Geodiversity as a potential indicator of stream health in ecological quality assessment systems

Stream geodiversity is a novel concept that is used to describe the variety of geological, geomorphological and hydrological features at a reach‐scale level. In this study, we investigated the relationship between geodiversity and ecological indicators based on fishes, benthic macroinvertebrates and diatoms used for the ecological classification of rivers in line with the Water Framework Directive (WFD) 2000/60/EC. We examined whether geodiversity can be used as a proxy indicator of ecological quality, and we further tested if geodiversity can explain a significant amount of taxa richness variation. We hypothesized that undisturbed or minimally disturbed rivers will be more hydrologically and geomorphologically diverse, and as such, the ecological quality of streams will improve with geodiversity. We also attempted to quantify the probability of achieving ‘good’ ecological quality in relation to geodiversity levels through ordinal regression analysis. Although we did not find a significant relationship between geodiversity and taxonomic richness for any of the three freshwater groups, our results showed positive effects of geodiversity on all three ecological indicators. We also found that the likelihood of achieving the WFD target increases significantly with higher geodiversity. Our findings indicate that geodiversity can potentially serve as a proxy of ecological quality and highlight the need to consider the inclusion of geodiversity measures in ecological quality assessment systems.


| INTRODUCTION
European rivers are threatened by multiple anthropogenic stressors that occur at various spatial scales (Cooper et al., 2013;Grill et al., 2019;Jorda-Capdevila et al., 2019;Kattel, 2019). Hydromorphological alteration and physicochemical degradation are the most common pressures that lotic systems are facing today, affecting the aquatic biodiversity and ecosystem functions (Branco et al., 2016;Filipe et al., 2013;Grizzetti et al., 2017;Lemm et al., 2021;L opez-Doval et al., 2013;Stefanidis et al., 2018). To fight this, the Water Framework Directive (WFD) 2000/60/EC, being the cornerstone of European Union's water policy, aims to protect all the surface waters, including rivers, by assessing their ecological status and thereby providing a better basis for promoting and applying mitigation and restoration measures (Carvalho et al., 2019;Giakoumis & Voulvoulis, 2018). The WFD requires from every member state that all water bodies should obtain a good ecological status, which is described as slight but no significant deviations from the natural and undisturbed conditions (Birk et al., 2012). In rivers, particularly, the ecological status is assessed by considering the responses of four discrete groups of aquatic organisms to major environmental disturbances. These four groups, fishes, benthic macroinvertebrates, benthic diatoms and aquatic macrophytes, are defined by the WFD as biological quality elements (BQEs). Ecological quality assessment methods in rivers are based on various biological attributes of the four BQEs that can quantify the deviation from the natural undisturbed conditions. Remarkably, despite the fact that the directive was issued more than 20 years ago, approximately 60% of European surface water bodies have failed to achieve a good ecological status (European Environment Agency, 2018).
There are several reasons behind this failure, but a general consensus is that there are more environmental problems that act in concert, hampering the effectiveness of management and restoration actions for most ecosystems (Hering et al., 2015;Spears et al., 2021). It might not be sufficient for managers to just deal with the dominant stressor, but they also need to control additional stressors that likely impose threats at a given ecosystem. For instance, besides nutrient control, riverine ecosystems may require additional mitigation measures that address multiple stressors in parallel, such as the re-establishment of riparian zones that may reduce nutrients and improve hydromorphological conditions (Birk et al., 2020). Recently, the importance of hydromorphology on the ecological status of rivers has been highlighted by several studies (Birk et al., 2020;Lemm et al., 2021), bringing forth the need to revisit hydromorphological assessment methods and restoration efforts that aim to improve hydromorphological conditions. Although various hydromorphological assessment methods have been developed and proposed for use in-line with the WFD requirements, most of them may require extensive effort and time at the field.
Additional expert training for the assessors, suitable field conditions and, in some cases, substantial knowledge of mapping and analysis of geographic information may also be required (Belletti et al., 2015;Rinaldi et al., 2017;Stefanidis et al., 2022). Furthermore, several hydromorphological assessments may be region-specific, which means that their implementation is practically restricted at rivers of certain geographic areas or specific hydrological conditions (Belletti et al., 2015).
An alternative option for assessing the hydromorphological conditions of rivers and streams could rely on the estimation of the stream geodiversity. Geodiversity describes the variety within abiotic environment, and it may include various geological, geomorphological and hydrological features and processes, such as soils, erosion related processes, river meanders, flow types and depositional features (Gray, 2013;Gray et al., 2013). Hence, geodiversity is considered likely to affect biodiversity and ecosystem functions and services (Alahuhta et al., 2018;Hjort et al., 2015;Schrodt et al., 2019). Stream geodiversity may refer not only to the physical factors, such as hydraulic and geomorphological features, but also to non-living biological features, such as woody debris, exposed bankside roots and fallen trees (Kärnä et al., 2018). Such features shape the physical characteristics of the aquatic habitats, which in turn may prove crucial for the survival, foraging and breeding of biotic communities (Archer et al., 2019;Deane et al., 2021;Kalogianni et al., 2020). A relatively simple approach for assessing the in-stream geodiversity is to count the richness of the physical, hydrological and geomorphological features (Hjort & Luoto, 2010;Kärnä et al., 2018) in situ. Such approaches have been used in studies that attempt to explore the relationships between geodiversity and biodiversity Toivanen et al., 2019;Tukiainen et al., 2019) and ultimately promote biodiversity conservation actions through the protection of the abiotic environment (Alahuhta et al., 2018;Crisp et al., 2022;Hjort et al., 2022). However, studies that investigate whether geodiversity metrics can be used as proxies for ecological quality of freshwater ecosystems are still lacking.
The main objective of this article is to identify whether there are significant relationships between geodiversity and various ecological indicators, including the taxonomic richness of three different groups of aquatic organisms (i.e., benthic diatoms, benthic invertebrates and fishes). We hypothesize that river reaches with increased variety of stream geomorphological and hydrological features (i.e., high geodiversity) will achieve a better ecological quality than reaches with lower stream geodiversity. To this end, we compiled datasets that include ecological descriptors (i.e., taxonomic richness and ecological indices) and geodiversity, which expresses the total number of the different geomorphological and hydrological features counted in situ. To disentangle possible confounding effects of water quality and geodiversity, we further evaluated the effects of a water quality index on the ecological quality. Finally, we attempted to quantify the probability of a given site to achieve 'good' ecological quality in relation to geodiversity levels in order to highlight the importance of geodiversity as a proxy measure of ecological quality and overall ecosystem health.

