of the most intense tropical cyclones, increase the amount of rainfall produced by tropical cyclones near the cyclone centre and increase the poleward distribution of tropical cyclones (Knutson et al., 2010, 2015; Christensen et al., 2013; Sobel et al., 2016; Kossin et al. 2017; Patricola and Wehner, 2018).
Tropical cyclones affect nearly every aspect of the ecosystem services that mangroves provide; for example, in protecting infrastructure, in provisioning of wildlife habitat and in carbon sequestration. Yet, storm influences on this ecosystem have not received comprehensive review. Here, we will define tropical cyclones, explain where they occur, describe their influences on mangrove ecosystem properties and relevant disturbance mechanisms, and describe cyclone–mangrove interactions with surge suppression, sea-level rise and climate change. We will end this review with a conceptual diagram detailing mechanistically what it means to be ‘disturbance-adapted’ and suggest avenues for impactful future tropical cyclone research on mangroves.
DEFINING TROPICAL CYCLONES
Across the world, tropical cyclones are described using several different names and classification systems. As a result, the literature describing cyclone effects on mangrove forests uses different terms that are location dependent (e.g. hurricane, typhoon or cyclone). Hurricane is the name used for powerful tropical cyclones (i.e. winds speeds that exceed 119 km h–1) in the Atlantic Ocean and north-eastern Pacific Ocean. In contrast, typhoon is used to describe powerful tropical cyclones in the north-western Pacific Ocean. In the Indian Ocean and south-western Pacific Ocean, powerful tropical cyclones are called tropical cyclones, with the addition of a location-specific modifier to indicate intensity (e.g. severe or very intense). In many areas, less powerful tropical cyclones are called tropical depressions (low wind speeds) or tropical storms (moderate wind speeds). Cyclone classification systems typically use maximum sustained wind speeds to designate classification categories. For example, the Saffir–Simpson hurricane wind scale uses wind speeds to categorize hurricanes into five groups (nhc.noaa.gov): Category 1 (119–153 km h–1), Category 2 (154–177 km h–1), Category 3 (178–208 km h–1), Category 4 (209–251 km h–1) and Category 5 (≥252 km h–1). In this review, we use the term ‘tropical cyclone’ in a general manner. However, when referring to intensity, we typically use the Saffir–Simpson categories.
WHERE DO TROPICAL CYCLONES AFFECT MANGROVES?
There are many regions of the world where mangrove forests are never affected by tropical cyclones due to climatic conditions that prevent tropical cyclone formation (Fig. 1; note the blue areas not crossed by red, orange or yellow tropical cyclone tracks). For example, tropical cyclones rarely affect mangrove forests in Africa, South America, Indonesia and Papua New Guinea. In contrast, there are certain cyclonic hot spots, where mangrove forests are greatly influenced by recurrent tropical cyclones (Fig. 1; note the blue areas surrounded by red, orange and yellow cyclone tracks). Hamilton and Casey (2016) provided a ranked list of the 20 countries with the largest mangrove areas. Our global map of tropical cyclone pathways (Fig. 1) indicates that the following seven of these 20 mangrove-rich countries are also affected by tropical cyclones: Australia, Mexico, Myanmar, Philippines, Bangladesh, Cuba and the USA (ranked in descending order according to mangrove area). In terms of aerial cover, these are the seven countries where the effects of tropical cyclones on mangroves are potentially largest. However, in terms of local effects, there are many smaller countries (e.g. Bahamas, Guadeloupe, Honduras, Belize and Haiti) where mangrove forest effects can be large due to their position within common cyclonic pathways.
Fig. 1.
Global distribution of tropical cyclone storm tracks in the northern and southern hemispheres along with Saffir–Simpson scale categories. The distribution of mangroves is depicted by blue shading (from Giri et al., 2011), and reveals those mangrove areas that are subjected to tropical cyclone impact on a recurring basis. Saffir–Simpson scale wind speeds are as follows: Category 1, 119–153 km h–1; Category 2, 154–177 km h–1; Category 3, 178–208 km h–1; Category 4, 209–251 km h–1; Category 5, ≥252 km h–1. Tropical cyclone track data are from Knapp et al. (2010).
INFLUENCES ON SYSTEM PROPERTIES
General descriptions of structural effects on forests from maximum sustained winds of specific intensities from tropical cyclones have been developed (Myers and van Lear, 1998). We have modified those for application to mangroves (Table 1). While difficulties arise when generalizing effects because of variation in geomorphology, tornadic wind downbursts, mangrove forest height and stand condition, two things should be noted from this table.
Table 1.
Generalized mangrove forest damage descriptions relative to Saffir–Simpson tropical cyclone intensity ratings (after Myers and van Lear, 1998 ).
| Category | Wind speed, km h–1 | Description |
|---|---|---|
| 1 | 119–153 (74–95 miles h–1) | Shrubby mangroves, trees and foliage are damaged, but lightly, with few being fully susceptible to mortality through stem breakage or tip-ups. Litter and small branch fall limited to weaker structures or species with less dense wood, and is probably compensatory to what would fall ordinarily in the coming months. Seedlings and saplings with minor damage, but not generally from wind vs. woody debris, tree/branch fall and surge. Surge to 1.2–1.5 m above normal. Wrack deposition limited but notable, probably containing re-deposited mangrove seeds (when applicable) and propagules. Sedimentation low with minimal contributions from re-suspended estuarine components. Coastal built structures (e.g. piers and signs) sustain minor damage. Roads that are low lying are inundated, especially if storm impacts at high tide. No peat collapse expected after the storm. |
| 2 | 154–177 (96–110 miles h–1) | Shrubby mangroves, trees and foliage damaged considerably, with many large trees blown down through stem snapping (especially trees of lighter wood densities) and tip-ups. Litter and branch fall extensive and includes compensatory and additive components to what would fall ordinarily in the coming months. Seedlings and saplings with moderate damage, sustaining only small portion of that damage from wind vs. woody debris, tree/branch fall and surge. Surge to 1.8–2.4 m above normal. Wrack deposition extensive, containing re-deposited mangrove seeds (when applicable), propagules, wood, seagrass, shells, human-sourced debris, etc. Sedimentation moderate and would include re-suspended estuarine components, especially if surrounding estuaries are shallow (<2–3 m deep). Considerable damage to coastal built structures, with many roads being inundated 2–4 h in advance of the approaching storm. Peat collapse expected after the storm only in forests of lower productivity from previous human impact. |
| 3 | 178–208 (111–130 miles h–1) | Mangroves of all sizes highly susceptible to wind fall as they reach a wind velocity threshold established by Category 3 designation (Doyle et al., 1995), especially species with lighter wood densities; nearly all large trees blown down or snapped, and smaller trees and shrubs sustaining moderate to heavy damage, often having damage to vascular structures if they do not break. Litter and branches ripped from trees, with extensive additive volumes of debris fall, affecting growth phenology of surviving mangrove trees into the next few growing seasons. Seedlings with moderate damage and saplings with heavy damage from wind, woody debris, tree/branch fall and surge. Surge to 2.7–3.6 m above normal. Wrack deposition extensive, containing similar elements to Category 2 storms but also including larger debris (e.g. trees) and larger human-sourced debris capable of physically damaging the mangrove forest. Sedimentation extensive (up to 40 mm) from upland, resuspended and autochthonous sources capable of covering emergent root lenticels. Great damage to coastal built structures, with debris from that damage infiltrating mangrove forests from wind throw and water. Most roads are inundated 3–5 h in advance of the approaching storm. Peat collapse expected in some forests after the storm, and imminent in forests of lower productivity from previous human impact. |
| 4 | 2109–251 (131–155 miles h–1) | All trees are affected extensively, with most trees and shrubs blown down or snapped, and saplings and smaller trees often sheared off (Craighead and Gilbert, 1962) in lieu of simply sustaining damage to vascular structures. Canopy loss of foliage and small branches extensive, both leading to delayed mortality of small surviving trees from foliage loss and affecting growth phenology for surviving mangrove trees into the next few growing seasons. Seedlings and saplings with moderate to heavy damage from all sources, often having to re-adjust to post-storm light intensities through new leaf development in order to ensure survival. Surge to 3.9–5.5 m above normal, submerging small statured mangrove forests aligning fringe locations with ocean water and serving (paradoxically) to buffer those small trees from wind damage in some cases. Terrestrial flooding extending to approx. 10 km inland. Wrack deposition extensive, containing similar elements to Category 2 storms but also including larger debris (e.g. trees) and larger human-sourced debris that physically damages the mangrove forest. Sedimentation even more extensive (up to 40–60 mm) from upland, resuspended and autochoctanous sources capable of covering emergent root lenticels and, in some cases, facilitating deposition of intact sections of soil from nearby damaged wetlands on top of the forest floor. Catastrophic damage to coastal built structures (including many homes), with debris from that damage infiltrating mangrove forests from wind throw and water. Roads are inundated 3–5 h in advance of the approaching storm. Beach erosion extensive, with new channels being forged through creeks and barrier islands, especially those areas without significant mangrove colonization. Peat collapse imminent after storm. |
| 5 | ≥252 (> 155 miles h–1) | Catastrophic damage incurred by mangrove forests, with 90–100 % destruction being noted in eyewall and immediate right quadrat areas in the northern hemisphere or left quadrat areas in the southern hemisphere. Shrubby mangroves, trees and saplings destroyed with only a few remaining seedlings, probably damaged by debris falling from broken canopies and debris in surge. Surge to >5.5 m above normal, submerging small statured mangrove forests aligning fringe locations with ocean water and serving (paradoxically) to buffer trees from wind damage in some cases. Terrestrial flooding extending to approx. 16 km inland. Wrack deposition extensive, containing similar elements to Category 2 storms but also including larger debris (e.g. trees) and larger human-sourced debris that physically damages the mangrove forest. Sedimentation extensive (>60 mm) from upland, resuspended and autochoctanous sources capable of covering lenticels on large segments of prop roots and entire pneumatophore structures, and facilitating deposition of intact sections of soil from nearby damaged wetlands deposited on top of the forest floor. Catastrophic damage to coastal built structures (including many homes, municipal structures and roads), with debris from that damage infiltrating mangrove forests from wind throw and water. Roads are inundated 4–7 h in advance of the approaching storm. Extensive beach erosion with and without established vegetation, with new channel formation through barrier islands and mangrove forests. Extensive surge and wave scouring evident, undermining mangrove roots. Peat collapse guaranteed after storm. |
First, tropical cyclones with Category 3 windspeeds or higher (≥178 km h–1) cause extreme visible effects on mangroves, almost without exception. Mangroves are positioned in wet soils and many species have dense wood (Chapman, 1976). Critical wind speeds >178 km h–1 often force tip-ups along with more prominent breakage of large trees (Roth, 1992; Aung et al., 2013), which subsequently fall on small trees and break them. In spite of this, smaller trees and saplings missed by falling trees and branches, and sometimes inundated by storm surge, can survive cyclones. Thus, smaller statured mangrove forests and trees in canopy gaps can appear more resistant to structural effects because they are positioned lower within the vertical boundary layer to allow some wind buffering (Smith et al., 1994). However, if Category 4 wind speeds (>210 km h–1) or higher are realized, dominant forest trees are often laid down (broken or tipped-up) and buffering of smaller trees from the wind is no longer ensured. While smaller statured forests may appear less affected than taller forests at windspeeds of 178–208 km h–1, post-storm recovery may be slower in forests with reduced growth potential.
Secondly, visible tropical cyclone effects on mangroves can be striking before they become extreme (Fig. 2A, B). The physical change imposed by storm-generated surges vs. the protection provided by complete submergence from surge can lead to swathes of visible effects in some fringing locations but not others, confounding spatial resolution of impact. While damage to built structures might even be light in some cases, injury to mangroves, especially those in poor condition prior to the cyclone (Odum and Johannes, 1975; Lewis et al., 2016), can be overwhelming to the casual observer (Fig. 2C). Thus, reactions are often negative (Smith et al., 1994), but herein we offer context; most mangroves within the repetitive storm-track areas shown in Fig. 1 have sustained cyclone impacts previously, and they have already recovered to what they are now. When left in their natural and unaltered state, mangroves often recover from tropical cyclones.
Fig. 2.
Examples of the visual damage incurred by mangrove forests from tropical cyclones. (A) Depicts a Bruguiera gymnorrhiza forest at the mouth of Saddok Tasi Stream on the Pacific island of Saipan approx. 15 months after Typhoon Soudelor affected the island on 2 August 2015 showing individual tree damage, some re-sprouting and seedling regeneration (photo: K. W. Krauss, U.S. Geological Survey). (B and C) Photographs depict predominantly Rhizophora mangle forests along the fringes of Princess Bay (Hurricane Hole) and Leinster Bay (Annaberg Swamp), respectively, on the Caribbean island of St. John (US Virgin Islands) approx. 14 months after Hurricane Irma affected the island on 7 September 2017, showing a range of responses from (B) defoliation, some mid-storey tree survival and limited regeneration, to (C) large structural effects, ponding and few signs of regeneration (photos: Andrew S., From U.S. Geological Survey).
Tree mortality
Tree mortality was 25–100 % over a mangrove area of approx. 40 000 ha from Hurricane Donna, which made landfall in south Florida in 1960 (Craighead and Gilbert, 1962). Likewise, mangrove tree mortality from tropical cyclone events ranged from zero to 100 % among published accounts since (Table 2). Tree mortality can continue for months (delayed mortality) after a cyclone makes landfall (Smith et al., 1994), so post-assessment dates need to be considered. Notable among mortality accounts are the muddled correlations between the intensity of a tropical cyclone based on simple descriptors (e.g. maximum sustained wind speeds) and realized mortality in any specific mangrove stand. For example, mortality ranged from 0 to 86 % from Typhoon Haiyan as it affected the Philippines with some of the strongest winds ever recorded in mangroves (Primavera et al., 2016). Structural complexity, hydrogeomorphic setting, antecedent environmental change, orographic positioning and angle of cyclone trajectory relative to forest location generate a spatial signature of response to (or effects of) tropical cyclones that not only translates into differences in recovery and regeneration, but also confounds cause–effect predictability. That a single site in Primavera et al. (2016) registered no visible effects from such an intense storm is notable. Modelling has sorted some of this complexity by incorporating better storm wind characteristics at finer resolutions (e.g. Doyle and Girod, 1997), but it is clear that generalizing degrees of responses from tropical cyclones should be cautioned.
Table 2.
