Woody Plant Encroachment of Grasslands: a Comparison of Terrestrial and Wetland Settings

University of Wollongong Research Online

Faculty of Science, Medicine and Health - Papers: part A Faculty of Science, Medicine and Health

1-1-2015

Woody plant encroachment of grasslands: a comparison of terrestrial and wetland settings

Neil Saintilan University of Wollongong

Kerrylee Rogers University of Wollongong, [email protected]

Follow this and additional works at: https://ro.uow.edu.au/smhpapers

Part of the Medicine and Health Sciences Commons, and the Social and Behavioral Sciences Commons

Recommended Citation Saintilan, Neil and Rogers, Kerrylee, "Woody plant encroachment of grasslands: a comparison of terrestrial and wetland settings" (2015). Faculty of Science, Medicine and Health - Papers: part A. 2676. https://ro.uow.edu.au/smhpapers/2676

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: [email protected] Woody plant encroachment of grasslands: a comparison of terrestrial and wetland settings

Abstract A global trend of woody plant encroachment of terrestrial grasslands is co-incident with woody plant encroachment of wetland in freshwater and saline intertidal settings. There are several arguments for considering tree encroachment of wetlands in the context of woody shrub encroachment of grassland biomes. In both cases, delimitation of woody shrubs at regional scales is set by temperature thresholds for poleward extent, and by aridity within temperature limits. Latitudinal expansion has been observed for terrestrial woody shrubs and mangroves, following recent warming, but most expansion and thickening has been due to the occupation of previously water-limited grassland/saltmarsh environments. Increases in atmospheric CO2, may facilitate the recruitment of trees in terrestrial and wetland settings. Improved water relations, a mechanism that would predict higher soil moisture in grasslands and saltmarshes, and also an enhanced capacity to survive arid conditions, reinforces local mechanisms of change. The expansion of woody shrubs and mangroves provides a negative feedback on elevated atmospheric CO2 by increasing carbon sequestration in grassland and saltmarsh, and is a significant carbon sink globally. These broad-scale vegetation shifts may represent a new stable state, reinforced by positive feedbacks between global change drivers and endogenic mechanisms of persistence in the landscape.

Disciplines Medicine and Health Sciences | Social and Behavioral Sciences

Publication Details Saintilan, N. & Rogers, K. (2015). Woody plant encroachment of grasslands: a comparison of terrestrial and wetland settings. New Phytologist, 205 (3), 1062-1070.

This journal article is available at Research Online: https://ro.uow.edu.au/smhpapers/2676

Woody plant encroachment of grasslands: a comparison of terrestrial and wetland settings

Running Head: Woody plant encroachment of grasslands

Neil Saintilan1,2

1NSW Office of Environment and Heritage

Tel: +612 9995 5631

Fax: +612 9995 2342

Email: [email protected]

Kerrylee Rogers 2

GeoQuest, University of Wollongong

Tel: +612 4239 2512

Fax: +612 42214250

Email: [email protected]

WORD COUNT: 4325 words

TABLES 1

FIGURES 2

ABSTRACT

A global trend of woody plant encroachment of terrestrial grasslands is co‐incident with woody plant encroachment of wetland in freshwater and saline intertidal settings. There are several arguments for considering tree encroachment of wetlands in the context of woody shrub encroachment of grassland biomes. In both cases, delimitation of woody shrubs at regional scales is set by temperature thresholds for poleward extent, and by aridity within temperature limits. Latitudinal expansion has been observed for terrestrial woody shrubs and mangroves, following recent warming, but most expansion and thickening has been due to the occupation of previously water‐limited grassland/saltmarsh environments. Altered fire regime cannot be invoked as a driver of change in most wetland settings. Increases in atmospheric CO2, may facilitate recruitment of trees in terrestrial and wetland settings. Improved water relations, a mechanism that would predict higher soil moisture in grasslands and saltmarshes, and also an enhanced capacity to survive arid conditions, reinforces local mechanisms of change, such as sea‐level rise in the case of saltmarsh. The expansion of woody shrubs and mangroves provides a negative feedback on elevated atmospheric CO2 by increasing carbon sequestration in grassland and saltmarsh, and is a significant carbon sink globally. These broad‐scale vegetation shifts represent a new stable state, reinforced by positive feedbacks between global change drivers and endogenic mechanisms of persistence in the landscape.

KEYWORDS: atmospheric CO2; saltmarsh, shrub, encroachment, fire,

INTRODUCTION

A recent trend towards semi‐arid and mesic grassland conversion to shrub dominance is global in scale and common to all vegetated continents (Fig 1). Woody shrub encroachment and thickening has converted grassland to woodland in the savannahs and prairies of North America (Bai et al., 2009, Van Auken, 2009), South America (Silva et al., 2001), semi‐arid and tropical Australia (Brook & Bowman, 2006, Fensham et al., 2005), Africa (Sankaran et al., 2008), Europe (Maestre et al., 2009), India (Misra, 1983) and China (Peng et al., 2013), where over 5 million hectares of the inner Mongolian grassland has been encroached by Caragana microphylla (Zhang et al., 2006). In cold climates, alpine grasslands have been encroached by shrubs in the high Andes in South America (Matson & Bart, 2013), and the Iberian Peninsula and Pyrenees in Europe (Montané et al., 2010), while Tape et al. (2006) make a case for pan‐arctic shrub expansion into tundra on the basis of aerial photography dating to the 1940’s. Shrub thickets are expanding into grass dominated swales on Atlantic US coast barrier islands (Young et al., 2007).

Shrub encroachment of arid and semi‐arid grasslands in the United States has been estimated to affect 220‐ 330 million ha (Houghton et al., 1999, Pacala et al., 2001). In North America alone Knapp et al. (2008) describe a shift in graminoid to shrub dominance across eight sites spanning 5550 km, and climatic gradients extending from 200‐1100 mm mean annual precipitation and ‐12.5‐22.0°C mean annual temperature. For example, observations stretching back to 1858 in the latter case confirm a steady conversion of the prairie grass Black grama (Bouteloua eriopoda) to shrub dominance in the Jornada Basin in New Mexico (Gibbens et al., 2005). Grasses covered 98% of the Jornada Experimental Range in 1858, and less than 1% by the turn of the millennium. These changes were well advanced by 1928, and shrub eradication measures, including clearance and grazing regulation, have at best slowed the advance.

There are several published accounts of woody vegetation expansion into freshwater wetlands dominated by reeds and/or grasses. Wet Sphagnum–sedge fens have suffered an accelerating rate of loss to woody encroachment since the 1950’s in locations as diverse as the Alaskan Kenai Peninsula (Berg et al., 2009); small Appalachian mountain peatlands (Stine et al., 2011, Warren et al., 2007), and the Biebrza and Narew river valleys of Poland (Kotowski & Piórkowski, 2005). Trees had previously been excluded from the Alaskan fens for 18000 years (Berg et al., 2009). Wetland grasslands dominated by the C4 graminoid Pseudoraphis spinescens have been progressively replaced by river red gum Eucalyptus camaldulensis in Australia’s largest inland forested wetland, the Barmah‐Millewa forest (Bren, 1992, Colloff et al., 2014), and by Melaleuca quinquenervia on large coastal floodplains within SE Australia (Johnston et al., 2003) over a period of at least 70 years. Sawgrass (Cladium jamaicense) has been extensively colonised by red maple (Acer rubrum) in Florida (Duever, 2005, Knickerbocker et al., 2009) since the 1940’s. It is difficult to ascertain whether these are representative of broader trends across wetlands globally.