| Study area and data description
The study area comprises 143 river reaches that belong to the Greek national network of sampling sites that are systematically assessed according to the guidelines and requirements of the WFD (Figure 1).
The studied sites are distributed among 12 major water districts and are classified into six intercalibration river types following the typology of the Mediterranean intercalibration group (Lazaridou et al., 2018), reflecting not only a wide range of biogeographical and geomorphological characteristics but also anthropogenic disturbances.
At each site, geodiversity was assessed by identifying and counting in situ different types of hydrological and geomorphological features during surveys that were conducted during the summers of 2018 to 2021. Surveys were carried out across a 500 m reach by wading the channel or walking onshore if the water was too deep. More specifically, we determined the number of different stream flow and F I G U R E 1 Map with the location of the 143 studied sites. substrate types, geomorphological features and wood structures in the channel, following a classification system that is included in the river habitat survey (RHS) method (Table 1) (Raven et al., 1998). The sum of the four categories was used as a proxy metric that expresses total geodiversity of the reach, adopting a similar concept of geodiversity as described by Kärnä et al. (2018). The hydromorphological modification of the sites was determined according to an ordinal scale with five levels that are characterized as pristine, unmodified, obviously modified, significantly modified and severely modified hydromorphological conditions (Raven et al., 1998). We also obtained information on benthic diatoms, benthic macroinvertebrates and fishes taxa richness sampled at each site during the same time period with the hydromorphological assessment. Benthic macroinvertebrates were identified at family levels and as such, we used the family richness as a measure of the community alpha diversity. Three ecological indicators, one for each BQE, were used as measures of ecological quality. The Specific Pollution Sensitivity index (IPS) was used for the assessment of biological quality based on benthic diatoms (Karaouzas, Smeti, et al., 2019) and the Hellenic Evaluation System 2 (HESY2) index was used for the benthic macroinvertebrates (Lazaridou et al., 2018). For the fishes, we used the Hellenic Fish Index (HeFI), which is a model-based fish bioassessment index specifically developed for the rivers of Greece (Zogaris et al., 2018). Besides the ecological indicators, ecological quality classes of each BQE were also used as a dependent variable in further analyses. Finally, we included in our analysis as an independent variable the Nutrient Classification System (NCS), which is a water quality index for evaluating water quality degradation caused by excessive nutrient pollution (Skoulikidis et al., 2006). The calculation of the NCS was based on the measurements of nutrients (N-NO 2 , N-NO 3 , NH 4 and P-PO 4 ), total phosphorus and total nitrogen of water samples that were collected in parallel with the biological sampling. Using both geodiversity and NCS as independent variables, we intended to assess the joint effect of hydromorphology and water quality on the biological indices and determine whether geodiversity can have a stronger effect than water quality for certain biotic groups.