Summary of mangrove forest mortality from tropical cyclones.
| Location | Cyclone name | Category | Maximum sustained windspeed, km h–1 | Location* | Species† | Mortality, %‡ | Reference |
|---|---|---|---|---|---|---|---|
| Isla del Venado, Nicaragua | Joan | 4 | 217 | Transects 1–3 | AVGE | 37 | Roth (1992) |
| Transects 1–3 | LARA | 53 | |||||
| Transects 1–3 | RHMA | 56 | |||||
| Southwest Florida, USA | Andrew | 4 | 232 | Eyepath | AVGE, LARA, RHMA | 16–61 | Doyle et al. (1995) |
| Right quadrat | AVGE, LARA, RHMA | 3–28 | |||||
| Left quadrat | AVGE, LARA, RHMA | 2–15 | |||||
| South Florida, USA | Andrew | 4 | 232 | Everglades National Park | AVGE | 12–66 | Smith et al. (1994); Baldwin et al. (1995); McCoy et al. (1996) |
| Everglades National Park | LARA | 18–90 | |||||
| Everglades National Park | RHMA | 21–90 | |||||
| Guadeloupe Island | Hugo | 4 | 160–230 | Coastal fringe | RHMA | 59 | Imbert et al., 1996 |
| Mixed dwarf | RHMA, LARA, AVGE | 26 | |||||
| Tall mixed | RHMA | 78 | |||||
| Black | AVGE | 3 | |||||
| Dominican Republic | Georges | 3 | 194 | Los Haitises National Park | AVGE | 64 | Sherman et al. (2001) |
| Los Haitises National Park | LARA | 26 | |||||
| Los Haitises National Park | RHMA | 50 | |||||
| Pacific coast, Mexico | Rosa | 2 | 167 | Agua Grande | RHMA, LARA, AVGE | 27 | Kovacs et al. (2001) |
| Ilsa La Palma | RHMA, LARA, AVGE | 37 | |||||
| Canas | RHMA, LARA, AVGE | 31 | |||||
| Chantilla | RHMA, LARA, AVGE | 28 | |||||
| Punta Burro | RHMA, LARA, AVGE | 30 | |||||
| Bay Islands, Honduras | Mitch | 4 | 188–260 | High impact, Fringe | RHMA, LARA, AVGE | 100 | Cahoon et al. (2003) |
| High impact, Basin | RHMA, LARA, AVGE | 100 | |||||
| Medium impact, Fringe | RHMA, LARA, AVGE | 0–94 | |||||
| Medium impact, Basin | RHMA, LARA, AVGE | 0–50 | |||||
| Low impact, Fringe | RHMA, LARA, AVGE | 0–5 | |||||
| Low impact, Basin | RHMA, LARA, AVGE | 0–15 | |||||
| Yap, Micronesia | Sudal | 3 - 4 | 177 - 209 | Windard sites | RHAP, SOAL, BRGY | 10–27 | Kauffman and Cole (2010) |
| 119 - 151 | Leeward sites | RHAP, SOAL, BRGY | 3–8 | ||||
| Cape Bolinao, Philippines | Chan-hom | 2 | 85–150 | 11-year-old plantation | RHMU | 61 | Salmo et al. (2014) |
| 18-year-old plantation | RHMU | 69 | |||||
| Eastern Samar, Philippines | Haiyan | 5 | 305 | Maslog, Lawaan | RHAP, RHMU | 30 | Primavera et al. (2016) |
| Sto. Nino, Quinapondan | BRGY, RHAP, AECO | 7 | |||||
| Anahaw, General MacArthur | SOAL, AEFL | 20 | |||||
| Maliwaliw, Salcedo | RHAP, RHST | 73 | |||||
| Batang, Hernani | SOAL, AEFL, AVMA | 86 | |||||
| Abejao, Salcedo | SOAL, RHAP, LULI | 0 | |||||
| Bagonbaua, Guiuan | SOAL, AEFL, RHST | 72 | |||||
| Save River Delta, Mozambique | Eline§ | 3–4 | 168–212 | Creek forests | AVMA, RHMU, CETA, SOAL, BRGY, XYGR | 3–29 | Macamo et al. (2016) |
| Seaward forests | AVMA, RHMU, CETA, XYGR | 21–100 |
*Location refers to specific distinctions made by the authors of the manuscript. Please consult the original article for details.
†AECO, Aegiceras corniculatum; AEFL, Aegiceras floridum; AVGE, Avicennia germinans; AVMA, Avicennia marina; BRGY, Bruguiera gymnorrhiza; CETA, Ceriops tagal; LARA, Laguncularia racemosa; LULI, Lumnitzera littorea; RHAP, Rhizophora apiculata; RHMA, Rhizophora mangle; RHMU, Rhizophora mucronata, RHST, Rhizophora stylosa; SOAL, Sonneratia alba; XYGR, Xylocarpus granatum.
‡When the percentage mortality was unreported, reductions in stem densities were used for its estimation.
§Other cyclones also impacted this study area, including Japhet in 2003 (157 km h–1) and Favio in 2007 (222 km h–1), but Eline caused the most damage in 2000.
Differences in tolerances among mangrove species to wind throw also contribute to disparity of effects. This is expected, as tropical cyclones are not considered a stressor per se, but rather wind, rain, surge and excessive sedimentation are the stressors, and those are spatially delivered in vastly different ways during tropical cyclone events. Among neotropical mangroves, historical accounts suggested that Laguncularia racemosa was more susceptible to tropical cyclone effects than Rhizophora mangle (Wadsworth and Englerth, 1959); however, this trend has not always held up in the greater Caribbean region (Table 2). Accounts of differential tree species mortality from Cyclone Nargis, which affected the Ayeyarwady Delta in Myanmar as a Category 3 storm, were much more distinct than the multiple accounts from the Caribbean region. Bruguiera sexangula and Rhizophora apiculata trees experienced >90 % mortality, while mortality was far less for Excoecaria agallocha, Sonneratia caseolaris, Heritiera fomes and Avicennia officinalis, all of which had >85 % survival (Aung et al., 2013). Results from the Caribbean region suggest that the characteristics of the cyclone vs. site conditions greatly influence species-ranked susceptibility outcomes during tropical cyclones (Table 2). In contrast, results from the old world tropics generally support suggestions by Woodroffe and Grime (1999) that mangrove species in the family Rhizophoraceae (Rhizophora and Bruguiera) may be more susceptible to tropical cyclone effects than families such as Avicenniaceae. Perhaps the greater number of species in these families within the old world tropics allow for more prominent species-level trends to emerge at specific threshold wind intensities.
Forest structure
Plot-level assessments of mortality and basal area reductions after a tropical cyclone make use of at least three different approaches: (1) plots are established before the cyclone affects the mangrove forest, with effects being measured in time after the cyclone (e.g. Sherman et al., 2001); (2) plots are established after the cyclone affects the mangrove forest, with pre-effect structure surmised by re-constructing structure from tipped-up or broken stems (e.g. Smith et al., 1994); and (3) plots are established in different locations relative to storm trajectory, controlling for cyclone influences in space (e.g. Kauffman and Cole, 2010). Some assessments combine these approaches (e.g. Doyle et al., 1995). Those techniques that include spatial variability in cyclone effects are likely to be most useful overall because they will translate as new remote-sensing technologies are developed, possibly even leading to less reliance on post-cyclone field surveys in the future.
Mangrove forest structure can be substantially altered by cyclone-induced mortality. In some cases, for example stands with smaller statured trees, lower forest structure can buffer stands against major wind pulses. Taller trees were more likely than shorter trees to suffer breakage from Hurricane Joan in Nicaragua (Roth, 1992), and 70 % of larger adult mangrove trees suffered broken stems in lieu of being uprooted during Cyclone Nargis in Myanmar (Aung et al., 2013). Both McCoy et al. (1996) and Smith et al. (1994) suggested that taller trees were more susceptible to windthrow in their post-hurricane surveys, but Sherman et al. (2001) suggested that size influences among these same mangrove species were more difficult to discern in the Dominican Republic during Hurricane Georges. Species present on site also matter for recovery; snapped stems of R. mangle (and Rhizophora spp.) equate to mortality, as they are not able to coppice or sprout in other ways (Gill and Tomlinson, 1969), while mangroves of genera Avicennia, Sonneratia, Xylocarpus, Excoecaria, Lumnitzera, Laguncularia and Conocarpus readily coppice or sprout (Snedaker, 2002).
Roth (1992) described structural effects and tracked recovery of mangroves for 17 months after Hurricane Joan affected Nicaragua in 1988. There, 45 % of mangroves had broken trunks, and only 41 % of the mangroves began to re-foliate 17 months after the storm. Because larger trees had a greater frequency of broken stems than smaller trees, Hurricane Joan significantly altered the basal area of the mangroves. Re-foliation was registered by smaller trees, suggesting some wind resilience by smaller size classes (Roth, 1992), which would spark recovery. Doyle et al. (1995) assessed mangrove forest structure after Hurricane Andrew affected south Florida in 1992 by stratifying data collection into eyepath, right quadrat and left quadrat sections relative to the path of the hurricane. On average, basal area was reduced from 26.4 to 12.2 m2 ha–1, or by approx. 54 %, within eyepath plots compared with approx. 17 % in right and left quadrat plots (Doyle et al., 1995).
Basal areas were reduced from 18.3 to 7 m2 ha–1 on three of four mangrove forest sites on Guadeloupe Island when Hurricane Hugo made landfall in 1989. Imbert et al. (1996) divided the mangrove zone into coastal fringe, mixed dwarf, tall mixed and black (i.e. referring to the black mangrove species: A. germinans). While the black mangrove zone was nearly unaffected by the tropical cyclone (3 % reduction), overall basal area reductions to the other three zones averaged 23–68 %. The lower percentage was registered in the dwarf stand of lower height. Surveyed upland forests also had fewer visible effects than the mangroves, with Guadeloupe’s upslope rainforests averaging 11 % and semi-evergreen forests averaging 12 % basal area reductions (Imbert et al., 1996). Sherman et al. (2000) divided the mangrove forests of the Dominican Republic into R. mangle-, L. racemosa- and A. germinans-dominated zones along two transects. Plots were established in 1994 and Hurricane Georges made landfall 4 years later in 1998. Basal areas were reduced by 68 % in the three plots dominated by A. germinans (from 20.4 to 6.6 m2 ha–1), by 44 % in the 12 plots dominated by R. mangle (from 23.8 to 13.4 m2 ha–1) and by 19 % in the four plots dominated by L. racemosa (from 40.9 to 33.0 m2 ha–1) (Sherman et al., 2001). Visible effects were less on the L. racemosa forest in that instance, presumably because wind resistance was provided by an even L. racemosa canopy (Sherman et al., 2001). Canopy structuring potentially explains discrepancies in interspecies susceptibilities found among studies (e.g. Smith et al., 1994; Imbert et al., 1996). Collectively, these studies reiterate that the effects of tropical cyclones on mangroves can be highly case specific.
Kovacs et al. (2001) established five transects relative to the passage of Hurricane Rosa on the Pacific coast of Mexico and discovered that basal area was also reduced by 32–71 % depending on the location of the mangrove forest relative to the storm’s track. Again, species presence was important and few small diameter trees were uprooted; when stems were broken, all R. mangle died and, when larger trees were uprooted, L. racemosa fared better (39 % dead) than A. germinans (77 % dead) (Kovacs et al., 2001). On the Island of Yap (Micronesia), Typhoon Sudal affected the mangrove forests in 2004. Kauffman and Cole (2010) established study plots after the storm based upon cyclone trajectory and discovered that plots established on the leeward side of the storm had very high basal areas of 72 m2 ha–1 but, in comparison, plots established on the windward side of the storm had lower basal areas of 54 m2 ha–1. Tropical cyclone effects led to a 226 Mg ha–1 reduction in total above-ground stand biomass in Yap, or >52 % loss of what was standing before Typhon Sudal hit (Kauffman and Cole, 2010). Tropical cyclones do not affect Yap very often, so there was a lot of biomass available to be lost from a single event. Mangroves in Panama were noted to attain twice as much biomass as the mangroves of south Florida and Puerto Rico, which are affected by tropical cyclones more frequently (approx. 50 per century for Puerto Rico if cyclones of all intensities are tallied: Lugo, 2000). Cyclones can also pose a risk to commercial mangrove forestry operations in regions with high tropical cyclone frequencies. For example, Typhoon Chan-hom reduced above-ground biomass of mangroves growing in 11- and 18-year-old plantations in the Philippines by 70–79 % (Salmo et al., 2014). Commercial forestry operations often produce mangrove forests with more uniform canopies, which could theoretically reduce wind influences from lower intensity cyclones through this type of stand management.
Litter and woody debris fall
Annual litterfall of leaf, twig and reproductive debris in mangrove forests ranges globally from 1.3 to 27.6 Mg ha–1 year–1 (Saenger and Snedaker, 1993; Krauss et al., 2014), the full balance of which can be deposited on the forest floor from any single tropical cyclone event in less than a few hours. Adame et al. (2013) even suggested that, with all else being equal, mangrove stands with moderate frequency of tropical cyclone effects develop higher rates of litterfall production over time vs. stands unaffected or heavily affected by cyclones. Litter and coarse woody debris stores can be high after a cyclone and, particularly in the case of woody debris, can persist for decades. Ten years after Hurricane Andrew, woody debris volume on the mangrove forest floor still ranged from 20 to 181 m3 ha–1 in south Florida (Krauss et al., 2005). Fine woody debris (<7.5 cm diameter) represented approx. 36 % of that total woody debris volume, which would have been deposited in the years after the cyclone. These woody debris volumes equate to a biomass of approx. 23.8 Mg ha–1. In comparison, woody debris biomass ranged from 9.5 to 33.9 Mg ha–1, respectively, in an Australian mangrove forest without major disturbance (Robertson and Daniel, 1989) and a Micronesian mangrove forest undergoing heavy harvesting pressure (Allen et al., 2000).
The ecological role of litter and woody debris after a tropical cyclone has not been systematically assessed in mangrove forests. On the one hand, woody debris movement during a cyclone can cause physical damage to seedlings, saplings and trees. However, on the other hand, woody debris decomposes more slowly than leaf litter, suggesting a role in modulated delivery of nutrients to forests as woody debris decomposition ensues over time (Harmon and Hua, 1991). Different stores decompose at different rates, and all of these components would release nutrients akin to slow-release fertilizers. Leaves on the mangrove forest floor decompose the fastest (0.57 ± 0.08 % d–1) followed by wood (0.10 ± 0.02 % d– 1), and both of these sources of nutrients and mineralized carbon are augmented by root decomposition (0.15 ± 0.02 % d–1) (averages from Middleton and McKee, 2001; Romero et al., 2005; Poret et al., 2007), which would mineralize nutrients from below-ground as their supporting stems and canopies are also killed and soils become compacted (Lang’at et al., 2014). Woody debris further serves a role in erosion control, soil pedogenesis, nursery beds for seeds and propagules, structure to trap propagules in large quantities near debris stores, and habitat for heterotrophic communities (Harmon et al., 1986; Allen et al., 2000; Pinzón et al., 2003).
