Original ArticlesAssessing ecosystem collapse risk in ecosystems dominated by foundation species: The case of fringe mangroves
Introduction
Ecosystems have immense intrinsic value whilst also providing vital ecosystem services on which human life depends (Millennium Ecosystem Assessment, 2005). Human activities have however led to degradation of many ecosystems globally (Hansen et al., 2013, Davidson, 2014, Haddad et al., 2015), reducing their capacity to support life. Degradation can eventually lead to ecosystem collapse, a state in which ecosystems lose their defining abiotic and biotic features to the extent that their identity has been irremediably changed. Ecosystem collapse amounts to a transition into a novel ecosystem, characterised by different biota and mechanisms of organisation and/or altered abundance, interactions, and ecological functions of the remaining original biota (Jackson et al., 2001, Folke et al., 2004, Bland et al., 2016). Ecosystem collapse can have severe consequences for biodiversity, ecosystem services and subsequently human welfare (Dobson et al., 2006; but see Raudsepp-Hearne et al., 2010), meaning that there is currently increasing interest in being able to avoid them whenever possible.
Predicting where and when transitions into novel ecosystems may occur is however often difficult (Keith, 2015) since the number of species and processes that can change before an ecosystem loses its original identity has rarely been quantified (Boitani et al., 2015). Three years ago, the International Union for the Conservation of Nature (IUCN) adopted the Red List of Ecosystems (RLE) Categories and Criteria as a robust and consistent tool for monitoring the risk status of ecosystems in order to plan appropriate conservation actions (Keith et al., 2013, Bland et al., 2016). Key to the RLE assessments is the concept of ecosystem collapse, defined there as the endpoint to ecosystem decline, “when it is virtually certain that its defining biotic or abiotic features are lost from all occurrences, and the characteristic native biota are no longer sustained” (Bland et al., 2016). Two of the risk assessment criteria assess spatial symptoms of ecosystem collapse through declines in distribution (Criterion A) and restricted distribution (Criterion B); two criteria assess functional symptoms of ecosystem collapse through environmental degradation (Criterion C) and biotic disruption (Criterion D); the final criterion (Criterion E) evaluates quantitative estimates of the risk of collapse through the integration of multiple threats and symptoms into models of ecosystem dynamics (Bland et al., 2016). The RLE assessment is based on a conceptual model which summarises the most important biotic and abiotic components of a given system, as well as significant ecosystem functions and processes. This model facilitates characterising all relevant pathways to collapse, as well as choosing appropriate variables to monitor ecosystem degradation.
But is this general approach robust enough to help detect ecosystem collapse risk for ecosystems dependent on a few dominant, so-called “foundation species”, to perform most of their functions? Relatively minor reductions in the abundance of a foundation species could indeed have critical consequences for the functioning of these ecosystems, with significant impacts on associated biota, even before the characteristic native biota is entirely lost (Dayton, 1972, Ellison et al., 2005). In this situation, monitoring the foundation species provides a robust estimate of collapse risk only if changes that affect ecosystem functioning are captured. For instance, merely monitoring tree cover to assess the condition of a forest will not be enough to robustly assess collapse risk if the functions supported by trees (e.g. as habitat for other biota) vary with stand or age structure (Burns et al., 2015). Non-foundation species or the abiotic environment may moreover change in a way that fundamentally alters ecosystem functioning, but unless the foundation species are significantly affected, the ecosystem may not appear to have changed. For instance, defaunation of structurally intact forests alters processes such as seed dispersal and tree seedling recruitment (Stoner et al., 2007, Terborgh et al., 2008), resulting in changes in the relative abundance of tree species at the seedling stage (Terborgh et al., 2008, Effiom et al., 2013). Due to the long generation time of trees, recently defaunated forests are difficult to distinguish from forests with a full set of large vertebrates, in terms of tree distribution and composition (Harrison et al., 2013).
