Puccinia striiformis f.sp. tritici (Pst) is the causal agent of wheat yellow (stripe) rust, that is one of the most devastating diseases of wheat worldwide and is largely dispersed by air currents that transmit asexual urediniospores over long distances (Zadoks, 1961; Hovmøller et al., 2011). Wheat yellow rust is not transmitted through seed although under high infection intensities Pst urediniospores can be observed on host plant ears (Chen et al., 2014). Changing climatic conditions drive the continuous breeding of new varieties for increased fitness in response to these rapidly shifting environments. However, new Pst races also regularly emerge that can overcome resistance in introduced wheat varieties making the control of yellow rust challenging. Furthermore, recent shifts in the geographic distribution of Pst, particularly in North America, have also been linked to temperature-specific adaptation (Lyon and Broders, 2017), reflecting the high adaptability of Pst to environmental pressures. Consequently, changes in pathogens populations affected by changes in climatic conditions, selection pressure of resistant varieties and applied pesticides (Ash and Brown, 1990; Pritsch et al., 2000; Ahmad et al., 2010; Ma et al., 2013; Xi et al., 2015; El Jarroudi et al., 2017; Carmona et al., 2019; Chen, 2020) makes yellow rust control challenging (Chen, 2005; Ali et al., 2014; Hovmøller et al., 2016; Bueno-Sancho et al., 2017).

The potential for pathogens to better adapt to changing environmental conditions are also influenced by the type of reproduction, giving high adaptation abilities to pathogens with sexual reproduction stages in their life cycles. Pst is heteroecious, needing two hosts to complete its life cycle; it reproduces asexually on wheat and sexually on Berberis spp (Jin et al., 2010). Pst diversity is centered in the Himalayan region, where sexual reproduction and thus significant genetic variation occurs, in contrast to the predominant clonal population structure found in Europe, America, and Australia (Hovmøller et al., 2011; Ali et al., 2014). While mutation is known to serve as a major source of genetic variation, generating new alleles and genotypes, sexual, parasexual, and somatic recombination also influence genotype diversity (Hovmøller et al., 2011).

Before 2011, the population of Pst in Europe was mainly clonal, relying on mutations with minimal influence from sexual recombination (Hovmøller et al., 2002). Consequently, new virulent Pst strains arose due to stepwise mutation within these clonal populations (Linde et al., 2002; de Vallavieille-Pope et al., 2012). However, this situation abruptly changed in 2011, with the discovery of two new Pst races, termed ‘Warrior’ and ‘Kranich’, that were detected across numerous European countries and had greater genetic variability than previous clonal populations (Hovmøller et al., 2016). These new exotic Pst lineages then rapidly replaced the previously known Pst population in Europe (Hubbard et al., 2015) and harbored virulence against varieties that were previously resistant to the prevailing Pst races, making rust control much more challenging (Sørensen et al., 2014). The Warrior race has also spread to North Africa since its appearance in Europe, challenging wheat production (Ali et al., 2014; Hovmøller et al., 2016; Shahin, 2020). These findings underscore the reality that, despite ongoing efforts to develop wheat varieties resistant to existing Pst populations, the emergence of new Pst races can trigger outbreaks of epidemic proportions, even on a continental scale (Brown and Hovmøller, 2002).

The first observations of yellow rust in Serbia occurred at Rimski Šančevi in 1997 within a genetic collection comprising 630 genotypes of wheat and wild relatives (Jevtić et al., 1997). The first warning of its heightened threat was issued in 2007 as part of the ADAGIO Project (Adaptation of Agriculture in European Regions at Environmental Risk Under Climate Change) (Jevtić et al., 2009; Mihailovic et al., 2009; Jevtić and Župunski, 2023). This increase in yellow rust infections in 2007 was associated with increased drought and heat likely due to climate change and coincided with expanded geographic distribution of Pst. However, until 2014, the related wheat leaf rust pathogen, Puccinia triticina, remained the predominant wheat rust species in Serbia. Yet, during the 2013/2014 production season, winter temperatures exceeded average temperatures since 1964 leading to yellow rust predominance over leaf rust and jeopardizing wheat production (Jevtić et al., 2017, 2020). The outbreak of yellow rust in Serbia in 2014 was attributed to influx of the Pst Warrior race (Jevtić et al., 2017, 2020), with disease indices (DI) ranging from 40% to 60% (Jevtić et al., 2020).

The studies primarily focus on monitoring shifts in the yellow rust population and its effects on wheat production, but the number of model genotypes used per study to assess the impact of yellow rust on agronomic traits is usually limited to a few well-known cultivars categorized as either susceptible or resistant. However, the variation in responses among different susceptible cultivars to yellow rust infection is often overlooked, as is the combined effect of biotic and abiotic stressors on disease occurrence (White et al., 2011; Juroszek and von Tiedemann, 2013). The lack of information on factors that affect diversity of reactions of susceptible varieties to yellow rust infection, and how they compensate for yield losses makes predictions of yellow rust occurrence more challenging and limits the effectiveness of methods aimed at preventing agronomic losses.

