The Economic Value of the Saltmarsh Habitat in the UK Using Benefit Transfer: A Methodology-Consistent Meta-Analysis

Abstract

This study presents a comprehensive economic valuation of UK saltmarsh habitats, utilising a benefit transfer approach. The core of this research underscored the necessity for consistency in the selection of primary studies for meta-regression models (MRMs) to mitigate potential inaccuracies. A commodity-consistent, methodology-consistent meta-regression model was established based on the existing literature that only used the stated preference methods for saltmarsh valuation in the UK. This research is distinct in its concentration on UK-based studies, aiming to provide a valuation that is not only more reflective of the region-specific importance of these habitats but also contributes to the formulation of more informed policies. The results from the unweighted Ordinary Least Squares (OLS) model, which accounted for approximately 61% of the variance in LnWTPHA (logarithm of per hectare willingness to pay), were particularly revealing. These findings suggest a higher valuation for saltmarshes when a spectrum of benefits is presented for valuation purposes. Incorporating the economic valuation derived from this research, the estimated existence value of saltmarsh habitats in the UK stands at GBP 991 per hectare per year. These findings offer region-specific insights critical for formulating effective conservation strategies, emphasising balanced approaches that consider diverse saltmarsh sizes and socio-economic factors. The study’s UK-focused, consistent methodology and commodity and variable analysis provide policymakers and environmental managers with robust tools to ensure sustainable preservation of saltmarsh habitats.

1. Introduction

The dynamic coastlines of the UK, stretching from the rugged edges of Scotland to the rolling shores of southern England, are enriched with a patchwork of saltmarsh habitats. These ecosystems, whilst expansive, have witnessed significant historical declines. Since the mid-1800s, a staggering 85% of the UK’s saltmarshes have been lost [1,2]. Land reclamation, especially during the 17th to 19th centuries, saw vast tracts of these marshes repurposed for agriculture. Subsequent industrial developments, the establishment of coastal infrastructure, and urban expansions further encroached upon and degraded these habitats. The detrimental effects of pollution from urban and agricultural areas compounded their decline. In recent times, these habitats face the impending threat of sea-level rise, a consequence of global climate change, which is inundating marshes that cannot keep pace with the rising waters. As of the most recent assessment in 2022, England accounted for 35,504.85 hectares of saltmarsh. Reflecting a promising trend, this represents an increase of 2342.75 hectares (7%) compared to baseline figures from 2006 to 2009 [1]. A substantial segment of this gain, precisely 869.64 hectares, can be ascribed to both managed and unmanaged realignment projects, as well as to regulated tidal exchange sites. Scotland has 7076 hectares of saltmarsh, 77.4% of which lie within Sites of Special Scientific Interest (SSSIs) [3]. Welsh saltmarshes make up an estimated 6950 ha of this coastline, often located within low-energy enclosed bays and estuaries [4].

Saltmarshes’ ecological and socio-economic contributions are multifaceted. They act as sanctuaries for diverse marine life, stopovers for migratory birds, and refuges for saline-adapted vegetation. Additionally, they absorb wave energy, serving as natural bulwarks against coastal erosion, and stand as potent carbon sinks, vital in the global effort against climate change. On the socio-economic front, these ecosystems nurture local fisheries, offer recreational outlets, and bear significant cultural significance for many coastal communities.

The pressing need to restore and recreate saltmarsh habitats is evident. They are key to climate change mitigation, effectively trapping and storing atmospheric carbon. Beyond carbon sequestration, their restoration also promises enhanced coastal defence, richer biodiversity, and improved water quality. Moreover, financing such large-scale restoration efforts is becoming increasingly feasible, primarily due to businesses showing a greater inclination towards offsetting CO₂ emissions through the voluntary carbon market [1]. Given the historical degradation, current dynamics, and potential of these habitats, a consolidated economic valuation becomes vital.

The increasing demand for using ecosystem values in decision-making and the time and resource constraints on conducting primary valuation studies have led to the widespread use of benefit transfer to quantify these values [5,6,7,8]. Benefit transfer leverages existing research to predict welfare measures such as willingness to pay (WTP) and predicts values at policy sites where primary valuation studies are not conducted [9,10]. These analyses frequently use meta-regression models (MRMs) where the dependent variable is often taken from primary valuation studies and the independent variables are either taken from primary studies or secondary sources. This method can signify observed factors that can explain variations in the dependent variable. MRMs have become increasingly significant for estimating benefits related to non-market goods and are often used in cost–benefit analysis [6,10].

Despite the increasing adoption of these approaches, there is a lack of consistent and methodologically sound meta-analyses focused on the UK context. The scarcity of primary valuation studies in the UK, compounded by the expertise, time, and financial resources required to conduct them, limits the availability of reliable, region-specific data. Existing meta-analyses often incorporate inconsistent commodities and methodologies or non-UK data, leading to inaccurate benefit transfers. This research seeks to address a critical gap in the economic valuation of saltmarsh ecosystems in the UK. This study addresses this gap by asking the following questions: Why is it necessary to use region-specific data and methodology- and commodity-consistent methods for benefit transfer in economic valuation? And what is the economic value per hectare of UK saltmarsh habitats when applying this approach?

To address these questions, the study tests the following hypotheses: A methodology-consistent and commodity-consistent meta-regression model using UK-based stated preference studies will produce a more accurate economic valuation of saltmarsh habitats, with a narrower range of WTP estimates, compared to models incorporating global or methodologically inconsistent data, as evidenced by a higher R-squared, lower prediction errors (MAE, MASE), and a narrower confidence interval. These hypotheses are grounded in the literature, which highlights the inaccuracies of global benefit transfers [11] and the importance of consistent methodologies [12]. A UK-focused, consistent MRM should be able to improve valuation accuracy, reduce estimate variability, identify key WTP drivers, and reflect regional differences compared to global studies. To test these hypotheses, the research first evaluates existing meta-analyses in the literature. Based on the insights gained, it then develops and applies a tailored meta-analysis focused on UK saltmarsh ecosystems, ensuring both methodological and commodity consistency. To the best of the author’s knowledge, this is the first study to focus exclusively on UK-based primary studies, ensuring that the valuations reflect local ecological and socio-economic contexts. By restricting the MRM to stated preference methods, it achieves unparalleled methodological and commodity consistency, overcoming inaccuracies in prior studies that combine diverse valuation approaches or ecosystem services.

2. Literature Review

2.1. Meta-Regression Analysis (MRM)

Benefit transfer can be either in the form of value transfer (a single value from a study site to a policy site) or a function transfer (calibrates the estimated function for a policy site). Meta-regression analysis (MRM) is a type of function transfer where the results of primary valuations from multiple sites are considered [13]. In MRMs, the dependent variable is often taken from primary valuation studies, and the independent variables are either taken from primary studies or secondary sources. Benefit transfer is a very popular and fast method for valuation; however, using values from one locale for another can lead to significant inaccuracies due to disparities in context and how local communities perceive the worth of services from their ecosystems [14,15].

