The first step of the cascade model is the assessment of the biophysical structure and processes of ecosystems. The biophysical structure of ecosystems includes the species composition, the structural and genetic diversity of flora and fauna and the description of the physical environment that supports life (Fig. 1). In the original definition of the cascade model of Haines-Young and Potschin (2010), processes are simply the precursor of functions and they are the result of the activities and dynamics of each ecosystem component without an explicit consideration of their interactions (e.g. vegetation growth and river channel formation in our case study) (see also De Groot et al. (2010)). The distinction between processes and functions is ambiguous in literature. However, this distinction does not have high relevance in the determination and valuation of ESs (Spangenberg et al. 2014; Baró et al. 2016; Czucz et al. 2020). For this reason, we only discussed here how biophysical structure has been considered in Tamarix control studies and treated processes and functions together in the next section. Thus, we followed the recommendation by Potschin-Young et al. (2018) for adapting the ES cascade model to our case study.

Goetz et al. (2024) exhaustively reviewed the outcomes of Tamarix control through monitoring using vote count and a meta-analysis of 96 studies published from 1990 to 2020. They provided a list of indicators and ecosystem components that have been monitored and noted an over-representation of vegetation monitoring and a paucity of studies examining the response of other biotic and abiotic ecosystem components, such as fauna, physicochemical properties of water and soil and geomorphic characteristics of fluvial landforms that riparian vegetation occupies. We agree with the conclusions of Goetz et al. (2024) that more research on effects of Tamarix control beyond the vegetation component is necessary and essential to provide information for the next steps of the cascade model and achieve an integrative evaluation of riparian social-ecological systems across the American West.

Describing ecosystem functioning is the second step in the cascade model (Fig. 1). Ecosystem functions are the subset of interactions between the biological and physical structure and processes that govern the flow of matter and energy across ecosystems (Potschin and Haines-Young 2016; Raimundo et al. 2018; Hu et al. 2022). Recommendations for integrating ecosystem functioning into evaluation of management of natural resources, including invasive species management (e.g. International Standards for Ecological Restoration, Gann et al. (2019)), have not been as widely implemented as those related to biophysical structure (see Palmer et al. (2014) in the field of ecological restoration and González et al. (2015) for restoration of riparian vegetation specifically). Ecosystem functioning has been overlooked for multiple reasons. First, structural indicators are usually sufficient to evaluate compliance of management projects, which is often the only goal of monitoring (Matzek 2018). Second, there is a tendency to remain at the “structural phase” of evaluation because of the common belief that, if the biophysical structure is restored, recovery of processes and functions will follow (the ‘Field of Dreams’ hypothesis; see Palmer et al. (1997) and Suding (2011)). Finally, ecological functions are harder to conceptualise and monitor, despite efforts to simplify their quantification (e.g. Meyer et al. (2015)). Advances in functional ecology, such as the emergence and application of functional traits and functional diversity to understand ecosystem dynamics, can help to better characterise ecosystem functions (Díaz et al. 2007; Haines-Young and Potschin 2010; Funk et al. 2014). We are aware of only one study that used functional traits and functional diversity to assess the effectiveness of Tamarix control. Henry et al. (2023) used specific leaf area, plant height and seed mass to explore the response of the riparian plant community to Tamarix biocontrol, but their choice of traits was intended to reflect responses to environmental change (“response” traits) instead of to reflect effects on ecosystem functioning (“effect” traits). Effect traits are still underutilised in monitoring outcomes of management interventions (for example, in ecological restoration, see Loureiro et al. (2023)).

