Changes in Agricultural Soil Quality and Production Capacity Associated with Severe Flood Events in the Sava River Basin

Abstract

Intensifying urbanization in Central Europe is increasingly pushing flood retention areas onto private farmland, yet the agronomic and socio-economic trade-offs remain poorly quantified. We conducted a narrative review of published field data and post-event assessments from 2014–2023 along the transboundary Sava River. Information was collected from research articles, case studies, and environmental monitoring reports, and synthesized in relation to national and EU regulatory thresholds to evaluate how floods altered soil functions and agricultural viability. Water erosion during floods stripped up to 30 cm of topsoil in torrential reaches, while stagnant inundation deposited 5–50 cm of sediments enriched with potentially toxic elements, occasionally causing food crops to exceed EU contaminant limits due to uptake from the soil. Flood sediments also introduced persistent organic pollutants: 13 modern pesticides were detected post-flood in soils, with several exceeding sediment quality guidelines. Waterlogging reduced maize, pumpkin, and forage yields by half where soil remained submerged for more than three days, with farm income falling by approximately 50% in the most affected areas. These impacts contrast with limited public awareness of long-term soil degradation, raising questions about the appropriateness of placing additional dry retention reservoirs—an example of nature-based solutions—on agricultural land. We argue that equitable flood-risk governance in the Sava River Basin requires: (i) a trans-boundary soil quality monitoring network linking agronomic, hydrological, and contaminant datasets; (ii) compensation schemes for agricultural landowners that account for both immediate crop losses and delayed remediation costs; and (iii) integration of strict farmland protection clauses into spatial planning, favoring compact, greener cities over lateral river expansion. Such measures would balance societal flood-safety gains with the long-term productivity and food security functions of agricultural land.

1. Introduction

Soil quality, land use, and land value are closely interconnected. Higher soil quality allows for more diverse land uses, increasing competition for the land. As climate change and urbanization intensify, governments are increasingly implementing public flood protection measures on private land, including agricultural areas [1,2,3]. River discharge data analysis across the EU found that floods in the central Balkans decreased over most of the study period (1960–2010), due to declining precipitation and soil moisture [4]. However, this decreasing trend started to reverse during the 1990s. Land use changes, such as soil compaction, abandonment or removal of terraces, and shifts in land cover, have likely increased flood discharges in small catchments. Additionally, warming-induced hydrothermal anomalies [5] and variations in catchment size may further contribute to the rising frequency of floods in small catchments in southern Europe in recent years [4].

Water needs space, and as the public areas are often unavailable, the solutions necessitated flood retention areas to be allocated on private agricultural land [6,7]. Various small water retention measures help retain water on agricultural soils, predominantly to improve soil structure and thus water infiltration [8,9]. Whereas these measures increase drought resilience and contribute to retaining water within catchments, they are not considered effective flood protection measures due to their limited storage capacity and they mainly delay rather than significantly reduce peak discharges [10]. In highly urbanized catchments, where impervious surfaces accelerate runoff and flood waves are large and sudden, such small-scale measures cannot offset the threatening water volumes [11]. There are examples of landowners implementing various measures to protect their assets from flood damage and respond to flood events, often relying solely on their own resources and funding [12]. In Italy, landowners offered their land to be used for a dry retention reservoir after two consecutive flood events in 2023 Emilia-Romagna floods [13]. Rural landowners are aware of the risk brought by extreme precipitation events and want to contribute more to protecting their land through natural small water retention measures. However, they face multiple barriers, including bureaucratic limitations, financial constraints, insufficient administrative help, lack of equipment, shortage of knowledge, and lack of time [14]. Farmers emphasize the need for expert advice on appropriate flood mitigation measures and the establishment of demonstration sites to clearly illustrate the effectiveness of such interventions [8,12,15]. However, farmers’ perception of flood risk mitigation measures can differ or even directly oppose from institutionally proposed solutions [16]. Several studies show that previous flood experiences increase risk awareness, which in turn makes farmers more likely to adopt control measures [17]. Perceived control over decisions increases farmers’ engagement in flood risk mitigation [18]. Transparent accountability mechanisms, such as compensation schemes and participatory planning, build trust and cooperation [19]. Incentives like subsidies or damage compensation encourage participation, while their absence or perceived unfairness reduces motivation [20]. Well-designed measures reduce long-term damage and promote sustainable management, whereas poorly designed ones may cause mistrust, administrative burdens, and conflicts [21].

Dry retention basins (dry reservoirs) are flood protection structures that temporarily store rainwater and gradually release it after the flood peak. In contrast to conventional reservoirs, they remain empty during dry periods and are only activated during extreme rainfall events [6,7,22]. They are usually built in low-lying rural or peri-urban areas where land is available for temporary flooding and are generally cheaper to build and maintain than conventional reservoirs. This is because they do not require permanent water storage or associated infrastructure [7,22,23]. Their drainage and water control systems typically include inlets, adjustable outlets and spillways that regulate inflow and outflow to avoid flood peaks downstream [7,22]. Their main benefits include reducing flood peaks, improving water quality through sediment and pollutant deposition, and enabling multifunctional land use (e.g., agriculture or recreation) after drainage [7,22,24]. However, regular flooding can affect soil and crop quality and productivity, and sediments can accumulate potentially toxic elements (PTEs), persistent organic pollutants (POPs) or pathogens [25,26,27,28]. Private agricultural land is often considered as a flood retention area for protecting downstream urbanized zones, assuming that the water is retained for a short time (e.g., 2–3 days, with depths up to 2 m) before gradually draining, infiltrating or evaporating [29,30].