Contemporaneous with the expansion and thickening of woody shrubs in grassland biomes across the globe has been the expansion of mangrove into saltmarsh environments. Mangroves are woody trees and shrubs that occur exclusively in the marine environment, distributed by tides and colonising intertidal soft‐sediment

environments, primarily in tropical and subtropical latitudes. Along many of the world’s coastlines, mangroves co‐occur with saltmarshes, a floristically diverse group of low‐stature plants tolerant of saline, periodically saturated conditions, though not exclusively intertidal in distribution. Interactions between mangroves and saltmarshes have been the subject of growing interest; with widespread reports of transitions in boundaries corresponding to regional and global environmental change (Saintilan et al., 2014). On the US Gulf coast, mangrove advance into saltmarsh has been periodically arrested by frost in SE Louisiana (Giri et al., 2011), but further south in Texas and Florida stratigraphic and photographic evidence suggests a consistent expansion and thickening of mangrove over the past century (Bianchi et al., 2013, Smith III et al., 2013). Indeed, mangrove area in Florida has doubled at the northern end of their distribution since 1984 (Cavanaugh et al., 2014). Similar expansion has been observed on the Pacific coast of Mexico (López‐Medellín et al., 2011) and Peru (Saintilan et al., 2014).

Figure 1: Global distribution of grasslands, as sourced from Bailey’s Ecoregions Map of the World (Bailey, 1989). Pink points denote locations of cited shrub encroachment into grassland; citations are included in the table below.

Changes in the distribution of mangrove and saltmarshes on the Australian mainland further illustrate this trend. Mangrove encroachment of saltmarsh is near‐ubiquitous in the estuaries of SE Australia, occurring within estuaries along 4000 km of coastline from central Queensland to South Australia, encompassing the range of geomorphic and climatic settings within which the two communities co‐exist (Saintilan & Williams, 1999). Historical documents, including paintings, maps and plans of early Sydney suggest that the expansion of

mangrove into saltmarsh environments may have commenced in the 19th Century (McLoughlin, 1987, McLoughlin, 2000) when saltmarsh dominated now mangrove‐fringed shorelines. Mangroves are also encroaching saline and freshwater wetlands in the tropical north, prompting Williamson et al. (2011) to propose that mangrove encroachment of saltmarsh might be analogous to shrub encroachment of savannahs in the same region.

Interactions between woody vegetation and grasslands in terrestrial and wetland contexts have hitherto been considered in discrete bodies of literature reflecting the disciplinary boundaries of terrestrial, freshwater and marine ecology, with only two studies (D'Odorico et al., 2012, Williamson et al., 2011) drawing comparisons across the wetland‐terrestrial divide. This might not be surprising given the obvious differences in the growth habits of the species and the characteristics of the environments being considered, and that a strong prima facie case might be made for fire and hydrology as drivers of change in these contrasting settings. However, there are several compelling similarities between the woody shrub encroachment in common across these settings. For the most part, saltmarshes and floodplain wetlands are grasslands, and one of the few instances where C4 grasses extend into higher latitudes (Collatz et al., 1998). Many of the dominant saltmarsh and floodplain wetland species across the globe are members of the family Poaceae (including species of saltmarsh genera Spartina, Distichlis, Pragmites, Puccinellia, and Sporobolus, and in freshwater wetlands; Pseudoraphis and Paspalum). Though coastal wetlands are regularly inundated by tidal water, physiological water stress is imposed by the presence of salt, carried on tidal waters and concentrated through evaporation between inundation and rainfall events. Shrubs growing in saltmarsh environments therefore share many traits associated with plants in semi‐arid climates, including succulence (Chenopodiaceae) and tightly regulated transpiration promoting high water use efficiency in mangroves (Ball, 1996, Krauss et al., 2014), and stunted growth. On floodplain wetlands, hydrological drawdown can induce sharp competition over water between grasses and tree seedlings (Bren, 1992).

This review poses the question of whether tree expansion in freshwater and intertidal wetlands is best understood in the context of woody encroachment of grasslands observed to have occurred over the same period across nearly every other important grassland biome. If so, the case of wetlands may provide some important insights into global drivers of grassland loss, in that many of the putative drivers of change in coastal wetlands cannot apply to terrestrial grasslands and vice versa. For example, much of the debate over the drivers of tree encroachment in terrestrial grassland has focussed on the role of fire (Higgins et al., 2007, Wakeling et al., 2012), and yet fire is absent from many wetland types subject to tree encroachment considered in this review. Our thesis is that global drivers (elevated CO2 and temperature) work by modulating the response of woody plants to local drivers (such as fire and hydrology). Furthermore, similar feedbacks between encroachment, persistence and CO2 sequestration may apply across the spectrum of terrestrial and wetland settings.

TREE RANGE EXPANSION AND TEMPERATURE

Kaufman et al. (2009) document a cooling trend over two millennia in the arctic abruptly terminated at the time of the industrial revolution. This cooling was most probably associated with decreases in summer insolation that commenced following the Holocene thermal maximum some 7500 years ago. Over the latter Holocene, the tree line retreated and herbaceous tundra expanded in Eurasia (Fisher et al., 1995, Seppä & Birks, 2002). This sequence provides the backdrop against which historical observations should be interpreted and it is reasonable to suppose that woody shrub encroachment in sub‐polar regions following 1850 is related to climatic warming. Supporting evidence is found in the increase in radial and vertical growth of woody plants since 1860, and accelerating since 1970 coinciding with acceleration of warming in the region (Hallinger et al., 2010).

The arctic is not the only region where temperature controls on shrub expansion may be important. D’Odorico et al. (2012) note that most of the shrubs encroaching on arid grasslands are limited by low temperatures. For example, cold sensitivity has been reported for Prosopis glandulosa in south‐west Texas and New Mexico (Felker et al., 1982), Acacia mellifera encroaching savannah in southern Africa (Sekhwela & Yates, 2007) and Larrea tridentata in the Sonoran Desert, the northern limit of which corresponds to the ‐180 C winter isotherm, the temperature at which freezing‐induced cavitation and mortality occurs (Pockman & Sperry, 1996, Pockman & Sperry, 2000). Holocene climate variability over the region has influenced the relative extent of grassland and shrubland, with shrubs expanding during the Medieval Warm Period (Bradley et al., 2003), then contracting during the “Little Ice Age” (Neilson, 1986). The late 19th century would represent a minimum northern extent for the Holocene for terrestrial shrubs and mangroves, and some expansion may be due to range extension within new thermal limits.