A variety of ecosystem functions have been evaluated in the context of Tamarix control; however, many of these functions have been relatively understudied. As increasing water yield (or ‘water salvage’) was a long-standing management goal for Tamarix control, the ecosystem function that has received most attention in the evaluation of Tamarix control outcomes is water cycling and evapotranspiration (Suppl. material 1: appendix S1). A growing body of literature has also been considering trophic relationships, directly or indirectly, via studies of the effects of the biocontrol beetle on Tamarix defoliation, dieback, plant physiology and cover and on other ecosystem components (biocontrol-related herbivory; Suppl. material 1: appendix S1). However, a paucity of studies reflecting ecosystem functions other than water cycling, evapotranspiration and biocontrol-related herbivory in Tamarix control evaluations has been explicitly acknowledged by the scientific community. For example, in a paper discussing the possible unintended consequences of the Tamarix biocontrol programme that was beginning to unfold by the time of its publication, Hultine et al. (2010) suggested that the decline in Tamarix may lead to reduced carbon storage and sequestration, at least in the short-term, a reduced carbon exchange in the ecosystem and a release of nitrogen through defoliation and downstream export following erosion of unstable landforms. They made a call to the scientific community to test these hypotheses. More than ten years later, however, little has been done to understand the changes in nutrient cycling that U.S. rivers have experienced after the biocontrol programme was put in place or as a result of Tamarix control efforts using other techniques (but see Uselman et al. (2011), Snyder et al. (2012) and Snyder and Scott (2020), in Suppl. material 1: appendix S1). At least two other studies have quantified the response of other ecosystem functions to Tamarix management. Kennedy et al. (2005) studied changes in aquatic food webs after mechanical clearing of Tamarix along a small creek in Nevada. Tredick et al. (2016) examined black bear scat to understand potential changes in bear diet after removal of Tamarix in Canyon de Chelly National Monument, Arizona.

A core function that remains overlooked and supports several ecosystem services is primary productivity (Fig. 1). In general, a more thorough understanding of ecological functions associated with Tamarix control evaluations could be achieved with more frequent implementation of the methodologies and experience developed in the evaluation of invasive species management and other types of restoration approaches in a riparian context. This includes assessing ecological functions other than water cycling and evapotranspiration, trophic relationships and nutrient cycling.

ESs and benefits are two sides of the same coin. ESs reflect what the ecosystem provides to human welfare in biophysical terms, while benefits represent the contributions to aspects of well-being, such as health and safety. As ESs and benefits commonly overlap and their distinction is not critical for the valuation of the latter, we will concentrate here on the definition and description of ESs and will not distinguish between ESs and benefits in the next sections of the article. ESs are distinguished from functions (step 2) in that there is a direct or indirect use of an ecosystem resource or property by ESs beneficiaries, while functions represent the “capacity” or ability of the ecosystem to generate ESs (Czucz et al. 2020). There are several classifications of ESs (e.g. MEA (2005); The Economics of Ecosystems and Biodiversity – TEEB – developed by De Groot et al. (2010); to name two of the most popular). One of the most used is the Common International Classification of Ecosystem Services (CICES, Haines-Young and Potschin (2013)). In its last published version (v.5.1, Haines-Young and Potschin (2018)), 90 “classes” of ESs are detailed and grouped hierarchically into “groups”, “divisions” and “sections”. At the highest level (sections), services are classed into: “a) the provisioning of material and energy needs, b) regulation and maintenance of the environment for humans or c) the non-material characteristics of ecosystems that affect physical and mental states of people”. These are three of the four main categories of ESs that the MEA (2005) originally referred to as “provisioning”, “regulating” and “cultural”, respectively. A consensus was reached to consider a fourth category “supporting” as intermediate services. Supporting services are integrated in the previous steps of the cascade model as ecosystem structure, processes and functions (Carpenter et al. 2009). This matching with the MEA framework ultimately reflects the intention of the CICES v.5.1 to cross-reference other classifications and facilitate international comparisons (Haines-Young and Potschin 2018).

The CICES v.5.1 classification particularly addressed the complexity in distinguishing between ESs and benefits (Haines-Young and Potschin 2018). The definition of each service is made up of two parts; one describing the biophysical output from the ecosystem (i.e., what the ecosystem delivers) and the other describing the contribution it makes to human well-being (i.e. how that output is used or enjoyed by people in terms of health, good social relations, security, basic needs etc.). While the CICES list is rather exhaustive, it is not practical to include all ESs in actual evaluations (Matzek 2018). Moreover, the contingent nature of the ES concept implies that establishing a universally applicable, final checklist of ecosystem-supported services is an unachievable (and unnecessary) objective. The list of services should be treated more as a “menu” of ESs and benefits themes, with steps one and two of the cascade model serving to examine how particular systems operate and provide information for the choice and quantification of ESs (Haines-Young and Potschin 2010; Potschin and Haines-Young 2016; Potschin-Young et al. 2018). We are unaware of studies evaluating the effectiveness of invasive riparian plant species management under the prism of an ES approach. However, assessments of ESs outcomes of river restoration have generated lists of ESs, based on project and system singularities (e.g. Acuña et al. (2013); Terrado et al. (2016); Vermaat et al. (2013); Gerner et al. (2018)).