As MRMs gain traction in real-world applications, concerns over their validity and reliability have surfaced. Johnston et al. (2017 and 2021) and Boyle et al. (2010) discuss this issue in detail, advising caution in selecting primary studies that are not only econometrically robust but also to take into account characteristics such as the sensitivity of the estimated values to scope, scale, and spatial dimension [12,13,16]. From an empirical perspective, Pendleton et al. (2016) mention that using studies carried out on a global scale (e.g., de Groot et al., 2012 or Costanza et al., 1997, 2014) for benefit transfer tends to overstate value estimates [11,17,18,19]. Using values from one locale for another can lead to significant inaccuracies due to disparities in context and how local communities perceive the worth of services from their ecosystems [14,15,20]. For example, Lindhjem and Navrud’s (2008) study examines the effectiveness of using meta-analytic benefit transfer (MA-BT) methods for forest conservation in Norway, Sweden, and Finland [20]. These countries were chosen due to their similar cultural, institutional, and economic conditions, and the homogeneity of their valuation methodologies. Despite these favourable conditions and the comprehensive explanatory power of the meta-analysis, they find that transfer errors can still be significant and that international MATs do not generally outperform simpler methods that average domestic study values [20]. Pascual et al. (2010) acknowledged the notable influence of individual preferences on willingness to pay (WTP) [21]. In light of this, they pioneered the application of the Bayesian benefit transfer method, which permits variations in both preferences and the distribution of characteristics at different sites [21,22]. Moreover, to modify the WTP from the policy site to the study site, they proposed adjustment factors for WTP, formulated based on the mean income across twelve Western European nations, as well as the USA and Canada [22].

One of the suggestions to improve the model’s validity was meeting a basic level of consistency across metadata observations, referred to as “commodity consistency” [23,24]. This means that the goods or services being valued should be roughly similar within and across different studies to prevent comparing vastly different things, like apples to Apple Inc. The EPA guidelines also highlight that, in order to effectively conduct a benefit transfer, certain criteria should be met between the study and policy site. Among these criteria, it is emphasised that the “baseline and extent of change should be similar” or the “affected populations should be similar” [25] p. 7. Not only commodity consistency but methodology consistency is advised in any benefit transfer study. Among others, Boyle et al. (2010) mention the difference between primary valuation methods; travel cost and hedonic price methods rely on both Marshallian and Hicksian demand functions, whereas stated preference methods depend solely on Hicksian demand functions [13]. Revealed preference methods, which are linked to use values, can calculate both non-marginal and marginal estimates, including those derived from hedonic approaches. Conversely, stated preference methods are employed to determine both use and non-use values, in addition to assessing marginal and non-marginal alterations. They raised attention that the values deduced from Marshallian and Hicksian demand functions cannot be consolidated under a single “y”. Smith and Pattanayak (2002) and Vedogbeton and Johnston (2020) demonstrate the same concern that in many meta-analyses and benefit transfer studies, different welfare measures, such as Hicksian or Marshallian consumer, producer surplus, or non-welfare theoretic values, are mixed and pooled with little attention to their econometric justification [23,24]. This issue is exacerbated when incorporating studies from countries with different viewpoints on values.

The importance of maintaining consistency is widely recognised, yet many MRMs often neglect this. This is because, in order to increase their sample sizes, they pool value estimates associated with different, usually dissimilar commodities, like recreation, flood control, fisheries, and aesthetics. Some examples of this are Woodward and Wui (2001), Brander et al. (2006) and (2012), and Reynaud and Lanzanova (2017) [26,27,28,29].

The concept of metadata consistency within meta-regression models is a widely debated subject in the literature, with numerous authors addressing the topic. Examples include studies by Brouwer et al. (1999), Smith and Pattanayak (2002), Johnston et al. (2017), Johnston and Bauer (2020), and Vedogbeton and Johnston (2020), among others [12,23,24,30,31]. These works recognise that achieving absolute consistency is a nearly unattainable goal. They therefore collectively acknowledge that, while achieving absolute consistency is nearly unattainable, ensuring a minimal degree of validity and credibility in analysis necessitates careful consideration of and adherence to commodity consistency.

The individual studies, though insightful, offer fragmented perspectives on saltmarsh value as a habitat and its biodiversity perspective in the UK. This study, therefore, utilises meta-regression analysis to merge data from various research efforts, striving to provide a comprehensive, standardised, and region-specific valuation of saltmarsh habitats in the UK. Through this analysis, the aim is to shape a profound understanding that could be instrumental in influencing future research trajectories and policy formulations. Following Johnston and Bauer (2020), the consistency of methods in valuing coastal saltmarsh meta-analyses will be explored [31].

2.2. Previous Meta-Analysis of Coastal Saltmarsh Studies

The economic assessment of ecosystem services in coastal and wetland environments has become an increasingly prominent field of academic study. However, in the existing body of literature, inconsistencies, methodological limitations, and empirical gaps are prevalent. Saltmarshes, a critical component of coastal ecosystems, are particularly underrepresented, exacerbating the challenges in this field. This underrepresentation is not merely a matter of academic oversight but has significant implications for policy formulation and resource allocation.

Firstly, the issue of saltmarsh underrepresentation is a glaring gap in the literature. Himes Cornell’s 2018 [32] review, which analysed 101 publications, found saltmarshes to be notably underrepresented, featuring in only 15 studies, with a mere 3 emanating from Europe. They also found that a significant 62% relied heavily on market price or benefit transfer methods for valuation. The study also highlighted a recurrent use of outdated data, with the majority being over 4 years old at the time of publication, raising concerns regarding the accuracy of the current ecosystem service values and urging the development of comprehensive methodologies to encapsulate the multifaceted benefits offered by these ecosystems [32].

Undoubtedly the most famous meta-analysis of the monetary value of ecosystems was conducted by Costanza et al. (1997) [18]. The research, though subject to much criticism, reshaped views on the significance of ecosystem services. The estimated value for wetlands including tidal marshes was at least USD 4.8 trillion annually, or 14,785 USD/ha/year. The authors revisited the critiques of the 1997 paper and provided a new estimate based on different assumptions regarding unit values and biome areas [19]. This new estimate showed a significant increase in the global value of ecosystems, with more than a USD 100 trillion increase, with tidal marsh/mangrove ecosystems increasing from around 14,000 to around 194,000 USD/ha/year [19]. The authors suggested that the rise in per hectare values from 1997 to 2007/2011 was due to improved valuation methods, increased demand for certain services, and changes in ecosystem functionality, with a significant portion resulting from more accurate and numerous ecosystem service value estimates [18,19].