Spatial distribution of effects
When reporting structural effects on mangrove forests, maximum sustained winds and duration of sustainment from tropical cyclones are not always available for specific forests, which might be many kilometres from the nearest weather station. This is a major limitation as these are important details (Lugo et al., 1983) and, where available, would help to describe specific cyclonic influences on the spatial distribution of forest injury. The strongest sustained winds are typically registered in a cyclone’s eyewall (Jordan et al., 1960), although, in the northern hemisphere, effects can sometimes be most severe in the storm’s right quadrant (Shea and Gray, 1973), spawned by a higher probability of unpredictable wind downbursts and tornadoes in eyewall and immediate right quadrat areas (Novlan and Gray, 1974). Note that the right quadrant in the northern hemisphere corresponds to a left quadrant in the southern hemisphere due to hemispheric differences in cyclone rotation. During Hurricane Andrew in south Florida, the direction of mangrove tree fall corresponded very nicely to model-projected tropical cyclone wind directions from all except one location (0.26 < r2 < 0.75); plots located in right quadrat areas immediately fringing the Gulf of Mexico had substantially lower correlations perhaps because of tornadic activity (r2 = 0.05) (Doyle et al., 2009). Aerial transects revealed greater storm effects along the fringe than from an inland transect, related to greater structural development of mangroves along the fringe, with the majority of those trees being felled by the forewinds of Hurricane Andrew (Doyle et al., 2009). For subsequent tropical cyclones in south Florida, spatially consistent wind damage was most evident among trees over 4 m tall and nearest the coast (Zhang et al., 2016).
Shielding of cyclone winds can occur. For example, mangroves located on the leeward side of Yap were shielded from Typhoon Sudal’s major effect relative to windward stands (Kauffman and Cole, 2010), while tree mortality and major pulses of sedimentation were reduced through wind shielding on Roatan, Honduras, during Hurricane Mitch (Cahoon et al., 2003). However, the 3-D structure of mountains can cause wind vortices and vertical sheer that contribute to wind-related injury. Although this type of effect may be more common in upland forests (e.g. Lugo et al., 1983), even there, aspect and exposure were not strong predictors of forest response among Caribbean hurricanes (Brokaw and Walker, 1991). Windward slopes were, however, typically more affected than leeward slopes, but it is uncertain how this would relate uniformly to mangroves developing on flat intertidal planes corresponding to those slopes.
In addition to slope and aspect, mangrove trees developing in canopy gaps tended to be less affected than trees positioned in the canopy during Hurricane Andrew (Smith et al., 1994). This related to average tree size in that trees in gaps were smaller. Such gaps have also been shown to contribute to spatial variability in structure and diversity in upland tropical forests (Brokaw and Grear, 1991; Vandermeer et al., 2000). Considering that many mangrove forests within the Caribbean and other parts of the world have a number of lightning gaps in some stage of recovery (Whelan, 2005; Amir, 2012), gaps would provide significant legacies that contribute to resulting spatial structure of mangrove forests after a tropical cyclone.
The spatial signature of mangrove forest effects from tropical cyclones may be less when viewed across the landscape vs. through the lens of individual plots. Dutta et al. (2015) described the spatial effects of three tropical cyclones on a series of mangrove islands in the Sundarbans of India and Bangladesh from 2007 to 2009. Cyclones SIDR (windspeeds of 46–212 km h–1), Rechmi (windspeeds of 46–83 km h–1) and Aila (windspeeds of 37–111 km h–1) all created major structural change in very small areas of the mangrove forest, but, overall, these tropical cyclones affected the majority of the forest with <10 % effects (Fig. 3). Individual plots would have revealed greater structural effects (>60 %) in some locations, but, overall, visible effects were moderate and varied spatially by cyclone trajectory and size (Dutta et al., 2015). Among the approx. 8568 ha of mangrove area affected by Typhoon Haiyan in the Philippines as it made landfall with windspeeds >252 km h–1 (Long et al., 2016), 870 ha were initially severely affected (i.e. trees uprooted and broken, or complete defoliation), 1820 ha were moderately affected (i.e. broken branches, partially broken tree trunks still standing but defoliated) and 5900 ha were minimally affected (i.e. mostly intact with minor defoliation). These damage areas were reduced significantly within 18 months of the initial post-cyclone assessment from recovery (Long et al., 2016).
Fig. 3.
Spatial distribution of damage (% area disturbed) caused by three separate tropical cyclones (SIDR, Reshmi and Aila) affecting the Sundarbans mangrove forest of India and Bangladesh from 2007 to 2009. Approximate tropical cyclone storm tracks are depicted by red-dashed arrows. The figure modified is from Dutta et al. (2015). Reprinted by permission from Springer Nature Customer Service Centre GmbH: Natural Hazards [Assessment of ecological disturbance in the mangrove forest of Sundarbans caused by cyclones using MODIS time-series data (2001–2011)], D. Dutta, 2015.
Hurricane Vance, the strongest storm ever to affect the Australian mainland, removed 5700 ha of the 12 800 ha of mangroves in one region, but within 6 years 68 % of the affected area was recovering (Paling et al., 2008). Finally, 47.8 % of the mangrove forests of Mozambique were affected by Cyclone Eline across a spatial landscape, but mangroves dominated by Rhizophora mucronata along the seaward fringes experienced upwards of 100 % mortality, while those forests of different species located along tidal creeks were often intact and recovering (Macamo et al., 2016). Therefore, the development of mangroves along different ecogeomorphic settings (sensuLugo and Snedaker, 1974) seems to have a role in tropical cyclone resilience to wind energy, making it even more difficult to develop surveys that isolate the spatial effects of singular cyclone events on spatially diffuse mangrove forests. Tropical cyclones are often more frequent along some coasts (Fig. 1). Typhoon Eline caused the most effects in Mozambique, but two subsequent cyclones contributed to additional effects (Typhoon Japhet in 2003 and Typhoon Favio in 2007) and, collectively, the effects from these tropical cyclones are difficult to separate (Macamo et al., 2016). Other cases of frequent effects from tropical cyclones include, but are not limited to, the greater Caribbean region (including Florida), parts of Central America, the Philippines and the Northern Marianas Islands.
Biogeochemical effects
The biogeochemical effects of tropical cyclones on mangrove forests are diverse and often context dependent. In some cases, tropical cyclones can improve abiotic conditions, increase nutrient availability and stimulate biotic productivity (Lovelock et al., 2011). However, in extreme cases, tropical cyclones can lead to forest mortality, rapid carbon losses and, while uncommon, the conversion of mangrove forests to an entirely different ecosystem (Wanless and Vlaswinkel, 2005; Smith et al., 2009). Positive biotic effects of cyclones are most pronounced in hypersaline and/or nutrient-poor environments, where the freshwater and nutrient subsidies provided by tropical cyclones have the potential to alleviate physiological stress and stimulate productivity. For example, in an arid, hypersaline mangrove forest in north-western Australia, Cyclone Pancho increased nutrient availability and enhanced plant growth (Lovelock et al., 2011). Similarly, in the oligotrophic, phosphorus-limited Florida Everglades (USA), phosphorus-rich storm sediments transported by tropical cyclones can increase forest productivity (Chen and Twilley, 1999; Davis et al., 2004; Castañeda-Moya et al., 2010). Karstic-mangrove ecosystems, like those in the Everglades and the Yucatan Peninsula (Mexico) that have developed on calcareous substrates without large terrigenous sediment inputs, are often phosphorus limited (Rovai et al., 2018; Twilley et al., 2018); hence, storm-derived, phosphorus-rich sediments can trigger comparatively large biotic responses in these nutrient-limited systems (Rivera-Monroy et al., 2011; Adame et al., 2013). In contrast, the nutrient-enrichment effects of storm sediments are expected to be less pronounced in mangrove forests that have developed in deltas and other sediment- and nutrient-rich environments. Although storm-derived sediment additions are often beneficial, too much sediment has the potential to bury aerial roots and limit oxygen exchange, which, in extreme cases, can lead to anoxia and forest mortality (Ellison, 1998; Paling et al., 2008). Tropical cyclone floodwaters that are impounded and retained by anthropogenic barriers or cyclone-transported sediments can also produce anaerobic conditions, high sulfide concentrations and/or hypersaline soils, which may lead to forest stress and mortality if the flooding persists or natural tidal connectivity is hindered for an extended period (i.e. months or years) (Lewis et al., 2016). In an arid mangrove system in Baja California (Mexico), a tropical cyclone-transported sand dune hindered tidal exchange and resulted in mangrove degradation; hydrological restoration was subsequently used to restore tidal connectivity and improve abiotic conditions for mangrove recovery (Bashan et al., 2013).
Tropical cyclone effects on carbon cycling in mangrove forests are largely understudied. In the most extreme cases, carbon may be permanently lost and exported from the system if mangrove peat collapses following forest mortality (Cahoon et al., 2003; Wanless and Vlaswinkel, 2005; Smith et al., 2009). Wind-induced leaf defoliation and loss of vegetation structure can result in temporary losses of above-ground carbon, due to reductions in above-ground biomass and productivity; however, these short-term carbon losses are often followed by periods of rapid carbon gains during forest regeneration and recovery (Odum, 1969; Barr et al., 2012; Danielson et al., 2017; Kominoski et al., 2018). The tree heights and above-ground carbon stocks of mangrove forests in cyclone-free regions are often much larger than their counterparts along coasts that are affected by frequent tropical cyclones (Simard et al., 2019), but we expect that the below-ground carbon stocks and soil burial rates in cyclone-active regions vs. cyclone-free regions would be more difficult to generalize.
MECHANISMS OF EFFECT
Structural effects on individual mangrove trees during tropical cyclones occur in three ways: (1) stems are broken; (2) trees are knocked over (tip-ups); or (3) trees remain standing but are whipped by the wind and sustain internal structural injuries and/or are defoliated. The first two categories are often considered together in post-cyclone effect surveys, as both are often, but not always, lethal. Some insight into the distribution of each category from a major tropical cyclone event in south Florida can be inferred from McCoy et al. (1996), who separated effects according to uprooted vs. broken stems. Across three mangrove species, 49–81 % of the trees were uprooted or broken, with 43–62 % of those trees being uprooted. A few of the trees with broken stems appeared to be still in good condition; however, most were judged to be dead or in poor condition (McCoy et al., 1996). Roth (1992) reported that 37, 53 and 56 % of A. germinans, L. racemosa and R. mangle trees, respectively, were broken from Hurricane Joan in Nicaragua. Indeed, certain species do tend to have more wind-resistant structural characteristics as rated in survival terms. The mangrove species Ceriops tagal, Rhizophora stylosa, Bruguiera parviflora and Excoecaria agallocha were very susceptible to windthrow in Australia, while Xylocarpus moluccensis, Aegiceras corniculatum, Aegialitis annulata and Lumnitzera racemosa were not (Stocker, 1976).
Ecophysiology of individual trees
What characteristics of individual mangrove trees dictate relative susceptibility to tropical cyclone winds, or is this left more to chance? In a review, Saenger (2002) suggested that trees susceptible to windthrow are likely to be those with weakly developed cable root systems, or those with structures weakened by erosion, slumping or biological agents. Aerial root structures have long been hypothesized to provide storm resilience to forested wetland trees (Lamborn, 1890), and aerial root structures are diverse and abundant in mangroves (Tomlinson, 2016).
Crown alteration is the most common type of tropical cyclone effect described globally, as the morphological structure of many mangrove leaves (i.e. thick and large) often catch the wind and are ripped off trees (Saenger, 2002). Even during cyclones with lower wind velocities, substantial crown alteration can occur. For example, broken branches and twisted stems were found in 33–71 % of mangrove trees occurring in the Indian River Lagoon, Florida, from a combination of Hurricanes Frances (Category 2) and Jeanne (Category 3) (Vogt et al., 2012). In order to ensure survival of the tree, the canopy must quickly re-grow leaves. A characteristic of Avicennia following a tropical cyclone is prolific stem re-sprouting on decurrently growing primary branches that once supported dominant crowns, perhaps serving as a mechanism to quickly re-establish photosynthesis and assist the tree with further structural repairs. Yet, we must keep in mind that in instances where all of the crown is lost, mortality can sometimes occur even in trees with the capacity for secondary lateral bud formation.
New growth from stems must be supported by at least a partially intact vascular structure (xylem) within the stem, and winds can affect the internal anatomical structuring (e.g. induce embolism) of mangrove trees to the point of mortality. Some mangrove trees show recovery of part of their canopy only, potentially related to the parts still supported by an intact vascular structure through the stem in the months following a tropical cyclone. Furthermore, individual branches and trees continue to die long after a cyclone hits, especially among those trees that are uprooted (Walker, 1995). Sherman et al. (2001) discovered that 6–9 % of the mangrove trees recorded to be alive 7 months after Hurricane Georges in the Dominican Republic were dead by 18 months. Additional mortality and morbidity occur from a combination of defoliation stress, reduction or alteration of vascular structures and limited carbohydrate stores, which are sometimes exacerbated by greater soil anaerobiosis from prolonged flooding or sediment smothering post-cyclone. Steinke and Ward (1989) even discovered that physical stress of flooding from two cyclones contributed to the loss of a 75 m shoreline fringe of Avicennia marina, with larger trees being directly killed through scour along with the injury from wind. Prolonged flooding sufficient to result in barnacle growth on some mangroves after the cyclone also contributed to delayed mortality from altered soil conditions (greater anaerobiosis). Certainly, natural and human-sourced debris from flooding and storm surge routinely affect or injure mangrove stands and are especially prone to killing seedlings and saplings (Table 1).