Mangrove range expansion provides an interesting illustration of this trend. The tropical mangrove family Rhizophoraceae dominated the mangrove flora of the Richmond River, northern New South Wales, Australia, during the Holocene thermal maximum (Hashimoto et al., 2006). At the time of the first historical surveys following European colonisation the species was rare in the region (West et al., 1985), though has shown strong population growth and southward expansion over the past half‐century (Wilson, 2009), corresponding to changing temperature clines (Hennessy et al., 2004), and appears to be well within thermal limits on the basis of leaf phenology (Wilson & Saintilan, 2012). This observation concords with a noted global conservatism in mangrove range expansion in relation to temperature thresholds (Quisthoudt et al., 2012) suggesting dispersal constraints limit latitudinal range extension, particularly on wave‐dominated coastlines.

The mangrove genus Avicennia is able to withstand mild frosts and extends to higher latitudes than Rhizophora. A. marina contracted in latitudinal range during cooling that followed the mid‐Holocene thermal maxima in New Zealand (Mildenhall, 1994) and, in spite of strong population growth over the past four decades has not reclaimed former latitudes due probably to dispersal constraints (de Lange & De Lange, 1994). Geomorphic controls on dispersal are less of an issue on low‐wave energy open coastlines, such as the US Gulf coast. Here, A. germinans is periodically subjected to intense freezing‐induced mortality. The last major event

occurred more than 25 years ago, and A. germinans populations have shown strong recovery since, expanding into and replacing Spartina‐dominated saltmarsh (Giri et al., 2011), particularly at poleward limits (Cavanaugh et al. 2014). Thermal modelling based on IPCC projections suggests that in a medium‐high gas emissions scenario (A2) with 2°C‐4°C increase in mean annual minimum temperature, nearly 10 000 km2 of the herbaceous saltmarsh of the Gulf coast will be vulnerable to mangrove encroachment (Osland et al., 2013). This represents, as a percentage of the state total for saltmarsh, 95% for Louisiana, 100% for Texas, and 60% for Florida. Avicennia spp. expansion into herbaceaous saltmarsh at poleward limits has also been observed in Australia, South Africa and South America (Saintilan et al., 2014).

CO2 AND TREE ESCAPE FROM DISTURBANCE

While temperature is an important control of shrub/mangrove distribution at latitudinal and altitudinal limits, most expansion and thickening occurs well within the thermal limits of the species involved. Regional and sub‐ continental studies of shrub encroachment into grassland stress the importance of aridity gradients in determining shrub cover. For example, phases of woody shrub proliferation and contraction in semi‐arid Australia correspond to phases of wet and dry respectively (Fensham et al., 2005). Similarly, woody cover in the African savannah is strongly associated with Mean Annual Precipitation to a threshold of 650 mm per year, above which other factors appear to influence thickening (Sankaran et al., 2008), and alternate states of tree cover may correspond to fire frequency (Staver et al., 2011) . Frequent fire may induce a recruitment bottleneck, preventing saplings from “escaping” to a height immune from fire mortality (Kgope et al., 2010). Fire suppression, often a consequence of grazing (and reduced fuel loads), is often cited as a driver of tree encroachment and thickening in savannahs (Scholes & Archer, 1997, Van Auken, 2000).

Experimental studies have suggested that elevated CO2 may assist saplings in escaping the fire trap by increasing growth rate to a height above the fire threat (Kgope et al., 2010). Under elevated CO2, most notably between sub‐ambient and ambient concentrations, saplings accumulate greater below‐ground carbon stores, which, facilitated by post‐fire nutrient release; stimulates faster re‐sprouting; and increases the likelihood that the sapling will recruit to fire resistant height (Kgope et al., 2010). Fire regimes that excluded trees from grasslands a century ago may no longer do so (Wigley et al., 2010). Models of the relationship between CO2 concentration, sapling growth to “escape height” and fire frequency are consistent with paleo records in South Africa documenting a commencement of woody encroachment following the last glacial maximum and acceleration over the historical period (Bond et al., 2003). It is also consistent with lower recruitment in cooler, upland settings (Wakeling et al., 2012).

The “escape height” model (Bond & Midgley, 2000, Wakeling et al., 2012) developed explains the interaction between altered fire regime, CO2 fertilisation and tree recruitment for terrestrial grasslands can be adapted for freshwater wetlands (Fig 2b). In the absence of fire, flood return period is an analogous disturbance. E. camaldulensis encroachment of P. spinescens grasslands has been attributed to alterations in flooding regime,

as floods of 6‐month duration above seedling foliage height have been sufficient histtorically to repress red gum germination (Bren, 1992, Colloff et al., 2014). River regulation for irrigation has reduced depth and duration of large floods while providing more reliable base flow. We propose that both alterations to flooding regime, seedling survival, and growth rate, may combine to fundamentally alter recruitment success across the wetland. In this case, the race to achieve “escape height” between successive flood peeaks would be aided by

CO2 fertilisation, an effect facilitated by higher soil water nutrient loads resultting from agricultural developments in the catchment (Fig 2b).

represented as pre‐inddustrial, low CO2 (solid green line), and current (dotted line), adapted from Bond, Midgley and Woodward (2003).. (b) Conceptual model of the relationship between sapling growth rate and inundation regime on Pseudoraphis spinescens (Moira grass) plains. IInundation regime is represented as the return interval of 6‐monnth continual inundation at depths 0‐2.5 m, for natural (solid blue lline) and post‐regulation (dashed blue line) conditions. Growth rate is represented as pre‐development (solid

green line), and currennt (dotted line), following nutrient and CO2 fertilisation. (c) Concepptual model of the relationship between sapling rooting depth rrate and groundwater depth regime for a groundwater dependent wetland. In this case, current groundwater levels (sollid line) increase on historic norms (dashed line) because of sea‐level rise in coastal wetlands, and/or reduced transpiration losses. The recruitment bottleneck is avoided when rooting depth exceeds ssaturated groundwater depth. Growth rate is represented as currennt (dotted line), following nutrient and CO2 fertilisation, and previous (dashed line).

Studies across three continents have shown woody shrub encroachment in areas subject to a range of disturbance regimes from heavy fire and grazing to ungrazed and rarely burnt (Fensham et al., 2005, Hastings & Turner, 1965, Lenzi‐Grillini et al., 1996, Silva et al., 2001, Wigley et al., 2010). Woody encroachment in semi‐ arid Australian savannah showed no relationship with fire, or the interaction of fire and grazing, over large spatial and temporal scales (Fensham et al., 2005). Woody encroachment of tallgrass prairie has been noted to occur where fire intensities have increased (Briggs et al., 2005, Knapp et al., 1998). Shrub removal programs on the whole fail to stem the progression of shrub encroachment (D'Odorico et al., 2012, Rango et al., 2005), even when combined with changes in land‐use aimed at ameliorating these threats (Gibbens et al., 1992). Though fire is important on a local scale in many settings, the more pervasive effect of global drivers seem to override these effects at broader spatial and temporal scales.