Woodward and Wui’s (2001) study was based on 39 previous pieces of research on wetlands, including coastal wetlands, and contained the valuation of 10 different ecosystem services, with mean values ranging from USD 3 per acre per year for amenity services to 1212 USD/acre for birdwatching and 1747 USD/acre for flood control (all prices at 1990 price level) [26].

Brander et al. (2006), based on a review of the literature, examined 80 papers that contained sufficient information for their meta-analysis. The study examined the feasibility of using meta-analysis results for value transfer in wetland valuation. The study concluded that the errors were significantly influenced by geographic location and valuation methods. The average value was highest for unvegetated sediment (9000 USD/ha/year) and lowest for mangroves (400 USD/ha/year), with salt/brackish marshes under 4000 USD/ha/year (all at 1995 price level) [27]. A subsequent paper by Brander et al. (2010) introduced a new methodology for transferring and scaling up ecosystem service values over large geographic areas for policy and academic applications. Their methodology used meta-analysis and a geographic information system to create a value function. While the paper addresses a complex issue, the study did not adequately explain the inclusion of statistically insignificant or negative variables in their estimations [33].

Of the above literature, many coefficients in Brander et al. (2006, 2010, and 2012) and Woodward and Wui’s (2001) meta-analysis functions were not statistically significant, which raises concerns about the robustness of the functions for predicting values in specific policy contexts [26,27,28,33]. The lack of significance suggests that some of the variables included in the model may not have a consistent impact on the value of wetland benefits across different studies, which could lead to unreliable value transfers when applied to new contexts. This raises concerns about the reliability of the results found when using this function for benefit transfer, especially given the high error rates. Johnston et al. (2021) particularly express concern regarding the limitations of “scaling up” benefit transfers, stating that this process “can jeopardise validity…” and reliability, and hence cannot be used for the large-scale linear scaling of benefits [16]. For example, look at studies such as that of Eftec (2010: case study 3), who conducted a valuation study in the Lower Derwent Flood Risk Management Strategy Project [34]. A meta-analysis function derived from Brander et al. (2008) was employed for value transfer to estimate the economic value of wetland creation and restoration. Importantly, this study indeed took values from the Brander et al. (2008) meta-analysis function that included variables which were not statistically significant. Their valuation yielded a conservative estimate of GBP 425 per hectare per year for wetland benefits, which is significantly lower than the £3109 per hectare per year obtained when applying the function without adjustments for the policy site’s context [35]. Incorporating non-significant variables from a meta-analysis function into a value transfer analysis, as happened in this research, is a methodological choice that can be contentious. This raises concerns about the robustness of the resulting valuation.

In light of this, while taking a meta-analysis function from other studies offers a comprehensive dataset for ecosystem service valuation, its application in value transfer should be approached with caution. It is essential to selectively use variables that are significant to put forward for monetary valuations which are also contextually relevant to the specific policy goal in question. Adjustments may be necessary to align the function with the unique attributes of the policy context. This also effectively requires a detailed sensitivity analysis and, where appropriate, the creation of a bespoke meta-analysis function that is tailored to the available data and the specific environmental policy or project being evaluated, ensuring both precision and credibility in the valuation process.

Similarly, de Groot et al.’s (2012) study on the global value of ecosystems included a wide range of ecosystems and services but suffered from the same limitations of broad assumptions and potential inaccuracies. Their valuation of coastal wetlands, which presumably includes saltmarsh ecosystems, was estimated at a mean value of 193,845 Int.USD/ha/year (2007 price levels). However, the pooling of various methodologies from different geographic locations raises questions about the validity of such estimates [17].

Ghermandi and Nones’ (2013) meta-analysis focused on the recreational value of various global and coastal ecosystems. Their research pooled data from 79 primary valuation studies, resulting in 253 distinct observations. The study reported a wide range of recreational values spanning from 0.13 Int.USD/ha/year to 59,533 Int.USD/ha/year. They did not report the value based on ecosystem type. The work left open questions about the feasibility of transferring these benefits internationally across diverse geographical and cultural settings [36].

Wang et al. (2022) conducted an economic evaluation of global saltmarsh ecosystems’ restoration [37]. The study does not provide any meta-analysis of previous valuation studies, but it delivers interesting findings. The authors would rather investigate the cost–benefit dynamics of saltmarsh restoration projects of varying scales. Contrary to the economy of scale theory, which suggests larger projects should be more cost-efficient, the study finds no simple linear relationship between scale and cost per hectare. Using the natural breaks method, the research reveals that as the restoration scale increases, the mean total cost rises significantly. However, the mean benefit per hectare initially decreases, hits a low between 471 and 1,400 hectares, and then rises again. This complexity is attributed to factors like ecological services, landscape scale, and public willingness to pay. Small-scale projects incur higher unit costs and offer lower benefits, suggesting a longer time to recover their initial expenses [37].

Liu and Stern’s (2008) meta-analysis is noteworthy for its methodological consistency. It analysed 39 contingent valuation papers and found that over 75% of the variation in willingness to pay could be attributed to factors like commodity type, methodology, and study quality. However, their findings indicate a geographical disparity, with the mean willingness to pay for saltwater wetlands, marshes, or ponds (USD 2189 per household) being much higher than the values estimated at the UK level [38].

This review of the literature shows that there is a lack of consistent meta-analyses. It also shows that many studies, including some of those mentioned above, suffer from over-parameterised models. Furthermore, while one can justify the inclusion of insignificant variables in monetary valuation, we should not ignore the noise or bias that they bring into estimates. It is common knowledge that if a variable is statistically insignificant, its coefficient is not reliably different from zero, which means it may not have a meaningful impact on the dependent variable. The question that remains is as follows: why do benefit transfer studies include insignificant variables in their monetary valuation?

It is also obvious that previous meta-analyses of ecosystem services in the UK are lacking, clearly, because of the lack of primary studies. Quite a large number of studies have taken into account global valuation studies while issues regarding international benefit transfer are still unresolved. See, for example, the literature review in the report by Brander et al. (2008) that demonstrates the difficulty of international benefit transfer (even within neighbouring countries) because of socio-economic, cultural, and political differences. Page 19 of that report states that “There also seems to be consensus on the relatively poor performance of international value transfer in comparison to domestic value transfer…” and “Value transfer between dissimilar countries is even more problematic (for example, angler benefits between Iceland and Norway).”

Numerous investigations, encompassing those by consultancy agencies, have frequently utilised benefit transfer in their research. This pattern has motivated us to implement a benefit transfer approach that adheres to uniformity in commodity classification and methodological application.