Considering the ability of Pst to rapidly overcome host plant resistance, and the complexity of signal cross-talk controlling plant responses to abiotic and biotic environmental factors, we set out to examine the yield response of winter wheat varieties to Pst infection. In particular, we decided to assess how different levels of Pst infection affect yield responses of commercial winter wheat varieties, and the potential for different host genotypes to compensate for yield losses after delayed fungicide application under high established Pst infection levels. Our results, indicate that the level of pathogen pressure that triggers susceptible reactions is cultivar-specific, with diversity in the ability for commercial winter wheat varieties to overcome damage caused by high levels of Pst infection. Furthermore, through population genetic analysis we also uncovered a potential recent change in the Pst population in Serbia that could have an impact on disease severity in the region.

In this study, we demonstrated that yield achievements of susceptible wheat genotypes are not uniform in years favorable for Pst infection. It appears that for every wheat variety, there exists a specific level of yellow rust infection that triggers yield decrement. As little was known about the potential of fungicides to reduce yield decrement if applied later in the growing season when yellow rust disease severity index (DI) exceeds 20%, we analyzed fungicide effects on yield responses of susceptible genotypes. These results indicate that this relationship is complex and highly associated with the aggressiveness of the dominant Pst race(s) and the capabilities of the host plant to overcome damage caused by high levels of Pst infection. This indicates that more research is needed to elucidate the factors affecting plant responses to high levels of Pst infection and the potential of fungicide treatments to reduce yield losses under these circumstances.

Beyond the presence of the pathogen, climatic conditions are an important factor influencing pathogen outbreaks, alongside the susceptibility of host plants. It is known that elevated temperatures in January and February favor Pst infection in Serbia’s agro-ecological conditions. In Serbia, from 2014 to 2023, temperatures in January varied by almost 10°C, and in February by 7°C. Total precipitation also ranged from 1 to 65.5 mm in March and 11.1 to 74.5 mm in April. In 2014, yellow rust predominated over leaf rust for the first time in Serbia, causing enormous damage in wheat production areas (Jevtić et al., 2017, 2020), with the ‘Warrior’ race of Pst causing the epidemic in Serbia in 2014 (Jevtić et al., 2017, 2020). Whereas, in 2023 we detected a diversification in the Pst isolates identified, with all samples analyzed clustering in a clade branching from the PstS10 race group. This indicates a potential shift in the dominant Pst race present in Serbia in the past decade.

Virulence phenotypes of yellow rust are categorized into genetic lineages and designated with “Pst” followed by a sequential number (Ali et al., 2014; Hovmøller et al., 2016; Thach et al., 2016). Yellow rust races in the Northwestern European population, identified before 2011, belonged to a single clonal lineage known as PstS0 (Ali et al., 2014). From 2011 onwards, yellow rust races from the PstS7 (Warrior), PstS8 (Kranich), and PstS10 lineages became widespread in Europe, accounting for over 80% of the examined isolates (Ali et al., 2014; Hubbard et al., 2015; Hovmøller et al., 2016). After its appearance in Europe, PstS7 also emerged in North Africa (Ali et al., 2014; Hovmøller et al., 2016; Shahin, 2020). PstS10, previously referred to as Warrior (-), has been the dominant lineage in most parts of Europe since 2014 (Ali et al., 2014). Virulence pathotypes that predominates in other parts of the world are: PstS, closely related to PstS2, predominates in North America while PstS2 in East Africa (Hovmøller et al., 2008; Ali et al., 2014; Walter et al., 2016); PstS3 is widespread across West Asia, southern Europe, and North Africa (Ali et al., 2014); PstS4 comprises races primarily found on triticale in Northern Europe (Hovmøller et al., 2008; Hovmøller et al., 2016); Two races with distinct microsatellite profiles fall under PstS5 (Ali et al., 2014; Hovmøller et al., 2016; Thach et al., 2016) and PstS6 is another lineage common in East Africa (Ali et al., 2014; Hovmøller et al., 2016; Thach et al., 2016).

The occurrence of a more aggressive Pst race in Serbia in 2023 was also associated with lower yields in untreated plots compared to 2014. Additionally, it has been reported that varietal resistance to yellow rust prevents yield losses, and that yield responses of single susceptible varieties could differ in years conducive to yellow rust (Zhou et al., 2022). The study by Jevtić et al. (2017) yielded similar results, where the diversity in yield loss responses of susceptible cultivars to leaf rust and powdery mildew ranged from 20% to 50% over a ten-year period, despite the same level of leaf rust and powdery mildew infection at 20%. However, our study found that there might be a specific level of yellow rust infection that triggers yield decrement individually for each variety, and that susceptible varieties, although infected with high levels of yellow rust, can respond differently to the damage caused. This is consistent with similar analysis conducted for Fusarium head blight (FHB), where each wheat genotype was shown to tolerate a specific range of FHB disease pressure without showing a significant difference in expression of yield potential. Furthermore, it has been shown that defense mechanisms towards FHB are genotype-specific, indicating that resistant genotypes exhibit unique transcriptome profiles during Fusarium-host interactions (Pan et al., 2018).