Floodwater can have a positive impact on soil quality in the case of sediments devoid of pollutants [31] compared to pre-flooding soil quality [32]. However, the potential negative effects on soil and crop properties within these dry retention reservoirs are often overlooked. Flood waves are often destructive; they carry not only silt and sediments but also a wide range of pollutants, PTEs (e.g., Cd, Cu, Pb, Zn), nutrients (e.g., excess nitrogen and phosphorus), organic contaminants, and POPs (e.g., pesticides) [25,33,34,35]. These substances originate from multiple sources, such as eroded agricultural soils, industrial effluents, sewage overflows, and urban runoff. PTEs can bind to fine sediment particles and subsequently leach into soil or groundwater, posing long-term risks to soil fertility, drinking water quality, and ecosystem health [33]. Nutrient-rich flood deposits can lead to eutrophication of downstream water bodies. At the same time, organic pollutants and POPs can also accumulate in crops and aquatic food chains, endangering human and animal health [25,34].

Climate change-induced shifts in flood regimes may further exacerbate sediment and contaminant loads, amplifying the ecological and economic damage to affected areas [35]. While such measures protect adjacent land outside the retention area from flooding and associated soil degradation, uncontrolled flood events without protective infrastructure result in widespread soil damage across all inundated areas. These damages include the physical removal of fertile topsoil by erosion, which reduces soil organic matter and nutrient stocks [27,31]. Moreover, the deposition of coarse sediments that can bury productive horizons and disrupt soil structure [36,37] leading to soil compaction and oxygen depletion due to prolonged waterlogging [27]. Besides contamination with PTEs, pPOPs and pathogens transported with floodwaters [25,26,28,33,38,39] may occur. Such processes may lead to long-term decline in soil fertility, lowering agricultural productivity and crop quality in affected areas [27,28,40,41]. Thus, from a broader point of view, dry detention reservoirs bring more economic, sociological, and environmental benefits than the absence of such measures. Nevertheless, there is a cost for land quality and consumer safety that needs to be addressed in planning such measures [22,25].

In this integrative review, we assess the impacts of major flood events on agricultural soils in the transboundary Sava River basin, focusing on empirical evidence from the 2014 and 2023 floods. We review flood-induced changes in soil quality, yield performance, and contaminant accumulation, particularly in areas affected by industrial legacy pollution and intensive agricultural use. The review highlights typical risks for soil degradation and food safety under varying geomorphological and land use conditions. Drawing from the reviewed studies, we outline key challenges and considerations that should be addressed before designating agricultural land for flood retention. These findings are discussed in the context of proposed flood risk management strategies for the upper Sava River basin, where dry detention reservoirs are increasingly promoted as part of nature-based solutions.

2. Materials and Methods

This study is based on a comprehensive narrative review of flood events affecting agricultural soils in the Sava River basin (861 km long, 95,719 km², 46°29′33.0″ N, 13°44′15.3″ E–44°49′22.1″ N, 20°27′ E)—extending from the sub-alpine region in Slovenia across much of the Balkan peninsula (Figure 1).

The review covered the period 2014–2023, with particular attention to two major flood events (2014 and 2023). All relevant sources related to flooding and its impact on soil contamination and agriculture in the Sava River basin were reviewed and analyzed. The review included both scientific articles and technical/government reports published between 2014 and 2023, with occasional inclusion of earlier references when they provided essential background. Studies published in English and local languages (Slovenian, Croatian, Bosnian, Serbian) were considered.

Articles were selected based on their explicit focus on soil contamination, agricultural impacts, or management strategies in flood-affected areas of the Sava River basin. Potential biases in the reviewed literature mainly relate to opportunistic sampling, which in some cases did not follow predefined grids and, therefore, limited spatial representativeness; this was taken into account when drawing conclusions. Most papers focus on chemical hazards to soil originating from industrial activities, mining, or other pollution sources [36,37,42,43,44,45,46,47,48,49,50,51], while only a few have addressed yield reduction and crop quality [48,52].

The main sub-catchments considered were Savinja, Sana, Vrbas, Bosna, Drina, Janja, and Kolubara (Figure 1). These catchments include both heavily polluted areas (according to national regulations [53,54]) and zones with minimal industrial or agricultural activity. The review also assessed protective measures that should be implemented before agricultural land is designated for temporary flood retention, with particular emphasis on contamination risks, long-term soil fertility, and implications for food production.

3. Results and Discussion

3.1. Flood Event Development & Response

The first studied flood event occurred in the mid and lower part of the Sava River basin in May 2014 and affected three countries: Croatia, Bosnia and Herzegovina and Serbia (44°50′ N, 19°03′ E–44°50′ N, 20°27′ E). In April and May 2014, a series of cyclones brought more than twice the average precipitation (1961–1990), with record-breaking rainfall of up to 150 L m⁻² within 48 h (13–14 May) in Bosnia and Herzegovina [42,43]. Such precipitation had a return period of 100–500 years in parts of the Bosna River basin [42,43], causing prolonged flooding and catastrophic damage across the three countries.