The structure of the mangrove stand itself can dissipate tropical cyclone winds. Laguncularia racemosa stems experienced uncharacteristically greater survival during a tropical cyclone in the Dominican Republic because of the occurrence of L. racemosa within a uniform canopy (Sherman et al., 2001). While the stand provides structural protection, individual L. racemosa trees in these stands may also have had a higher stem length to diameter ratio characteristic of many forest plantations; this ratio is suspected in reducing sway frequency during high winds (Sellier and Fourcaud, 2009). Thus, the structure of the mangrove stand can influence individual tree characteristics and survival during a tropical cyclone. Many naturally seeded mangrove forests over a certain height (>5 m), however, have uneven structural characteristics composed of emergent trees potentially more susceptible to initial tropical cyclone winds. While we know very little about patterns of anatomical injury to individual mangrove trees during wind events, studies from other forest types have indicated that material properties of wood (e.g. modulus of elasticity) can have only a limited influence on wind resistance; rather, small differences in individual tree morphology can cause extreme variation in stem oscillations during wind events that compound structural failures in various ways (Sellier and Fourcaud, 2009). Whether a mangrove tips-up or is broken may also be related to these small anatomical variations, position in a stand and substrate, and perhaps less on wood density per se. In firm soils in upland forests (Peltola et al., 2000), wind effects occurred more often from stem breakage vs. tip-ups, such as when soils were looser (Coutts, 1986). Mangrove soils are often wet, but soils around mangrove trees can be very firm, providing structural support to mangroves far beyond storm resilience (e.g. McKee et al., 2007). Appreciable soil strength may also contribute in some way to the percentages of broken vs. tipped-up mangrove trees. Crown characteristics also play a role. Trees with larger crowns and leaves that are more resistant to breakage at the petiole cause greater mechanical stress to stems during wind events (Sellier and Fourcaud, 2009).
Ecophysiology of mangrove stands
Mangrove canopies that do survive a tropical cyclone would be expected to encounter different environmental conditions after the storm as sunlight, heat and humidity are also altered. Hurricane Wilma, which passed through the mangrove forests of Everglades National Park in October of 2005 with maximum sustained winds of 176–190 km h–1, caused complete defoliation of the upper canopy comprised of A. germinans, L. racemosa and R. mangle surrounding an established eddy covariance tower. After the cyclone, soil temperatures down to 50 cm into the soil increased, variability in air temperature increased with higher peak temperatures and greater transport of water vapour from soil to the upper canopy drove latent heat flux increases and sensible heat flux decreases post-cyclone (Barr et al., 2012). These conditions continued through at least 2009 (i.e. 4 years). Residual mangrove canopy trees would not only have to sustain and overcome structural reduction and injury incurred during the tropical cyclone, but would also have to survive a new transpirational diffusion gradient imposed by higher vapour pressure deficits to the canopy. This could press the need for greater stomatal control over canopy water losses.
Recall that the primary driver of water use in forests is the vapour pressure diffusion gradient between the canopy and atmosphere, and, as water is transpired from leaves, pressure builds in stem xylem, which prompts a greater need for water extraction through the stem and from the roots (Cohesion–Tension Theory; Dixon and Joly, 1894). Mangrove stands almost universally have two primary characteristics that allow them to survive these higher vapour pressure deficits after a cyclone: (1) mangroves have strong stomatal control of transpirational water losses (Clough and Sim, 1989); and (2) mangroves have very high rates of water use efficiency, meaning that they need to use very little water to support photosynthesis (Farquhar et al., 1982; Reef and Lovelock, 2015), a characteristic being realized among mangroves at the leaf, stem and stand level (Lovelock et al., 2016). Despite lower total water use of mangrove stands from structural losses after Hurricane Wilma (O’Halloran et al., 2012), surviving mangrove trees might have still been pressed to use more water because of higher vapour pressure deficits in the months following the cyclone as many of the surrounding trees were broken or blown over. Therefore, residual trees may have to deal with additional desalination and less efficient use of water. For example, after Hurricane Wilma, sap flow from the outer sapwood (1.5–2.0 cm) of individual R. mangle, A. germinans and L. racemosa trees averaged 68–179 % higher in surviving trees of broken canopies vs. these same species growing in intact canopies a distance away (cf. Krauss et al., 2015) (Fig. 4), perhaps at least partly reflecting a greater physiological demand for water loss and stomatal regulation from persistently high vapour pressure deficits or enhanced nutrient conditions post-cyclone.
Fig. 4.
Box plots with maximum daily individual tree sap flow rates as measured from the outer sapwood of three Neotropical mangrove tree species in Florida, USA. Plot (A) shows relatively high sap flow rates in June and July from individual trees left standing approx. 21 months after Hurricane Wilma damaged mangroves along the Shark River in Everglades National Park on 24 October 2005 (n = 2–3 trees per species; 16 d), and plot (B) shows lower sap flow rates from the same mangrove species in May and June of a previous year (2003) within an intact, undamaged canopy at Rookery Bay National Estuarine Reserve, Florida (n = 15–18 trees per species; 49 d). U.S. Geological Survey data (V. Engel and K. W. Krauss; after Krauss et al., 2015).
Tropical cyclones impose changes to the net ecosystem exchange of CO2 and radiative forcing from mangrove forests in line with what has been documented from other forest disturbances (Amiro et al., 2010). Injury to the mangrove canopy from Hurricane Wilma caused a significant decrease in overall photosynthesis, an increase in respiration and a reduced albedo from the underlying dark peat-based soils, altering radiative forcing over the years following the tropical cyclone (O’Halloran et al., 2012). This is not necessarily what happens in other forests as lighter understorey soils are exposed after passage of a cyclone to increase albedo. Tropical cyclone influences on mangrove air and soil temperatures are not consistent though. Among six tropical cyclones to influence mangroves in China, air and soil temperatures were not reduced after the storm, and in only one stand was evapotranspiration higher after the cyclone (Typhoon Megi). All other tropical cyclones imposed relative, though not significant, decreases in evapotranspiration after the cyclone (Chen et al., 2014). While these study sites from China suggest caution when interpreting atmospheric change and stress to the residual canopy in mangroves post-cyclone, both stands assessed in China were located lower in the boundary layer, with canopy heights of 3–4 m, than sites in Everglades National Park, with canopy height averaging 19 m. The climate zone along with underlying soil type, state of forest succession and site index no doubt matter to this assessment (O’Halloran et al., 2012; Chen et al., 2014).
Regeneration systems
Recovery of mangroves after a tropical cyclone by various means is often assured; however, challenges exist. For example, tree mortality from repetitive cyclone strikes can lead to rapid decomposition and compaction of peats, or ‘collapse’ (Snedaker, 1995; Lang’at et al., 2014), and the increased standing water levels that result may stymie regeneration (Sherman et al., 2000; Cahoon et al., 2003). One obvious mechanism to overcome this (and previously covered) involves the ability of certain mangrove trees to re-sprout either from coppiced stems or from lateral stem sprouts of trees felled by cyclonic winds (Snedaker, 2002). However, data on the percentage of mangrove trees coppicing or re-sprouting after individual storms and their subsequent contribution to the regenerating forest are difficult to find, perhaps because coppicing or re-sprouting can take time and most post-cyclone assessments occur within 2 years of the event. Sherman et al. (2001) did report that most of the L. racemosa seedlings present 7 months after a tropical cyclone in the Dominican Republic were from basal sprouts, but there were very few A. germinans sprouts to be found at that point in time. Also, profuse re-sprouting from dormant epicormic buds was recorded for L. racemosa and A. germinans after Hurricane Andrew in south Florida (Baldwin et al., 1995), but differentiation between sprout and seedling in-growth to the sapling or young tree stage was not made (Baldwin et al., 2001).
Surge and high tides can serve to buffer and protect seedlings and saplings from wind effects during tropical cyclones. When Cyclone Tracy affected the mangrove forests around Darwin, Australia in 1974, forests were on a rising tide, leaving most saplings and small trees covered in water and unaffected by winds (Stocker, 1976). Ferwerda et al. (2007) suggested that flooding of the regeneration pool contributed significantly to a near-complete recovery of the mangrove forest within 10 years of Cyclone Tracy. Short-statured mangroves in Exmouth Gulf may have also been spared some wind and wave effects by being inundated up to one half of the average canopy height during Cyclone Vance (Paling et al., 2008).
Many post-cyclone surveys provide counts of seedlings from two primary seedling-based regeneration systems inherent to mangroves. The first seedling-based regeneration system includes all of the seedlings or saplings that are rooted in the understorey before a tropical cyclone event. This is referred to as advance regeneration (Helms, 1998). Advance regeneration provides an opportunity for seedlings or saplings surviving in the understorey on seed or hypocotyl reserves, sparse side or overhead light, or sunflecks to grow into the mid-storey and eventually replace the mangrove forest that was blown over. While the concept of shade tolerance in mangroves has been questioned (Snedaker and Lahmann, 1988), it is clear that mangrove seedlings and, in some cases, saplings can survive in the understorey for a period of time overlapping cyclonic activity. Because mangrove seedlings and saplings must also balance salinity tolerance, high light levels are eventually required for growth and development in most mangroves (Lugo, 1986). For example, seedling survival in A. germinans was higher at low salinity, and increased light levels promoted greater survival under high vs. low salinity in a Venezuelan mangrove forest (López-Hoffman et al., 2007). Mangrove seedlings benefit from higher light levels almost universally and would be stifled long term in the understorey (Krauss et al., 2008).
Paradoxically, seedlings surviving in the understorey after a tropical cyclone would have to adjust to new and more intense light levels once the overstorey is partially or completely removed. This is a common issue in tropical forests as seedlings present in newly formed gaps or larger canopy openings must tolerate photosynthetic light induction and altered cellular metabolism as light levels increase acutely (Press et al., 1996). Avoiding photoinhibition in high irradiance, saline environments is enabled through inherently low stomatal conductance, high photosynthetic water use efficiencies and low light-saturated rates of photosynthesis at the leaf level in order to protect photosystem II proteins, all of which are characteristics of mangroves (Farquhar et al., 1982; Ball and Farquhar, 1984). Thus, photoinhibition is rarely noted in mangroves (Cheeseman, 1991); however, mangrove leaves developing in shaded environments often have high quantum yields (Björkman et al., 1998; Krauss and Allen, 2003), which may indicate that such leaves would not be optimally designed for the higher light environments encountered immediately after a tropical cyclone removes overstorey shading.
The second seedling-based regeneration system includes dispersal of seeds and propagules to understorey locations within cyclone-affected mangroves, which can continue for months to years after a tropical cyclone. Roth (1992) even suggested that the production of propagules in some Caribbean mangrove species might coincide by design with higher hurricane frequencies in August through October, thus recognizing the potential evolutionary importance of dispersing mangrove propagules to post-cyclone recovery. Furthermore, dispersal of propagules can occur over considerable distances as some mangrove propagules can float for many months and remain viable. For example, propagules of R. apiculata and R. mucronata from Micronesia floated for up to 3 and 5 months, respectively (Drexler, 2001), and propagules of B. sexangula from introduced populations in Hawaii floated for >2 months (Allen and Krauss, 2006). Propagules of R. mangle, L. racemosa and Pelliciera rhizophorae floated for >100, 31 and 6 d, respectively, in experiments conducted in Panama (Rabinowitz, 1978), and, among 13 species of Australian mangroves subjected to flotation trials, most floated for >15 d (Clarke et al., 2001). Indeed, some mangroves exhibit a greater capacity for dispersal than is realized in nature, e.g. the majority of Ceriops tagal propagules in northern Australia did not disperse >3 m from parent trees (McGuinness, 1996), despite the potential of having >80 % of propagules floating beyond 15 d (Clarke et al., 2001). Thus, much of the seed and propagule base might also be very locally sourced after a tropical cyclone.
As a result of these two regeneration systems, regeneration of mangrove forest trees can be high after a tropical cyclone, and proceed rapidly in some cases (Ferwerda et al., 2007). Some 17 months after Hurricane Joan affected mangrove forests in Nicaragua, 0.68–1.82 seedlings m–2 (6800–18 200 seedlings ha–1) were established and represented all four mangrove species present there even though seedlings had to contend with shading from the mangrove-associated fern, Acrostichum sp. (Roth, 1992). A general theme among post-cyclone surveys in mangrove forests is variability in recovery (Stoddart, 1969), which might not always be predictable across an affected landscape. For example, in south Florida, mixed-species mangrove forests in Everglades City regenerated more heavily with A. germinans vs. its pre-hurricane distribution, but Broad River mangrove forests regenerated more heavily with R. mangle and L. racemosa approx. 9 years after Hurricane Andrew affected south Florida (Fig. 5A, C). As further contrast, mixed-species mangrove forests along the Lostmans River regenerated with more equal representation among all three species (Fig. 5B); all three sites sustained large amounts of visible effects because of their location in the eyewall and right quadrats of tropical cyclone circulation (Doyle et al., 1995). On two fringe mangrove forests in south Florida also affected by Hurricane Andrew, much of the regeneration pool for R. mangle was present before the cyclone, while densities of L. racemosa seedlings increased exponentially for 7 years after the cyclone (Baldwin et al., 2001). Over this same period, regeneration of A. germinans increased on one site, but not on the other. In-growth to the sapling stage was prolific but variable by species on each site, with sapling heights increasing considerably between 1995 and 1999 (or 3–7 years after the tropical cyclone). Baldwin et al. (2001) concluded that moderate tropical cyclone effects on fringe mangrove forests tended to promote release via advance regeneration of R. mangle seedlings, while heavy effects necessitate more reliance on seedling recruitment, which may lead to mixed-species stands upon full recovery of these fringe mangrove forests (Baldwin et al., 2001). Likewise, mangrove forests assigned as coastal fringe and interior fringe in south-eastern Florida were dominated by R. mangle stems after the same tropical cyclone in 1992, but both sites differed greatly in seedling numbers at specific heights by 2001 (Ross et al., 2006). Piou et al. (2006), who assessed the recovery of mangrove forests in Belize some 41 years after Hurricane Hattie, suggested that early post-cyclone regeneration trajectories greatly influence the stand structure observed well into the future. In some extreme cases, mangrove seedlings are completely stripped from the understorey during tropical cyclones. For example, fringing mangrove forests within one study block were denuded of seedlings (≤40 cm tall) from Cyclone Eline in Mozambique (Macamo et al., 2016). Likewise, Cyclone Vance took a massive toll on A. marina forests of north-western Australia in 1999, with the majority of initial post-cyclone regeneration being dominated by R. stylosa in lieu of A. marina (Paling et al., 2008).
Fig. 5.
