In agricultural land-use management, flood hazard can be defined as the probability of exceeding critical peak discharge, resulting in economic damage and raising food security concerns (O’Connell et al., 2007). Usually, runoff generated from small streams is considered less severe. Still, there is a high chance that these insignificant changes will propagate, resulting in higher runoff at a larger catchment scale. The runoff changes on the local scale need to be monitored, as they can result in flooding. Modern flood management is typically designing a system that prefers allowing water to be stored by restoring flood plains. However, these restorations significantly affect the agricultural areas, ultimately burdening productivity (Brémond et al., 2013).

Prior research has studied the effects of different conservation practices on flooding. Practices referring to reduced or no-till farming methods minimized soil disruption and allowed soil biota and structure to reorient. The presence of soil biota often led to the formation of micro and macro sized pores (Sun et al., 2020; Hovis et al., 2021). These pores at larger scales absorbed the excess water during rainfall, thus delaying the peak flows and improving soil infiltration capacity (De Laney, 1995; Hovis et al., 2021). Practices such as bioretention, vegetated grass strips, and filter strips reduced runoff through accelerating storage, evapotranspiration, and infiltration. Wild and Davis (2009) reported that these practices successfully reduced peak volume by 42%. Dakhlalla and Parajuli (2016) reported grassed waterways were effective in reducing peak flow (up to 8.4%) by increasing the roughness of the surface. Similarly, a study found grassed waterways were effective in attenuating muddy flooding (Evrard et al., 2008). There was a 69% reduction in peak discharge. The study also reported that the efficacy of grassed waterways could be increased with the combination of alternating farming practices, such as reduced tillage.

Studies have also referred to cover crops’ effectiveness in reducing the peak flow during high flood frequency scenarios (Muenich et al., 2016; Antolini et al., 2020). Since these crops were planted once primary harvesting is complete, they were most effective at controlling flood peaks during winters or early springs (Hovis et al., 2021). It was also found that in the post-flooding scenario, applying a cover crop immediately after the soil was dry promotes the growth of microorganisms, which are essential for nutrient cycling (Al-Kaisi, 2019). Wetlands could be one potential solution to counter peak flow problems as it delays flood water movement and reduces downstream water flow. A study identified two types of wetlands, wetland retention basins and forested wetlands, in countering stream velocity and sediment transport (Hovis et al., 2021). The wetland’s location in the watershed significantly impacts its effectiveness in lowering the flood peaks. A global study cited the efficacy of multiple small freshwater marshes at the watershed’s upper reach in reducing the peak flow (Mitsch and Gosselink, 2000). These small multiple units mitigated the chances of wetlands’ destruction from abnormal precipitation. It can also act as a detention basin which can delay and reduce peak flows. Similarly, Ogawa and Male (1983) reported the effectiveness of wetlands when located downstream.

Apart from the practices cited above, this review has identified other practices that have effectively reduced the peak flow (magnitude based on spatial scale at which studies were carried out). The practices included utilizing afforestation (up to 80%) (Huang et al., 2003), detention ponds (up to 6%) (Dakhlalla and Parajuli, 2016), grade stabilization structures (up to 25%) (Kaini et al., 2008), grasslands (40%–45%) (Gerla, 2007), parallel terraces (up to 3.1%) (Dakhlalla and Parajuli, 2016), soil bunds infused with grasses and short trenches (up to 55%) (Sultan et al., 2018), and urban forests (3%–8%) (Lormand, 1988).

Surface Runoff - Surface runoff refers to water movement on the land surface due to excessive precipitation, stormwater, snowmelt, or any other source that is no longer infiltrated in the soil. Excessive surface runoff often results in flooding and is one of the major sources of non-point source pollution (Environmental Protection Agency, 2018). These pollutants can be sediments, pesticides, and nutrients, which heavily affect off-site water bodies (Wallace et al., 2017). The variables that control surface runoff include temperature, precipitation, relative humidity, and wind speed (Kurothe et al., 2014; Adimassu et al., 2017; Luo et al., 2017).

