Assessment of the Kalaji watershed health based on hydrological and geomorphological criteria: relative and absolute approaches

Watersheds provide services to living organisms, referred to as watershed services (Porras 2008). Watershed services have a vast wide scope and can be classified into four general categories: providing services, regulating services, cultural services, and supporting services (Hamel et al. 2018). Watershed health is a term used to describe the ecological condition of a watershed (Jones et al. 2002). A healthy watershed has a structure and function that supports its ecosystems and has appropriate services for the relevant biological communities (Ahn & Kim 2019). In a healthy watershed, the quality of water and habitat is suitable to support native species, and the amount of natural cover is such that its geomorphological and hydrological processes are within normal limits (Ahn & Kim 2017). Since the level of watershed services has a direct relationship with the level of watershed health, checking the health status of watersheds is very important (Jones et al. 2002).

During the last few decades, the rapid growth of the population and the need to provide food have put excessive pressure on all types of renewable resources, including water, soil, and natural vegetation (Hou et al. 2016). Excessive pressure on watershed resources has taken the hydrogeomorphic processes of watersheds out of the natural range and reduced the health of watersheds. The drying up of wetlands (Tussupova et al. 2020), the drop in the level of underground water tables (Madani 2014), severe erosion (Mosaffaie & Talebi 2014), the occurrence of numerous floods (Karimi Sangchini et al. 2022), the intensification of desertification (Rashvand et al. 2013), the increase of dust centers (Soltani et al. 2023), and the reduction of the area of natural habitats (Holland et al. 2016) are among the signs of the deteriorating health of watersheds (Mosaffaie & Salehpour Jam 2021; Jam et al. 2022).

To apply effective management to watersheds, first of all, it is necessary to determine the correct management policy including maintaining the status quo or revitalization based on the health status of the watershed (Jam & Mosaffaie 2023). In recent years, a variety of research has been conducted on watershed health assessment (Sadeghi & Hazbavi 2017; Hazbavi et al. 2018; Alilou et al. 2019; Hazbavi et al. 2019; Sadeghi et al. 2019; Mirchooli et al. 2021; Mosaffaie et al. 2021; Chamani et al. 2023; Gatgash & Sadeghi 2024). Two features are noteworthy in past research: (1) wide variety in sub-indices of watershed health and (2) relativity of the assessments.

Finding indicators that will be both valid and feasible is often the most challenging design issue in a monitoring system or evaluation (Gari et al. 2015). Watershed health sub-indices used in past research can be divided into two general categories: (1) indicators that directly represent the status of watershed health (hydrological, geomorphological, and habitat criteria) and (2) indicators that do not directly represent the status of watershed health but can potentially affect it. Sub-indices such as soil erosion rate, flood potential, groundwater loss, water quality, and habitat condition are among the indicators that can directly provide an overview of different criteria for watershed health (Tsai et al. 2021). Among the sub-indices of the second category are mean annual precipitation, mean annual evapotranspiration, population density, area of agricultural land with slope >25%, the area under anthropogenic activity, environmental sensitive area index, the slope of contribution area in runoff generation, landscape diversity index, normalized difference vegetation index, landscape dominance index, rangelands area ratio, landscape fragmentation index, and drainage density, which do not directly represent the status of watershed health criteria but can potentially affect them. In addition to confusion, the large number and variety of watershed health sub-indices may prevent obtaining a correct view of the watershed health status if they are not related or if there is colinearity with other sub-indices.

In addition to the wide variety of sub-indices, another characteristic of past research has been the relativity of assessments. In almost all past research, temporal or spatial variations in watershed health have been assessed relatively. In the relative approach, the detailed assessment of the situation of the subject under study (watershed health) is not relevant, but different treatments are only compared in terms of the subject in question (Linkov et al. 2006; Podviezko & Podvezko 2014). Although the information obtained from relative assessment is suitable for prioritizing watersheds for protection programs, there is a need to employ an absolute approach, the results of which indicate the actual health status of the watershed, rather than measuring the status of the sub-units relative to each other. Therefore, this research aims to assess the health of the Kalaji watershed (as a pilot watershed) using the key and direct indicators based on two relative and absolute approaches and to highlight the differences in the results obtained from the application of these two approaches. The results of this research provide a better understanding of the relative and absolute assessments of watershed health and can be considered as a preliminary for the development of absolute models of watershed health assessment.