Besides Sava itself, six rivers overflowed the river beds in the Sava river basin (Sava: 46°29′33.0″ N, 13°44′15.3″ E–44°49′22.1″ N, 20°27′ E; Una: 45°16′11.6″ N, 16°55′04.2″ E; Vrbas: 45°06′28.8″ N, 17°30′53.3″ E–44°20′15.3″ N, 17°16′15.6″ E; Bosna: 45°03′60.0″ N, 18°28′01.6″ E–43°49′08.2″ N, 18°16′04.3″ E; Drina: 44°53′27.7″ N, 19°21′20.5″ E–43°20′54.8″ N, 18°50′21.4″ E; Kolubara: 44°39′42.8″ N, 20°14′53.9″ E–44°15′40.2″ N, 19°52′22.3″ E, Janja: 44°31′58.5″ N, 18°52′37.0″ E–44°40′01.5″ N, 19°15′55.1″ E), affecting an area of 342 km² [42]. Three million people were affected, with 139 fatalities. Damage estimation exceeded 4 billion € (with more than 100,000 houses and 200 schools damaged, including roads, farm assets, industrial plants and coal mines).

Such detrimental floods demanded a broader environmental impact, inter alia agricultural and soil quality impact [42,43]. About 32,495 farmers were affected, the majority of them (26,286) with farm sizes less than 2 hectares. Soil degradation was observed in 72% of flood-affected areas, 1638 landslides were observed [43] and approximately 12,000 ha of arable land was excluded from production. These floods also triggered extensive soil contamination and degradation [36,37,47,48,49,50,51].

Another severe flood event occurred only nine years later, in the summer of 2023 and affected the agricultural land in the upper stream of the Sava River basin in Slovenia. These floods caused 145 million euros of damage to agriculture, and more than 2700 farms were affected [44]. In Slovenia, seven fatalities were reported during the extreme hydrometeorological event and in the immediate post-flood recovery phase. Between 3–6 August 2023 Slovenia experienced a prolonged heavy rainfall storms and downpours [44,45]. During the night of 3–4 August, 150–200 mm of rain fell within 6–12 h on soils already saturated from July precipitation, causing record stream discharges, extensive flooding of residential and commercial areas, landslides, and severe damage to bridges and road infrastructure. Rainfall of lower intensity persisted on 4–5 August and spread throughout the country [44,45,46]. According to [44], out of the more than 12,000 hectares of agricultural land in the Upper Savinja valley and Meža valley, 1500 hectares have been permanently damaged.

In 2014, soil and sediment samples were mainly collected through targeted sampling on flooded agricultural lands, riverbanks, and areas with visible sediment deposits [36,37,42,43,47,48,49,50,51,52,55,56]. National and regional agencies conducted rapid emergency sampling within days to weeks after water receded, generally using opportunistic rather than predefined grid designs [42,43,55,56,57,58]. For PTE contamination studies, transects perpendicular to river channels were occasionally used to capture contaminant gradients [36,47,51], with samples typically taken from 0–5 cm or 0–30 cm depths [36,48,50,52]; sediment layers were collected separately where present [47,50,51]. However, detailed sampling registers and spacing information were rarely reported, limiting spatial comparability [55,56,57]. Sampling during the 2023 floods in the upper Sava River basin followed a similar, largely opportunistic approach focused on rapid hazard and contamination assessment [44,45,46,59,60]. Samples were taken from visibly flood-affected areas, particularly near infrastructure and heavily silted fields, again without predefined grids. As in 2014, soil and sediment were collected from surface deposits and buried soil layers [44].

3.2. Soil Loss, Sedimentation and a Decrease in Soil-Water Holding Capacity

First focusing on the flood event, which occurred between 31 May and 16 July 2014, multiple research campaigns were carried out on agricultural land flooded by the rivers Sava (Figure 2) and its tributaries Drina, Janja, Bosna, Vrbas, and Sana (Figure 2). The extent and severity of soil degradation varied, depending on the flood characteristic. The aftermath of flood events was observed on soil several weeks after near or complete water retreat (Figure 3).

In areas affected by torrential floods from the Bosna, Drina, Vrbas, and Janja rivers, the high-energy flood waves caused significant erosion of the arable soil layer, which had been prepared for sowing prior to the event (Figure 4).

In some parts, due to the destructive power of the flood wave, an entire soil profile was eroded to the parent substrate. Since it was a large area affected by the flood talus, the parts of the soil, sand and gravel carried away were deposited in other places in a layer of up to 1 m (Figure 5).

Stagnant flooding occurred along some parts, where the overflow was gradual, and the water stagnated for up to 25 days, resulting in a 15 cm thick deposits of sludge and other materials (Figure 6).