Different conservation practices were found to be effective in reducing surface runoff. Straw mulch was effective in improving the infiltration rate and the cohesive strength against erosive forces (Olkeba et al., 2012). Antolini et al. (2020) found the effectiveness of cover crops in reducing runoff as they provided a protective layer to the surface, which reduced surface runoff and runoff velocity. Conservation practices referring to no-tillage, ridge-farming tillage, and stubble mulch farming tillage effectively reduced surface runoff (Kurothe et al., 2014). The study concluded that no-tillage resulted in the minimum disturbance in soil, which promoted filtration and reduced runoff (up to 16%). Similarly, ridge formation along a contour and tillage along the contour created a barrier to overland flow, which increased the surface detention time and, thus, the infiltration period. This reduced surface runoff by up to 69%.

Farahani et al. (2016) discussed the effectiveness of contour farming in reducing surface runoff. Contour cultivation reduced the runoff by 10% compared to when crops were planted perpendicular to the slope. The study also found that the effectiveness of contour farming was further increased with the combination of a no-tillage system. Bunds were also found to be effective in controlling runoff because of their ability to retain water across the slope. The studies found different soil bunds, such as grade, level, and stone, to reduce runoff by 25.2%, 60.1%, and 34.4%, respectively (Herweg and Ludi, 1999; Araya et al., 2011; Adimassu et al., 2014). Similarly, Blanco-Canqui et al. (2004) discussed the effectiveness of buffer strips in reducing runoff through changing hydrologic pathways, resulting in a higher infiltration period and reduced surface flow. However, studies have observed that buffer strips were only effective for reducing sediment and nutrient load (Chaubey et al., 2010; Wagena and Easton, 2018). Such cases indicated that conservation practice effectiveness changes with weather, soil, and cropping system.

The deterioration in water quality severely impacts living beings, as it adversely affects crop productivity and environmental sustainability (Hurni, 1996; Adimassu et al., 2017). The upcoming generation of farming and landscape requires multifunctional and ecological sustainability (Kassam et al., 2014). In achieving these requirements, conservation practices can play a crucial role (Santhi et al., 2006; Kassam et al., 2014; Adimassu et al., 2017; Singh et al., 2018). The impact of conservation practices on water quality in this review was studied by analyzing their influence on soil erosion, soil sediments, and soil nutrients.

Soil Erosion —Soil erosion refers to the removal of the topmost layer of soil on the earth’s surface. These fine layers have the highest amount of plant nutrients and have a significant concentration of agrochemicals, soil biota, soil carbon, and organic matter. Soil erosion reduces the effective mass and depth of in situ soil, thus reducing the storage capacity of moisture and nutrients (Rickson et al., 2015). Several factors affecting soil erosion are existing vegetative cover, deforestation rate, soil and water conservation measure, population growth, and climate change (Zhao et al., 2013). Soil erosion has caused severe damage to the food production system (Adimassu et al., 2014). It degrades the land quality as it deteriorates soil conditions. Additionally, soil biota, which is severely affected by soil erosion, reduces fertility and infiltration (Pimentel et al., 1995). Different literature has provided their input on how soil erosion affected crop productivity. For instance, Young (1989) found that eroded soil decreased crop productivity by 50%–75%. Similarly, Bakker et al. (2004) found a loss of 4.3%–26.6% productivity per 10 cm in the soil layer.