3. Method

3.1. The Coastal Marsh Habitat Metadata

As stated, the focus of this study was on UK-based studies. Advanced searches both on Google Scholar and Scopus were used to find the relevant research. There were also a number of theses and grey reports that were obtained either through a broader Google search, ResearchGate, and colleagues. All the relevant research on coastal saltmarsh valuation was then examined to decide inclusion/exclusion in the final analysis. The number of studies reported in this paper is by far the most comprehensive for saltmarsh valuation studies in the UK (

Table 1. Relevant studies on coastal valuation in UK.

Number	References	Year	Valued Commodity	Region	MRM	Mean WTP
[39]	Rendón et al.	2018	Saltmarsh as nature-based solution	Wales	Yes	Multiple entries
[40]	Needham et al.	2019	Managed realignment	Scotland, Fife, Tay Estuary	Yes	GBP 42–43
[41]	Jones et al.	2015	Valuation of existing hard defences	England	No	-
[42]	Deely et al.	2020	Blue-green and grey coastal protection	Ireland	Yes	Multiple entries
[43]	Christie et al.	2012	SSSIs including saltmarsh	England and Wales	Yes	Multiple entries
[32]	Himes-Cornell	2013	Saltmarshes, sea grass beds, mangrove forests	Worldwide, including UK	Yes	Multiple entries
[44]	Torres et al.	2016	Coastal and marine ecosystems	Global	No	Does not have economic valuation
[45]	Barbier et al.	2011	Estuarine and coastal ecosystems	Global	No	Huge values
[46]	Birol et al.	2007	Coastal ecosystem	England	Yes	Multiple entries
[47]	Warren & Dawson	2005	Saltmarsh	Scotland	Yes	GBP 18.11
[48]	Taherzadeh and Howley	2017	Biodiversity	England	No	Not primary valuation
[49]	Armstrong et al.	2023	Managed realignment	England	No	Not primary valuation
[50]	Jobstvogt et al.	2014	Marine protected area	UK	No	Not primary valuation
[51]	Beaumont et al.	2008	Marine biodiversity	UK	No	Not primary valuation
[52]	Balmford et al.	2008	Ecosystems and biodiversity	Global	No	Not primary valuation
[53]	Hames et al.	2012	Coast	UK	No	Not primary valuation
[54]	Lockwood and Drakeford	2021	Saltmarsh	England	No	Carbon value
[55]	Beaumont et al.	2014	Coastal habitats	UK	No	Carbon value
[56]	Grilli et al.	2022	Saltmarsh	England	Yes	Multiple entries
[57]	Sen et al.	2014	Recreation value of ecosystems	England and Wales	No	Used TCM
[58]	McCormick et al.	2019	Saltmarsh	UK	No	Fish species
[59]	Luisetti et al.	2014	Coastal ecosystem	England	Yes	Multiple entries
[60]	Luisetti et al.	2011	Managed realignment	England	Yes	Multiple entries
[9]	Turner et al.	2007	Managed realignment	England	No	Not primary valuation
[27]	Brander et al.	2006	Wetland	Global	No	Benefit transfer
[61]	da Silva et al.	2014	Managed realignment	England	Yes	Multiple entries
[62]	Davis et al.	2019	Saltmarsh	England	No	Not primary valuation
[63]	Doherty et al.	2014	Water ecosystems including coastal saltmarsh	Ireland	Yes	Multiple entries
[64]	Stithou et al.	2012	River catchment	Ireland	No	Not relevant
[65]	Norton et al.	2018	Coastal	Ireland	No	No primary valuation
[66]	Bullock et al.	2007	Biodiversity	Ireland	No	Not primary valuation
[67]	Van der Biest et al.	2017	Coastal dune	Belgium	No	Not saltmarsh
[68]	Rao et al.	2015	Shoreline	Global	No	-
[69]	Kenter	2016	Coastal area	Scotland, Inner Forth	Yes	Multiple entries
[70]	Moran et al.	2007	Agri-environment ecosystems	Scotland	No	Not primary valuation
[71]	Abone et al.	2014	UK ecosystems	UK	No	Not primary valuation
[72]	Jacobs	2004	Natura 2000 sites	Scotland	No	Not primary valuation
[73]	Norton et al.	2014	Coastal, estuarine, and marine ecosystems	Ireland	No	Not primary valuation
[74]	ATKINS	2017	Managed realignment	England	No	Not primary valuation
[75]	Riegel et al.	2023	Saltmarsh	Scotland	Yes	Multiple entries
[32]	Himes-Cornell et al.	2018	Blue forests	Global	No	No economic valuation
[76]	Bhatia	2012	Managed realignment	England	Yes	Multiple entries
[77]	Shepherd et al.	2007	Managed realignment	England	No	Not primary valuation
[78]	Polyzos & Minetos	2007	Coastal area	England	No	Not primary valuation
[79]	Tinch and Ledoux	2006	Managed realignment	UK	No	Not primary valuation
[34]	Eftec	2010	Coastal area	UK	No	Not primary valuation
[80]	Wade	2018	Coastal	Scotland, Eden Estuary	Yes	Multiple entry
[81]	Birol and Cox	2007	Estuary	England	Yes	Multiple entry
[82]	Blakemore et al.	2008	Coastal defences	Wales	No	Relied on hard defences only
[83]	Christie and Gibbons	2011	Biodiversity	UK	Yes	Multiple entries
[84]	Fairchild et al.	2021	Coastal wetlands	England	No	Not primary valuation
[85]	Blakemore and Williams	1998	Beach ecosystem	Wales	No	Not primary valuation
[28]	Brander et al.	2012	Wetlands	Global	No	Not primary valuation
[86]	Taherzadeh and Howley	2018	Biodiversity	England	No	Not primary valuation
[87]	Rendon et al.	2019	Coastal wetlands	England	No	Not economic valuation

).

Of the 116 reviewed pieces of literature, only 59 included economic analysis. Of these, only 16 studies employed primary valuation approaches (contingent valuation or Choice Experiment) that could be entered into the meta-analysis.

Those studies that only concentrated on the carbon sequestration value of saltmarshes were excluded. Therefore, only studies that used stated preference (SP) methods for saltmarsh valuation were included. The primary purpose was to concentrate solely on primary studies that valued saltmarsh as an ecosystem. Nonetheless, the scarcity of primary studies conducted in the UK necessitated a certain level of inconsistency in the metadata to achieve the sample size essential for statistical analysis.

The analysis was carried out using 32 observations. From these, 3 variables were considered outliers and had to be removed from further analysis. The willingness to pay values were first converted to 2023 values and then, using GIS software (ArcGIS Pro version) and the study areas, converted to per hectare WTP. Income or household size data were not available for all studies, so supplementary sources had to be used to obtain these.

Table 2. List of variables used in MRM.