We identified the ESs and associated benefits that riparian systems dominated by native species could provide compared to those dominated by Tamarix, as replacing Tamarix with native vegetation is one of the main goals of Tamarix control (Shafroth et al. 2008). We modified the list of ESs and benefits provided by riparian systems in Riis et al. (2020) for four types of dominant vegetation. We present an abbreviated version of the list in Fig. 1 and an annotated, extended version in Suppl. material 1: appendix S2. Dickie et al. (2014) also listed the ESs provided by Tamarix, but we chose to use the classification by Riis et al. (2020) because Dickie et al. (2014) did not use vegetation categories or compare between control/impact or before/after Tamarix control. Dickie et al. (2014) simply enumerated the ESs provided by Tamarix trees: visual amenity/ornamental (cultural ESs); timber, building materials, poles, posts, pulp, crafts and firewood and charcoal (provisioning ESs); habitat for wildlife, protection from predators (supporting ESs); erosion control, including windbreaks and temperature regulation via shading (regulating ESs).

Some have described ESs provided by riparian systems that were affected by Tamarix control (even though virtually none of them used the term “ecosystem service” in their assessments). Dykstra (2010) enumerated the multiple potential uses of Tamarix biomass obtained from removal efforts, including its transformation into composite wood, its use as biofuel in the form of wood pellets, bio oil and charcoal and for artistic creations (“Timber”, “Biomass for fuel”, “Indirect interaction – artistic”, Fig. 1; Suppl. material 1: appendix S2). Bateman et al. (2012) assessed fire regulation by reduction of fuel loads (“Fire regulation”, Fig. 1; Suppl. material 1: appendix S2). Wieting et al. (2023) and references therein showed that Tamarix removal promotes erosion (“Erosion control” and “Buffering and attenuation of mass movement”, Fig. 1; Suppl. material 1: appendix S2) by reducing the stability of riverbanks and hydraulic roughness. This is typically perceived as a “disservice” by managers (Suppl. material 1: appendix S3). The ES that has received more attention in the context of Tamarix control assessments is “Maintaining populations and habitats” (Fig. 1; Suppl. material 1: appendix S2). Several publications have compared the suitability of Tamarix-dominated and Tamarix-restored sites as habitat for birds (e.g. Shanahan et al. (2011); Darrah and van Riper (2018); Mahoney et al. (2022)) and for herpetofauna (Bateman et al. 2012, 2015; Mosher and Bateman 2016). To our knowledge, there are no other publications that discuss and quantify the other ESs provided by forest patches and/or fluvial features dominated by Tamarix or where Tamarix has been controlled, listed in Fig. 1. We see this as an avenue for further research. A variety of methods for mapping and modelling the supply and demand of ESs were summarised by Harrison et al. (2018).

Once the ESs/benefits have been identified, the final step of the cascade model is to conduct economic and socio-cultural valuations of the ESs/benefits (Fig. 1). Values in general can be defined as the criteria by which we assign importance to something and valuation is the process of expressing or quantifying that value for a particular action or object (Farber et al. 2002; Potschin and Haines-Young 2016; IPBES 2019). Different valuation methods and techniques exist to give an economic, typically monetary value, to ESs/benefits. They are divided into direct market (e.g. production-based, cost-based), indirect market or “revealed preference” (e.g. travel cost modelling, hedonic pricing) and simulated market or “stated preference” (e.g. contingent valuation, choice experiments or conjoint analysis, participatory mapping) methods. We present definitions and hypothetical examples of their use in the Tamarix control context in Suppl. material 1: appendix S3. See Harrison et al. (2018) for an exhaustive list of ESs valuation methods.