Studies have stated their findings regarding the effectiveness of different conservation practices at the watershed and field levels. It is important to note that the effectiveness of these practices varies with change in weather, soil, and cropping system. According to a study, the potential conservation practices that can counter the increasing soil erosion rate were crop rotation, strip cropping, grass strips, cover crops, mulches, living mulches, agroforestry, and windbreaks (Pimentel et al., 1995). Other research reported the effectiveness of terracing, afforestation, natural rehabilitation, and check-dams construction in reducing the impact (Zhao et al., 2013). Practice such as residue and tillage management addresses the amount, orientation, and distribution of crop and plant residue. Crops are grown in narrow slots or tilled strips to reduce soil disturbance, thus, reducing erosion (Hengsdijk et al., 2005). Merten et al. (2015) suggested that the presence of residue effectively controlled erosion as it dissipated rain-drop energy, improved infiltration capacity, and increased surface roughness, thus reducing the runoff velocity. Further, the mulching treatment was effective as it could minimize erosion up to 98% (Adimassu et al., 2014). Conservation practices such as graded and level soil bunds and grass strips were also found to be effective in retaining the water, thus minimizing erosion (Herweg and Ludi, 1999). The combination of different practices was found to be effective in countering erosion. Farahani et al. (2016) found the effectiveness of contour farming in reducing erosion when combined with conservation tillage. Likewise, tied ridge and field yard manure effectively reduced soil erosion (Araya et al., 2011). Implementation of these practices can be effective. However, demographic and socioeconomic characteristics are among the factors that influence conservation measures and, subsequently, erosion. Effective planning for applying soil conservation measures, new techniques, and modifications in educational/training programs can effectively increase awareness among the population regarding soil erosion (Ersedo, 2021).

Soil Sediments —The movement of soil sediments can be attributed to soil disturbance. Runoff carries these sediments from a non-point source are one of the major sources of surface water pollution (Muñoz-Carpena et al., 1999). Sediment transport is often interlinked with soil erosion and nutrient loss. Soil erosion on a larger scale result in sediment deposition in water bodies (Vanacker et al., 2003). These sediments are often accompanied by nutrients that degrade water quality and soil health (Zhao et al., 2013).

Although conservation practices have diversified benefits, some were highly effective in controlling sediment loading. Conservation practices such as filter strips, field borders, and grassed waterways effectively trapped sediment movement (Dillaha et al., 1986; Magette et al., 1989; Muenich et al., 2016). These practices reduced sediment transport from runoff through different processes such as filtration, deposition, adsorption, and decomposition (Muñoz-Carpena et al., 1999). These processes reduced flow resulting in reduced sediment transport capacity. Studies have found that these practices, on average, can reduce sediment transport by up to 40%–45% (Santhi et al., 2006; Her et al., 2016). Studies have reported the effectiveness of the water and sediment control basin, which reduced sediment loading (Her et al., 2016; Turnbow, 2021). Water and sediment control basins (WASCoBs) reduced the erosive potential of surface water by holding it with berms and allowing it to flow through stable outlets (Turnbow, 2021). Water exit from stable outlets helped in reducing sediment transport by allowing more deposits. Her et al. (2016) found that WASCoBs effectively reduced sediment transport by up to 47% at the field level. Similarly, slash-and-burn shifting cultivation with long rotation (1-year farming and 8 years uncultivated) effectively reduced sediment loss. Long rotation helped increase soil cover, thus reducing sediment movement (Valentin et al., 2008).

The literature also showed that the no-tillage condition in farming helped in reducing sediment deposition (Kurothe et al., 2014; Her et al., 2016). The study concluded that minimum disturbance under no-tillage conditions and mulch surface protection effectively reduced sediment concentration. Cover crops were effective in reducing sediment load. Her et al. (2016) found that sediment transport was reduced by 48%–72% with the implementation of a cover crop as they provided continuous cover to the soil. Contour farming also effectively controlled sediment transport (Santhi et al., 2006; Farahani et al., 2016). Growing crops perpendicular to the slope direction prevented sheet and rill erosion and, thus, sediment movement (Santhi et al., 2006). Further, check dams were also considered an effective way to reduce sediment movement. On-stream check-dams were often built upstream of a watershed, which ceases the infiltration of course- and medium-sized sediments (Mishra et al., 2007). The study concluded that the presence of check-dams could reduce sediment transport by up to 64% at a watershed level. These results indicated the effectiveness of practices in reducing sediment transport.