Spatial depiction of individual mangrove tree diameter at breast height (DBH: cm) distributions within 10–13 m radius circular plots prior to sustaining damage from Hurricane Andrew (graphs labelled 1992) and 9 years after sustaining damage from Hurricane Andrew and showing regeneration (graphs labeled 2001). Hurricane Andrew barrelled through the Florida Everglades on 24 August 1992, with (A) Everglades City plots oriented to the north and outside of the eyewall but within the storm’s right impact quadrat, and (B) Lostmans River and (C) Broad River plots oriented within the northern eyewall and central eyewall, respectively, of the storm. U.S. Geological Survey data (T. W. Doyle and T. J. Smith).
A large threat to post-cyclone recovery for mangroves globally is hydrological modifications incurred before a tropical cyclone event, which can often limit or altogether prevent advance regeneration or successful seeding after a storm because of improper hydrodynamic conditions established before the cyclone (Lewis et al., 2016). For example, a greater number of seedlings were recruited in a tidally unrestricted (69.4 % change in seedling density) vs. tidally restricted (31.2 % change in seedling density) impoundment over a 2 year period following Hurricane Charley on Sanibel Island, Florida (Harris et al., 2010).
SURGE SUPPRESSION AND COASTAL PROTECTION
Identifying the role of mangroves in surge suppression and damage abatement to anthropogenic structures is part of the underlying theme of coastal protection (Gedan et al., 2011; Lee et al., 2014); surge can also affect those protective forests, as has been documented during past tropical cyclone and major tsunami events (Smith et al., 1994; Kamthonkiat et al., 2011). On the other hand, mangrove forests along estuarine fringes that are flooded by surge from approaching tropical cyclones are afforded some protection from wind (Smith et al., 1994), perhaps even lessening the suppression role of mangroves as waters overtop small stands. Indeed, assigning a role to vegetation in suppression and identifying suppression rates use two very different approaches. The first set of evidence includes a series of empirical assessments from actual tropical cyclone events, whereby water level recorders are present in a cyclone’s path. These are uncommon, and the first of these assessments probably did not include many mangroves; however, beach ridges and any marsh-dwelling scrub species would have contributed to an estimated storm surge suppression rate of approx. 6.9 cm km–1 (i.e. centimetres of surge suppression per linear kilometre of wetland) through the coastal wetlands of Louisiana along the northern US Gulf Coast (House Document, 1965). Additional published rates of surge suppression from a range of other coastal wetland types, including some that merely re-state previous sources, expand suppression to 4.2–18.9 cm km–1 (Krauss et al., 2009), with very few observation points aligned to evaluate specific habitat types. In contrast, water level recorders were positioned just right for two tropical cyclone events affecting south Florida mangrove forests in 2004 (Hurricane Charley) and 2005 (Hurricane Wilma). One series of recorders was aligned inside mangrove forests along the Shark River, recording surge suppression of 4.2 cm km–1, while the other series of recorders was aligned within interior mangrove forests not associated with a river, recording greater surge suppression of 9.5 cm km–1 (Krauss et al., 2009).
The second set of evidence includes a series of models of both theoretical scenarios and actual tropical cyclone events. Three-dimensional vegetation structure, different plant types (woody vs. herbaceous), surge height, surge velocity, coastal geomorphology and storm characteristics (Resio and Westerink, 2008) all contribute to surge suppression and can be modelled with some degree of accuracy (Mazda et al., 1997a, b; Sheng et al., 2012). A meta-analysis using wave attenuation data from four mangrove and ten marsh environments indicated that wave height attenuation was similar in marshes and mangroves, despite the more complex vegetation structure in mangroves (Gedan et al., 2011). Actual measurements of surge suppression from storm events assessed over larger distances were universally lower than suggested by modelled rates (Gedan et al., 2011), but the immediate seaward fringe consistently provided the largest surge suppression (Narayan et al., 2010; Zhang et al., 2012).
Suppression of waves by mangroves is sensitive to the speed of the approaching tropical cyclone; reducing that speed by 6.7 m s–1 to 2.2 m s–1 is equivalent to the difference between a Category 3 and a Category 5 cyclone in terms of flooding (Liu et al., 2013). The faster the storm moves inland the more effective mangroves are in attenuating surge. Thus, there are a large number of simulation scenarios to consider, so the most effective way to handle modelling is to consider actual scenarios of coastline geomorphology or characteristics of past tropical cyclone events. Mangroves (6–10 m tall) were found to decrease cyclone wave heights by 60 % under simulated conditions, suggesting that bands of mangroves just 300–800 m wide placed in front of surge will greatly reduce damage to built structures (Narayan et al., 2010). Zhang et al. (2012), using hindcasts from storm readings, suggested that storm surge suppression from Hurricane Wilma could have been 20–50 cm km–1 for bands of mangrove 6–30 km wide. Much of the suppression documented by their modelling occurred within the first 3 km of the mangrove band from open water inland. Indeed, for the series of recorders aligned during Hurricane Charley’s approach to the Ten Thousands Islands region of Florida, suppression was highest in the 2.3–3.2 km band (15.8 cm km–1), intermediate in the 3.2–4.5 km band (9.4 cm km–1) and lowest in the 4.5–5.5 km band (4.1 cm km–1) from the open estuary, with the latter two bands dominated by marsh (Krauss et al., 2009).
The specific differences associated with cyclones vs. normal waves and tsunamis have created some confusion in modelling. Tropical cyclone-generated storm surges act like a high tide event fostering a temporary increase in sea level (Lowe et al., 2001). As that surge moves inland, barometric pressure decreases commensurate with the intensity of the tropical cyclone to allow for water volume expansion to add to surge intensity, and strong winds can create additional ‘normal waves’ superimposed on the top of surge to compound effects (Wells, 2012). Therefore, it is not appropriate to simply apply a normal wave model to surge suppression; energy, wave amplitude, pressure and velocity are different (von Storch and Woth, 2008). It is uncertain how some of the more notable models of surge waves account for this difference (see Baird et al., 2009), or how 3-D vegetation structure is accounted for as water levels increase to higher vegetation layers during the surge, sensuSheng et al. (2012). For example, using normal wave models, wave height is suppressed by 20 % over 100 m of mangrove (Mazda et al., 1997a) and by 50 % over 150 m of mangrove when simulating Rhizophora-style prop roots (Brinkman et al., 1997). This is not a reasonable expectation of suppression for tropical cyclone-generated surge events through mangroves. Likewise, a tsunami creates a different scenario. Despite the publication of 54 papers on the topic of coastal protection by mangroves between 1972 and 2005 (Dahdouh-Guebas and Jayatissa, 2009), protection offered by tropical cyclone surge suppression for specific events may still be overstated by modelling (Baird et al., 2009).
SEDIMENTATION, SURFACE ELEVATION CHANGE AND RESPONSES TO SEA-LEVEL RISE
In the face of accelerated-sea level rise, the fate of mangrove forests will depend upon their ability to adjust their vertical and horizontal position in the landscape via local surface elevation change (Krauss et al., 2014; Woodroffe et al., 2016) as well as via horizontal migration into upslope and/or upriver ecosystems (Enwright et al., 2016; Borchert et al., 2018). The sedimentation and surface elevation changes that are triggered by tropical cyclones can affect local, soil surface elevation in positive and negative ways. Mangrove ecosystems are, in general, highly resilient ecosystems that can often tolerate or quickly recover from tropical cyclones, as discussed in previous sections. Nevertheless, there are a few examples throughout the Holocene where potent tropical cyclones have resulted in mass mangrove mortality followed by peat collapse, soil elevation losses and conversion to mud flats (Jones et al., 2019). One example of cyclone-induced peat collapse comes from the Cape Sable region of Florida’s Everglades National Park (USA), where forest mortality and peat collapse triggered by the powerful 1935 Labor Day Hurricane – the most intense tropical cyclone to make landfall in the USA (Landsea et al., 2015) – resulted in the local loss of mangrove forests, which were replaced by mudflat ecosystems (Bischof, 1995; Wanless and Vlaswinkel, 2005; Smith et al., 2009). Another example of peat collapse occurred in the Bay Islands of Honduras, where mass mangrove mortality caused by Hurricane Mitch resulted in rapid elevation losses 18–33 months after the storm (11 mm year–1) due to decomposition of root material and sediment compaction, and which were modelled to continue for at least another 8 years at a reduced rate (7 mm year–1) (Cahoon et al., 2003).
In contrast to these examples of negative cyclone effects, there are many examples of positive effects of tropical cyclones on soil surface elevation change. Tropical cyclones have the potential to deliver nutrient-rich sediments to mangrove forests (Castañeda-Moya et al., 2010, 2013; Lovelock et al., 2011; Smoak et al., 2013; Breithaupt et al., 2014, 2017). Although excessive sedimentation can bury roots and lead to mortality (Ellison, 1998; Paling et al., 2008), moderate sediment deposits can stimulate plant growth and produce immediate gains in soil surface elevation. For example, in the Florida Everglades, Hurricane Wilma deposited up to 8 cm of sediment onto the forest surface; sediment deposition during this storm was negatively correlated with distance upstream and was estimated to be equivalent to between 18 000 and 27 000 dump truck loads of sediment (Smith et al., 2009; Whelan et al., 2009). While a portion of this storm sediment was lost due to erosion, much of it was quickly stabilized by the upward growth of mangrove roots. Using pre- and post-Wilma measurements at surface elevation table-marker horizon (SET-MH) stations, Whelan et al. (2009) found that some of the gains in elevation due to the storm sedimentation were later offset by compaction below the soil surface; however, the net elevation effect of the storm during the measurement period was positive, resulting in a net rate of change that was greater than the rate of relative sea-level rise (Whelan et al., 2009; Feher et al., 2019) (Fig. 6). In the Everglades, most of the cyclone-derived sediments come from the ocean; however, in watersheds that are more sediment rich, tropical cyclones also have the potential to deliver terrigenous sediments to mangroves from upstream ecosystems. Given the potential for future increases in storm intensity, it is possible that sediment delivery and elevation change in some mangrove forests could increase. However, long-term elevation change data from mangrove forests are rare, and there is clearly a need to advance understanding of the contribution of storm sediments to long-term elevation changes relative to rising sea levels.
Fig. 6.
Sedimentation patterns relative to various soil depths from Hurricane Wilma in a mangrove forest in Everglades National Park, Florida, USA (after Whelan et al., 2009). ‘Accretion’ represents the deposition of sediment on the soil surface, and ‘Shallow RSET’, ‘Original SET’ and ‘Deep RSET’ represent the influence that sediment deposition had on different segments of the soil profile to include depths of approx. 0–0.35, 0–4 and 0–6 m, respectively. Reprinted by permission from Springer Nature Customer Service Centre GmbH: Wetlands [Hurricane Wilma’s impact on overall soil elevation and zones within the soil profile in a mangrove forest], K. R. T. Whelan, 2009.
In addition to affecting local responses to sea-level rise via elevation changes, tropical cyclones also have the potential to facilitate landward migration and adaptation to sea-level rise by transporting mangrove propagules long distances, into upslope and/or upriver locations. For example, in the south-eastern USA and the Caribbean, a portion of the hurricane season coincides with the timing of mangrove propagule production. As a result, cyclones and other tropical storms can transport propagules longer distances than possible under normal tidal fluctuations. Peterson and Bell (2012, 2015) demonstrate the critical role of high water events, such as storm surges, for lifting mangrove propagules above vegetative barriers and across the mangrove ecotone, where they may become entrapped and established within an adjacent plant community. These cyclone-mediated, long-distance dispersal events can facilitate mangrove landward migration in response to sea-level rise.
CLIMATE CHANGE AND TROPICAL CYCLONE EFFECTS ON MANGROVES
In the coming century, the global frequency of tropical cyclones is expected either to decrease or remain essentially unchanged in response to climate change (Christensen et al., 2013; Kossin et al., 2017). However, at the global scale, climate change is expected to: (1) increase the frequency of the most intense tropical cyclones; (2) increase the amount of rainfall produced by tropical cyclones near the cyclone centre; and (3) increase the poleward distribution of tropical cyclones (Knutson et al., 2010, 2015; Christensen et al., 2013; Sobel et al., 2016; Kossin et al. 2017; Patricola and Wehner, 2018). Collectively, these projections indicate that tropical cyclone effects on mangrove forests in some regions could be amplified in the future. However, future changes in cyclone activity are less consistent, and challenging to predict at the regional scale. Increases in cyclone intensity may increase the frequency of extreme events that result in local mangrove mortality, peat collapse and mangrove conversion to open water. Tropical cyclone effects will also be modulated by interactions with other aspects of climate and land use change (Osland et al., 2018; Sippo et al., 2018). For example, rising sea levels and increased precipitation are expected to increase storm surge effects, and the poleward expansion of tropical cyclone pathways may enhance the long-distance poleward dispersal of mangrove propagules (Van der Stocken, 2019a, b; I. C. Feller, pers. commun.), which could facilitate the poleward expansion of mangrove forests in response to warming winter temperatures. Moreover, anthropogenic activities, such as urban and agricultural encroachment, can lead to habitat fragmentation and nutrient enrichment, which can affect mangrove responses to tropical cyclones (Feller et al., 2015). As noted throughout this review, mangrove forests are highly resilient ecosystems that have continuously adjusted to dynamic, ever changing abiotic conditions at the land–sea interface. In response to past changes in climate and sea level, mangrove forests have used long-distance dispersal mechanisms to migrate to new positions in the landscape as well as biogeomorphic feedbacks to adapt locally. In the coming century, mangrove forests will use these same mechanisms to adapt to future change. From a coastal management perspective, humans can help facilitate mangrove adaptation to climate change by restoring favourable abiotic conditions in existing mangrove forest as well as ensuring that landward migration corridors are available for future migration into adjacent upland and upriver ecosystems (Enwright et al., 2016; Borchert et al., 2018).
CONCLUSIONS AND PERSPECTIVES FOR ADDITIONAL STUDY
The influence of tropical cyclones on structural characteristics of mangrove forests depends not only on storm characteristics, but also on ecogeomorphic location, stand structure, previous wind events, regeneration present before the storm, seed and propagule dispersal potential after the storm, nutrient conditions, the characteristics of individual species present in the mangrove forest being affected and whether hydrological processes have been altered before the cyclone (Fig. 7). With continued study, we may one day improve predictions of mangrove ecosystem responses to individual tropical cyclone events; however, we might also anticipate that wide-scale convergence in responses will be elusive given the variable trajectories of responses to tropical cyclones created by legacies inherent to any one forest (Fig. 7). Indeed, mangroves have an ability to recover from damage and reduce impacts of tropical cyclones, which limits many of the perceived negative effects over longer time frames contributing to their reputation for being ‘disturbance adapted’. Cyclones are also a source of potentially important sediment amendments and can facilitate dispersal of mangroves far inland. It is evident that having mangroves along a coastline provides important protective roles from surge and winds. A greater understanding of surge suppression by mangroves could be facilitated by additional empirical assessments to include water-level observations after multiple tropical cyclone events to improve modelling.