Economic valuations are frequently used in cost-effectiveness, cost-benefit analyses and damage assessments. In the context of Tamarix control, cost-effectiveness and cost-benefit analyses could be used to combine monetary valuation of improvements on ecosystem status and ESs, respectively, with the cost of restoration actions (sensu Terrado et al. (2016)), while damage assessments value the loss of ESs (Unsworth and Petersen 1995; NPS 2005) and are more frequently used to investigate the negative effects of invasive species (Marbuah et al. 2014). Cost-effectiveness and cost-benefit analyses are key to assessing economic viability of management interventions and are particularly relevant for invasive species management (Hanley and Roberts 2019). Great Western Research (1989) analysed the economic, environmental and social effects of Tamarix control in the western United States and northern Mexico and estimated annual beneficial effects of $22 million and $40–62 million ($ are not inflation-corrected) for 50% and 90% control, respectively. Economic benefits outweighed adverse economic effects, but the study did not factor in the control programme costs (Barz et al. 2009). Zavaleta (2000) compared the monetary cost of water consumption and sediment retention by Tamarix with removal costs to conclude that the economic gains of potential eradication were considerable. The work by Zavaleta (2000) was frequently used to justify Tamarix control in the region, but her estimates of water consumption by Tamarix have been discredited by some (Stromberg et al. (2009) and references therein). McDaniel and Taylor (2003) estimated the cost of several removal methods and compared them in terms of their cost-effectiveness using Tamarix mortality as an indicator of project compliance. Hart et al. (2005) provided detailed costs of Tamarix removal during 1999–2003 along the Lower Pecos River (Texas), as well as estimates of percent mortality, changes in salinity of the river water, changes in water flow and estimates of water salvage. However, they did not calculate cost-effectiveness ratios. Barz et al. (2009) conducted more explicit cost-benefit analyses of Tamarix control efforts along the Middle Pecos River (New Mexico). They concluded that attempting Tamarix eradication was not worthwhile, based on consideration of different scenarios: direct costs of herbicide spraying, removal and revegetation; indirect costs of increased bank erosion and reservoir sediment accumulation following the eventual reduction of Tamarix; and benefits such as water salvage and associated groundwater recharge. O’Meara et al. (2010) and Bateman et al. (2012) provided detailed estimates of costs of different control methods, but they did not report cost-effectiveness or cost-benefit ratios. More recently, Albers et al. (2018) used bioeconomic modelling to consider the trade-off in terms of costs and positive effects on the ecosystem between controlling invasive Tamarix and restoring habitats with native species. All these examples show that, in the Tamarix control case, indirect and simulated market methods are underutilised and that the ES concept has been rarely, if ever, invoked. We believe a more systematic use of the ES cascade concept in cost-effectiveness, cost-benefit analyses and damage assessments would facilitate the comparison of results across studies.

Not all ESs and related benefits can be valued economically. While conceptual and methodological developments in economic valuation have aimed to cover a broad range of ESs, including cultural ESs, it can be argued that socio-cultural values (symbolic, aesthetic, ethical, relational etc.) cannot be fully captured by economic valuation techniques (Schröter et al. 2014). Socio-cultural values in invasive species management can be represented in more simple terms by the degree of satisfaction of different interested parties. For example, the aesthetic appreciation of the ecological condition of riparian zones by different groups of people has been evaluated with photo-elicitation surveys (e.g. Le Lay et al. (2013); Chin et al. (2014); Arsénio et al. (2020)). Other methods for understanding preferences or social values for ESs, such as deliberative valuation methods, preference ranking methods and multi-criteria analysis methods (Harrison et al. 2018), have been used in the evaluation of invasive species management planning more often than for monitoring outcomes (e.g. Liu et al. (2011); Japelj et al. (2019)). Perceptions and preferences of different interested parties are important because even the perceptions of success by environmentally-informed sectors of the population such as restoration practitioners do not necessarily align with abiotic and biotic parameters measured in the field (Jähnig et al. 2011) and public acceptance of outcomes is key for restoration success (Heldt et al. 2016). There is currently a dearth of studies that describe and quantify public opinion about Tamarix control and what society perceives as successful riparian ecosystem restoration along rivers in the American West. We are unaware of any studies of this kind. Only Sher et al. (2020) have explored how the human component (manager characteristics and decisions) may help explain Tamarix control outcomes in terms of vegetation structure and composition. Clark et al. (2019) previously showed the high degree of collaboration between restoration practitioners and scientists in Tamarix control contexts.