Soil Nutrients —Soil nutrients are a complex fusion of organic remnants in varying stages of decomposition (Hofman, 2004). Conventional practices such as continuous cropping and tillage farming lead to the exhaustion of organic matter in the soil. This deteriorates soil structure, moisture, and nutrient retention (Stockdale et al., 2002). The movement of nutrients poses a severe threat to the agricultural system. The studies have indicated that the movement of nutrients can also occur through an irrigation network resulting in non-point source pollution (Quemada et al., 2013). In the context of soil nutrients, conservation practices aim to reduce nutrient losses (organic matter, nitrogen content, and phosphorus content). Surface runoff and leaching are often considered primary sources of nutrient loss. In such scenarios, practices that effectively control runoff can reduce nutrient loss.

In the context of conservation practices, cover crops were effective in reducing nutrients’ leaching. A meta-analysis found that the winter cover crop has the potential to reduce leaching by up to 70% (Tonitto et al., 2006). Similarly, one study showed nitrate leaching reduction by up to 56% through nonleguminous cover crops (Thapa et al., 2018). They concluded that the effect varied with soil type, cover crops, and climate conditions. It is observed that a combination of practices often increased the efficiency in the performance of practices. Studies have reported the effectiveness of no-tillage with the cover crop, which significantly decreased movement of organic nutrients in the soil (Rickson et al., 2015). Similarly, Baker (2001) found the effectiveness of crop rotation, no tillage, and nutrient management practices in reducing leaching by up to 30%. Retaining the crop residue can also be one feasible way to restore the organic content in the soil, as the top layer has a higher concentration of soil nutrients, making it necessary to conserve the top layer of the soil (Rickson et al., 2015).

Conservation practice referring to filter strips has also shown the ability to slow the runoff rate, allowing the infiltration process to reduce nutrient losses (Magette et al., 1989). Studies have found a reduction of up to 80% in nutrient transport post-implementation of filter strips (Dickey and Vanderholm, 1989). Likewise, in a qualitative study, Gilley and Risse (2000) reported the effectiveness of manure in reducing nutrient loss, while Adimassu et al. (2014) reported the efficacy of graded soil bunds in reducing the organic matter loss, nitrogen, and phosphorus content. These cases indicated that nature-based solutions have the potential to reduce nutrient loss.

One of the challenges associated with the assessment of conservation practices is the limited availability of spatially and temporally extensive conservation practice data (Her et al., 2016; Singh et al., 2018). The literature review demonstrated that this is a global challenge for researchers. For instance, Sullivan and Batten (2007) created geographic databases of conservation practices based on the information listed in previous files and maps. Jha et al. (2010) and Tomer et al. (2008) extracted the information from field surveys. Gowda et al. (2003) carried out aerial imaging to detect different practices applied in their region, while Pennington et al. (2008) conducted interviews with farmers. Notably, all these studies could not find complete information on conservation practices associated with their region, which limits their ability to carry out an exhaustive assessment. One way to overcome the challenges referring to conservation practices information, land use, soil, and weather data, can be through the advanced application of remote sensing. Although, several studies have considered the application of remote sensing for identifying the type and location of conservation practices to assess their effectiveness (Gowda et al., 2003; Tomer et al., 2008; Osmond et al., 2012b). Future studies can work more towards high-resolution mapping of different conservation practices over time, enhanced land use–land cover (LULC) information, and improved remotely sensed meteorological datasets such as cloud cover, surface temperature, snow, ice cover, etc., across higher temporal and spatial scales. Further, survey data can be considered a reliable source of information as it directly comes from the individuals who have implemented conservation practices and have seen its impact on flooding and water quality (Carletto et al., 2021). They can be used as a validation instrument for other data sources. The information obtained from the opinion of farmers is also helpful in cost-benefit analyses and understanding individuals’ attitudes toward nature-based solutions (Traoré et al., 1998). This review finds that application of such datasets is not frequently seen across the literature, though it can potentially improve the accuracy of assessments. Survey/feedback data can be helpful for assessments at field and watershed scales. However, they may be less effective on a larger scale due to a lack of spatial and temporal resolution in information, leading to sizable errors.