CO2, ARIDITY AND DROUGHT TOLERANCE

Soil water has also been identified as a limiting factor regulating shrub encroachment mesic grasslands. Although precipitation is sufficient to allow the development of a closed forest, grassland water use in the upper soil imposes water limitation on seedling establishment (Scholes & Archer, 1997). A similar mechanism has been invoked to explain the dieback of E. camaldulensis seedlings in competition with Pseudoraphis spinescens wet grassland during hydrological drawdown (Bren, 1992). However, with an increase in the ambient to intracellular CO2 diffusion gradient comes a greater uptake of CO2 with a minimal reduction in stomatal conductance (Polley et al., 1997). Photosynthesis can thereby be maintained with lower water loss, and this mechanism has been proposed as a driver of shrub encroachment into grassland along aridity gradients (Idso, 1992).

Elevated CO2 reduces leaf conductance of rangeland shrubs (Morgan et al., 2004) and Spartina saltmarshes (Mateos‐Naranjo et al., 2010a, Mateos‐Naranjo et al., 2010b, Rozema et al., 1991), though this is offset to some extent by a negative feedback between transpiration rate, leaf temperature and ambient humidity (Morgan et al., 2004). None‐the less, evapotranspiration may be reduced by up to 20% in savannah ecosystems (Nelson et al., 2004, Polley et al., 2011) and saltmarsh sedgelands (Li et al., 2010). That decreased evapotranspiration is occurring over broad regional scales is supported by evidence of increased soil moisture (Carol Adair et al., 2011, Hungate et al., 2002) and run‐off (Labat et al., 2004) in areas where woody thickening is occurring (Eamus & Palmer, 2008). These improved soil moisture conditions may improve percolation to the lower rooting zone of rangeland shrubs, and allow survival of seedlings in mesic settings in competition with shallow‐rooting grasses. Elevated CO2 tends to facilitate higher growth rate and a higher proportional allocation of plant mass to roots (Nie et al., 2013), providing for an increased rate of root growth to critical depths for drought resistance, or the maintenance of the high root:shoot ratios required of mangroves in saline and hypersaline environments (Saintilan, 1997).

The “escape height” model describing CO2 effects on fire resistance therefore finds a parallel in “escape depth” to drought resistance (Fig 3c).

Improved mangrove seedling survival may also be promoted, especially following the depletion of cotyledon reserves (Farnsworth et al., 1996) when elevated CO2 has been shown to improve growth of Avicennia germinans (McKee & Rooth, 2008). The proportion of mangrove and saltmarsh can be predicted in northern Australia on the basis of rainfall (Bucher & Saenger, 1991), and field studies of seedling establishment suggest that mangroves are delimited by desiccation in the saltmarsh (Clarke & Myerscough, 1993). Mangrove encroachment of saltmarsh has been associated with higher water levels in Florida (Krauss et al., 2011, Smith III et al., 2013), eastern Australia (Rogers et al., 2006), Mexico (López‐ Medellín et al., 2011), and tropical Australia (Williamson et al., 2011). Phases of higher rainfall and sea‐level are often coincident, associated with inter‐annual and inter‐decadal phases of atmosphere‐oceanic cycles such as the El Nino Southern Oscillation and the Inter‐decadal Pacific Oscillation (Zhang & Church, 2012). These phases often correspond to phases in vegetation transition (Kim et al., 2010, Rogers et al., 2014). However, while some die‐back of shrubs during drought has been noted (Fensham et al., 2005, Sankaran et al., 2008), more commonly shrubs are retained and provide a platform for a successive phase of colonisation. This directional trend requires a press driver, and gives credence to the global effect of elevated CO2 in sustaining the effect of drivers operating in pulses, such as rainfall and sea‐level.

FEEDBACKS BETWEEN SHRUB ENCROACHMENT, LOCAL PERSISTENCE AND CO2 SEQUESTRATION

While exogenic factors operating at regional and global scales provide opportunities for woody shrub recruitment, several endogenic feedback mechanisms may promote persistence and further thickening (D'Odorico et al., 2012) in an alternative stable state. For example, the shrub architecture promotes the concentration of rainfall to the base of the stem, and improved infiltration and soil moisture near the base of trees (Eldridge & Freudenberger, 2005), conferring a competitive advantage over surrounding grasses. Soil nitrogen and carbon is enhanced in most settings by the presence of shrubs (Eldridge et al., 2011, Maestre et al., 2009), providing opportunities for clonal expansion (Ratajczak et al., 2011). In cold‐limiting environments, woody encroachment provides an insulating microclimate in both savannah and saltmarsh (D'Odorico et al., 2012), increasing winter nocturnal temperatures. The competitive replacement of grasslands reduces fuel loads, which improves recruitment of adult bushes (D’Odorico et al., 2006). Exogenic drivers interact with these feedbacks: CO2 reducing evapo‐transpiration and further enhancing soil moisture; warmer winters leading to greater snow coverage further insulating soils (Sturm et al., 2005).

The triggering of endogenic feedback mechanisms by exogenic global change drivers may lead to more persistent ecosystem states in the long‐term. For example, elevated CO2 may help intertidal wetland plants respond to the pressure of sea‐level rise by facilitating the production of below‐ground biomass, and through this, the building of marsh surface elevation. Experimental manipulation of CO2 in a mixed Schoenoplectus americanus (C3 sedge) and Spartina alterniflora (C4 grass) saltmarsh field site demonstrated increased growth of S. americanus, in elevated ambient CO2 with this species dominating surface elevation gain (Cherry et al.,

2009, Langley et al., 2009). Experimental studies on CO2 and N effects on root mass suggest similar responses might be expected in mangroves. Avicennia germinans root mass increases under elevated CO2, and this effect was enhanced under enhanced levels of N (McKee & Rooth, 2008). However, these responses may be mediated by environmental conditions including inter‐specific competition (McKee & Rooth, 2008) and salinity (Ball et al., 1997).

The higher rate of soil carbon accumulation in mangrove, and improved sediment trapping observed in US and Australian comparisons of sympatric mangrove and saltmarsh (Comeaux et al., 2012, Lovelock et al., 2011, Saintilan et al., 2013), suggest that mangrove encroachment of saltmarsh may promote elevation‐building and long‐term persistence at higher rates of sea‐level rise. Relatively higher mangrove soil strength might also improve marsh resistance to wave attack (Comeaux et al., 2012). Increased drought resistance may also be conferred in the transition from saltmarsh to mangrove. In a comparison of water use efficiency of the invasive mangrove Avicennia germinans and the dominant Gulf Coast saltmarsh, Spartina alterniflora, Krauss et al. (2014) demonstrated a 53‐65% greater transpiration from Spartina, concluding that mangrove encroachment would conserve water in the substrate, helping to explain the persistence of mangrove in areas of saltmarsh dieback in a recent Gulf Coast drought (McKee et al., 2004).