Finally, the value of ecosystems also has an ecological component that may be represented by fundamental properties of ecosystems, such as resilience, stability, health, complexity and integrity (De Groot et al. 2010). These are ecological values (or intrinsic values of nature) that cannot be expressed in economic or socio-cultural terms because they are not based on human preferences or principles, as they go beyond the anthropocentric approach of ESs (Kretsch and Stange 2016; Potschin and Haines-Young 2016). The quantification of these critical ecosystem properties and the subsequent integration into the evaluation of natural resources management, is still in its infancy and is subject to intense debate and study in academic circles (Jaunatre et al. 2013; Moreno-Mateos et al. 2020; Rohwer and Marris 2021; Dakos and Kéfi 2022; Ren and Coffman 2023). No efforts to value such fundamental ecological properties of systems responding to invasive species management, including Tamarix-dominated systems, have been made. Functional traits can be used to value resilience and stability of plant communities through measurable properties, such as functional redundancy, dispersion and response diversity (Laliberté et al. 2010). Other approaches to measure ecological values include ecological networks (Raimundo et al. 2018) and genome sequencing that incorporates eco-evolutionary processes in ecosystem recovery (Moreno-Mateos et al. 2020).

An important consideration when assessing ESs and associated benefits in the context of Tamarix control is the definition of the smallest spatial scale at which ESs will be examined (i.e. the grain of the spatial scale, Turner et al. (1989)). For example, to quantify the contribution of river restoration to ES provision, Vermaat et al. (2013) determined that the grain should be forest patches or fluvial features (e.g. sand or gravel bar, secondary channel, terrace) no larger than 100 m². Cassiano et al. (2013) also used a 100 m² resolution to assess the contribution of remnant riparian forest patches to water-related ESs in an agricultural landscape of south-eastern Brazil. Rather than determining an optimal value for the grain size, Riis et al. (2020) defined ESs provided by riparian systems using study units based on a classification of four different vegetation types. Determining a spatial scale that can discriminate between Tamarix- and native species-dominated units, possibly the forest patch or fluvial features (see, for example, Scott et al. (2022) for criteria to determine relevant geomorphic units), will be key for a fair socio-economic valuation of Tamarix control projects using an ES approach. It will also be critical to determine the ecosystem biophysical structure, processes and functions of riparian systems (pre- or post-Tamarix control) that constitute the first two steps of the cascade model.

Unfortunately, ESs provided by riparian corridors have usually been overlooked precisely because they have been quantified using an inappropriately large grain where ESs are assigned to general land-use categories, such as agricultural, urban and natural (e.g. Felipe-Lucía and Comín (2015)). Lumping natural areas into one category simplifies the heterogeneity of ecosystems and ignores important differences in dominant vegetation, which can strongly influence some attributes of ecosystem structure, such as biodiversity, that ultimately determine supporting functions and final ESs (e.g. wildlife use: Tamarix-dominated, native-dominated and mixed riparian forests can support different avian communities, Van Riper et al. (2008)). For example, the aesthetic appreciation (class service “3.1.1.1” in CICES v.5.1) of a mixed riparian forest dominated by healthy native cottonwoods (Populus spp.) may not be the same as the one provided by a defoliated Tamarix monoculture, even though they both may be designated as “natural forests” when compared to lands occupied by urban sprawl or agricultural fields. Evaluating the steps of the cascade model at the appropriate scales is important so that resource management actions are likewise implemented and monitored at the appropriate scale.

The authors have declared that no competing interests exist.

No ethical statement was reported.

U.S. Geological Survey, Ecosystems Mission Area, Invasive Species Program.

EGS conceived the idea and wrote the article with the help of PBS and FB.

Eduardo González-Sargas https://orcid.org/0000-0001-6485-0088

Patrick B. Shafroth https://orcid.org/0000-0002-6064-871X

Francesc Baró https://orcid.org/0000-0002-0145-6320

This article does not include or analyze any new data.