Fig. 7.
An integrated, disaggregated view of the effects of tropical cyclone disturbances on mangrove ecosystems (adapted from Peters et al., 2011). This conceptual framework includes the effects of an individual tropical cyclone event (large blue box) as well as the contributions of subsequent events and other environmental factors towards the ecosystem reaching a quasi-equilibrium state. In the large blue box, an individual tropical cyclone event consists of three measurable components: (1) the environmental drivers; (2) the initial system properties; and (3) the biological and physical mechanisms of effect. Together, these three components result in a change in ecosystem properties (red symbol and text) and a mangrove ecosystem state with disturbance legacies. Following the cyclone disturbance event (areas outside of the large blue box), subsequent environmental drivers and successional processes (blue symbol and text) will produce a recovering mangrove ecosystem. Ultimately, recurrent disturbance events may produce a quasi-equilibrium ecosystem state with natural variability in ecosystem dynamics (orange symbol and text). Examples are provided as bullets throughout the conceptual framework. Mangroves are especially adept at tolerating legacy effects of past disturbance through this disaggregated view, creating the notion that mangroves are disturbance-adapted ecosystems.
Mangrove studies should also focus on establishing, measuring and re-measuring long-term forest plots after singular and repetitive tropical cyclones to understand recovery and resilience with iterative effect, a condition that may become more common in the future with climate change. Along with this is a fundamental need to measure, at the mangrove scale of study, the wind energy being dissipated by the event; not knowing the wind speed in relation to measurements of response will keep ecologists guessing forever. Efforts should be made to incorporate these long-term forest and wind observations into modelling efforts of post-cyclone recovery such that society can understand their requirements for recovery and availability for future storm shielding. Likewise, if society is to become increasingly dependent on the ecosystem services and recovery of mangrove ecosystems during and after tropical cyclones, more effort must be made to identify mangrove stress on large spatial scales far in advance of tropical cyclone impact so that they can be rehabilitated and remain resilient to storm effects (sensuLewis et al., 2016). Stress comes in many forms, intended and unintended, but all chronic stress affects recovery from acute events in mangroves.
It is also a point of curiosity to understand the unique microclimatic conditions faced by residual standing mangrove trees after a tropical cyclone, created by increased temperatures, reduced or enhanced albedo, greater vapour pressure deficits and larger fluxes of sunlight. Surviving the short-term effects of the tropical cyclone might not be enough. While mangroves are extremely tolerant of harsh environmental conditions within a forest stand setting, there may be thresholds to individual seedling, sapling or tree tolerances when short-term pulses of sunlight or alterations in microclimatic conditions occur, potentially facilitating delayed mortality of residual plants, which would need to be included in tropical cyclone recovery models. Spatially expansive mangrove forests may be far more resilient to tropical cyclones than smaller stands because the effect is aligned with cyclone trajectory and is partial, with surviving areas serving as seed and propagule sources for affected areas. More information is also needed to determine whether sedimentation events from tropical cyclones have a true long-term positive influence on elevation trajectories in mangrove forests in partially offsetting sea-level rise. Finally, some effort should be directed toward understanding what causes mangroves to be most resilient to cyclones. Does maintaining hydrological connectivity matter most, or are disturbance legacies or other aspects of soil biogeochemistry as important? If the right aspects of the tropical cyclone response and recovery are known among multiple mangrove ecosystems throughout the world, then perhaps management action can target greater resilience of those ecosystems to tropical cyclone events.
FUNDING
This review was supported by the USGS Land Change Science R&D Program, USGS LandCarbon Program, USGS Environments Program and USGS Greater Everglades Priority Ecosystem Science Program.
ACKNOWLEDGEMENTS
We thank Mick E. Hanley (University of Plymouth) and Tjeerd J. Bouma (Utrecht University) for the invitation to develop this review article, and Ariel E. Lugo (USDA Forest Service, Puerto Rico), Brita J. Jessen (Rookery Bay National Estuarine Research Reserve), Hannah L. Mossman (Manchester Metropolitan University) and one anonymous referee for reviewing a previous draft. We thank Thomas W. Doyle, Thomas J. Smith III, Kevin R. T. Whelan and Vic Engel for assistance with table and figure review and/or data referenced in this review.
LITERATURE CITED
- Adame MF, Zaldívar-Jimenez A, Teutli C, et al. 2013. Drivers of mangrove litterfall within a karstic region affected by frequent hurricanes. Biotropica 45: 147–154. [Google Scholar]
- Allen JA, Krauss KW. 2006. Influence of propagules flotation longevity and light availability on establishment of introduced mangrove species to Hawaii. Pacific Science 60: 367–376. [Google Scholar]
- Allen JA, Ewel KC, Keeland BD, Tara T, Smith TJ III. 2000. Downed wood in Micronesian mangrove forests. Wetlands 20: 169–176. [Google Scholar]
- Allen JA, Ewel KC, Jack J. 2001. Patterns of natural and anthropogenic disturbance of the mangroves on the Pacific Island of Kosrae. Wetlands Ecology and Management 9: 279–289. [Google Scholar]
- Alongi DM. 2008. Mangrove forests: resilience, protection from tsunamis, and responses to global climate change. Estuarine, Coastal and Shelf Science 76: 1–13. [Google Scholar]
- Amir AA. 2012. Canopy gaps and the natural regeneration of Matang mangroves. Forest Ecology and Management 269: 60–67. [Google Scholar]
- Amiro BD, Barr AG, Barr JD, et al. 2010. Ecosystem carbon dioxide fluxes after disturbance in forests of North America. Journal of Geophysical Research 115: G00K02. doi.org/ 10.1029/2010JG001390. [DOI] [Google Scholar]
- Aung TT, Mochida Y, Than MM. 2013. Prediction of recovery pathways of cyclone-disturbed mangroves in the mega delta of Myanmar. Forest Ecology and Management 293: 103–113. [Google Scholar]
- Baird AH, Bhalla RS, Kerr AM, Pelkey NW, Srinivas V. 2009. Do mangroves provide an effective barrier to storm surge? Proceedings of the National Academy of Sciences, USA 106: E111. doi: 10.1073/pnas.09008799106. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Baldwin AH, Platt WJ, Gathen KL, Lessmann JM, Rauch TJ. 1995. Hurricane damage and regeneration in fringe mangrove forests of southeast Florida, USA. Journal of Coastal Research SI 21: 169–183. [Google Scholar]
- Baldwin A, Egnotovich M, Ford M, Platt W. 2001. Regeneration in fringe mangrove forests damaged by Hurricane Andrew. Plant Ecology 157: 149–162. [Google Scholar]
- Ball MC, Farquhar GD. 1984. Photosynthetic and stomatal responses of two mangrove species, Aegiceras corniculatum and Avicennia marina, to long term salinity and humidity conditions. Plant Physiology 74: 1–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Barr JG, Engel V, Smith TJ III, Fuentes JD. 2012. Hurricane disturbance and recovery of energy balance, CO2 fluxes and canopy structure in a mangrove forest of the Florida Everglades. Agricultural and Forest Meteorology 153: 54–66. [Google Scholar]
- Bashan Y, Moreno M, Salazar BG, Alvarez L. 2013. Restoration and recovery of hurricane-damaged mangroves using the knickpoint retreat effect and tides as dredging tools. Journal of Environmental Management 116: 196–203. [DOI] [PubMed] [Google Scholar]
- Bischof BG. 1995. Aerial photographic analysis of coastal and estuarine mangrove system dynamics of the Everglades National Park, Florida, in response to hurricanes: implications for the continuing sea-level rise. MS Thesis, University of Miami. Coral Gables, FL. [Google Scholar]
- Björkman O, Demmig B, Andrews TJ. 1988. Mangrove photosynthesis: response to high-irradiance stress. Australian Journal of Plant Physiology 15: 43–61. [Google Scholar]
- Borchert SM, Osland MJ, Enwright NM, Griffith KT. 2018. Coastal wetland adaptation to sea-level rise: quantifying the potential for landward migration and coastal squeeze in northern Gulf of Mexico estuaries. Journal of Applied Ecology 55: 2876–2887. [Google Scholar]
- Bouillon S, Borges AV, Castañeda-Moya E, et al. 2008. Mangrove production and carbon sinks: a revision of global budget estimates. Global Biogeochemical Cycles 22: GB2013. doi: 10.1029/2007GB003052. [DOI] [Google Scholar]
- Breithaupt JL, Smoak JM, Smith TJ III, Sanders CJ. 2014. Temporal variability of carbon and nutrient burial, sediment accretion, and mass accumulation over the past century in a carbonate platform mangrove forest of the Florida Everglades. Journal of Geophysical Research: Biogeosciences 119: 2032–2048. [Google Scholar]
- Breithaupt JL, Smoak JM, Rivera-Monroy VH, et al. 2017. Partitioning the relative contributions of organic matter and mineral sediment to accretion rates in carbonate platform mangrove soils. Marine Geology 390: 170–180. [Google Scholar]
- Brinkman RM, Massel SR, Ridd PV, Furakawa K. 1997. Surface wave attenuation in mangrove forests. In: Proceedings of the 13th Australasian coastal and ocean engineering conference and the 6th Australasian port and harbour conference, Volume 2 Christchurch, NZ: Centre for Advanced Engineering, University of Canterbury. [Google Scholar]
- Brokaw NVL, Grear JS. 1991. Forest structure before and after Hurricane Hugo at three elevations in the Luquillo Mountains, Puerto Rico. Biotropica 23: 386–392. [Google Scholar]
- Brokaw NVL, Walker LR. 1991. Summary of the effects of Caribbean hurricanes on vegetation. Biotropica 23: 442–447. [Google Scholar]
- Cahoon DR, Hensel P, Rybczyk J, McKee KL, Proffitt CE, Perez BC. 2003. Mass tree mortality leads to mangrove peat collapse at Bay Islands, Honduras after Hurricane Mitch. Journal of Ecology 91: 1093–1105. [Google Scholar]
- Castañeda-Moya E, Twilley RR, Rivera-Monroy VH, Zhang K, Davis SE III, Ross M. 2010. Sediment and nutrient deposition associated with Hurricane Wilma in mangroves of the Florida Coastal Everglades. Estuaries and Coasts 33: 45–58. [Google Scholar]
- Castañeda-Moya E, Twilley RR, Rivera-Monroy VH. 2013. Allocation of biomass and net primary productivity of mangrove forests along environmental gradients in the Florida Coastal Everglades, USA. Forest Ecology and Management 307: 226–241. [Google Scholar]
- Chapman VJ. 1976. Mangrove vegetation. Germany: J. Cramer. [Google Scholar]
- Cheeseman JM. 1991. The analysis pf photosynthetic performance in leaves under field conditions: a case study using Bruguiera mangroves. Photosynthesis Research 29: 11–22. [DOI] [PubMed] [Google Scholar]
- Chen H, Lu W, Yan G, Yang S, Lin G. 2014. Typhoons exert significant but differential impacts on net ecosystem carbon exchange of subtropical mangrove forests in China. Biogeosciences 11: 5323–5333. [Google Scholar]
- Chen R, Twilley RR. 1999. A simulation model of organic matter and nutrient accumulation in mangrove wetland soils. Biogeochemistry 44: 93–118. [Google Scholar]
- Christensen JH, Kumar KK, Aldrian W, et al. 2013. Climate phenomena and their relevance for future regional climate change. In: Stocker TF, Qin D, Plattner GK, et al, eds. Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change. Cambridge, UK: Cambridge University Press. [Google Scholar]
- Clarke PJ, Kerrigan RA, Westphal CJ. 2001. Dispersal potential and early growth in 14 tropical mangroves: do early life history traits correlate with patterns of adult distribution? Journal of Ecology 89: 648–659. [Google Scholar]
- Clough BF, Sim RG. 1989. Changes in gas exchange characteristics and water use efficiency of mangroves in response to salinity and vapour pressure deficit. Oecologia 79: 38–44. [DOI] [PubMed] [Google Scholar]
- Coutts MP. 1986. Components of tree stability in Sitka spruce on peaty gley soil. Forestry 59: 173–195. [Google Scholar]
- Craighead FC, Gilbert VC. 1962. The effects of Hurricane Donna on the vegetation of southern Florida. Quarterly Journal of the Florida Academy of Science 25: 1–28. [Google Scholar]
- Dahdouh-Guebas F, Jayatissa LP. 2009. A bibliometrical review on pre- and post-tsunami assumptions and facts about mangroves and other coastal vegetation as protective buffers. Ruhuna Journal of Science 4: 28–50. [Google Scholar]
- Dahdouh-Guebas F, Jayatissa LP, Di Nitto D, Bosire JO, Lo Seen D, Koedam N. 2005. How effective were mangroves as a defence against the recent tsunami? Current Biology 15: R443–R447. [DOI] [PubMed] [Google Scholar]
- Danielsen F, Sørensen MK, Olwig MF, et al. 2005. The Asian tsunami: a protective role for coastal vegetation. Science 310: 643. [DOI] [PubMed] [Google Scholar]
- Danielson TM, Rivera-Monroy VH, Castañeda-Moya E, et al. 2017. Assessment of Everglades mangrove forest resilience: implications for above-ground net primary productivity and carbon dynamics. Forest Ecology and Management 404: 115–125. [Google Scholar]