Limited monitoring stations often result in the unavailability of precise spatial data, especially when it comes to modeling non-point source pollution (Sharpley et al., 2002). This inconsistency in information across different sub-watersheds within a watershed presents an enormous task to handle. The extrapolation of these datasets is challenging, as most models are developed for specific soil and climatic conditions (Hofman, 2004). Moreover, with spatial variability, the nature and properties of soil and water samples show dynamic inconsistency (Adimassu et al., 2017). In the absence of sufficient observations, studies usually rely on different calibration parameters from watersheds where similar modeling efforts are performed, which increases the uncertainty associated with assessment (Antolini et al., 2020). The lack of available datasets also limits the model’s ability to capture different conservation practices’ responses to water quality and soil health (Sharpley et al., 2002). In such conditions, there is a need to focus on improving the data quality, which can be achieved by increasing the number of observation networks and improving the efficiency of monitoring devices to detect changes occurring in the watershed (O’Donnell, 2012). Different studies in this context have come up with various methods like the Internet of Things (IoT), such as global positioning-based remote-controlled monitoring (Suma et al., 2017) and smart sensors (Mathurkar et al., 2014). Future studies can focus on improving the monitoring network capacity and performance to increase assessment capabilities.

The evaluation of conservation practices requires multiple sub-assessments of different management and environmental components related to hydrology, crop growth, pesticide, and more. These sub-assessments are carried out for multiple basins. In such situations, it is essential to include different hydrological, geomorphological, and landscape processes, which are necessary to understand the interactions among them and assess the effectiveness of those practices. One way to understand multiple processes is through the applications of coupled modeling. Studies in the past have applied it in different cases; for example, Cai et al. (2020) applied a coupled rainfall-runoff-erosion model (regional climate model, SCS-CN model, and MUSLE model) to assess the effectiveness of conservation practices in reducing runoff and erosion. Naseri et al. (2021) applied SWAT and a genetic algorithm to see the tradeoff between the implementation of conservation practices and the reduction in sediment yield and flow. Studies can apply such models to understand the effectiveness of conservation practices in sustaining environmental and economic objectives. Future research can also focus on implementing an integrated modeling approach to tackle this problem. This can be an integration of model-free and model-based approaches based on the availability of methods for simulating different hydrologic processes (precipitation, runoff, groundwater, surface water inflow, etc.), tracking water quality parameters, and analyzing the interactions of conservation practices with different ecosystem processes. The results of one approach can be input for another and provide assessments at different scales, i.e., field-to-watershed scale, watershed-to-regional scale, and regional-to-global scale. Furthermore, multi-scale assessments can provide significant value. All these approaches can increase the realism of the assessments and result in the development of more effective conservation plans.

In general, models apply conservation practices homogenously to their resolution capacity. However, actual management can have spatial and temporal heterogeneities at even smaller scales (Antolini et al., 2020). This challenge was observed in many studies; for instance, in Singh et al. (2018), the effectiveness of the filter strip varied spatially; however, it was not accommodated in the model. Further, pastures were simulated homogenously over barren lands. Though there were rocky formations in a few barren lands, the model does not provide the flexibility to account for that. This homogeneity restricts the accurate assessment of the effectiveness of conservation practices. In such cases, studies can look for integrating high-resolution conservation practice information (e.g., remote sensing sources) with existing lithological and hydrological information to improve the simulation ability of models. Likewise, application of machine learning (ML) can be helpful when it comes to the data-intensive assessment of conservation practices’ effectiveness. In the context of agricultural management, ML algorithms have been applied in several past studies (e.g., Liakos et al., 2018; Meshram et al., 2021). These algorithms can be convenient for predicting yield, disease detection, weed detection, crop quality, soil composition, soil moisture, and species recognition, providing rich recommendations to farmers and decision-makers (Liakos et al., 2018; Meshram et al., 2021). However, ML has not been implemented extensively in the assessment of conservation practices and their interactions with hydrology. These algorithms are well poised to address several crucial problems related to the impacts of conservation practices through exploratory data analysis, pattern recognition, feature selection, analysis of spatial-temporal changes and variabilities, prediction of conservation-relevant processes, and understanding of the effectiveness and efficiency of different conservation practices.