Variable	Description	Data Type
ln YEAR	The year the original study was conducted.	continues
lnWTP	The willingness to pay, converted to per ha/year.	continues
ln QUAREA	The total size of the area in hectares where the questionnaire was distributed and implemented.	continues
ln SSAREA	The study area (hectares) of saltmarsh/coastal area under consideration for valuation purposes.	continues
ln INCOME	The income either reported in the study or extracted from secondary data.	continues
CE	A binary variable for the method. The Contingent Valuation Method (CVM) was the default value, set to zero, and Choice Experiment (CE) was set to 1.	Dummy
Multi	A binary variable that denoted if only one good or service represented the saltmarsh (= 0) or multiple products or services did (= 1).	Dummy
ln HHSN	The number of households in the questionnaire area.	continues
JOURNAL	A dummy variable for if the research was published in a journal (1), otherwise 0.	Dummy
REGION	The region where the study was performed. As adding national/regional as separate dummy variables caused so much noise in the model, the variable UK-wide also coded as part of the region variable. Case study = 1, regional = 2, UK = 3	nominal
ln HCH	The hectare area of habitat change proposed during the valuation scenario.	continues
ln SSIZE	The sample size.	continues
Ln REL	The relative size of the area of saltmarsh, in hectares, divided by the market size (here questionnaire area).	continues
Ln Percent	The percentage change from the status quo.	continues

summarises the variables used for analysis.

WTP per hectare estimation was carried out as below:

• WTP Data per Household: Existing data was used to find out how much each household was willing to pay for saltmarsh conservation.
• Total Number of Households: The total number of households that were impacted by or benefit from the saltmarsh was identified. This concentrated on the specific geographic area reported by the author(s).
• Total WTP for All Households:

Total WTP for All Households = WTP per Household × Total Number of Households

4.

Total Area of Saltmarsh in Hectares: The total area of the saltmarsh under consideration was obtained from the original study or private communication with the authors.

5.

WTP per Hectare of Saltmarsh:

$$\mathrm{WTP} \mathrm{Per} \mathrm{Hectare} \mathrm{of} \mathrm{Saltmarsh}={\displaystyle \frac{\mathrm{Total} \mathrm{Area} \mathrm{of} \mathrm{saltmarsh} \mathrm{in} \mathrm{Hectares}}{\mathrm{Total} \mathrm{WTP} \mathrm{for} \mathrm{All} \mathrm{Households}}}$$

Those values that were not in the form of an annual value were converted to annual values using the perpetuity model. The reason for this was that the benefits from saltmarshes are expected to last indefinitely. In this model, the annual WTP was assumed as simply the one-off WTP multiplied by the discount rate, assuming that the benefit lasts forever. One should bear in mind that this is a simplification and may not be appropriate for all cases. Using the perpetuity model with a discount rate of 3%, the annual willingness to pay (WTP) equivalent for a one-off WTP was calculated.

3.2. Statistical Model

A meta-regression was used to estimate the potential impact of the variables on WTP. The adopted Equation (1) for saltmarsh valuation MRM is as below:

$$y_i=\alpha +{\beta }_{1 }\left(x_i\right)+\dots +{\beta }_n\left(x_n\right)+{\gamma }_1{D}_1+\dots +{\gamma }_m{D}_m+ϵ$$

(1)

where

• $\alpha$ is the Constant;
• $y_i$is the welfare measure (WTP) from each study;
• $x_i$, …, $x_n$ is a set of independent variables that determines the systematic change in welfare estimation;
• ${\beta }_1$, …, ${\beta }_n$ are the coefficients to be estimated;
• $D_1$, …, $D_m$ are dummy variables for different commodities or other categorisations;
• ${\gamma }_1$, …, ${\gamma }_m$ are the coefficients for these dummy variables;
• $ϵ$ is an error term.

Different assumptions were tested, such as a linear Ordinary Least Square (OLS) model and unweighted random effect OLS, where logarithmic dependent and independent variables, logarithmic dependent, and non-logarithmic independent variables were assumed. The analysis was carried out using Nlogit statistical package.

Meta-Regression Model for Saltmarsh Valuation:

The meta-regression model (MRM) used to estimate the impact of the variables on willingness to pay (WTP) is expressed as follows:

ln(WTP_i) = α + β₁ln(YEAR_i) + β₂ln(QUAREA_i) + β₃ln(SSAREA_i) + β₄ln(INCOME_i) + β_5ln(HHSN_i) + β₆ln(HCH_i) + β₇ln(SSIZE_i) + β₈ln(REL_i) + β₉ln(Percent_i) + γ₁CE_i + γ₂Multi_i + γ₃JOURNAL_i + γ₄*REGION_i + ε_i

where

• ln(WTP_i): The dependent variable, the natural logarithm of willingness to pay (in per hectare per year).
• ln(YEAR_i): The natural logarithm of the year the original study was conducted.
• ln(QUAREA_i): The natural logarithm of the total area (in hectares) where the questionnaire was distributed and implemented.
• ln(SSAREA_i): The natural logarithm of the study area of saltmarsh/coastal area (in hectares) considered for valuation.
• ln(INCOME_i): The natural logarithm of income, either reported in the study or extracted from secondary sources.
• ln(HHSN_i): The natural logarithm of the number of households in the questionnaire area.
• ln(HCH_i): The natural logarithm of the hectare area of habitat change proposed in the valuation scenario.
• ln(SSIZE_i): The natural logarithm of the sample size.
• ln(REL_i): The natural logarithm of the relative size of the saltmarsh area (hectares divided by the market/questionnaire area).
• ln(Percent_i): The natural logarithm of the percentage change to the status quo.
• ε_i: The error term, capturing unobserved heterogeneity or random noise.

4. Results

The unweighted OLS model produced an R-squared value of approximately 0.61, indicating that around 61% of the variability in the dependent variable could be explained by the model. The regression model quantifies the influence of various independent variables on the dependent variable, denoted as lnWTP (per hectare WTP). Of the 14 independent variables considered, 9 exhibited statistical significance. The study highlighted that the MULTI variable, indicative of the diverse benefits of saltmarshes, had a significantly positive impact on LnWTPHA. This finding suggests a higher valuation for saltmarshes that offer a spectrum of benefits. In contrast, the “REGION” variable, reflecting the study’s geographical scope, demonstrated a significant negative impact, revealing that an expansion in geographical scope could lead to a decrease in the economic valuation of saltmarsh benefits. The detailed results, as highlighted in

Table 3. Results of MRM.