The average annual precipitation equal to 434.9 mm and the average annual temperature equal to 16.9 have brought a semi-arid climate (according to the Demartini method) to this watershed. According to the ambrothermic curve, the driest months range from May to October. The lithological units of the basin include green to gray shale (K_s), glauconitic sandstone (K_at), Quaternary loess (Q_geh), new river sediments (Q_al), and old alluvium (Q_f1). Due to the large extent of thick-layered loess, various geomorphological facies and erosion types, including sheet, rill, piping, badland, and gully, can be seen in the area. Agricultural activities and livestock grazing, regardless of soil limitations and ecological capacity, have aggravated erosion in the region. These features have also caused many landslides in the study area. Sediments caused by severe soil erosion have decreased the water quality of waterways and also reduced the useful capacity of the Bustan Dam reservoir located downstream of the watershed. The Soil and Water Conservation Bureau and the Council of Agriculture both focus on increasing vegetative land cover and stream conservation.

The establishment of indicators depends on the definition of what one wants to indicate (Bosch 1967). The choice of key indicators can make a critical difference in the results of an evaluation (Gari et al. 2015; Czúcz et al. 2021). In this research, watershed health indicators were selected according to the concept of watershed health. Therefore, according to the definition of a healthy watershed, five criteria, including erosion status, flood status, landslide status, water quality, and habitat status, were chosen to explain the geomorphological, hydrological, water quality, and habitat conditions of the sub-watersheds.

Good indicators are simple, variable, valid, clearly defined, measurable, reliable, and quantifiable (Bosch 1967). The number of indicators should not be too numerous to clutter the overview nor too few to prevent the provision of sufficient information (Czúcz et al. 2021). Therefore, in this research, six indicators, including specific erosion, specific peak flood discharge (return period = 50 years), landslide density ratio (Ldr) (used only in the relative approach), the area percentage of landslide hazard class higher than medium (used only in the absolute approach), specific sediment yield, and the percentage of natural cover were selected as watershed health indicators. In the following, the data collection method for each of the indicators is explained.

The Ldr indicator is only applicable in the relative approach because it represents the density of landslides in each sub-watershed only compared to the entire study area. So, for the absolute approach, the landslide hazard map of the watershed was prepared using the Haeri–Sameie method (Armin et al. 2019). This method has been developed for the Golestan and Mazandaran provinces of Iran where the study watershed is located. The formula of this method calculates the landslide hazard index from the combination of thematic maps of seven factors, including lithology, slope, fault length, road and river length, precipitation, rainfall intensity, and earthquake acceleration. The resulting landslide hazard map was classified into seven classes (Sarfaraz et al. 2021). The other factors that are required by this method are documented in the literature (Armin et al. 2019; Sarfaraz et al. 2021).

The biggest threat to biodiversity to date has been anthropogenic activities, which have transformed natural habitats into agricultural, industrial, residential, mining, and other unnatural land uses (Prakash & Verma 2022). In this research, the percentage of natural cover was used as an indicator to show the status of the habitat criterion. In this regard, among the land uses in the watershed, ranges, and river terraces were considered land uses with natural cover, and agricultural lands, orchards, and residential areas were considered areas without natural cover.

The analytic hierarchy process (AHP) was used to calculate the contribution (weight) of each indicator (Saaty 1980). The AHP technique has a particular application in weighing factors based on multiple criteria and pairwise comparisons (Saaty 2008). The weighting process of this research can be summarized in four steps: creating a matrix to compare indicators; making pairwise comparisons of indicators using a scale of 1–9 (conducted by the opinions of 10 environmental experts); calculating the weights of indicators, and eventually checking the consistency of the evaluations.