Additional carbon sequestration resulting from the conversion of grassland and saltmarsh to shrub‐dominance represents a potentially significant negative feedback on increasing atmospheric CO2. Eldridge et al. (2011) reviewed 51 studies reporting on soil carbon in the upper 15cm of shrub and grassland communities, and found a highly significant increase in soil carbon with shrub encroachment, one of the most consistent ecosystem effects of shrub encroachment measured. The net sequestration benefit of mangrove encroachment on saltmarsh has seen less attention though potentially represents an important carbon sink. The progressive removal of cold‐temperature limitations on mangrove northern expansion in the Gulf of Mexico exposes over a million hectares of saltmarsh to mangrove encroachment over the coming century (Bianchi et al., 2013). While posing many challenges to ecosystem services provided uniquely by saltmarsh (Osland et al., 2013, Saintilan & Rogers, 2013), mangrove expansion may facilitate feedbacks promoting the long‐term resilience of wetlands to climate change and their enhanced mitigation of radiative forcing. Bianchi et al. (2013) estimated that conversion of the coastal marshes of the Gulf of Mexico to mangrove dominance would represent a net increase in carbon sequestration of 1.29 ± 0.32 Tg C yr‐1 (Bianchi et al., 2013).

CONCLUSIONS AND FURTHER RESEARCH

The breadth of grassland settings subject to tree encroachment, from intertidal and freshwater wetland through semi‐arid savannah, and the range of disturbance regimes associated with these, strongly implicates common global drivers. Recent poleward expansion has been observed in both terrestrial and intertidal settings in relation to warmer minimum temperatures. However, most woody shrub/mangrove encroachment and thickening occurs well within species thermal limits, and in these situations elevated CO2 may act in

concert with local drivers to facilitate woody encroachment (Higgins & Scheiter, 2012). The “escape height” model is a good illustration of this: that elevated CO2 promotes higher growth rate and greater likelihood of escape (sapling recruitment) between successive fires. Fire is absent from most wetland systems, but the “escape height” model works equally well in explaining woody recruitment into wetland grasslands between successive floods.

In other situations CO2 may facilitate woody sapling persistence by conferring greater drought tolerance. A 5‐ year experiment on the Colorado shortgrass steppe (Morgan et al., 2007) in which CO2 was doubled with respect to controls, led to an approximately 40‐fold increase in aboveground biomass and a 20‐fold increase in plant cover of the shrub Artemisia frigida, common to North American and Asian grasslands. For a variety of reasons tree species may be prevented from recruiting into grasslands by periodic drought, either because of competition for water by grasses (in both rangelands and wetlands) or by the combination of periodic aridity and high salinity (“physiologically dry” in the case of saltmarshes). A shift to tree dominance may occur when droughts are no longer sufficient to induce mortality. This may occur because of local hydrological changes (more consistent base flows in regulated rivers, higher eustatic sea‐levels), but will be aided by the effects of elevated CO2 on drought resistance, including the potentially higher water availability resulting of lower transpiration losses, or an improved rate of root growth to access more dependable groundwater at depth (Fig. 3c).

There are, however, important information gaps and new hypotheses to be tested. We do not have global, consistent estimates of rates of grassland and wetland loss to woody encroachment. Few CO2 enrichment studies have been of sufficient duration to properly consider CO2 acclimation effects on photosynthesis and carbon partitioning (Smith & Dukes, 2013). Most experiments to date have tested the response of shrubs

(mangrove and terrestrial shrubs) to anticipated levels of CO2 (e.g. ~700 ppm) rather than sub‐ambient concentrations relevant to a consideration of historical responses (though see Kgope et al., 2010, Polley et al.,

1997). Research on the response of mangroves to elevated and sub‐ambient CO2 is limited to a handful of laboratory studies. These are important questions in the context of adaptation and mitigation measures globally. The shift in vegetation dominance from grassland to shrub with the onset of elevated atmospheric

CO2 is consistent with a trend to higher carbon sequestration potential, a negative feedback that may have regulate atmospheric CO2 concentrations in periods of less rapid increase than at present. The shift from grassland/saltmarsh to shrub/mangrove dominance also has implications for vegetation persistence in altered climatic and hydrological conditions associated with climate change.

REFERENCES

Bai Y, Colberg T, Romo JT, Mcconkey B, Pennock D, Farrell R (2009) Does expansion of western snowberry enhance ecosystem carbon sequestration and storage in Canadian Prairies? Agriculture, Ecosystems & Environment, 134, 269-276.

Bailey RG (1989) Explanatory Supplement to Ecoregions Map of the Continents. Environmental Conservation, 16, 307-309.

Ball MC (1996) Comparative ecophysiology of mangrove forest and tropical lowland moist rainforest. In: Tropical forest plant ecophysiology. pp 461-496. Springer.

Ball MC, Cochrane MJ, Rawson HM (1997) Growth and water use of the mangroves Rhizophora apiculata and R. stylosa in response to salinity and humidity under ambient and elevated concentrations of atmospheric CO2. Plant, Cell & Environment, 20, 1158-1166.

Berg EE, Hillman KM, Dial R, Deruwe A (2009) Recent woody invasion of wetlands on the Kenai Peninsula Lowlands, south-central Alaska: a major regime shift after 18 000 years of wet Sphagnum–sedge peat recruitment. Canadian Journal of Forest Research, 39, 2033-2046.

Bianchi TS, Allison MA, Zhao J, Li X, Comeaux RS, Feagin RA, Kulawardhana RW (2013) Historical reconstruction of mangrove expansion in the Gulf of Mexico: Linking climate change with carbon sequestration in coastal wetlands. Estuarine, Coastal and Shelf Science, 119, 7-16.

Bond WJ, Midgley GF (2000) A proposed CO2-controlled mechanism of woody plant invasion in grasslands and savannas. Global Change Biology, 6, 865-869.

Bond WJ, Midgley GF, Woodward FI (2003) The importance of low atmospheric CO2 and fire in promoting the spread of grasslands and savannas. Global Change Biology, 9, 973-982.

Bradley RS, Hughes MK, Diaz HF (2003) Climate in Medieval Time. Science, 302, 404-405.

Bren LJ (1992) Tree invasion of an intermittent wetland in relation to changes in the flooding frequency of the River Murray, Australia. Australian Journal of Ecology, 17, 395-408.

Briggs JM, Knapp AK, Blair JM, Heisler JL, Hoch GA, Lett MS, Mccarron JK (2005) An Ecosystem in Transition: Causes and Consequences of the Conversion of Mesic Grassland to Shrubland. BioScience, 55, 243-254.

Brook B, Bowman DJS (2006) Postcards from the past: charting the landscape-scale conversion of tropical Australian savanna to closed forest during the 20th century. Landscape Ecology, 21, 1253-1266.

Bucher D, Saenger P (1991) An Inventory of Australian Estuaries and Enclosed Marine Waters: An Overview of Results. Australian Geographical Studies, 29, 370-381.

Carol Adair E, Reich PB, Trost JJ, Hobbie SE (2011) Elevated CO2 stimulates grassland soil respiration by increasing carbon inputs rather than by enhancing soil moisture. Global Change Biology, 17, 3546-3563.

Cavanaugh KC, Kellner JR, Forde AJ, Gruner DS, Parker JD, Rodriguez W, Feller IC (2014) Poleward expansion of mangroves is a threshold response to decreased frequency of extreme cold events. Proceedings of the National Academy of Sciences, 111, 723- 727.