| Statistical analysis
Since we hypothesized that geodiversity reflects the hydromorphological stress at the river site, first, we tested whether geodiversity varied among sites with different degrees of hydromorphological T A B L E 1 Variety of flow types, geomorphological features, wood structures and substrate types counted as geodiversity. The description of all features is based on the river habitat survey field manual (Environment Agency, 2003) and the paper of Kärnä et al. (2018).

Flow types Description Geomorphological features Description
Free fall Vertically falling water Exposed bedrock Exposed bedrock  (Ogle et al., 2022).
To investigate the relationship between geodiversity, biodiversity and the ecological indicators, we employed generalized additive models (GAMs). GAMs are an extension of the generalized linear models that incorporate non-linear relationships between the response variable and the predictors (Hastie & Tibshirani, 1990), and they are often used in ecology for fitting non-linear relationships between biotic responses and environmental predictors (Guisan et al., 2002;Pedersen et al., 2019). In our case, models were fitted using the ecological indicator for each of the three BQEs as a response and the geodiversity with the NCS as independent variables (predictors). The models were fitted with the 'mixed GAM computation vehicle (mgcv)' package in R environment (Wood, 2017) using cubic smoothing splines.
We also employed ordered logistic regression to model the probability of the ecological quality to meet a specific class (i.e., bad, poor, moderate, good or high) in relation to the geodiversity and the NCS.
The ordered logistic regression or ordered logit model is a regression model that is used for analysing ordinal responses (Agresti, 2006). It is defined by a set of equations where the cumulative probabilities are related to linear predictors through a logit function. Thus, we used the ecological quality class for each BQE as an ordinal response with five levels ranging from bad to high (bad, poor, moderate, good and high) and the geodiversity and the NCS as predictors. The models were developed in R environment using the proportional odds logistic regression (polr) function of the 'Modern Applied Statistics with S (MASS)' package (Venables & Ripley, 2002).

| Geodiversity across a hydromorphological modification gradient
We found statistically significant differences of geodiversity among the five classes of hydromorphological modification ( p ≤ 0.001). Pairwise comparisons showed significant differences of geodiversity between the most impaired class (severely modified) and all other classes of hydromorphological modification. The geodiversity median for the 'severely modified' class was notably lower than any other class ( Figure 2).
Similarly, we found significant differences of geodiversity among the five classes of ecological quality for all BQEs (Figure 3). Medians of geodiversity were higher at classes of good and high quality compared with poor or moderate for the benthic diatoms and invertebrates. For the fishes, geodiversity was higher at classes of good and high quality than bad or poor.

| Relationships between biodiversity, ecological indicators and geodiversity
Our results did not show strong relationships between geodiversity and taxa richness for any of the three aquatic groups. All three models showed very low but similar R 2 adj values (<0.1). The effect of geodiversity on the taxa richness varied among the three BQEs, being positive for the benthic macroinvertebrates, negative for the fishes and slightly negative for the benthic diatoms, although the relationship was non-significant ( Figure 4).
To explore the type of the relationship between the geodiversity and the ecological indicators, we fitted GAMs that showed significant positive effects of both geodiversity and NCS for all three indicators ( Figure 5). Our results showed that the model for the benthic diatoms demonstrated higher adjusted R-squared with more significant effects of geodiversity and NCS ( p ≤ 0.001) than those for the macroinvertebrates and fishes (Table 2). Concerning the macroinvertebrates and fishes, the results were quite similar, albeit less significant than those for diatoms. The IPS and HESY2 indices increased with geodiversity and NCS, which implies that sites that are richer in geodiversity features and have better water quality index will achieve better ecological quality. HeFI, that is, the index for fish, showed a humpbacked curve with increasing geodiversity with values larger than 20, the index remaining rather stable. Regarding the effects of the NCS on HeFI and HESY2, it seems that it was stronger at values above 3, which indicate 'good' physicochemical quality.