In this research, the health of the 25 sub-watersheds of Kalaji was assessed based on two relative and absolute approaches. In this regard, the Watershed Health Index (WHI) was calculated according to relative and absolute methods (below subsections). To better discriminate the differences of WHI resulting from each approach, the ArcGIS 10.8 software was applied to categorize the range of WHI [0, 1] using an equal interval method into five classes where (0.00–0.20), (0.21–0.40), (0.41–0.60), (0.61–0.80), and (0.81–1.00), respectively, represent unhealthy, relatively unhealthy, moderately healthy, relatively healthy, and healthy watersheds.

By these equations, the maximum and minimum values of each indicator were standardized to 1, and zero, respectively, and the rest of the values were standardized between zero and one.

In the absolute approach of watershed health assessment, the range of index values is not limited to the values within the studied watershed and includes values that are recorded even outside the study area. Therefore, in this approach, the values of the sub-indices of the study area are compared with the values of a larger scale (in this research with the values recorded in the whole of Iran). For this purpose, Table 1 was developed to assign a health score for the indicator values of the watershed compared to national ranges of indicators. In this regard, the results of flooding rate classification (Porhemmat 2018; Sharifi 2022), soil erosion rate classification (Sharifi 2015), and sediment yield classification (Sharifi 2017) were collected for Iran's watersheds. The percentage of area under more than medium classes of landslide hazards (Haeri–Sameie method) was also used to index the landslide criterion.

Some of the main parameters and indicator values at 25 sub-watersheds of the study area are presented in Table 2.

Table 3 shows normalized values and partial health scores of sub-indicators that have been used to calculate the relative and absolute WHI, respectively.

In this research, the selection of watershed health criteria and indicators were established based on indicators that directly represent the status of various aspects (geomorphological, hydrological, and habitat) of watershed health. In addition to confusion, employing numerous sub-indices may prevent a correct view of the health status of the watershed if they are not related or if there is colinearity with other sub-indices. So, indicators such as population density, water consumption, climate change, and other indicators that do not directly represent the health status of the watershed were not included in the analysis. This is while most previous studies (Hazbavi & Sadeghi 2017; Hazbavi et al. 2020; Sadeghi et al. 2023) employed a complex of direct and indirect indicators to investigate watershed health. Entering indirect indicators can cause the results to deviate from the reality of watershed health, although these indicators can affect watershed health in the future. Since the establishment of indicators should be based on the concept of watershed health (what one wants to indicate (Gari et al. 2015)), the authors suggest the use of indicators that only directly represent the status of watershed health.

Due to the diversity and multiplicity of watershed health criteria and indicators, extensive environmental data and information (temporal and spatial) were collected. Also, to cover the data gap, various methods such as direct measurement, results of previous studies, experimental models, multicriteria decision-making, remote sensing, and statistical generalization from areas with data to areas without data were used. However, due to the lack of data and information, some criteria were indexed by indicators that do not directly express the status of these criteria. Among the most important of these criteria were water quality and habitat condition, which in this research were indexed by the sediment load and the percentage of natural land indicators, respectively. This limitation can potentially affect the results of watershed health assessment in both the relative and absolute approaches. So, designing and establishing a WHI monitoring system can remove such limitations and, as a result, promote the accuracy of the overall index of watershed health.