Similar consequences were observed in the 2023 flood event in the upper Sava river basin. Sampling during the 2023 floods in the upper Sava River basin followed a similar approach to that used after the 2014 events, relying predominantly on non-systematic, opportunistic sampling aimed at rapid hazard and contamination assessment [44,45,46,59,60]. Samples were collected from visibly flood-affected areas, particularly near infrastructure and heavily silted agricultural fields, without the use of predefined grids or standardized spacing. As in 2014, soil and sediment were taken from sediment and buried soil layers [44]. Figure 7 shows soil siltation and affected agricultural land, buildings, and machinery at the upper Sava river basin after a significant flood event in August 2023.

Sampling after both the 2014 and 2023 floods was carried out under emergency conditions by national agencies and research teams. The primary aim was rapid assessment of immediate hazards rather than statistically representative surveys, which explains the reliance on opportunistic sampling [42,43,44,45,46]. Visibly affected areas were defined by clear signs of inundation and sediment deposition, such as thick silt layers, debris, or crop burial. Later, additional analyses were conducted within regular soil and environmental monitoring networks in the affected countries, using more grid-based or systematic approaches [55,56,57,58,59,60].

3.3. Soil Contamination

Despite good agricultural practices and afforestation efforts in the upper parts of the sub-catchments area of the Savinja in Slovenia, Sana, Vrbas, Janja, in Bosnia and Hercegovina or Zapadna Morava rivers in Serbia—areas not burdened by industrial or intensive agricultural pollutants- extreme floods caused by heavy rainfall, transported fragments of parent-material (i.e., Mesozoic ultramafic serpentinite and peridotite) from the upper catchments, as evidenced by studies linking serpentine-derived sediments to Ni- and Cr-enrichment in downstream floodplains [37,48,49]. Although the water was originally clean, these particles are naturally rich in PTEs, which were carried to the other parts of the catchment [36,37,48]. This pattern was confirmed by a study in the Bosna River catchment [37], which analyzed residual flood deposits to assess PTE concentrations and their risk for agricultural soils. Among all analyzed elements (Pb, Cd, Cr, Ni, Zn, Cu), only Ni exceeded the limit value (100 mg kg⁻¹) in all soil and sediment samples, classifying these soils as contaminated with Ni [37].

Analysis conducted in the period from 2014–2019 by [57], on the agricultural land on the right bank in the lower reaches of the Spreča River, which flows into the Bosna River, revealed increased content of Ni (195.60–559.00 mg kg⁻¹) and Cr (45.80–208.20 mg kg⁻¹), Hg (0.08–3.17 mg kg⁻¹), As (2.43–16.83 mg kg⁻¹), Cd (0.15–2.30 mg kg⁻¹) and Pb (16.97–213.30 mg kg⁻¹) which is linked to the geological substrate and in the upper course of the river Spreča. Besides that, increased content of polycyclic aromatic hydrocarbons (PAHs) (0.04–37.45 mg kg⁻¹) connects with the industry of the Tuzla basin through which the river Spreča flows (coal mine, chemical plant Lukavac, Salt mines Tuzla, Sodium factory Lukavac, Asbestos mine Petrovo Selo) [57]. On the other hand, research conducted by [58] on the left bank of the Spreča River confirmed an increased content of Ni (<532 mg kg⁻¹) and Cr (<191 mg kg⁻¹), but did not confirm an increased content of Pb (<31.1 mg/100g), Cd (<0.1 mg/100g) and Zn (<100 mg kg⁻¹) which are associated with industrial waste from the Tuzla Basin [57].

In the sub-catchments of the West Morava River (Serbia), additional analysis of PTEs in sediment and agricultural soil affected by flood in 2014, found an increase in Ni, Cr, As, Pb, and Cu [55]. About 95% of soil and sediment samples were found to be extremely polluted by Ni and about 65% were slightly polluted by Cr, whereas about 90% of soils and sediments were not polluted by As, Cd, Pb, Cu, and Zn according to the pollution index. The minimum and maximum Ni contents in the flood sediments was 91 mg kg⁻¹ and 372 mg kg⁻¹ and in the soil 32 mg kg⁻¹ and 468 mg kg⁻¹. For Cr, the minimum and maximum values in the sediments were 51 mg kg⁻¹ and 207 mg kg⁻¹ and in the soil 30 mg kg⁻¹ to 250 mg kg⁻¹. In some samples, the minimum and maximum values for Pb was 15 mg kg⁻¹ and 127 mg kg⁻¹ in the sediments and 14 mg kg⁻¹ and 126 mg kg⁻¹ in the soil [55]. In the Vrbas River catchment, near the riverbed, Ni concentrations reached up to 151 mg kg⁻¹ in mud deposits, confirming earlier findings of elevated Ni levels in this area prior to the 2014 floods [49]. Similarly, along the Drina and Janja rivers, Ni concentrations in sediments reached up to 136.0 mg kg⁻¹ (Drina) and 195.5 mg kg⁻¹ (Janja), while Zn concentrations in the Janja River sediments reached up to 123.2 mg kg⁻¹, indicating widespread contamination [49,55].

A study conducted by [47] in sub-catchments with intensive industrial activity (e.g., Kolubara rivers) found elevated pseudo-total concentrations of Cu and Ni that could limit agricultural use. Enrichment factor analysis indicated that Ni concentrations in 59% of soil samples and 68% of flood sediment samples were classified within impact classes, suggesting a notable anthropogenic contribution to Ni contamination [47].

