Introduction

Science is a collective endeavour that responds to social imperatives. In the 1970s, as the emerging sustainability mega-trends of extinction, population pressure and climate change were increasingly discussed, survival sciences emerged (Egan 2017); and conservation biology was coined as a crisis discipline (Soulé 1985). Eliminating anthropogenic pressures and separating human activity from nature to halt the loss of biodiversity were the focal points of conservation science during the advent of the discipline (Mace 2014). This was followed by a realization of failure in the 1980s due to continued evidence for unacceptably high extinction rates (Mace 2014). By the early 2000s, studies quantifying economic and societal benefits of nature came to the fore and were quickly utilized to promote nature conservation (Millennium Ecosystem Assessment 2003; Costanza et al. 1997). Here, we demonstrate that today scientific articles increasingly highlight the benefits of, rather than the threats to, habitats, especially for coastal ecosystems. We further discuss whether this signals a shift in conservation science from a crisis to a services narrative, and how this may affect conservation action.

Emergence of the ecosystem-services concept

Ecosystem services developed into a widely accepted framework for assessing nature’s benefits to society in the early 2000s, with the publication of the Millennium Ecosystem Assessment (2003). The notion of nature’s services first appeared in the late 1970s (see, e.g. Force 1978), exemplifying that ‘inexorable trend of policy makers in the Global North seeking to monetize items and qualities of nature formerly regarded as priceless’ (Westman, 1977, p. 197). In a bid to identify and quantify ‘natural capital’, ecosystem services have emerged as a favoured means to communicate the value of nature amongst the ‘new conservationists’ (Soulé 2013). The link between conservation and utilitarian economic thinking has strengthened, so that implicit and explicit links between science and policy are now routinely made using the ecosystem services concept (McKinley et al. 2020). This is despite apprehensions that the new ‘humanitarian focus’ may not always ‘trickle down’ to benefit biodiversity (Soulé 2013). Cultural ecosystem services that cannot be easily monetized are still underappreciated (Hirons et al. 2016) which fuels an ongoing resistance to the monetizing of all ecosystem services in favour of more values-based assessments (Díaz et al. 2018).

The arguments over whether an ecosystem-services framing helps or hinders the protection of biodiversity are continuing (e.g. Dee et al. 2017, Potschin et al. 2016, Silvertown 2015), as is the debate about the extent to which biodiversity (i.e. species richness) directly correlates with the delivery of ecosystem services (Winfree et al. 2015; Isbell et al. 2015). There is also a risk that management and restoration motivated only by the delivery of a single ecosystem service, such as monospecific tree planting for carbon sequestration, will damage biodiversity.

As authors, we can choose how to frame our studies on conservation biology, biodiversity and ecosystem functions. However, many scientists entering the field post 2000 may already take the ecosystem-services narrative for granted (Kull et al. 2015) instead of reflecting on how the framing of their research may affect conservation objectives.

Text mining for crisis versus ecosystem-services framing

To understand how perceptions in the ecology and conservation-biology research community have changed over the past five decades, we analysed over 90,000 publications using a text-mining approach (detailed methods are available from the supporting information 1). We analysed publications on different habitats for the occurrence of words and word combinations (also referred to as n-grams in text mining) that either relate to threats and stresses (i.e. crisis framing; derived from the IUCN threats classification scheme (IUCN n.d.-a) and the IUCN stresses classification scheme (IUCN n.d.-b)) or to ecosystem services. The latter was derived from the World Resources Institute’s categorisation of ecosystem services (Hanson et al. 2012) which builds on the original Millennium Ecosystem Assessment (2003) categories. The list of n-grams is available in the supporting information 2. We calculated the percentage of papers directly related to crisis vs. services (i.e. the research topic defined as the proportion of publications per year with at least one of the respective n-grams in the title or author keyword list). We also considered the research context or framing of publications that are not directly about crisis or services (i.e. calculated as the annual percentage of publications that contain at least one crisis vs. services n-gram in their abstract minus the publications with crisis vs. services n-grams in title or keyword). The proportion of publications with ecosystem services as the main topic has been increasing across all habitat types especially in the past 10 years (Fig. 1 left panel). For publications on seagrass, mangrove and salt marsh, the percentage of published papers adopting an ecosystems service research context now exceeds those with a crisis context (i.e. threats/stresses) (Fig. 1). Publications on rainforest and grassland habitats have seen similar trends; however, the crisis context still exceeds the services-context proportion.