To investigate the robustness of the current RLE approach for collapse risk assessments in ecosystems dependent on a few foundation species to perform most of their functions, we here apply the RLE protocol to fringe mangrove forests in the Philippines. Fringe mangrove forests are tide-dominated mangrove forests (i.e. they have the highest tidal inundation frequency), as opposed to river-dominated or inland (basin) mangrove forests (Ewel et al. (1998). Fringe mangrove forests have distinct abiotic settings, as well as distinct ecosystem composition and functioning compared to riverine and basin mangroves; distinctive features include consistently high salinity, relatively higher abundance of migratory birds, and the highest carbon export values among all mangrove forest types (Ewel et al., 1998). Taken together, this suggests they are best conceptualised as a separate ecosystem. In the Philippines, fringe mangrove forests are dominated by only two true mangrove species (Avicennia marina and Sonneratia alba; Ricklefs and Latham, 1993, FAO, 2007, Sinfuego and Buot, 2008, Sinfuego and Buot, 2014), making them an ideal case study of ecosystems dominated by few foundation species. Over 50% of the total mangrove area in the Philippines has reportedly been lost in the last century, and mangrove forests in general are continuing to disappear from South East Asia at an estimated rate of 3.6–8.1% per year (Polidoro et al., 2010, Long et al., 2014, Hamilton and Casey, 2016); this suggests that fringe mangroves, like other types of mangroves, could be at an increased risk of collapse. As they provide vital ecosystem services including coastal protection, provision of raw materials, and carbon sequestration, this would not only result in loss of biodiversity, but would likely have devastating consequences for humans, both locally and worldwide (Garcia et al., 2014).
Section snippets
Methods
A detailed description of the RLE assessment process is provided by Bland et al. (2016) and Murray et al. (2016). In short, this entails describing the fringe mangrove ecosystem, identifying suitable variables to assess ecosystem degradation, and defining ecosystem collapse as bounded thresholds in these variables. These are used to assign one of eight risk categories to each criterion: Collapsed (CO), Critically Endangered (CR), Endangered (EN), Vulnerable (VU), Near Threatened (NT), Least
Results
Fringe mangroves were successfully identified by our land cover classification approach, with an overall classification accuracy of 88.1% (Kappa coefficient: 0.8414; Table 2). The extent of the Philippines’ fringe mangrove forests in 2000, 2010, and 2016 was estimated as 2038 km2, 1924 km2, and 2538 km2 respectively. The resulting estimates of percentage area change range from a 28% loss to a 99% gain (Table 3). A 28% loss is within 10% of the threshold for the Vulnerable category, qualifying
Discussion
The RLE risk assessment provides support for classifying fringe mangroves in the Philippines as ‘Least Concern’, based on small projected changes in geographic distribution (Criterion A) and their currently large extent (Criterion B; Table 4). Given that mangroves in the Philippines have generally undergone deforestation and degradation (Garcia et al., 2014), this may seem surprising. The large variability in projected distribution changes might in part be underpinned by differences in
Conclusion
Integrating different types of data via satellite remote sensing, GIS and modelling approaches is a promising way to meet the unique demands of risk assessments for ecosystems depending on a few foundation species. The conceptual ecosystem model at the heart of the RLE assessment protocol is central to structuring both the identification of relevant data sets and for the integration process, including the development of quantitative models to address particular indicators. Gathering new data
Data statement
The fringe mangrove maps are available from the corresponding author upon request.
The R code used in the land cover classification and NDVI time series analysis is available from the corresponding author upon request.
Acknowledgement
We would like to thank Calvin Lee and Clare Duncan for valuable advice on mangrove ecology and for sharing technical expertise.
Conflicts of interest
The authors declare no conflicts of interest.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Data accessibility
The fringe mangrove maps are available from the corresponding author upon request. The R code used in the land cover classification and NDVI time series analysis is available from the corresponding author upon request.
References (87)
Mangrove forests: resilience, protection from tsunamis, and responses to global climate change
Estuar. Coast. Shelf Sci.
(2008)- et al.
A birds-eye view of biological connectivity in mangrove systems
Estuar. Coast. Shelf Sci.
(2015) - et al.
A 7 year record of above-ground net primary production in a southeastern Mexican mangrove forest
Aquat. Bot.
(1996) - et al.
Rehabilitating mangrove ecosystem services: a case study on the relative benefits of abandoned pond reversion from Panay Island, Philippines
Mar. Pollut. Bull.
(2016) Holocene palynology and sea-level change in two estuaries in Southern Irian Jaya
Palaeogeogr. Palaeoclimatol. Palaeoecol.
(2005)- et al.
Occurrence of species-rich crab fauna in a human-impacted mangrove forest questions the application of community analysis as an environmental assessment tool
Estuar. Coast. Shelf Sci.
(2012) - et al.
Threats to mangroves from climate change and adaptation options: a review
Aquat. Bot.
(2008) - et al.
Biology of mangroves and mangrove Ecosystems
Adv. Mar. Biol.
(2001) - et al.
Environmental drivers in mangrove establishment and early development: a review
Aquat. Bot.
(2008) - et al.
Organic carbon dynamics in mangrove ecosystems: a review
Aquat. Bot.
(2008)