| Probability of good ecological quality class increases with geodiversity
The results of the ordered logistic regression analysis showed that both variables (geodiversity and NCS) had a significant positive effect on the probability of achieving a 'good' ecological quality class regardless of the assessed BQE (Table 3). Although it is obvious that NCS has a stronger effect than geodiversity in increasing the probability of a better ecological quality class, for every one unit increase in geodiversity, the odds ratio for a certain site to improve from moderate to good quality class increases by 9.4%, 6.4% and 10% for diatoms, macroinvertebrates and fishes, respectively (Table 3). The effect of geodiversity is slightly stronger for the fishes than the other two BQEs; whereas, the effect of NCS is stronger for the benthic diatoms.
Hence, the probability of good and high ecological quality class increases; whereas, the probability of moderate, poor and bad classes decreases with geodiversity, regardless of the BQE (Figure 6a-c). In particular, the probability of high quality class clearly increases rather exponentially with geodiversity. For the good quality class, the probability increases and remains stable at geodiversity values around 20 for the macroinvertebrates and fishes; but for the benthic diatoms, it follows a u-shaped curve, and it appears to decrease at geodiversity beyond 15. Nevertheless, the cumulative probability of achieving the WFD requirements clearly increases with geodiversity, and in highly diverse environments, there is around 70% chance for a site to meet the WFD target if assessed based on macroinvertebrates and fishes.
For benthic diatoms, the respective likelihood is even higher and estimated to be around 90% (Figure 6d-f).