At the Jamena site (Vojvodina Province, Serbia), where the Sava River flooded agricultural soils following an embankment failure, the average Ni content in the topsoil (0–30 cm) of flooded fields reached ~93 mg kg⁻¹, exceeding the maximum permissible concentration set by Serbian soil quality standards [52]. Compared with background values, elevated concentrations of several analyzed PTEs were observed, with soil pollution status ranging from moderate to highly polluted [52]. Along the River Sava at several localities in Slovenia, Croatia, and Serbia, contaminated sediments deposited during flooding also contained increased levels of PTEs [51]. Seven PTEs (As, Cd, Cr, Cu, Ni, Pb, and Zn) were analyzed in sediments along the River Sava. Mean concentrations of As, Cd, Cr, Cu, Ni, and Zn generally increased downstream, with approximate mean ranges of As (5–13 mg kg⁻¹), Cd (0.1–0.5 mg/kg), Cr (20–90 mg kg⁻¹), Cu (10–60 mg kg⁻¹), Ni (15–85 mg kg⁻¹), and Zn (50–250 mg kg⁻¹) [52]. In contrast, Pb concentrations ranged from 25 to 70 mg kg⁻¹ and indicated a medium environmental hazard in most areas and a high hazard in the upper reaches, exceeding local sediment quality guideline thresholds [51].

Significant amounts of pesticides can enter retentions via surface flows, with flood-affected sediments in Vojvodina showing 0.01–0.25 mg kg⁻¹ of mainly atrazine, terbuthylazine, and metolachlor [50]. Detailed analysis of PAHs, polychlorinated biphenyls (PCBs), pesticides, plasticizers, and other emerging compounds in water and sediments from 10 rivers and canals in Northern Serbia revealed moderate pollution levels in river sediments within Vojvodina Province, which partly belongs to the Sava catchment area [50].

During the 2014 floods in Bosnia and Herzegovina, erosion also displaced explosive remnants of war from former minefields along riverbeds into agricultural and inhabited areas [58]. Similar risks associated with landmine displacement due to erosion were documented in other post-conflict regions [23,61]. Post-2023 flood assessments in the upper Sava River basin found that previously contaminated soils in Celje (Savinja River) were overlaid by new deposits with Zn concentrations up to 69 mg kg⁻¹ [43]. Agricultural soils at Sava Sneberje (110 mg kg⁻¹ Pb) and Litija (120 mg kg⁻¹ Pb) initially exceeded national warning values but were later covered by sediments with ~47–53 mg kg⁻¹ Pb, reducing concentrations through dilution [44]. In historically contaminated areas such as Mežica and Žerjav, deposited sediments contained up to 26 mg kg⁻¹ Cd, 590 mg kg⁻¹ Pb, and 4400 mg kg⁻¹ Zn, reflecting legacy mining contamination [44].

3.4. Yield Loss

It was assessed that the extreme floods in the Balkans in 2014 caused a rapid and profound decrease in soil function manifested by reduced yields [37,47,48]. In the Kolubara River Basin, approximately 482,700 ha of land was flooded [47], with flood sediment requiring removal from 335,400 m² of agricultural areas [47], resulting in a complete yield loss on about 25% of the flooded arable land (mainly maize and wheat) [47], while yields in the remaining areas were reduced by 30–50% due to waterlogging and sediment deposition [47].

In the northern Republic of Srpska (Bosnia and Herzegovina), floods inundated 2845 ha at the Bosna–Sava confluence [48], leaving 0.5–10 cm thick sediment–mud layers for up to 22 days [48], resulting in the destruction of crops on a large proportion of affected land and substantial yield reductions elsewhere [48].

According to [44], the floods in the upstream Sava river basin in 2023 caused significant damage to crops, especially young and delicate plants like maize, pumpkins, and soybeans. Crops that remained underwater for more than three to four days were severely affected and required reseeding. The yield of crops in areas with standing water was reduced by up to 50% [44]. Some areas still needed to sow maize and pumpkins, but it was difficult due to flooded and soggy soils. There was an increase in cereal diseases, and the grasslands were polluted with mud and became unsuitable for animal fodder [44]. Vegetable cultivation was also affected by diseases, especially lettuce, which could not be sown due to water-saturated soil [44]. The farm income in that year decreased by 50% [44]. Some vegetables did not germinate, and there was a greater rotting of onions and garlic. The orchards were flooded, and cherry and plum crops suffered a 70% loss. The berry crops suffered a 40% loss due to rot. The vineyards were also severely affected by landslides, which prevented the use of mechanization and protection of the vines due to the steep and wet terrain [44]. Similar difficulties in agricultural production and yield losses are expected in regulated flooding within dry detention reservoirs.

3.5. Food Contamination

Authors of [37] investigated the impact of flood-deposited pollutants on crop quality and found that elevated nickel (Ni) concentrations in soil did not necessarily translate into higher Ni accumulation in crops grown on alkaline soils, despite these soils being classified as contaminated. However, in slightly acidic soils, prolonged water saturation and high groundwater levels created unfavorable redox conditions that enhanced the mobility and bioavailability of Ni, resulting in increased uptake by plants [37]. This was reflected in maize leaves, where Ni concentrations reached 7.89 mg kg⁻¹—six times higher than in plants grown on alkaline soils, where Ni content remained at 1.3 mg kg⁻¹, despite the soil Ni concentration being twice as high (270 vs. 129 mg kg⁻¹). Similar patterns were observed in other crops, confirming that soil pH and hydrological conditions strongly influence the transfer of Ni from soil to plants, regardless of the total soil Ni content.