The results of watershed health assessment with two relative and absolute approaches indicate a significant difference in the results of these two approaches. Figure 6 implies that in the relative approach, the health class of the sub-watersheds was distributed in three classes so that healthy and unhealthy classes were not calculated for any of the sub-watersheds. This is while, in the absolute approach, the health class of the sub-watersheds was distributed in four classes, and the healthy class was not calculated for any of the sub-watersheds. The highest frequency in relative and absolute approaches belongs respectively to the relatively healthy (64%) and relatively unhealthy (80%) classes. The average health index for sub-watersheds in relative and absolute approaches were also calculated as 0.601 and 0.308, respectively. These findings imply that the relative approach has resulted in higher values for the overall WHI. The reason for this is the ranges of values of the sub-indices, which in the relative approach belong to the studied watershed (Kalaji), but in the absolute approach belong to a wide region (in this research, the whole country of Iran).

It should be kept in mind that the purpose and function of these two approaches are also different. The goal of the absolute approach, which can be used for even a single watershed, is to gain a correct understanding of the health status, and if there are several watersheds or sub-watersheds, they can also be prioritized based on the results. This is despite the fact that the function of the relative approach is only for prioritizing several sub-watersheds and with this approach, it is not possible to achieve a correct understanding of the health status unless the area under study is large enough to include both unhealthy and healthy sub-watersheds (reference watershed).

Therefore, choosing a disproportionate approach can cause deviation from the goal, and the difference in the results of watershed health can even affect the adoption of a suitable policy for watershed management. For example, based on the results of the absolute approach of this research, the health status of the Kalaji watershed is relatively unhealthy (0.308) which is not satisfactory and an improvement policy may be considered for this situation. Meanwhile, the results of the relative approach indicate a relatively healthy status (0.601) which is almost satisfactory, and the policy of maintaining the status quo may be considered for this situation. So, the difference in the results of the two approaches can lead to changes in the policies adopted for watershed management. The results of this research emphasize that before assessing the health of the watershed, the proportionate approach should be chosen according to the intended goals.

Finally, considering the unsatisfactory health status of the Kalaji watershed, it is suggested that necessary measures be taken to improve this situation. In this regard, the reasons for such status should be analyzed by a cause and effect model (such as drivers, pressures, state, impact, and response (or DPSIR) framework) and appropriate responses can be identified (Salehpour Jam et al. 2021b, 2021a).

There are two categories of watershed health assessment approaches: relative and absolute. Past research has often focused on the relative approach of watershed health. This research aimed to highlight the difference between the results obtained from the relative and absolute approaches of watershed health assessment. So the health status of 25 sub-watersheds of the Kalaji watershed (as a pilot watershed) was assessed based on these two approaches using the key indicators. However, the lack of a WHI monitoring system caused some limitations in the selection of indicators for some watershed health criteria, so some criteria (such as water quality and habitat condition) were indexed with somewhat indirect indicators.

The findings indicated a significant difference in the results of the two approaches. In general, the relative approach provided a higher health index and more healthy class than the absolute approach. In the relative assessment approach, a clear vision of the watershed health is not provided and the health of watersheds is only compared to each other. The results of the absolute approach imply that the health of the Kalaji watershed is not satisfactory and should be improved. So, it is suggested that future studies seek to determine appropriate responses to improve this situation by applying cause and effect analysis (such as DPSIR).

Therefore, the results of this research emphasize that due to the huge difference in the results obtained from the two approaches, it is necessary to be careful in the watershed health assessment process, so that the appropriate approach should be chosen according to the desired goal. Future research can also apply the two approaches presented in this research in other watersheds with different geographical and climatic conditions and highlight the differences in the results of the two approaches. Overall, this study provides a better understanding of the relative and absolute approaches of watershed health assessment and can be considered as a preliminary for the development of these models, especially the absolute ones.

This research which was conducted as a research project with code 124-29-29-018-01033-010622 was jointly supported by the Soil Conservation and Watershed Management Research Institute (SCWMRI) and Gorgan University of Agricultural Sciences and Natural Resources. The authors sincerely appreciate the members of the watershed management group at the SCWMRI for their valuable suggestions. The authors are also grateful to the anonymous referees for their valuable comments.

All relevant data are included in the paper or its Supplementary Information.

The authors have declared that there is no conflict of interest.