| DISCUSSION
Stream geodiversity, which describes the variety of hydrological and geomorphological features located at a site, was not associated significantly with taxa richness of benthic diatoms, benthic macroinvertebrates or fishes. In general, the role of habitat heterogeneity in promoting species diversity is one of the most common theories and highly cited concepts in ecology (Ricklefs & Dolph, 1993). Areas that have a greater habitat variety are more likely to provide an increase in habitat complexity and availability of resources and niches for community members. However, as shown in the present study, whether geomorphic and in-stream heterogeneity strongly influences taxa diversity remains an open question. Yet, we found that geodiversity was positively related with the three ecological indicators that are used for the ecological assessment of rivers in accordance with the WFD. Our results showed that all three indicators increased with both the geodiversity and the nutrient quality classification system index, which implies an improvement of the ecological quality as the hydromorphological conditions become more diverse, and the water quality gets higher. The effect of water quality appeared to be stronger than geodiversity, which validates the strong response of the biological indicators to stressors such as organic pollution, hypoxia and nutrient enrichment. Nevertheless, the positive effect of geodiversity indicates that geodiversity can reflect environmental changes caused by hydromorphological stressors. Hence, we suggest that geodiversity can be used as a proxy of hydromorphological modification caused by various human-induced stressors that may act simultaneously at the site level.
For instance, hydrological changes that cause low flow or even stagnant water conditions, will probably relate to low flow type richness affecting total geodiversity. Other anthropogenic interventions that are more harmful for the riparian zone and the channel F I G U R E 2 Boxplots of geodiversity (richness) for each class of hydromorphological modification. geomorphology, such as sand mining, channel overdeepening and channel realignment, are likely to have an enormous impact on riverine geodiversity. Practices that involve the alteration of the physical geometry of the channel, such as realignment, widening and deepening, often transform the natural complex channels to simple forms causing a simplification of the river network (Anim et al., 2018;Elosegi & Sabater, 2013;Stefanidis et al., 2020). Particularly in Greece, channelization is a common source of hydromorphological modifications that results to a loss of physical and morphological complexity of the riverine ecosystems (Stefanidis et al., 2020), possibly explaining the observed reduced geodiversity at 'severely modified' sites. Further studies could possibly investigate whether the loss of geomorphological and hydrological complexity is closely linked to low geodiversity and reduced habitat heterogeneity and ascertain if it affects the aquatic biota and the overall ecosystem functioning.
Several studies have shown, so far, a positive link between heterogeneous habitat conditions and freshwater biodiversity, but studies that directly link geodiversity with freshwater biodiversity are still scarce. For instance, macroinvertebrate communities are known to be affected by local habitat features, such as the channel dimensions, the water flow velocity and the substrate type (Buffagni, 2021;Graeber et al., 2017;Karaouzas, Theodoropoulos, et al., 2019;Theodoropoulos et al., 2020), but only a few studies have examined the relationship between stream geodiversity and macroinvertebrate diversity (Kärnä et al., 2018(Kärnä et al., , 2019. Kärnä et al. (2019) found a strong indication of a positive effect of catchment and local geodiversity on benthic macroinvertebrate and diatom species richness and suggested that geodiversity could be used as a proxy indicator for assessing stream species richness, with particular emphasis placed on the role of geodiversity as an additional tool for biodiversity conservation. Contrary to the findings of Kärnä et al. (2019), we did not find a strong relationship between geodiversity and diatom or macroinvertebrate richness, but we did find a good relationship between ecological indicators and total geodiversity, which implies that geodiversity could be related to changes of diatom and macroinvertebrate communities caused by anthropogenic disturbance. Our results are partially in agreement with those of Feld et al. (2014), who, even though they found weak relationships between macroinvertebrate species richness and hydromorphological alterations, found strong relationships between two ecological metrics (German Fauna Index and number of Ephemeroptera, Plecoptera and Trichoptera families) and the gradient of hydromorphological change.
Although there is a sheer amount of research that investigates the responses of riverine fishes to habitat heterogeneity and hydromorphological changes, including in-stream barriers, hydropeaking and droughts, we could not find any study that has previously examined the potential effect of geodiversity on fish richness or on fish related ecological indicators. Fishes are considered a good indicator of hydromorphological change Sun et al., 2022), but the relationship between fish richness and hydromorpological gradients remains rather ambiguous. In particular, studies that explored the responses of fishes to hydromorphological restoration efforts did not find any significant change in the total richness but highlighted significant changes in the functional community structure that could be associated with an overall improvement of the ecological quality (Haase et al., 2013;Schmutz et al., 2016).
F I G U R E 3 Boxplots of geodiversity (richness) per quality class per biological quality element (BQE).
Restoration studies have also improved our understanding on which stream features are more important for certain groups of aquatic organisms and thus can provide substantial information on how geodiversity may improve ecological quality, given the fact that stream geodiversity studies are lacking. For instance, Mueller et al. (2014) found that gravel introduction as a measure for restoring reaches to a natural condition had a strong effect on macroinvertebrate communities, including an increase in the abundance of species with high conservation value (e.g., Plecoptera). They also verified that the constructed gravel habitats acted as spawning habitats for several fish species, which highlights that stream features such as the occurrence of point and mid-channel gravel bars is of paramount importance for fish spawning. There are several examples of restoration projects that aimed to recover the natural characteristic of an altered river reach by restoring the natural planform and recreating physical habitats and ultimately exhibited a significant improvement of ecosystem functioning (Dyste & Valett, 2019;Muhar et al., 2016;Poppe et al., 2016). Another study highlighted the role of introducing large woody debris in promoting macroinvertebrate diversity (Deane et al., 2021). Although the effect of geodiversity on aquatic communities was not explicitly investigated in the aforementioned F I G U R E 4 Relationships between (a) diatom species richness, (b) macroinvertebrate family richness and (c) fish species richness with geodiversity, fitted with generalized additive models (GAMs). The shaded area represents the 95% confidence intervals. studies, their results implied that more diverse conditions in terms of substrate and features might be better for macroinvertebrates and fishes.
As a general finding, increased geomorphological and hydrological diversity might not have an effect on species richness, but it appears that it has a substantial impact on the habitats promoting favourable conditions for certain species. This change in the community composition that is caused by the increased geodiversity may justify the improved ecological quality in natural-like sites compared to heavilymodified ones . Since we identified a significant decline of geodiversity at hydromorphologically heavily impaired sites and we also associated higher geodiversity with higher ecological quality, we would likely consider that hydromorphological modifications are responsible for geodiversity loss, which in turn increases the likelihood for lower ecological quality. A possible explanation is that the geodiversity loss is linked with community changes and loss of ecosystem functions associated with good ecological conditions. To this end, further beta diversity analyses that employ detailed community data could provide insight on how species composition changes across a geodiversity gradient.
Concerning our findings about weak relationships between geodiversity and diatom and macroinvertebrate and fish richness, it possibly shows that other environmental descriptors play a more important role than the richness of hydrogeomorphological features in shaping species richness. Indeed, freshwater species richness and endemism patterns are the result of climate, productivity and biogeographical T A B L E 2 Generalized additive model results on adjusted R 2 and p-value effects for each predictor.