Investigation of residual hazardous elements in potatoes, carrots, and onions grown on the soil after flooding in Serbia [52] was conducted using wet digestion of edible parts followed by atomic absorption spectrophotometry (AAS) to determine metal concentrations. The results revealed that the average levels of Pb in potato and carrot samples (0.3 mg kg⁻¹) were higher than the maximum allowable concentrations established by EU and Serbian regulation. Potential health risk was assessed using the USEPA hazard index (HI) approach based on estimated daily intake, resulting in values of 1.16 for adults and 1.60 for children. Therefore, floods might be considered a long-term secondary and diffuse source of soil contamination in the affected areas [52].

4. Discussion

4.1. Increasing Use of Agricultural Land for Flood Retention

With increasing extreme weather events, urban areas are being prioritized over agricultural land when it comes to flood protection [24,62,63,64,65]. Consequently, constructing flood retention measures on agricultural land to protect existing urbanized areas is becoming more common [22,63]. Dry retention reservoirs proposed on agricultural land are frequently promoted to increase flood safety, supported by a misconception of minor to no adverse impacts on agricultural production [7,22,24]. However, we can learn from past flood events that flooding brings several uncertainties to agricultural production; instead, forests, grasslands, and reservoirs function as the primary suppliers of flood regulation services [64]. With the introduction of flood retention, arable farming, horticultural production, and orchards gradually turn into grasslands, forests and pastures due to increased risk for agricultural production and adverse impact on yield quality and quantity. However, forage production on flood-prone grasslands can pose health risks to livestock due to potential contamination with pathogenic Clostridium species (e.g., C. botulinum, C. perfringens), which thrive in anaerobic, organic-rich conditions typical after flooding [26,65]. Contaminated silage or haylage may cause botulism or enterotoxemia in ruminants, impacting animal health and productivity [65]. Additionally, Clostridium tyrobutyricum can lead to late blowing defects in hard cheeses, reducing quality and causing economic losses [66].

4.2. Flood-Induced Soil Degradation, Yield Instability and Contaminant Accumulation

Flooding of agricultural land may lead to nitrate leaching and hypoxia [67] resulting in yield loss. Besides that, flood-induced water saturation and field inaccessibility frequently hinder sowing operations, leading to delays or complete failure to establish crops [40]. Rapid and profound changes to soil function manifested by reduced yields due to changing redox potential, nitrogen leaching, and changing greenhouse gas emissions after water withdrawal, from CH4 emissions to N2O emission, were observed in other catchments [27]. However, with the recovery of the land after a year or two, the balance of production functions is established (ignoring the potential contamination with pollutants, such PTEs or pesticides), and the usual crop yields are achieved. Therefore, to mitigate the recovery period of the productive functions of the soil, ref. [27] recommends the cultivation of crops in the retention field, that better tolerate wet conditions. Pesticide residues are monitored in water bodies in the EU; however, soils of the retention areas are not systematically included in the monitoring system, although soil can act as a source or sink for pollutants [68]. In Slovenia, groundwater is monitored for PTE, pesticide residues, PCBs, halogenated organic compounds (AOX), pharmaceuticals, and from 2025, per- and polyfluoroalkyl substances (PFAS) [69]. The program follows national regulations aligned with the EU Water Framework Directive (2000/60/EC) [70] and the Groundwater Directive (2006/118/EC) [71]. Soils in retention areas are not systematically monitored, though some national soil points may coincidentally fall within these zones. This points out the need to institutionalize a comprehensive soil quality monitoring framework for flood retention areas, incorporating chemical (e.g., PTEs, POPs), biological (e.g., pathogens), and hydro-pedological indicators. Such a system would provide robust surveillance of food safety and public health risks and help guide adaptive management strategies [22,28,68]. Beyond detection, the long-term effects of POPs and PTEs on soils and crops deserve particular attention. Their affinity for fine particles and organic matter enables repeated transport during floods and remobilization under fluctuating redox conditions [25,33,50,68,72]. Compounds such as atrazine, terbuthylazine, PCBs, and PAHs remain chemically stable and can bioaccumulate in crops, posing chronic risks to food safety [25,34,50]. In contrast to PTEs, which often remain spatially localized in sediments, POPs are more mobile within the soil–water system and degrade slowly depending on soil type, pH, and organic carbon [25,33,34,35].