Although wheat pathogens evolve in response to agro-ecological conditions, the combined effect of biotic and abiotic stressors on disease occurrence has frequently been overlooked in research (White et al., 2011; Juroszek and von Tiedemann, 2013). Heeb et al. (2019) advocated for a climate-smart pest management (CSPM) approach, yet they acknowledged the formidable challenge of creating a comprehensive model to predict short-term climate change-induced pest outbreaks at the local level. Understanding that the regulatory network of plant responses to abiotic and biotic stressors and their effects on yield consist of many components that may function antagonistically or synergistically (Glazebrook, 2005; Yasuda et al., 2008; Kissoudis et al., 2014). Our study highlights that when assessing yield achievements and cultivar responses to different levels of yellow rust pressure, more attention should be given to the combined effects of abiotic and biotic stressors. Furthermore, although the majority of studies have addressed fungicide efficacy in lowering host plant infection under diverse levels of pathogen pressure and have also investigated the most efficient timing for fungicide application, little is known about host plant responses to fungicide treatments when disease development is aborted later in the growing season and how it affects final yield achievements.

We found that yield loss between different genotypes in untreated plots ranged from 3% to 80% under yellow rust infection of 50%. This indicates that fungicide efficacy in reducing yield losses, if applied later in the growing season on susceptible genotypes, is complex. Generally, it can be expected that yield losses would be greater if susceptible varieties are infected in the early growth stages (Ash and Brown, 1990; Gaunt and Cole, 1991; Murray et al., 1995), due to the complexity of effects of Pst infection on all yield parameters (Ash and Brown, 1990). In our study, the first visible symptoms were recorded at the 30-32 BBCH (Stem elongation) in 2014, and at 20-23 BBCH (Tillering) in 2023. The higher aggressiveness of Pst races and much earlier infection in 2023 could have led to the lower yield achievements in untreated plots in 2023 compared to 2014. However, our study also indicated a significant difference in varietal responses to delayed fungicide application when high levels of Pst infections were already established. Although fungicide treatment resulted in the abortion of yellow rust development in later growth stages, varieties in our study showed different potential to recover from the damage caused by Pst infection and reimburse yield distortion. Consequently, our study indicates that more research should be done to elucidate mechanisms affecting host plant recovery from high levels of Pst infection, which would directly influence studies dealing with fungicide efficacy in preventing yield losses, especially when a high level of yellow rust infection is reached.

Establishing correlations between genotype and phenotype is essential for breeding wheat that is better adapted to the changing climate. This challenge is compounded by the fact that plant growth in natural habitats is influenced by a combination of abiotic and biotic stressors. Most existing research has concentrated on plant responses to individual stressors, leading to a limited understanding of adaptation to multiple stress factors. Given that 60% of expression changes under combinatorial stress conditions could not be predicted based on responses to individual stressors (Kissoudis et al., 2014), it can be inferred that the differences between susceptible and resistant plant reactions may be more closely linked to variations in the timing and magnitude of responses to combined stressors, rather than solely to the expression of individual genes (van Loon et al., 2006). There are documented instances of disease resistance being compromised by elevated temperatures and humidity, evidence suggests that temperature variations—rather than fixed thresholds—affect resistance mechanisms against yellow rust in wheat (Bryant et al., 2014). Additionally, drought and salinity have been shown to impact pathogen resistance responses, and the interaction between plants and pathogens can further influence plant responses to abiotic stress (Kissoudis et al., 2014). As a result, analyzing the association between morphological and physiological resistance traits conferred by adult plant resistance (APR) genes against obligate pathogens, alongside abiotic stressors and the expression of abiotic stress-related genes under field conditions, could yield a more detailed genotype profile. This, in turn, would aid in making informed decisions for breeding programs aimed at improving resistance to both biotic and abiotic stressors.

Given that phenotyping in plant breeding will always be an important practice, as determination of genetic variability may not always explain the complexity of phenotype markers (Velu and Singh, 2013), this study filled relevant knowledge gaps to facilitate the improvement of screening methods for winter wheat varieties’ responses to yellow rust infection. Our results indicated that the capabilities of commercial winter wheat varieties to overcome damage caused by high levels of Pst infection (>50%) are different, which affects fungicide efficacy in preventing yield losses. Our findings also suggest that the level of pathogen pressure that triggers susceptibility reactions of varieties is cultivar-specific and needs to be investigated in more detail in order to ensure more reliable yellow rust risk prediction in the future.