Response
Adj T A B L E 3 Results from the ordered logistic regression analysis conducted for each BQE. The proportional odds ratios indicate the change of odds for a certain quality class for every one unit increase in geodiversity and NCS. They are derived following the proportional odds assumption that the odds ratios across the five classes are the same.  history, so factors such as river connectivity, upstream barriers and isolation are likely to provide a better understanding of the distribution of fish species at the regional scale than local scale factors, such as hydromorphology or pollution (Griffiths et al., 2014;Oikonomou et al., 2014;Sutela et al., 2020). Previous works on freshwater fishes have shown that total species richness was also influenced by drainage characteristics and factors related to area, energy availability (i.e., net primary productivity) and climatic parameters (annual rainfall and average annual temperature) (Boll et al., 2016;Griffiths et al., 2014;Pelayo-Villamil et al., 2015;Tedesco et al., 2017;Tisseuil et al., 2013). Otherwise, benthic diatoms are primarily driven by resource availability, mainly nutrients (both nitrogen and phosphorus) and light, but also by pH, grazing and elevation (Liess et al., 2009;Schneider et al., 2013;Taxböck et al., 2020). Furthermore, diatoms are microorganisms, and the estimation of the species richness at a given site is influenced by the sampling strategy, which may not ensure the collection of rare species (Taxböck et al., 2020). In addition, we must point out that we used macroinvertebrate family richness instead of species richness, and this could potentially explain the weak connection between geodiversity and macroinvertebrate richness.
Nevertheless, we found that ecological indicators for all three BQEs responded positively to geodiversity. Hence, more diverse river sites in terms of hydrological and geomorphological conditions are more likely to host aquatic communities that resemble those communities that are found at sites with slight deviation from the reference conditions. The results of the ordinal regression analysis allowed us to estimate the likelihood of achieving the WFD target (i.e., good quality) in relation to the geodiversity measured at each site. Although we showed that the likelihood of achieving the WFD target greatly increases by improving the overall water quality, we also found that increasing geodiversity has a substantial effect on improving the ecological conditions. This result is of great importance because it can further aid managers and decision makers to identify critical thresholds of hydromorphological conditions that need to be achieved in order to meet the WFD goal. Our results also highlight the valuable contribution of measures that enhance hydromorphology and improve geodiversity in restoration. Furthermore, we found that even at good water quality conditions, the likelihood of meeting the WFD target remains relatively low unless geodiversity increases. That is to say that measures that target exclusively nutrient pollution might not be sufficient. Considering that most European rivers are impacted by the joint effect of pollution and hydromorphological degradation and that despite the management efforts, the ecological quality has not reached the WFD target yet (Birk et al., 2020); it is crucial to highlight the need for establishing restoration practices targeted to the improvement of hydromorphology.

| CONCLUSIONS
Stream geodiversity is of vital importance for the aquatic communities and is often considered that it could promote aquatic biodiversity by increasing habitat heterogeneity. Although our findings did not show any clear connection between geodiversity and taxa richness for benthic diatoms, macroinvertebrates or fishes, we did find signs of positive relationships between geodiversity and the three associated ecological indices. An important finding was that stream geodiversity may reflect the impact of hydromorphological modifications, particularly for sites that have been severely modified. Furthermore, we showed that geodiversity is higher at sites with 'good' and 'high' ecological quality than at sites with worse ecological conditions, regardless of the assessed biological quality element. Hence, it is worthwhile to further explore the potential of geodiversity as a proxy indicator of ecological quality.
Our analysis revealed that the likelihood of a stream site to achieve 'good' ecological quality increases with geodiversity. Our results further highlighted the need for reconsidering restoration measures that may improve and enhance geodiversity features, since even at good water quality conditions, the likelihood of meeting the WFD target will remain low unless hydromorphology improves. Nevertheless, future research that focuses on finer spatial scales (e.g., habitat) might reveal stronger relationships between aquatic biota and geodiversity metrics.
To this end, it could be of the utmost interest to explore relationships between additional facets of aquatic biodiversity, including functional diversity and community turnover across gradients of geodiversity change in order to pinpoint the role of stream geodiversity as an indicator of ecosystem function and health.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.