- Das S, Vincent JR. 2009. Mangroves protected villages and reduced death toll during Indian super cyclone. Proceedings of the National Academy of Sciences, USA 106: 7357–7360. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Davis SE III, Cable JE, Childers DL, et al. 2004. Importance of storm events in controlling ecosystem structure and function in a Florida gulf coast estuary. Journal of Coastal Research 20: 1198–1208. [Google Scholar]
- Dixon HH, Joly J. 1894. On the ascent of sap. Philosophical Transactions of the Royal Society B: Biological Sciences 186: 563–576. [Google Scholar]
- Donato DC, Kauffman JB, Murdiyarso D, Kurnianto S, Stidham M, Kanninen M. 2011. Mangroves among the most carbon-rich forests in the tropics. Nature Geoscience 4: 293–297. [Google Scholar]
- Doyle TW, Girod GF. 1997. The frequency and intensity of Atlantic hurricanes and their influence on the structure of south Florida mangrove communities. In: Diaz HF, Pulwarty RS, eds. Hurricanes: climate and socioeconomic impacts. Berlin Heidelberg: Springer, 109–120. [Google Scholar]
- Doyle TW, Smith TJ III, Robblee MB. 1995. Wind damage effects of Hurricane Andrew on mangrove communities along the southwest coast of Florida, USA. Journal of Coastal Research Special Issue 21: 159–168. [Google Scholar]
- Doyle TW, Krauss KW, Wells CJ. 2009. Landscape analysis and patterns of hurricane impact and circulation on mangrove forests of the Everglades. Wetlands 29: 44–53. [Google Scholar]
- Drexler JZ. 2001. Maximum longevities of Rhizophora apiculata and R. mucronata propagules. Pacific Science 55: 17–22. [Google Scholar]
- Duke NC, Ball MC, Ellison JC. 1998. Factors influencing biodiversity and distributional gradients in mangroves. Global Ecology and Biogeography Letters 7: 27–47. [Google Scholar]
- Dutta D, Das PK, Paul S, Sharma JR, Dadhwal VK. 2015. Assessment of ecological disturbance in the mangrove forest of Sundarbans caused by cyclones using MODIS time-series data (2001–2011). Natural Hazards 79: 775–790. [Google Scholar]
- Ellison JC. 1998. Impacts of sediment burial on mangroves. Marine Pollution Bulletin 37: 420–426. [Google Scholar]
- Enwright NM, Griffith KT, Osland MJ. 2016. Barriers to and opportunities for landward migration of coastal wetlands with sea-level rise. Frontiers in Ecology and the Environment 14: 307–316. [Google Scholar]
- Farquhar GD, Ball MC, von Caemmerer S, Roksandic Z. 1982. Effect of salinity and humidity on 13C values of halophytes – evidence for diffusional isotopic fractionation determined by the ratio of intercellular/atmospheric CO2 under different environmental conditions. Oecologia 52: 121–137. [DOI] [PubMed] [Google Scholar]
- Feher LC, Osland MJ, Anderson GH, Vervaeke WC, Krauss KW, Whelan KRT, Balentine KM, Tiling-Range G, Smith III TJ, Cahoon DR. 2019. The long-term effects of hurricanes Wilma and Irma on soil elevation change in Everglades mangrove forests. Ecosystems doi: 10.1007/s10021-019-00446-x [DOI] [Google Scholar]
- Feller IC, Dangremond EM, Devlin DJ, Lovelock CE, Proffitt CE, Rodriguez W. 2015. Nutrient enrichment intensifies hurricane impact in scrub mangrove ecosystems in the Indian River Lagoon, Florida, USA. Journal of Ecology 96: 2960–2972. [DOI] [PubMed] [Google Scholar]
- Feller IC, Friess DA, Krauss KW, Lewis RR III. 2017. The state of the world’s mangroves in the 21st century under climate change. Hydrobiologia 803: 1–12. [Google Scholar]
- Ferwerda JG, Ketner P, McGuinness KA. 2007. Differences in regeneration between hurricane damaged and clear-cut mangrove stands 25 years after clearing. Hydrobiologia 591: 35–45. [Google Scholar]
- Friess DA, Rogers K, Lovelock CE, et al. 2019. The state of the world’s mangrove forests: past, present, and future. Annual Review of Environment and Resources 44: 16.1–16.27. [Google Scholar]
- Gedan KB, Kirwan ML, Wolanski E, Barbier EB, Silliman BR. 2011. The present and future role of coastal wetland vegetation in protecting shorelines: answering recent challenges to the paradigm. Climatic Change 106: 7–29. [Google Scholar]
- Gill AM, Tomlinson PB. 1969. Studies on the growth of red mangrove (Rhizophora mangle L.): I. Habitat and general morphology. Biotropica 1: 1–9. [Google Scholar]
- Giri C, Ochieng E, Tieszen LL, et al. 2011. Status and distribution of mangrove forests of the world using earth observation satellite data. Global Ecology and Biogeography 20: 154–159. [Google Scholar]
- Hamilton SE, Casey D. 2016. Creation of a high spatio-temporal resolution global database of continuous mangrove forest cover for the 21st century (CGMFC-21). Global Ecology and Biogeography 25: 729–738. [Google Scholar]
- Harmon ME, Hua C. 1991. Coarse woody debris dynamics in two old-growth ecosystems. BioScience 41: 604–610. [Google Scholar]
- Harmon ME, Franklin JF, Swanson FJ, et al. 1986. Ecology of coarse woody debris in temperate ecosystems. Advances in Ecological Research 15: 133–302. [Google Scholar]
- Harris RJ, Milbrandt EC, Everham EM III, Bovard BD. 2010. The effects of reduced tidal flushing on mangrove structure and function across a disturbance gradient. Estuaries and Coasts 33: 1176–1185. [Google Scholar]
- Helms JA, ed. 1998. The dictionary of forestry. Bethesda, MD: The Society of American Foresters. [Google Scholar]
- House Document 1965. Letter from the Secretary of the Army, United States Army Corps of Engineers, Morgan City and vicinity, Louisiana. United States Congress Serial Set, 1965–1966, Vol. 14-2, Serial 12690-2, House Document No. 167. [Google Scholar]
- Imbert D, Labbé P, Rousteau A. 1996. Hurricane damage and forest structure in Guadeloupe, French West Indies. Journal of Tropical Ecology 12: 663–680. [Google Scholar]
- Jones MC, Wingard GL, Stackhouse B, et al. 2019. Rapid inundation of southern Florida coastline despite low relative sea-level rise rates during the late-Holocene. Nature Communications 10: 3231. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Jordan CL, Hurt DA, Lowrey CA. 1960. On the structure of Hurricane Daisy on 27 August 1958. Journal of Meteorology 17: 337–348. [Google Scholar]
- Kamthonkiat D, Rodfai C, Saiwanrungkul A, Koshimura S, Matsuoka M. 2011. Geoinformatics in mangrove monitoring: damage and recovery after the 2004 Indian Ocean tsunami in Phang Nga, Thailand. Natural Hazards and Earth System Science 11: 1851–1862. [Google Scholar]
- Kauffman JB, Cole TG. 2010. Micronesian mangrove forest structure and tree responses to a severe typhoon. Wetlands 30: 1077–1084. [Google Scholar]
- Kennedy JP, Pil MW, Proffitt CE, Boeger WA, Stanford AM, Devlin DJ. 2016. Postglacial expansion pathways of red mangrove, Rhizophora mangle, in the Caribbean Basin and Florida. American Journal of Botany 103: 260–276. [DOI] [PubMed] [Google Scholar]
- Knapp KR, Kruk MC, Levinson DH, Diamond HJ, Neumann CJ. 2010. The International Best Track Archive for Climate Stewardship (IBTrACS): unifying tropical cyclone best track data. Bulletin of the American Meteorological Society 91: 363–376. [Google Scholar]
- Knutson TR, McBride JL, Chan J, et al. 2010. Tropical cyclones and climate change. Nature Geoscience 3: 157–163. [Google Scholar]
- Knutson TR, Sirutis JJ, Zhao M, et al. 2015. Global projections of intense tropical cyclone activity for the late twenty-first century from dynamical downscaling of CMIP5/RCP4.5 scenarios. Journal of Climate 28: 7203–7224. [Google Scholar]
- Kominoski JS, Gaiser EE, Baer SG. 2018. Advancing theories of ecosystem development through long-term ecological research. BioScience 68: 554–562. [Google Scholar]
- Kossin JP, Hall T, Knutson T, et al. 2017. Extreme storms. In: Wuebbles DJ, Fahey DW, Hibbard KA, Dokken DJ, Stewart BC, Maycock TK, eds. Climate Science Special Report: Fourth National Climate Assessment, Volume I. Washington, DC: U.S. Global Change Research Program, 375–404. [Google Scholar]
- Kovacs JM, Blanco-Correa M, Flores-Verdugo F. 2001. A logistic regression model of hurricane impacts in a mangrove forest of the Mexican Pacific. Journal of Coastal Research 17: 30–37. [Google Scholar]
- Krauss KW, Allen JA. 2003. Influence of salinity and shade on seedling photosynthesis and growth of two mangrove species, Rhizophora mangle and Bruguiera sexangula, introduced to Hawaii. Aquatic Botany 77: 311–324. [Google Scholar]
- Krauss KW, Doyle TW, Twilley RR, Smith TJ III, Whelan KRT, Sullivan JK. 2005. Woody debris in the mangrove forests of south Florida. Biotropica 37: 9–15. [Google Scholar]
- Krauss KW, Lovelock CE, McKee KL, et al. 2008. Environmental drivers in mangrove establishment and early development: a review. Aquatic Botany 89: 105–127. [Google Scholar]
- Krauss KW, Doyle TW, Doyle TJ, et al. 2009. Water level observations in mangrove swamps during two hurricanes in Florida. Wetlands 29: 142–149. [Google Scholar]
- Krauss KW, McKee KL, Lovelock CE, et al. 2014. How mangrove forests adjust to rising sea level. New Phytologist 202: 19–34. [DOI] [PubMed] [Google Scholar]
- Krauss KW, Barr JG, Engel V, Fuentes JD, Wang H. 2015. Approximations of stand water use versus evapotranspiration from three mangrove forests in southwest Florida, USA. Agricultural and Forest Meteorology 213: 291–303. [Google Scholar]
- Lamborn RH. 1890. The knees of the bald cypress: a new theory of their function. Science 15: 65–67. [DOI] [PubMed] [Google Scholar]
- Landsea C, Franklin J, Beven J. 2015. The Revised Atlantic Hurricane Database (HURDAT2). NOAA National Hurricane Center; Available online at: nhc.noaa.gov/data/#hurdat. [Google Scholar]
- Lang’at JKS, Kairo JG, Mencuccini M, et al. 2014. Rapid losses of surface elevation following tree girdling and cutting in tropical mangroves. PLoS One 9: e107868. doi: 10.1371/journal.pone.0118334. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lee SY, Primavera JH, Dahdouh-Guebas F, et al. 2014. Ecological role and services of tropical mangrove ecosystems: a reassessment. Global Ecology and Biogeography 23: 726–743. [Google Scholar]
- Lewis RR III, Milbrandt EC, Brown B, et al. 2016. Stress in mangrove forests: early detection and preemptive rehabilitation are essential for future successful worldwide mangrove forest management. Marine Pollution Bulletin 109: 764–771. [DOI] [PubMed] [Google Scholar]
- Liu H, Zhang K, Li Y, Xie L. 2013. Numerical study of the sensitivity of mangroves in reducing storm surge and flooding to hurricane characteristics in southern Florida. Continental Shelf Research 64: 51–65. [Google Scholar]
- Long J, Giri C, Primavera J, Trivedi M. 2016. Damage and recovery assessment of the Philippines’ mangroves following Super Typhoon Haiyan. Marine Pollution Bulletin 109: 734–743. [DOI] [PubMed] [Google Scholar]
- López-Hoffman L, Anten NPR, Martínez-Ramos M, Ackerly DD. 2007. Salinity and light interactively affect neotropical mangrove seedlings at the leaf and whole plant levels. Oecologia 150: 545–556. [DOI] [PubMed] [Google Scholar]
- Lovelock CE, Feller IC, Adame MF, et al. 2011. Intense storms and the delivery of materials that relieve nutrient limitations in mangroves of an arid zone estuary. Functional Plant Biology 38: 514–522. [DOI] [PubMed] [Google Scholar]
- Lovelock CE, Krauss KW, Osland MJ, Reef R, Ball MC. 2016. The physiology of mangrove trees with changing climate. In: Goldstein G, Santiago LS, eds. Tropical tree physiology: adaptations and responses in a changing environment. Cham, Switzerland: Springer, 149–179. [Google Scholar]
- Lowe JK, Gregory JM, Flather RA. 2001. Changes in the occurrence of storm surges around the United Kingdom under a future climate scenario using a dynamic storm surge model driven by the Hadley Centre climate models. Climate Dynamics 18: 179–188. [Google Scholar]
- Lugo AE. 1986. Mangrove understory: an expensive luxury? Journal of Tropical Ecology 2: 287–288. [Google Scholar]
- Lugo AE. 2000. Effects and outcomes of Caribbean hurricanes in a climate change scenario. The Science of the Total Environment 262: 243–251. [DOI] [PubMed] [Google Scholar]
- Lugo AE, Snedaker SC. 1974. The ecology of mangroves. Annual Review of Ecology and Systematics 5: 39–64. [Google Scholar]
- Lugo AE, Cintrón G, Goenaga C. 1981. Mangrove ecosystems under stress. In: Barret GW, Rosenberg R, eds. Stress effects on natural ecosystems. Chichester, UK: John Wiley and Sons Limited, 129–153 [Google Scholar]
- Lugo AE, Applefield M, Pool DJ, McDonald RB. 1983. The impact of Hurricane David on the forests of Dominica. Canadian Journal of Forest Research 13: 201–211. [Google Scholar]
- Macamo CCF, Massuanganhe E, Nicolau DK, Bandeira SO, Adams JB. 2016. Mangrove’s response to cyclone Eline (2000): what is happening 14 years later. Aquatic Botany 134: 10–17. [Google Scholar]
- Mazda Y, Magi M, Kogo M, Hong PN. 1997a Mangroves as a coastal protection from waves in the Tong King Delta, Vietnam. Mangroves and Salt Marshes 1: 127–135. [Google Scholar]
- Mazda Y, Wolanski E, King B, Sase A, Ohtsuka D, Magi M. 1997b Drag force due to vegetation in mangrove swamps. Mangroves and Salt Marshes 1: 193–199. [Google Scholar]
- McCoy ED, Mushinsky HR, Johnson D, Meshaka WE Jr. 1996. Mangrove damage caused by Hurricane Andrew on the southwestern coast of Florida. Bulletin of Marine Science 59: 1–8. [Google Scholar]