Cherry JA, Mckee KL, Grace JB (2009) Elevated CO2 enhances biological contributions to elevation change in coastal wetlands by offsetting stressors associated with sea-level rise. Journal of Ecology, 97, 67-77.

Clarke PJ, Myerscough PJ (1993) The intertidal distribution of the grey mangrove (Avicennia marina) in southeastern Australia: The effects of physical conditions, interspecific competition, and predation on propagule establishment and survival. Australian Journal of Ecology, 18, 307-315.

Collatz GJ, Berry JA, Clark JS (1998) Effects of climate and atmospheric CO2 partial pressure on the global distribution of C4 grasses: present, past, and future. Oecologia, 114, 441-454.

Colloff MJ, Ward KA, Roberts J (2014) Ecology and conservation of grassy wetlands dominated by spiny mud grass Pseudoraphis spinescens in the southern Murray– Darling Basin, Australia. Aquatic Conservation: Marine and Freshwater Ecosystems, 24, 238-255.

Comeaux RS, Allison MA, Bianchi TS (2012) Mangrove expansion in the Gulf of Mexico with climate change: Implications for wetland health and resistance to rising sea levels. Estuarine, Coastal and Shelf Science, 96, 81-95.

D'odorico P, Okin GS, Bestelmeyer BT (2012) A synthetic review of feedbacks and drivers of shrub encroachment in arid grasslands. Ecohydrology, 5, 520-530.

D’odorico P, Laio F, Ridolfi L (2006) A Probabilistic Analysis of Fire‐Induced Tree‐Grass Coexistence in Savannas. The American Naturalist, 167, E79-E87.

De Lange WP, De Lange P (1994) An appraisal of factors controlling the latitudinal distribution of mangrove (Avicannia marina var. resinifera) in New Zealand. Journal of coastal research, 539-548.

Duever M (2005) Big Cypress regional ecosystem conceptual ecological model. Wetlands, 25, 843-853.

Eamus D, Palmer AR (2008) Is climate change a possible explanation for woody thickening in arid and semi-arid regions? International Journal of Ecology, 2007.

Eldridge DJ, Bowker MA, Maestre FT, Roger E, Reynolds JF, Whitford WG (2011) Impacts of shrub encroachment on ecosystem structure and functioning: towards a global synthesis. Ecology Letters, 14, 709-722.

Eldridge DJ, Freudenberger D (2005) Ecosystem wicks: Woodland trees enhance water infiltration in a fragmented agricultural landscape in eastern Australia. Austral Ecology, 30, 336-347.

Farnsworth EJ, Ellison AM, Gong WK (1996) Elevated CO2 alters anatomy, physiology, growth, and reproduction of red mangrove (Rhizophora mangle L.). Oecologia, 108, 599-609.

Felker P, Clark PR, Nash P, Osborn JF, Cannell GH (1982) Screening Prosopis (Mesquite) for Cold Tolerance. Forest Science, 28, 556-562.

Fensham RJ, Fairfax RJ, Archer SR (2005) Rainfall, land use and woody vegetation cover change in semi-arid Australian savanna. Journal of Ecology, 93, 596-606.

Fisher DA, Koerner RM, Reeh N (1995) Holocene climatic records from Agassiz Ice Cap, Ellesmere Island, NWT, Canada. The Holocene, 5, 19-24.

Gibbens R, Beck R, Mcneely R, Herbel C (1992) Recent rates of mesquite establishment in the northern Chihuahuan Desert. Journal of Range Management, 585-588.

Gibbens RP, Mcneely RP, Havstad KM, Beck RF, Nolen B (2005) Vegetation changes in the Jornada Basin from 1858 to 1998. Journal of Arid Environments, 61, 651-668.

Giri C, Long J, Tieszen L (2011) Mapping and Monitoring Louisiana's Mangroves in the Aftermath of the 2010 Gulf of Mexico Oil Spill. Journal of Coastal Research, 1059- 1064.

Hallinger M, Manthey M, Wilmking M (2010) Establishing a missing link: warm summers and winter snow cover promote shrub expansion into alpine tundra in Scandinavia. New Phytologist, 186, 890-899.

Hashimoto TR, Saintilan N, Haberle SG (2006) Mid-Holocene Development of Mangrove Communities Featuring Rhizophoraceae and Geomorphic Change in the Richmond River Estuary, New South Wales, Australia. Geographical Research, 44, 63-76.

Hastings J, Turner RM (1965) The changing mile: an ecological study of vegetation change with time in the lower mile of a semi-arid region, Tucson, Arizona, University of Arizona Press.

Hennessy K, Page C, Mcinnes K, Jones R, Bathols J, Collins D, Jones D (2004) Climate Change in New South Wales Part 1: Past climate variability and projected changes in average climate.

Higgins SI, Bond WJ, February EC et al. (2007) Effects of four decades of fire manipulation on woody vegetation structure in Savanna. Ecology, 88, 1119-1125.

Higgins SI, Scheiter S (2012) Atmospheric CO2 forces abrupt vegetation shifts locally, but not globally. Nature, 488, 209-212.

Houghton RA, Hackler JL, Lawrence KT (1999) The U.S. Carbon Budget: Contributions from Land-Use Change. Science, 285, 574-578.

Hungate BA, Reichstein M, Dijkstra P et al. (2002) Evapotranspiration and soil water content in a scrub-oak woodland under carbon dioxide enrichment. Global Change Biology, 8, 289-298.

Idso S (1992) Shrubland expansion in the American Southwest. Climatic Change, 22, 85-86.

Johnston SG, Slavich PG, Hirst P (2003) Alteration of groundwater and sediment geochemistry in a sulfidic backswamp due to Melaleuca quinquenervia encroachment. Soil Research, 41, 1343-1367.

Kaufman DS, Schneider DP, Mckay NP et al. (2009) Recent Warming Reverses Long-Term Arctic Cooling. Science, 325, 1236-1239.

Kgope BS, Bond WJ, Midgley GF (2010) Growth responses of African savanna trees implicate atmospheric [CO2] as a driver of past and current changes in savanna tree cover. Austral Ecology, 35, 451-463.

Kim D, Cairns DM, Bartholdy J (2010) Environmental Controls on Multiscale Spatial Patterns of Salt Marsh Vegetation. Physical Geography, 31, 58-78.

Knapp AK, Briggs JM, Collins SL et al. (2008) Shrub encroachment in North American grasslands: shifts in growth form dominance rapidly alters control of ecosystem carbon inputs. Global Change Biology, 14, 615-623.

Knapp AK, Briggs JM, Hartnett DC, Collins SL (1998) Grassland dynamics: long-term ecological research in tallgrass prairie, Oxford University Press New York.

Knickerbocker CM, Leitholf S, Stephens EL et al. (2009) Tree Encroachment of A Sawgrass (Cladium jamaicense) Marsh within an Increasingly Urbanized Ecosystem. Natural Areas Journal, 29, 15-26.