The results from the modeling approaches have a degree of uncertainty. These uncertainties can stem from various sources. For example, imperfect or limited representation of conservation practices in models due to the lack of available information related to those practices can contribute to uncertainty. Limitations of models due to the underlying assumptions, model structure, and the associated parameters, calibration schemes, etc., can restrict their ability to accurately simulate the coupled processes. Errors propagating from the input and target datasets can also affect the performance of a model and bias its calibration and validation (Sharpley et al., 2002; O’Connell et al., 2007). Models are often restricted to a fixed resolution (point-scale, hillslope-scale, hydrologic response unit-scale, basin-scale, etc.), which further restricts their ability to carry out multi-scale simulations. Tackling these problems requires the development of new modeling schemes and parameterizations that can effectively simulate the coupled nature of the conservation practice impacts on hydrology. Oftentimes, simple structural improvements in a model can lead to a drastic improvement in its performance (Roy et al., 2017). Once an improved structure/parameterization is established, the sensitivity of the associated parameters needs to be thoroughly assessed to get more insights into the underlying physical processes. It is also important to understand how uncertainty propagates through these coupled parameterizations and how improvements in these parameterizations can result in the reduction of uncertainty at different levels. Focus also needs to be drawn toward understanding how the dynamics of these coupled processes vary across scales. Proper characterization of these uncertainties can better inform potential management interventions.

Most of the studies reviewed in our literature have not considered the geological aspects such as lithological properties of strata and their relative position, faults and folds, the orientation of surface, elevation, and natural drainage network conditions in landscapes. This may be due to a lack of information or the inability of the models to simulate them. It is important to note that landscapes are shaped by geomorphological (aeolian, biological, fluvial, etc.) and pedological processes (soil formulation, soil evolution, etc.) and significantly influence soil properties, erosion, and sediment movement (Conacher, 2002). Future studies can include information from geomorphic assessments carried out in the past to perform studies on historical, existing, and future conditions of channels and their response to land-use changes that can help in understanding the impact of geology on soil conditions. This assessment needs to be considered more frequently as findings can be helpful in understanding the underlying causal relationships.

Studies have found the application of conservation practices sometimes decreases the overall crop productivity (Hatibu et al., 2003; Kinama et al., 2011). One reason is that the application of nature-based measures often utilizes the lands intended for production. In such a case, it is challenging for policymakers, researchers, and farmers to develop different policies and plans to carry out sustainable production. Future studies can also provide helpful insights into developing and implementing government policies and initiatives such as commodity support programs, incentives, and compensation. Further, research can guide the government in including private sectors to promote the application of conservation practices.

The economic perspective of implementing conservation practices is yet to be explored in most of the literature, though some studies have identified this (Schilling et al., 2014; Telles et al., 2019). This requires attention as it plays a vital role in building a strong perception of implementing conservation practices in the mind of farmers, policymakers, and local people. Further, studies have reported the importance of the role of socio-economic factors, policies, and people’s perceptions in adopting conservation practices. Studying these factors can help decision-making be more legitimate, salient, robust, and practical (Bennett et al., 2017). Thus, it is essential to consider all such information while assessing the effectiveness of conservation practice. The social-economic factor also plays a vital role in adopting conservation practices. In such cases, studying social vulnerability can be required before formulating policies or quantifying its risk.

Most studies have directed their efforts toward reducing surface runoff, soil erosion, soil nutrient loss, and sediment transport. But to attain sustainability, it is essential to consider other aspects of the ecosystem. These include the study of quantitative links between conservation practices and ecological responses, the influence of conservation practices on physical habitat quality, etc. (Lizotte et al., 2021). Future studies can focus on those attributes by developing a framework that links conservation practices and programs to the benefits of the ecosystem at national and regional scales.