- McGuinness KA. 1996. Dispersal, establishment and survival of Ceriops tagal propagules in a north Australian mangrove forest. Oecologia 109: 80–87. [DOI] [PubMed] [Google Scholar]
- McKee KL, Cahoon DR, Feller IC. 2007. Caribbean mangroves adjust to rising sea level through biotic controls on change in soil elevation. Global Ecology and Biogeography 16: 545–556. [Google Scholar]
- Middleton BA, McKee KL. 2001. Degradation of mangrove tissues and implications for peat formation in Belizean island forests. Journal of Ecology 89: 818–828. [Google Scholar]
- Myers RK, van Lear DH. 1998. Hurricane–fire interactions in coastal forests of the south: a review and hypothesis. Forest Ecology and Management 103: 265–276. [Google Scholar]
- Narayan S, Suzuki T, Stive MJF, Verhagen HJ, Ursem WNJ, Ranasinghe R. 2010. On the effectiveness of mangroves in attenuating cyclone-induced waves. In: McKee-Smith J, Lynett P, eds. Proceedings of 32nd conference on coastal engineering, No. 32. Shanghai, China: American Society of Civil Engineers. [Google Scholar]
- Novlan DJ, Gray WM. 1974. Hurricane-spawned tornadoes. Monthly Weather Review 102: 476–488. [Google Scholar]
- Odum EP. 1969. The strategy of ecosystem development. Science 164: 262–270. [DOI] [PubMed] [Google Scholar]
- Odum WE, Johannes RE. 1975. The response of mangrove to man-induced environmental stress. In: Wood EFJ, Johannes RE, eds. Tropical marine pollution. Amsterdam: Elsevier, 52–62. [Google Scholar]
- O’Halloran TL, Law BE, Goulden ML, et al. 2012. Radiative forcing of natural forest disturbances. Global Change Biology 18: 555–565. [Google Scholar]
- Osland MJ, Enwright N, Day RH, Doyle TW. 2013. Winter climate change and coastal wetland foundation species: salt marshes versus mangrove forests in the southeastern United States. Global Change Biology 19: 1482–1494. [DOI] [PubMed] [Google Scholar]
- Osland MJ, Feher LC, López-Portillo J, et al. 2018. Mangrove forests in a rapidly changing world: global change impacts and conservation opportunities along the Gulf of Mexico coast. Estuarine, Coastal and Shelf Science 214: 120–140. [Google Scholar]
- Paling EI, Kobryn HT, Humphreys G. 2008. Assessing the extent of mangrove change caused by Cyclone Vance in the eastern Exmouth Gulf, northwestern Australia. Estuarine, Coastal and Shelf Science 77: 603–613. [Google Scholar]
- Patricola CM, Wehner MF. 2018. Anthropogenic influences on major tropical cyclone events. Nature 563: 339–346. [DOI] [PubMed] [Google Scholar]
- Peltola H, Kellomäki S, Hassinen A, Granander M. 2000. Mechanical stability of Scots pine, Norway spruce and birch: an analysis of tree-pulling experiments in Finland. Forest Ecology and Management 135: 143–153. [Google Scholar]
- Peters DPC, Lugo AE, Chapin FS III, et al. 2011. Cross-system comparisons elucidate disturbance complexities and generalities. Ecosphere 2: Article 81. [Google Scholar]
- Peterson JM, Bell SS. 2012. Tidal events and salt-marsh structure influence black mangrove (Avicennia germinans) recruitment across and ecotone. Ecology 93: 1648–1658. [DOI] [PubMed] [Google Scholar]
- Peterson JM, Bell SS. 2015. Saltmarsh boundary modulates dispersal of mangrove propagules: implications for mangrove migration with sea-level rise. PLoS One 10: e0119128. doi: 10.1371/journal.pone.0119128. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Pinzón ZS, Ewel KC, Putz FE. 2003. Gap formation and seedling regeneration in mangrove forests in Micronesia. Journal of Tropical Ecology 19: 143–153. [Google Scholar]
- Piou C, Feller IC, Berger U, Chi F. 2006. Zonation patterns of Belizean offshore mangrove forests 41 years after a catastrophic hurricane. Biotropica 38: 365–374. [Google Scholar]
- Poret N, Twilley RR, Rivera-Monroy VH, Coronado-Molina C. 2007. Belowground decomposition of mangrove roots in Florida coastal Everglades. Estuaries and Coasts 30: 491–496. [Google Scholar]
- Press MC, Brown ND, Barker MG, Zipperlen SW. 1996. Photosynthetic responses to light in tropical rain forest tree seedlings. In: Swaine MD, ed. The ecology of tropical forest tree seedlings. New York: The Parthenon Publishing Group, 41–58. [Google Scholar]
- Primavera JH, de la Crux M, Montilijo C, et al. 2016. Preliminary assessment of post-Haiyan mangrove damage and short-term recovery in Eastern Samar, central Philippines. Marine Pollution Bulletin 109: 744–750. [DOI] [PubMed] [Google Scholar]
- Rabinowitz D. 1978. Dispersal properties of mangrove propagules. Biotropica 10: 47–57. [Google Scholar]
- Reef R, Lovelock CE. 2015. Regulation of water balance in mangroves. Annals of Botany 115: 385–395. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Resio DT, Westerink JJ. 2008. Modeling the physics of storm surges. Physics Today 61: 33–38. [Google Scholar]
- Rivera-Monroy VH, Twilley RR, Davis SE III, et al. 2011. The role of the Everglades Mangrove Ecotone Region (EMER) in regulating nutrient cycling and wetland productivity in south Florida. Critical Reviews in Environmental Science and Technology 41: 633–669. [Google Scholar]
- Robertson AI, Daniel PA. 1989. Decomposition and the annual flux of detritus from fallen timber in tropical mangrove forests. Limnology and Oceanography 34: 640–646. [Google Scholar]
- Romero LM, Smith TJ III, Fourqurean JW. 2005. Change in mass and nutrient content of wood during decomposition in a south Florida mangrove forest. Journal of Ecology 93: 618–631. [Google Scholar]
- Ross MS, Ruiz PL, Sah JP, Reed DL, Walters J, Meeder JF. 2006. Early post-hurricane stand development in fringe mangrove forests of contrasting productivity. Plant Ecology 185: 283–297. [Google Scholar]
- Roth LC. 1992. Hurricanes and mangrove regeneration: effects of Hurricane Joan, October 1988, on the vegetation of Isla del Venado, Bluefields, Nicaragua. Biotropica 24: 375–384. [Google Scholar]
- Rovai AS, Twilley RR, Castañeda-Moya E, et al. 2018. Global controls on carbon storage in mangrove soils. Nature Climate Change 8: 534–538. [Google Scholar]
- Saenger P. 2002. Mangrove ecology, silviculture, and conservation. Dordrecht, The Netherlands: Kluwer Academic Publishers. [Google Scholar]
- Saenger P, Snedaker SC. 1993. Pantropical trends in mangrove above-ground biomass and annual litterfall. Oecologia 96: 293–299. [DOI] [PubMed] [Google Scholar]
- Saintilan N, Wilson NC, Rogers K, Rajkaran A, Krauss KW. 2014. Mangrove expansion and salt marsh decline at mangrove poleward limits. Global Change Biology 20: 147–157. [DOI] [PubMed] [Google Scholar]
- Salmo SG, Lovelock CE, Duke NC. 2014. Assessment of vegetation and soil characteristics in restored mangroves interrupted by severe tropical typhoon ‘Chan-hom’ in the Philippines. Hydrobiologia 733: 85–102. [Google Scholar]
- Sellier D, Fourcaud T. 2009. Crown structure and wood properties: influence on tree sway and response to high winds. American Journal of Botany 96: 885–896. [DOI] [PubMed] [Google Scholar]
- Shea DJ, Gray WM. 1973. The hurricane’s inner core region. I. Symmetric and asymmetric structure. Journal of Atmospheric Science 30: 1544–1564. [Google Scholar]
- Sheng YP, Lapetina A, Ma G. 2012. The reduction of storm surge by vegetation canopies: three-dimensional simulations. Journal of Geophysical Research 39: L20601. [Google Scholar]
- Sherman RE, Fahey TJ, Battles JJ. 2000. Small-scale disturbance and regeneration dynamics in a neotropical mangrove forest. Journal of Ecology 88: 165–178. [Google Scholar]
- Sherman RE, Fahey TJ, Martinez P. 2001. Hurricane impacts on a mangrove forest in the Dominican Republic: damage patterns and early recovery. Biotropica 33: 393–408. [Google Scholar]
- Sherrod CL, McMillan C. 1985. The distributional history and ecology of mangrove vegetation along the Northern Gulf of Mexico coastal region. Contributions in Marine Science 28: 129–140. [Google Scholar]
- Simard M, Fatoyinbo L, Smetanka C, et al. 2019. Mangrove canopy height globally related to precipitation, temperature and cyclone frequency. Nature Geoscience 12: 40–45. [Google Scholar]
- Sippo JZ, Lovelock CE, Santos IR, Sanders CJ, Maher DT. 2018. Mangrove mortality in a changing climate: an overview. Estuarine, Coastal and Shelf Science 215: 241–249. [Google Scholar]
- Smith TJ III, Robblee MB, Wanless HR, Doyle TW. 1994. Mangroves, hurricanes, and lightning strikes. BioScience 44: 256–262. [Google Scholar]
- Smith TJ III, Anderson GH, Balentine K, Tiling G, Ward GA, Whelan KRT. 2009. Cumulative impacts of hurricanes on Florida mangrove ecosystems: sediment deposition, storm surges and vegetation. Wetlands 29: 24–34. [Google Scholar]
- Smoak JM, Breithaupt JL, Smith TJ III, Sanders CJ. 2013. Sediment accretion and organic carbon burial relative to sea-level rise and storm events in two mangrove forests in Everglades National Park. Catena 104: 58–66. [Google Scholar]
- Snedaker SC. 1995. Mangroves and climate change in the Florida and Caribbean region: scenarios and hypotheses. Hydrobiologia 295: 43–49. [Google Scholar]
- Snedaker SC. 2002. Mangrove ecology, silviculture, and conservation. Dordrecht, The Netherlands: Kluwer Academic Publishers. [Google Scholar]
- Snedaker SC, Lahmann EJ. 1988. Mangrove overstory absence: a consequence of evolution? Journal of Tropical Ecology 4: 311–314. [Google Scholar]
- Sobel AH, Camargo SJ, Hall TM, Lee C-Y, Tippett MK, Wing AA. 2016. Human influence on tropical cyclone intensity. Science 353: 242–246. [DOI] [PubMed] [Google Scholar]
- Steinke TD, Ward CJ. 1989. Some effects of the cyclones Domoina and Imboa on mangrove communities in the St Lucia Estuary. South African Journal of Botany 55: 340–348. [Google Scholar]
- Stocker GC. 1976. Report on cyclone damage to natural vegetation in the Darwin Area after Cyclone Tracey, 25 December 1974 [Northern Territory]. Leaflet No. 127, Canberra, Australia: Forestry and Timber Bureau. [Google Scholar]
- Stoddart DR. 1969. Post-hurricane changes on the British Honduras reefs and cays: re-survey of 1965. Atoll Research Bulletin No. 131, Washington, DC: The Smithsonian Institute. [Google Scholar]
- von Storch H, Woth K. 2008. Storm surges: perspectives and options. Sustainability Science 3: 33–43. [Google Scholar]
- Tomlinson PB. 2016. The botany of mangroves, 2nd edn Cambridge, UK: Cambridge University Press. [Google Scholar]
- Twilley RR, Rovai AS, Riul P. 2018. Coastal morphology explains global blue carbon distributions. Frontiers in Ecology and the Environment 16: 503–508. [Google Scholar]
- Vandermeer J, de la Cerda IG, Boucher D, Perfecto I, Ruiz J. 2000. Hurricane disturbance and tropical tree species diversity. Science 290: 788–791. [DOI] [PubMed] [Google Scholar]
- Van der Stocken T, Carroll D, Menemenlis D, Simard M, Koedam N. 2019a Global-scale dispersal and connectivity in mangroves. Proceedings of the National Academy of Sciences, USA 116: 915–922. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Van der Stocken T, Wee AK, De Ryck DJ, et al. 2019. b A general framework for propagule dispersal in mangroves. Biological Reviews 94: 1547–1575. [DOI] [PubMed] [Google Scholar]
- Vogt J, Skóra A, Feller IC, Piou C, Coldren G, Berger U. 2012. Investigating the role of impoundment and forest structure on the resistance and resilience of mangrove forests to hurricanes. Aquatic Botany 97: 24–29. [Google Scholar]
- Wadsworth FH, Englerth GH. 1959. Effects of the 1956 hurricane on forests in Puerto Rico. Caribbean Forester 20: 38–51. [Google Scholar]
- Walker LR. 1995. Timing of post-hurricane tree mortality in Puerto Rico. Journal of Tropical Ecology 11: 315–320. [Google Scholar]
- Wanless HR, Vlaswinkel BM. 2005. Coastal landscape and channel evolution affecting critical habitats at Cape Sable, Everglades National Park, Florida. Final Report to Everglades National Park. [Google Scholar]
- Ward RD, Friess DA, Day RH, MacKenzie RA. 2016. Impacts of climate change on mangrove ecosystems: a region by region overview. Ecosystem Health and Sustainability 2: e01211. doi: 10.1002/ehs2.1211. [DOI] [Google Scholar]
- Wells NC. 2012. The atmosphere and ocean, 3rd edn, Chichester, UK: John Wiley & Sons, Ltd. [Google Scholar]
- Whelan KRT. 2005. The successional dynamics of lightning-initiated canopy gaps in the mangrove forests of Shark River, Everglades National Park, USA. PhD Dissertation, Florida International University, Miami, FL. [Google Scholar]
- Whelan KRT, Smith TJ III, Anderson GH, Ouellette ML. 2009. Hurricane Wilma’s impact on overall soil elevation and zones within the soil profile in a mangrove forest. Wetlands 29: 16–23. [Google Scholar]
- Woodroffe CD, Grime D. 1999. Storm impact and evolution of a mangrove-fringed chenier plain, Shoal Bay, Darwin, Australia. Marine Geology 159: 303–321. [Google Scholar]
- Woodroffe CD, Grindrod J. 1991. Mangrove biogeography: the role of quaternary environmental and sea-level change. Journal of Biogeography 18: 479–492. [Google Scholar]
- Woodroffe CD, Rogers K, McKee KL, Lovelock CE, Mendelssohn I, Saintilan N. 2016. Mangrove sedimentation and response to relative sea-level rise. Annual Review of Marine Science 8: 243–266. [DOI] [PubMed] [Google Scholar]
- Zhang K, Liu H, Li Y, et al. 2012. The role of mangroves in attenuating storm surges. Estuarine, Coastal and Shelf Science 102–103: 11–23. [Google Scholar]
- Zhang K, Thapa B, Ross M, Gann D. 2016. Remote sensing of seasonal changes and disturbances in mangrove forest: a case study from South Florida. Ecosphere 7: e01366. doi: 10.1002/ecs2.1366. [DOI] [Google Scholar]