Kotowski W, Piórkowski H (2005) Competition and succession affecting vegetation structure in riparian environments: Implications for nature management. International Journal of Ecohydrology & Hydrobiology, 5, 51-57.

Krauss K, From A, Doyle T, Doyle T, Barry M (2011) Sea-level rise and landscape change influence mangrove encroachment onto marsh in the Ten Thousand Islands region of Florida, USA. Journal of Coastal Conservation, 15, 629-638.

Krauss KW, Mckee KL, Hester MW (2014) Water use characteristics of black mangrove (Avicennia germinans) communities along an ecotone with marsh at a northern geographical limit. Ecohydrology, 7, 354-365.

Labat D, Goddéris Y, Probst JL, Guyot JL (2004) Evidence for global runoff increase related to climate warming. Advances in Water Resources, 27, 631-642.

Langley JA, Mckee KL, Cahoon DR, Cherry JA, Megonigal JP (2009) Elevated CO2 stimulates marsh elevation gain, counterbalancing sea-level rise. Proceedings of the National Academy of Sciences, 106, 6182-6186.

Lenzi-Grillini CR, Viskanic P, Mapesa M (1996) Effects of 20 years of grazing exclusion in an area of the Queen Elizabeth National Park, Uganda. African Journal of Ecology, 34, 333-341.

Li J, Erickson JE, Peresta G, Drake BG (2010) Evapotranspiration and water use efficiency in a Chesapeake Bay wetland under carbon dioxide enrichment. Global Change Biology, 16, 234-245.

López-Medellín X, Ezcurra E, González-Abraham C, Hak J, Santiago LS, Sickman JO (2011) Oceanographic anomalies and sea-level rise drive mangroves inland in the Pacific coast of Mexico. Journal of Vegetation Science, 22, 143-151.

Lovelock C, Bennion V, Grinham A, Cahoon D (2011) The Role of Surface and Subsurface Processes in Keeping Pace with Sea Level Rise in Intertidal Wetlands of Moreton Bay, Queensland, Australia. Ecosystems, 14, 745-757.

Maestre FT, Bowker MA, Puche MD et al. (2009) Shrub encroachment can reverse desertification in semi-arid Mediterranean grasslands. Ecology Letters, 12, 930-941.

Mateos-Naranjo E, Redondo-Gómez S, Álvarez R, Cambrollé J, Gandullo J, Figueroa ME (2010a) Synergic effect of salinity and CO2 enrichment on growth and photosynthetic

responses of the invasive cordgrass Spartina densiflora. Journal of Experimental Botany, 61, 1643-1654.

Mateos-Naranjo E, Redondo-Gómez S, Andrades-Moreno L, Davy AJ (2010b) Growth and photosynthetic responses of the cordgrass Spartina maritima to CO2 enrichment and salinity. Chemosphere, 81, 725-731.

Matson E, Bart D (2013) Interactions among fire legacies, grazing and topography predict shrub encroachment in post-agricultural páramo. Landscape Ecology, 28, 1829- 1840.

Mckee KL, Mendelssohn IA, D. Materne M (2004) Acute salt marsh dieback in the Mississippi River deltaic plain: a drought-induced phenomenon? Global Ecology and Biogeography, 13, 65-73.

Mckee KL, Rooth JE (2008) Where temperate meets tropical: multi-factorial effects of elevated CO2, nitrogen enrichment, and competition on a mangrove-salt marsh community. Global Change Biology, 14, 971-984.

Mcloughlin L (1987) Mangroves and Grass Swamps: Changes in the shoreline vegetation of the Middle Lane Cove River, Sydney, 1780’s–1880’s. Wetlands (Australia), 7, 13-24.

Mcloughlin L (2000) Estuarine wetlands distribution along the Parramatta River, Sydney, 1788–1940: implications for planning and conservation. Cunninghamia, 6, 579-610.

Mildenhall DC (1994) Early to Mid Holocene pollen samples containing mangrove pollen from Sponge Bay, East Coast, North Island, New Zealand. Journal of the Royal Society of New Zealand, 24, 219-230.

Misra R (1983) Indian savannas. In: Ecosystems of the World. (ed Bourliere F) pp 151-166. New York, Elsevier.

Montané F, Romanyà J, Rovira P, Casals P (2010) Aboveground litter quality changes may drive soil organic carbon increase after shrub encroachment into mountain grasslands. Plant and Soil, 337, 151-165.

Morgan JA, Pataki DE, Körner C et al. (2004) Water relations in grassland and desert ecosystems exposed to elevated atmospheric CO2. Oecologia, 140, 11-25.

Neilson RP (1986) High-Resolution Climatic Analysis and Southwest Biogeography. Science, 232, 27-34.

Nelson J, Morgan J, Lecain D, Mosier A, Milchunas D, Parton B (2004) Elevated CO2 increases soil moisture and enhances plant water relations in a long-term field study in semi-arid shortgrass steppe of Colorado. Plant and Soil, 259, 169-179.

Nie M, Lu M, Bell J, Raut S, Pendall E (2013) Altered root traits due to elevated CO2: a meta-analysis. Global Ecology and Biogeography, 22, 1095-1105.

Osland MJ, Enwright N, Day RH, Doyle TW (2013) Winter climate change and coastal wetland foundation species: salt marshes vs. mangrove forests in the southeastern United States. Global Change Biology, 19, 1482-1494.

Pacala SW, Hurtt GC, Baker D et al. (2001) Consistent Land- and Atmosphere-Based U.S. Carbon Sink Estimates. Science, 292, 2316-2320.

Peng H-Y, Li X-Y, Li G-Y et al. (2013) Shrub encroachment with increasing anthropogenic disturbance in the semiarid Inner Mongolian grasslands of China. CATENA, 109, 39- 48.

Pockman WT, Sperry JS (1996) Freezing-induced xylem cavitation and the northern limit of Larrea tridentata. Oecologia, 109, 19-27.

Pockman WT, Sperry JS (2000) Vulnerability to xylem cavitation and the distribution of Sonoran Desert vegetation. American Journal of Botany, 87, 1287-1299.

Polley HW, Mayeux HS, Johnson HB, Tischler CR (1997) Viewpoint: Atmospheric CO2, soil water, and shrub/grass ratios on rangelands. Journal of Range Management, 278- 284.

Polley HW, Morgan JA, Fay PA (2011) Application of a conceptual framework to interpret variability in rangeland responses to atmospheric CO2 enrichment. The Journal of Agricultural Science, 149, 1-14.

Quisthoudt K, Schmitz N, Randin C, Dahdouh-Guebas F, Robert ER, Koedam N (2012) Temperature variation among mangrove latitudinal range limits worldwide. Trees, 26, 1919-1931.

Rango A, Huenneke L, Buonopane M, Herrick JE, Havstad KM (2005) Using historic data to assess effectiveness of shrub removal in southern New Mexico. Journal of Arid Environments, 62, 75-91.

Ratajczak Z, Nippert JB, Hartman JC, Ocheltree TW (2011) Positive feedbacks amplify rates of woody encroachment in mesic tallgrass prairie. Ecosphere, 2, art121.