The more frequent the flooding of agricultural land, the greater the risks of accumulated pollutants in the flood sediments [73]. Distribution of contaminants in flood sediments also depends highly on naturally occurring/geo-genic metals [28,33,34,35] in the soils upstream and the complexity of land use (e.g., the presence of urban areas, industrial zones and mines), as was the case in the West Morava river [36,37,48,49]. In the Elbe River basin [68] detected 13 different pesticides, with atrazine, terbuthylazine, metazachlor, metolachlor, isoproturon, and chlorotoluron being the most frequent. Concentrations were typically <0.05 mg kg⁻¹ in the higher, less frequently flooded zones, but reached up to 0.22 mg kg⁻¹ in frequently flooded areas. Atrazine and terbuthylazine were the only compounds consistently detected across all sampling sites, regardless of flooding frequency [68]. The study carried out in Germany demonstrated that floodplain soils of the River Wupper are contaminated with metal(loid)s. The lower the redox potential, the lower the Co, Cu, Mn, Ni, Sb, and Zn concentrations [72], meaning that soil type, soil organic content and pH play an important role in food toxicity if plants are grown in flooded areas.

The impact of flooding of the Odra River (Poland) on soil pollution by PTEs and PAHs and the possibilities of the migration of these pollutants into the soil profile were investigated by [38]. They tested sludge samples of floodwater and soil. No abnormal concentrations of PTEs were found—for example, the detected concentrations in flooded and arable soils ranged for Zn from 3.7–156 mg/kg (regulatory limit: 300 mg/kg), Cu 0.4–75.3 mg/kg (limit: 150 mg/kg), Pb 0.9–22.0 mg/kg (limit: 100 mg/kg), Ni 0.9–16.1 mg/kg (limit: 100 mg/kg), Cd <0.02–0.39 mg/kg (limit: 4 mg/kg), Cr 0.7–13.7 mg/kg (limit: 150 mg/kg), and Hg 0.007–0.076 mg/kg (limit: 2 mg/kg). Thus, all measured values were well below Polish standards for arable soils, likely due to low organic matter (<5%) and coarse texture, which limits retention [38].

With regard to transboundary rivers, there is justified concern about the risk of soil and environmental contamination. Water storage in retention ponds can increase the risk of infectious disease outbreaks in humans and animals due to the uncontrolled release of untreated wastewater and livestock effluents [39]. In such cases, mitigation strategies such as soil disinfection (mainly applied on smaller scales, e.g., in greenhouses), liming, or full-scale soil remediation may be necessary. Where contamination is severe or persistent, containment measures and land use conversion (e.g., from agriculture to forestry) should be considered [29]. Afforestation, in particular, may offer co-benefits for delaying flood peaks and reducing surface runoff [29].

4.3. Socio-Economic Considerations and Farmer Engagement

From the landowners’ perspective, who are often also land users (i.e., farmers), flood protection solutions are associated with land use restrictions and uncertainties regarding yield quality and quantity [22]. Therefore, these measures should be accompanied by compensation for flood retention and any changes in property-related rights and obligations [74,75]. Changes in land production capacity must be fully considered, and agricultural landowners should be actively consulted whenever flood mitigation measures are planned on their land [22]. Farmers must be provided with all the necessary information before being asked to lease their land for flood protection to ensure the safety and protection of farmers and food consumers. This information should cover economic and developmental aspects, ecological factors, legal considerations, and social justice issues [9,74,75,76,77].

4.4. Policy Frameworks, Compensation Mechanisms and Future Viability

Given the risk of pollutant deposition and its implications for food safety, a critical question arises: how can agriculture sustainably develop in areas designated for periodic inundation, such as dry detention basins?

If farmers in flood-prone areas are expected to adapt to increasing yield instability, shorter planting windows, and contamination risks—by cultivating flood-tolerant crops or switching to entirely new land uses—a key question arises: what viable economic models exist for farming in dry retention reservoirs? This is especially pressing in the Balkan region, where farms are small (e.g., 7 ha in Slovenia, 5.4 ha in Serbia, 2.44 ha in Bosnia and Herzegovina, <7 ha in Croatia) and highly fragmented [78]. Such structural constraints hinder profitability, as larger and more homogeneous farms (>50 ha) are typically more efficient for conventional crop production [78,79,80]. In these contexts, transitions to high-value alternatives—such as organic crops, horticulture, or greenhouse systems—may be riskier and less economically feasible.

Although the EU Common Agricultural (2023–2027) [81] promotes Nature-Based Solutions (NBS) (In Balkan mainly seen as dry retention reservoirs), to mitigate drought, reduce erosion, and enhance agroecosystem services, their role in flood management remains marginal. Farmers cultivating land within dry detention reservoirs may face higher costs for soil remediation and climate adaptation, while experiencing reduced economic returns during flood-prone years [22,24,40,81,82,83]. Current area-based CAP payments [81] further disadvantage fragmented and irregular parcels, disproportionately affecting smallholders.

In the Balkans, NBS implementation is hampered by fragmented land ownership, which fuels local resistance to new land use restrictions, and limited awareness among farmers of the wider benefits of NBS. Many landowners distrust politicians, feel that their influence on decision-making is minimal, and see politicians and powerful businessmen as the main drivers of land use decisions, which further discourages them [74,84]. As extreme flood events become more frequent [22,41,63,82,83], the socio-economic viability of farming in such zones becomes increasingly precarious.

While the EU Floods Directive (2007/60/EC) [85] acknowledges NBS as tools for sustainable flood risk reduction, it does not resolve issues of land availability or potential conflicts with agricultural production, leaving such decisions to Member States. Small-scale retention measures [8,9] may be effective in headwater catchments, but their capacity to mitigate the destructive effects of large-scale flood events is limited—especially in extensive or low-lying watersheds, where technical infrastructure remains essential.