Variables	Coefficient	Standard Error	95% Confidence               Interval
ln WTPHA
ln YEAR	1.25	0.92	−0.70                       3.21
ln SSAREA	0.51 ***	0.11	0.27                       0.75
ln HCH	0.37 *	0.20	−0.06                       0.80
CE	−1.60 **	0.66	−3.01                       −0.20
Multi	4.26 **	1.45	1.17                       7.35
REGION	−3.40 ***	0.894	−5.14                       −1.63
JOURNAL	0.53	0.89	−1.23                       2.29
ln REL	0.24 **	0.10	0.02                       0.45
ln HHSN	0.10 ***	0.02	0.05                       0.14
ln SSIZE	0.01	0.39	−0.75                       0.77
ln INCOME	2.25 ***	0.57	1.05                       3.47
ln QAREA	0.081	0.35	−0.60                       0.76
Ln Percent	0.0005	0.00	−0.00                       0.00
Constant	−13.93 **	6.46	−27.57                   −0.28

Note: ***, **, * ==> significance at 1%, 5%, 10% level. R-squared = 0.73, Adj R-squared = 0.61.

, are as follows.

In the regression model, the variables ln SSAREA, ln HCH, MULTI, REGION, ln REL, ln HHSN, CE, ln INCOME, and the Constant were found to be statistically significant. Specifically, ln SSAREA, REGION, lnHHSN, and lnINCOME were significant at the 1% level. CE, MULTI, lnREL, and the Constant were significant at the 5% level, while ln HCH was significant at the 10% level.

The variable MULTI demonstrated the most substantial positive impact based on the magnitude of its coefficient. Its high positive coefficient of 4.26 suggests that when saltmarsh is presented as delivering multiple benefits, such as to birds, habitat, landscape, flood defence, and coastal walks, it is valued the most, with an approximately 4.26% increase in lnWTP per hectare, all else being equal.

Conversely, the variable “REGION,” exhibited a negative coefficient (−3.40) that was significant at the 1% level, suggesting the adverse effect of geographical scope on lnWTP, holding other variables constant. This finding implies that the larger the geographical scope (from case studies to UK-wide surveys), the greater the economic costs or lower the benefits associated with willingness to pay, all else being equal. This warrants further investigation to understand the mechanisms by which geographical scope influences LnWTP and could be highly relevant for policy decisions.

Another significant variable is the method, as a binary variable, which is negative (−1.60) and significant at the 5% level. This suggests that studies using CE, compared to the default valuation method (CVM), tend to report a lower willingness to pay per hectare/year.

The “lnINCOME” variable has a high positive coefficient of 2.25 at the 1% significance level, indicating that income levels have a robust positive impact on lnWTP, meaning the higher the income the higher the willingness to pay.

The variable “lnSSAREA,” representing the logarithm of the study area of saltmarsh or coastal area in hectares under consideration for valuation purposes, was positive (0.51) and statistically significant at the 5% level. The positive coefficient of “lnSSAREA” suggests that larger saltmarsh or coastal study areas might be associated with increasing economic returns, as reflected in lnWTP. This could potentially be due to factors such as enhanced ecosystem services, increased biodiversity, or greater recreational and aesthetic values being associated with larger conservation areas.

The variable “lnHCH,” representing the area in hectares of habitat change proposed during the valuation scenario, is statistically significant at the 10% level, indicating a measurable impact on respondents’ willingness to pay for marsh restoration. The positive coefficient suggests that a proposed increase in the hectare area of habitat change positively influences the economic value represented by lnWTP.

The variable “lnREL,” which represents the logarithm of the relative size of the saltmarsh area in hectares divided by the market size (in this case, the questionnaire area), has a coefficient of 0.23 and is statistically significant at the 5% level. The positive coefficient suggests that an increase in the relative size of the saltmarsh area to the market size has a positive but very modest influence on the economic value, encapsulated by lnWTP.

The variable “lnHHSN,” in this study, represents the logarithm of the number of households in the questionnaire area, which is a measure of the population size within the area being surveyed. This analysis found that this variable was statistically significant at the 1% level, meaning that changes in the number of households (as measured by the natural logarithm) have a statistically detectable impact on the dependent variable under study. Among all the significant variables, it has the smallest magnitude, with 0.10.

lnYEAR denotes the natural logarithm of the year the original study was conducted; while it is positive, it is not statistically significant. lnSSIZE, denoting the sample size, is insignificant with a very small positive magnitude (coefficient = 0.01). This suggests that the sample size does not impact WTP in a significant way. Journal (indicating research published journals), lnQAREA (logarithm of questionnaire area), and Percent (indicating the percentage change from the status quo) are other positive and insignificant variables.

The estimated value is GBP 990.88 per ha per year from this study (95% CI: GBP 800–GBP 1200). The average WTP per hectare per year from the primary studies was about GBP 609.25.

The average Mean Absolute Error (MAE) for Leave-One-Out Cross-Validation (LOOCV) is approximately 0.72. This provides an estimate of the model’s predictive accuracy on new, unseen data. The Mean Absolute Scaled Error (MASE) for the above model is approximately 0.5345. This indicates that, on average, the forecast errors from the model are about 53.45% of the size of the errors from a naive forecast model. A MASE less than one suggests that a model has better predictive accuracy than a naive forecast.

5. Discussion

The economic valuation of saltmarsh conservation and restoration, as explored in this study, encompasses a range of estimates that reflect the diversity and complexity of the ecosystem services provided by these habitats. By incorporating additional literature into the analysis, a deeper understanding of the economic implications of saltmarsh management is achieved, providing a broader context for the regression model results.

The Multiple Regression Model (MRM) analysis in this study offers significant insights into the willingness to pay (WTP) for saltmarsh conservation, situating these results within the broader scope of environmental economics research.

To fully understand the reasons behind the differences in WTP, a deeper analysis of the methodologies, sample characteristics, and types of wetlands or saltmarsh areas valued in various studies is necessary. Exploring the socio-economic profiles of respondents, the framing of valuation questions, and the ecological characteristics of study areas could provide further insights into divergent findings.

The findings for the income variable (lnINCOME) from the current study are consistent with the broader literature [26,27,30], which often shows that higher income levels are associated with a greater willingness to pay for environmental goods and services. This reflects a general trend in environmental economics, where a positive relationship between income and WTP is typically observed.

The significance of variables like the hectare area of habitat change (lnHCH) and the provision of multiple ecosystem services (MULTI) in the current study underscores the ecological and functional attributes of saltmarshes as critical determinants of their economic valuation. This aligns with the focus on ecological characteristics in Ghermandi (2013) and Eftec (2010) [34,36].

The positive coefficient of “lnHHSN” indicates that an increase in the number of households in the questionnaire area is associated with a rise in the economic value as captured by lnWTP. Due to the scarcity of green spaces, residents in densely populated areas may be more willing to pay for the preservation or improvement of existing saltmarshes.