Rogers K, Saintilan N, Woodroffe CD (2014) Surface elevation change and vegetation distribution dynamics in a subtropical coastal wetland: Implications for coastal wetland response to climate change. Estuarine, Coastal and Shelf Science, 149, 46- 56.

Rogers K, Wilton KM, Saintilan N (2006) Vegetation change and surface elevation dynamics in estuarine wetlands of southeast Australia. Estuarine, Coastal and Shelf Science, 66, 559-569.

Rozema J, Dorel F, Janissen R, Lenssen G, Broekman R, Arp W, Drake BG (1991) Effect of elevated atmospheric CO2 on growth, photosynthesis and water relations of salt marsh grass species. Aquatic Botany, 39, 45-55.

Saintilan N (1997) Above- and below-ground biomasses of two species of mangrove on the Hawkesbury River estuary, New South Wales. Marine and Freshwater Research, 48, 147-152.

Saintilan N, Rogers K (2013) The significance and vulnerability of Australian saltmarshes: implications for management in a changing climate. Marine and Freshwater Research, 64, 66-79.

Saintilan N, Rogers K, Mazumder D, Woodroffe C (2013) Allochthonous and autochthonous contributions to carbon accumulation and carbon store in southeastern Australian coastal wetlands. Estuarine, Coastal and Shelf Science, 128, 84-92.

Saintilan N, Williams RJ (1999) Mangrove transgression into saltmarsh environments in south-east Australia. Global Ecology and Biogeography, 8, 117-124.

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.

Sankaran M, Ratnam J, Hanan N (2008) Woody cover in African savannas: the role of resources, fire and herbivory. Global Ecology and Biogeography, 17, 236-245.

Scholes RJ, Archer SR (1997) Tree-Grass Interactions in Savannas. Annual Review of Ecology and Systematics, 28, 517-544.

Sekhwela MBM, Yates DJ (2007) A phenological study of dominant acacia tree species in areas with different rainfall regimes in the Kalahari of Botswana. Journal of Arid Environments, 70, 1-17.

Seppä H, Birks HJB (2002) Holocene Climate Reconstructions from the Fennoscandian Tree-Line Area Based on Pollen Data from Toskaljavri. Quaternary Research, 57, 191-199.

Silva JF, Zambrano A, Fariñas MR (2001) Increase in the woody component of seasonal savannas under different fire regimes in Calabozo, Venezuela. Journal of Biogeography, 28, 977-983.

Smith Iii TJ, Foster AM, Tiling-Range G, Jones JW (2013) Dynamics of mangrove–marsh ecotones in subtropical coastal wetlands: fire, sea-level rise, and water levels. Fire Ecology, 9, 66-77.

Smith NG, Dukes JS (2013) Plant respiration and photosynthesis in global-scale models: incorporating acclimation to temperature and CO2. Global Change Biology, 19, 45- 63.

Staver AC, Archibald S, Levin S (2011) Tree cover in sub-Saharan Africa: Rainfall and fire constrain forest and savanna as alternative stable states. Ecology, 92, 1063-1072.

Stine MB, Resler LM, Campbell JB (2011) Ecotone characteristics of a southern Appalachian Mountain wetland. CATENA, 86, 57-65.

Sturm M, Schimel J, Michaelson G et al. (2005) Winter Biological Processes Could Help Convert Arctic Tundra to Shrubland. BioScience, 55, 17-26.

Tape KEN, Sturm M, Racine C (2006) The evidence for shrub expansion in Northern Alaska and the Pan-Arctic. Global Change Biology, 12, 686-702.

Van Auken OW (2000) Shrub Invasions of North American Semiarid Grasslands. Annual Review of Ecology and Systematics, 31, 197-215.

Van Auken OW (2009) Causes and consequences of woody plant encroachment into western North American grasslands. Journal of Environmental Management, 90, 2931-2942.

Wakeling JL, Cramer MD, Bond WJ (2012) The savanna-grassland ‘treeline’: why don’t savanna trees occur in upland grasslands? Journal of Ecology, 100, 381-391.

Warren RJ, Rossell IM, Moorhead KK, Dan Pittillo J (2007) The Influence of Woody Encroachment Upon Herbaceous Vegetation in a Southern Appalachian Wetland Complex. The American Midland Naturalist, 157, 39-51.

West RJ, Thorogood CA, Walford TR, Williams RJ (1985) Estuarine inventory for New South Wales, Australia. Sydney, NSW Department of Agriculture and Fisheries Bulletin No. 2.

Wigley BJ, Bond WJ, Hoffman MT (2010) Thicket expansion in a South African savanna under divergent land use: local vs. global drivers? Global Change Biology, 16, 964- 976.

Williamson G, Boggs G, Bowman DJS (2011) Late 20th century mangrove encroachment in the coastal Australian monsoon tropics parallels the regional increase in woody biomass. Regional Environmental Change, 11, 19-27.

Wilson NC (2009) The distribution, growth, reproduction and population genetics of a mangrove species, Rhizophora stylosa Griff. near it southern limits in New South Wales, Australia. Unpublished PhD Australian Catholic University, North Sydney.

Wilson NC, Saintilan N (2012) Growth of the mangrove species Rhizophora stylosa Griff. at its southern latitudinal limit in eastern Australia. Aquatic Botany, 101, 8-17.

Young D, Porter J, Bachmann C et al. (2007) Cross-Scale Patterns in Shrub Thicket Dynamics in the Virginia Barrier Complex. Ecosystems, 10, 854-863.

Zhang X, Church JA (2012) Sea level trends, interannual and decadal variability in the Pacific Ocean. Geophysical Research Letters, 39, L21701.

Zhang Z, Wang SP, Nyren P, Jiang GM (2006) Morphological and reproductive response of Caragana microphylla to different stocking rates. Journal of Arid Environments, 67, 671-677.

Figure 2: (a) Conceptual model of the relationship between sapling growth rate and fire regime in the African savannah. Fire regime is represented as the return interval of fires capable of killing saplings at a specified height, for natural conditions (solid red line) and heavily grazed savannah (dashed red line). Growth rate is represented as pre‐industrial, low CO2 (solid green line), and current (dotted line), adapted from Bond, Midgley and Woodward (2003). (b) Conceptual model of the relationship between sapling growth rate and inundation regime on Pseudoraphis spinescens (Moira grass) plains. Inundation regime is represented as the return interval of 6‐month continual inundation at depths 0‐2.5 m, for natural (solid blue line) and post‐ regulation (dashed blue line) conditions. Growth rate is represented as pre‐development (solid green line), and current (dotted line), following nutrient and CO2 fertilisation. (c) Conceptual model of the relationship between sapling rooting depth rate and groundwater depth regime for a groundwater dependent wetland. In this case, current groundwater levels (solid line) increase on historic norms (dashed line) because of sea‐level rise in coastal wetlands, and/or reduced transpiration losses. The recruitment bottleneck is avoided when rooting depth exceeds saturated groundwater depth. Growth rate is represented as current (dotted line), following nutrient and CO2 fertilisation, and previous (dashed line).