As demonstrated in this review, which focuses on biophysical aspects of post-flood soil degradation, farmers in flood-prone areas, including those where flooding is newly introduced through the construction of measures such as dry retention reservoirs—face substantial uncertainty regarding future land use, production stability, and economic viability [22,63,74,75]. This includes the increasing frequency of flood events driven by climate change, higher post-flood remediation costs (e.g., cleaning, soil restoration, or switching to flood-resilient crops), and long-term yield reductions. Sustainable development in periodically inundated areas requires adapting land use by flood frequency: margins with low inundation can remain in conventional rotations, while frequently flooded cores are better suited to grasslands, forestry, or non-food uses [10,22,76]. Food production may continue if supported by pre-/post-flood soil testing and clear “go/no-go” thresholds for contaminants [22,28,68]. Viable economic models documented in Europe include public land purchase [74,86], flood easements with one-off or annual payments [86,87], cooperative schemes co-funded by beneficiaries [87], and CAP-based top-ups [24,81]. Insurance or disaster relief may complement, but not replace, baseline compensation [24,40].

Both the EU’s agricultural policy and flood risk framework must adapt to this evolving risk context by expanding compensation and insurance schemes that protect farmers’ incomes when their land is used for flood retention. Policy incentives should encourage land stewardship aligned with flood mitigation goals while avoiding the transfer of disproportionate economic burdens onto landowners. When private land is flooded, the benefits—such as urban flood protection—often accrue to third parties, while landowners incur the costs of agricultural disruptions, infrastructure degradation, and altered microclimates [22,63]. Although such interventions can deliver broader ecosystem services (e.g., improved biodiversity, carbon sequestration, recreation), landowners rarely receive direct compensation for these added values. To secure rural livelihoods, compensation must exceed direct and indirect costs to landowners, including transaction costs.

Compensation approaches vary significantly across central and western Europe [74,82,83,87,88]. In Hungary and Germany, land used for flood retention is often purchased outright by the government [74,82], with Hungarian landowners along the Tisza River receiving above-market payments to cover crop losses and operational disruptions [82,83]. In Switzerland, farmers affected by the Seymaz River restoration were offered the choice to either sell their land or retain it under near-natural management in exchange for annual compensation [87]. Austria uses a more innovative model in the Altenmarkt area, where annual compensation payments are co-funded by local beneficiaries through a water cooperative [88]. Although innovative policy frameworks are increasingly promoted, their implementation often faces practical challenges. Experiences from Hungary, Germany, and Austria illustrate that compensation schemes range from land purchase to cooperative annual payments, yet administrative complexity and limited trust among farmers reduce their acceptance [74,75,77,78,79,80,81,82,83,84,85,86,87,88,89]. These examples demonstrate that without transparent, adequately funded, and context-specific mechanisms, policy innovations remain difficult to translate into practice.

In Slovenia, farmer resistance to NBS—particularly dry retention reservoirs—is widespread due to acute land scarcity and the high opportunity cost of losing productive land, especially in regions with specialized agriculture such as hop cultivation [11,22,63]. Besides direct economic losses, dry retention infrastructure can alter local microclimates, reduce soil quality over time, and increase pest and disease pressures [22,26,27,28]. While recent policy changes allow compensation for crop damage within dry retention reservoirs, long-term concerns remain largely unaddressed. These include persistent soil contamination, declining productivity, and unresolved issues related to the safe handling and disposal of potentially contaminated sediments [22].

5. Conclusions

Detrimental floods in the upper Sava River basin in 2014 and 2023 show that flooding is a significant cause of damage in agriculture. Beyond immediate yield losses, floods trigger long-term declines in soil productive functions through erosion, physical soil loss, and contamination. Despite this, flood retention measures for protecting urban areas in the river basin increasingly involve using private agricultural land for the public benefit of increased flood resilience.

While agricultural land provides essential regulatory ecosystem services by temporarily storing floodwater, these functions often come at the cost of agronomic instability, soil quality deterioration, and gradual disinvestment or abandonment. Farmers and landowners should therefore not be regarded as merely passive providers of ecosystem services, but as central stakeholders who disproportionately bear environmental and economic risks. Their perspectives and interests must be explicitly integrated into flood governance frameworks. The successful use of nature-based solutions (e.g., dry retention reservoirs) for flood risk reduction depends on their capacity to reflect the socio-economic and environmental realities of affected communities. Compensation mechanisms must be transparent, fair, and multidimensional, covering not only immediate crop losses but also long-term remediation, productivity decline, shifts in cropping systems, and opportunity costs tied to restricted land use. Policy instruments should encompass targeted insurance schemes, incentives for adaptive and flood-resilient agricultural practices, and clear regulatory provisions for sediment management and property rights.

Finally, the sustainability of such interventions requires integrative governance approaches that align flood risk management with long-term agricultural productivity and rural resilience. This is particularly critical in transboundary river basins such as the Sava, where effective coordination across jurisdictions is essential to balance flood protection objectives with the preservation of soil functions, food security, and farmer livelihoods.