The positive coefficient associated with the variable “lnSSAREA,” representing the logarithm of the study area of saltmarsh, indicates increasing economic returns for larger areas. This finding supports economies of scale and aligns more closely with Brander et al. (2008), who emphasised the size of wetlands as a significant determinant in willingness to pay (WTP) [35]. These findings, however, contrast with Tinch et al. (2006), who advocated for a more localised, case-specific view of saltmarsh values [79]. Tinch et al. (2006) suggested that multiple smaller habitats might be more valuable, diverse, and resilient than a single large one of the same total size, emphasising the ecological importance of habitat edges and transitions and applying the same argument to the ecotourism value of managed realignments [79]. While both the absolute size (lnSSAREA) and the relative size (lnREL) positively influence valuation, the stronger impact of lnSSAREA indicates that larger saltmarsh areas are valued more highly, possibly due to their greater ecological and recreational potential, regardless of their size relative to the questionnaire area.

The positive and significant relationship between the size of saltmarsh areas and their economic valuation, as indicated by the coefficient for lnSSAREA, should not be interpreted as a reason to disregard the importance of smaller, localised saltmarshes. Smaller saltmarshes offer crucial ecosystem services, contribute to biodiversity, and play vital roles in their local contexts. They can be essential for local water quality, provide habitat for diverse wildlife, and hold cultural or recreational significance [27,32]. Additionally, networks of smaller saltmarshes collectively contribute to ecological connectivity and overall environmental resilience. Conservation efforts should recognise the unique and complementary values of both large and small saltmarsh areas, as each plays a significant role in coastal ecosystem health and sustainability. Therefore, advocating for the protection and restoration of smaller saltmarshes remains an important aspect of environmental management and policy.

Notably, the variable REGION was found to have a significant negative impact on WTP in the current analysis. This finding is particularly interesting when compared with the literature, where the emphasis tends to be on the quality or type of wetland rather than the geographical scope. The negative coefficient suggests that a broader geographical scope may dilute the perceived value of saltmarshes, potentially indicating a ‘distance decay’ effect in valuation, which is not extensively documented in the studies used for this analysis. These insights suggest that, unlike other ecosystem services, saltmarsh conservation strategies may benefit from a more localised and focused approach, catering to the unique ecological and economic dynamics of these environments.

Another significant variable is method, as a binary variable, which suggests that studies using CE, compared to the contingent method (CVM), tend to report a lower willingness to pay per hectare. Similarly, Brander et al. (2006) reported a positive coefficient for contingent valuation studies, which implies that when the CVM is used to estimate the economic value of wetlands, higher WTP values tend to be recorded [27]. The lower willingness to pay (WTP) per hectare reported in studies using Choice Experiment (CE) compared to the Contingent Valuation Method (CVM) could be due to CE’s design, which typically offers more complex choices and trade-offs. This approach might lead to more conservative, realistic valuations, as respondents consider various attributes and scenarios, potentially including options with no cost. Conversely, the CVM usually presents a single, specific scenario, potentially eliciting higher WTP estimates. Thus, the methodological differences between CE and CVM significantly influence WTP outcomes in environmental economic studies.

Woodward and Wui (2001) identified replacement costs and hedonic pricing as having positive and significant coefficients, emphasising the economic impact of wetland functions and the value addition to adjacent properties [26].

While the current study does not directly compare these exact variables, it does find significant variables such as household income (lnINCOME) and the number of households in the questionnaire area (lnHHSN), which could be reflective of a similar economic context where socio-economic factors influence WTP.

lnYEAR denotes the natural logarithm of the year the original study was conducted. While positive, suggesting that more-recent studies tend to have a higher willingness to pay per hectare, it is not statistically significant.

Our willingness to pay (WTP) estimate of GBP 990.88 per hectare per year (95% CI: GBP 800–GBP 1200) for UK saltmarsh habitats, derived from a methodology-consistent and commodity-consistent meta-regression model (MRM), provides a robust valuation for conservation policy.

This estimate is moderate compared to prior studies, proving that UK-specific valuations are lower than global estimates due to regional ecological and socio-economic contexts. For instance, Brander et al. (2006) reported a global wetland WTP range of USD 400–USD 9000 per hectare per year (GBP 559–GBP 12,590 in 2023) [27], with saltmarsh values under USD 4000 (~GBP 2295), significantly higher than our estimate. Similarly, Ghermandi et al. (2008) estimated an average value of EUR 4129 (GBP 3303) per hectare per year for European wetlands, equating to GBP 5155 in 2023, encompassing fresh and saltwater wetlands, mangroves, and peat bogs using international benefit transfer with diverse methods and goods [88]. Woodward & Wui (2001) reported global wetland values of USD 306–USD 981 per hectare per year (GBP 389.14–GBP 1249.52 in 2023) [26], with our estimate falling within but closer to the upper bound. Eftec (2010) provided a broader range of GBP 200–GBP 4500 per hectare per year for UK saltmarsh services, with an indicative value of GBP 1400 (GBP 2056.34 in 2023) [34], capturing water quality, recreation, biodiversity, and aesthetic benefits. Christie et al. (2011) valued mudflat and saltmarsh ecosystems at GBP 1035 per hectare (GBP 1455.26 in 2023) [83], slightly higher than our estimate, while Defra’s (2004) PAG guidelines reported GBP 195–GBP 525 per hectare per year (GBP 334.64–GBP 900.95 in 2023) [89], below our value.

Household-based estimates, such as Brouwer et al.’s (1999) GBP 83.65 per household per year (GBP 154 in 2023) for wetland regeneration [30] and Nature’s (2001) GBP 20 per household per year (GBP 35.94 in 2023) for managed realignment in England and Wales [90], are not directly comparable but suggest lower per-unit valuations, reinforcing our per-hectare estimate’s robustness.

The accuracy and precision of our MRM, as predicted by the hypotheses, are evidenced by its R-squared of 0.61, explaining 61% of the variance in LnWTPHA, surpassing Woodward & Wui’s (2001) adjusted R-squared of 0.36 [26] and Brander et al.’s (2010) adjusted R-squared of 0.48 [33]. Our Mean Absolute Error (MAE) of 0.72 and Mean Absolute Scaled Error (MASE) of 0.5345, validated by Leave-One-Out Cross-Validation (LOOCV), demonstrate strong predictive accuracy, metrics absent in Woodward & Wui (2001) [26], Brander et al. (2006, 2010) [27,33], and Eftec’s (2010) non-statistical approach [34]. The narrower confidence interval (GBP 800–GBP 1200) compared to Eftec’s GBP 200–GBP 4500 [34], Brander et al.’s (2006) GBP 559–GBP 12,590 [27], or Woodward & Wui’s GBP 389.14–GBP 1249.52 [26] estimates supports our hypotheses of the prediction of a more precise WTP estimate. This precision stems from our UK-specific, stated-preference-only dataset (16 studies), which mitigates errors from global, mixed-method data, as critiqued by Johnston et al. (2017) [12] and Lindhjem & Navrud (2008) [20]. The significance of 9 of 14 variables (e.g., MULTI = 4.26, lnINCOME = 0.63, p < 0.05) further enhances our model’s robustness compared to prior studies’ significant number of insignificant coefficients.