Fig. 1
figure 1

Left: Annual percentage of publications per year that contain at least one ecosystem service vs. threats/stresses (i.e. crisis) n-gram in title or keyword list (i.e. the research topic of the study). Right: Annual percentage of publications that contain at least one ecosystem services vs. threats/stresses n-gram in their abstract minus the number of publications with n-grams in title or keywords (i.e. the research context or framing). The line represents a LOESS (locally weighted smoothing) smoother and the grey area the confidence interval

Full size image

A change in language

Thematic coding of publications in the five decades from 1970–1979 to 2010–2019 on the four coastal habitats (i.e. coral reef, mangrove, seagrass and salt marsh; see supporting information 3) provided further insights into the observed research-framing trends. Selecting the five most-cited papers from each habitat and decade, we coded their habitat introductions for the qualitative themes ‘ecological descriptions and biodiversity’, ‘ecosystem services’ and ‘threats’, to capture the motivating conceptual context used to introduce and justify the work. In the 1970s papers, all habitats were introduced as sites of high diversity or as having interesting ecological dynamics. This was true also for the 1980s, apart from mangroves for which most introductions now included ecosystem services as the study justification (e.g. ‘it is almost an article of faith amongst estuarine scientists that coastal wetlands, such as mangrove forests, are important nursery sites for juvenile fish and crustaceans’ (Robertson and Duke 1987)). This pattern remained until the 2010s for mangrove habitats. By the 1990s, the ecosystem-services framing was also dominant for seagrasses, and it became the most common introductory justification for salt marshes during the 2010s. Coral reefs were an exception to this trend. By the 1990s, the most influential reef papers had switched from justifications based on pure science and description to warning about threats (e.g. ‘Why, then, should the coral reef bleaching and mortality events of the 1980s command great concern? Probably... because the frequency and scale of bleaching disturbances are unprecedented’ (Glynn 1993)). This theme remains dominant today (e.g. ‘The world’s tropical reef ecosystems… are increasingly affected by climate change … rising sea surface temperatures… have triggered unprecedented mass bleaching of corals’ (Hughes et al. 2017)).

The contrast in framing between coral reefs and coastal wetland studies is striking. Does it suggest that reefs are more severely threatened than other coastal habitats? Or that they provide less valued ecosystem services? The literature suggests that coral reefs and coastal wetlands are both still in crisis. Souter et al. (2020) report an estimated decline in annual global average hard coral cover of nearly 20% between 1978 and 2019, with a historic global loss of 42%. Global decline in mangrove forests was estimated at 18% between 1985 and 2016 (FAO 1994; Bunting et al. 2018). A global assessment of historic salt marsh extent change is to our knowledge not available. However, for the USA, China, and Europe, marsh extent has declined by approximately 35% between the 1970s and 2010s with losses predominantly occurring in the USA and China (Irving et al. 2011; Gu et al. 2018; Laengner et al. 2019). Waycott et al. (2009) report a seagrass decline globally of nearly 40% from the 1930s to 2006, although a revised study by Dunic et al. (2021) places estimated net global losses nearer to 19% since 1880. All coastal habitats provide ecosystem services such as shoreline protection and fisheries provision (Barbier et al. 2011). The continued crisis framing of coral reef studies, in contrast to the ecosystem-services framing of coastal wetland studies, may reflect the different scientific cultures in coastal wetland vs. coral reef communities (see language change assessment above). It may, however, also reflect the role that various ecosystem services currently receive in policy and practise (e.g. blue carbon and coastal protection for coastal wetlands) and the high confidence in the negative short-term predictions for coral reefs under climate change (IPCC 2018).

Consequences of crisis vs. ecosystem-services framing

Although objectivity is the prime value of science, it is always hard (many would say impossible) to obtain. The most important processes and practices of science—such as peer review and replication—uphold objectivity, and the social status of science largely derives from it. In practice, however, all scientific discourse incorporates, to a greater or lesser extent, subjective and societal values. This is partly because language is ineradicably metaphorical; apparently neutral words such as ‘growth’ carry different nuance depending on their context (compare economics with epidemiology). Subjectivity also arises because many applied disciplines are openly normative; welfare economics, for example, aims to maximise human welfare whilst conservation biology upholds the value of biodiversity. It follows that the language used in reporting science, particularly how results are framed in introductions and then understood in discussions, deserves careful thought since it will carry explicit or implicit value judgements or political standpoints.

Consider a study that finds mangrove coverage in a country has declined by 20% over the past decade. These findings could be reported as unadorned fact, perhaps in comparison to other similar areas. They could be framed as evidence for decline of threatened species, as an injustice to marginalised communities that depend on the forests, or as a result of unequal power dynamics between men and women (if, for example, women are particularly dependent on fuelwood collection). They could be seen as an assault on indigenous belief systems that hold the forests to be sacred, or they could be set in the context of loss of natural capital implicitly sending a message to economists, and policy makers influenced by them, that here is a problem to be addressed using the tools of cost-benefit analysis and economic trade-off. One could even take a conservation-optimism standpoint when introducing this study by highlighting the potential for recovery rather than emphasizing what has been lost. All these framings may be correct (even simultaneously), and all imply different audiences, sentiments and actions. Perhaps a balanced conservation narrative, acknowledging threats and opportunities, economic and cultural values, may best incentivise the transformative change we urgently need.

Conclusion

Ecosystem-services framing, rather than a crisis framing, is increasing and now predominant for mangroves, seagrasses, and salt marshes whereas this remains marginal for the coral reef literature. The reasons for this may be embedded in the different services and threats that these habitats feature (e.g., blue carbon, bleaching) and the culture and language within the respective research communities. Whereas we did not establish how each narrative may incentivise or hinder the transformative change that is required to halt biodiversity loss, we suggest that it is imperative to actively consider the conservation and ecology research framing we adopt in publications.