The policy relevance of our estimate is underscored by its alignment with cost-effective conservation strategies. The cost of habitat rehabilitation and creation, reported by East Midlands Environmental Consultants (1995) [91] and NOAA (1997) via Spurgeon (1998) [92], ranges from USD 2000 to USD 160,000 per hectare (GBP 6952.53–GBP 556,230.21 in 2023). The lower end, associated with managed realignment in the UK, suggests that the reported WTP of 990.88 GBP/ha/year in this paper justifies investment in such strategies, as benefits may accrue over time to offset costs. For example, Nature’s (2001) low household WTP (GBP 35.94) for managed realignment indicates public support for cost-effective measures [90], aligning with our findings. Compared to Eftec’s (2010) broader service bundle [34], Brander et al.’s (2006) high saltmarsh estimate [27], or Ghermandi’s (2008) elevated European value [88], our focus on existence value ensures a conservative, defensible valuation for UK policymakers prioritising sustainable saltmarsh preservation.

The current study’s result, which used a methodology- and commodity-consistent approach that only concentrated on UK saltmarsh habitats, is a more defendable estimate.

6. Conclusions

This study addressed the following research questions: Why is it necessary to use region-specific data and methodology- and commodity-consistent methods for benefit transfer in economic valuation? And what is the economic value per hectare of UK saltmarsh habitats when applying this approach? It was guided by the main hypothesis (H1) that a methodology- and commodity-consistent meta-regression model (MRM) using UK-based stated preference studies would produce a more accurate valuation.

The core of this research involved a thorough review of the existing literature on the benefit transfer. By focusing on UK-based stated preference studies, the MRM achieved methodological and commodity consistency, explaining 61% of the variance in LnWTPHA (logarithm of per hectare willingness to pay) with a high R-squared, low prediction errors (MAE, MASE), and a narrow confidence interval, as shown in Table 2 and Table 3 (original analyses requiring no external source citation). This accuracy, compared to models using global or inconsistent data, answers the first part of the research question, highlighting that region-specific data captures local ecological and socio-economic nuances, mitigating inaccuracies in benefit transfer [11]. The influence of regional factors, evidenced by the REGION variable, further supports this necessity, as local contexts drive WTP variations, ensuring that valuations reflect UK-specific realities [12].

This research is distinct in its concentration on UK-based studies, aiming to provide a valuation that is not only more reflective of the region-specific nuances of these habitats but also contributes to the formulation of more informed policies.

The second part of the research question is answered by the valuation of GBP 991 per hectare per year, representing the existence value of UK saltmarsh habitats. This robust estimate, consistent with the existing literature, was supported by findings such as the significant positive impact of the MULTI variable, which highlights the diverse benefits of saltmarshes, including biodiversity support, migratory bird habitats, coastal defence, and carbon sequestration. This multifunctionality, detailed in Table 2 and Table 3, underscores the inherent value of saltmarshes and the necessity of their preservation, contributing to the high WTP.

The practical implications of this research are extensive, informing environmental policy and the management of saltmarsh ecosystems with a level of detail and specificity that ensures relevance and efficacy.

A key finding of this study is the significant positive impact of the MULTI variable, highlighting the diverse benefits provided by saltmarshes. These habitats are not only crucial for biodiversity, serving as stopovers for migratory birds, but also play pivotal roles in coastal defence and carbon sequestration. This multifunctionality underscores the inherent value of saltmarshes and the necessity of their preservation.

The discerned correlation between higher income levels and a greater willingness to invest in environmental preservation underpins the necessity of economic strategies that account for demographic variations in income. Such strategies may include differential pricing for access to conserved areas or tiered tax incentives for contributions to conservation efforts, ensuring that the economic burden of preservation is fairly distributed and that conservation efforts are financially sustainable.

Moreover, the specific influence of the REGION variable on saltmarsh value highlights the importance of targeted conservation planning. It suggests that conservation efforts are likely to be most effective when they are custom-tailored to local communities, underscoring the direct benefits to these populations. This strategy may foster deeper community involvement and higher support for local conservation projects.

In addition, the implications for the management of saltmarsh areas are profound. The evidence that both large and small saltmarsh areas contribute significantly to ecosystem services and biodiversity suggests that conservation efforts should be scaled and diversified to include both expansive and smaller, fragmented habitats. This could lead to the implementation of a mosaic approach to habitat conservation, where both large-scale projects and smaller, community-led initiatives are valued for their contributions to the preservation of ecological diversity and the provision of ecosystem services.

Ultimately, this research emphasises the role of economic value as a pivotal tool in the environmental policymaking toolkit. Providing a clearer understanding of the factors that influence the public’s value of the conservation of saltmarshes enables the development of policies that are not only ecologically sound but also economically justifiable. This is crucial for ensuring that the conservation of saltmarshes is pursued in a manner that is both responsive to the economic realities of different communities and committed to the long-term sustainability of these vital ecosystems. The GBP 991 per hectare per year value of saltmarsh is representative of the existence value of saltmarsh habitats in the UK. The robustness of this value presented here is reinforced by its consistency with the existing literature and the application of a methodology that remains consistent with commodity-specific and geographical considerations, particularly within the UK context. These valuation figures not only provide a basis for gauging the economic significance of saltmarshes but also assist in setting benchmarks for compensation measures and resource allocation in habitat restoration initiatives.

The study’s reliance on a meta-analysis approach, while comprehensive, is limited by the available data in the existing studies, potentially overlooking some aspects of saltmarsh valuation not widely covered in the literature. This meta-analysis, while offering a broader perspective, relies on the depth and breadth of the existing research, which in this case, was somewhat constrained. A notable limitation in the lack of primary valuation studies significantly impacted the meta-analysis. The scarcity of consistent and comprehensive primary research in this area posed challenges in collating and comparing data across different studies. This lack of consistency in data among the studies included in the meta-analysis might have led to difficulties in drawing broad, generalizable conclusions. Another limitation is the dynamic nature of ecosystem services valuation. The study captures a moment in time, which may not fully represent the evolving values of these services due to changes in environmental conditions, policy landscapes, and public awareness. Moreover, the study presumes a certain level of public knowledge and appreciation of saltmarsh functions and benefits, which may vary across different demographics and regions, affecting the accuracy of the willingness to pay measurements.

In response to these challenges, future research should be directed towards collecting robust, detailed data with a uniform methodological approach. Such endeavours would not only augment and refine the findings from meta-analyses but also lead to more solid conclusions about the economic value of saltmarshes. Achieving consistency in data across studies will facilitate stronger comparative analyses, fostering a more comprehensive understanding of saltmarshes’ complex role in environmental economics. This work is crucial to supporting well-informed policymaking and the development of conservation strategies that accurately reflect the intrinsic value of these ecosystems, thereby ensuring their conservation for the benefit of future generations.