The holoparasitic plant species sunflower broomrape (Orobanche cumana Wallr.) is currently one of the main constraints for sunflower production in most sunflower-producing areas. This species is currently distributed across Europe and Asia (Fernández-Martínez et al., 2015) and is spreading to other areas, such as Africa (Amri et al., 2012; Nabloussi et al., 2018). Sunflower yield losses due to broomrape infestation can surpass 50% when susceptible hybrids are grown, reaching even 100% under conditions of heavy infestation (Cvejić et al., 2020). In the wild, sunflower broomrape has a narrow range of host plants restricted to the Asteraceae family, particularly species of Artemisia, across a geographic range stretching from Central Asia to South-eastern Europe (Pujadas-Salvà and Velasco, 2000). In agricultural production, sunflower broomrape is host-specific and only parasitizes on sunflower crops (Fernández-Martínez et al., 2015).

Sunflower broomrape produces numerous minuscule seeds that remain in a dormant state within the soil until their germination is triggered by specific compounds exuded by the host plant, primarily strigolactones and sesquiterpene lactones (Raupp and Spring, 2013). Upon germination, the germ tube elongates its cells towards the host plant’s roots. When it reaches them, its apical cells differentiate into papillae secreting a mucilaginous substance that promotes adhesion to the host. After attachment, the germ tube produces a swollen structure at the apex, the prehaustorium, whose intrusive cells progress through the host root tissues (Delavault, 2015). Once the pre-haustorium penetrates the endodermis, parasite cells progress toward host vessels and then initiate a specialized endophytic organ, the haustorium sensu stricto, that proceeds to the establishment of vascular connections and enables the parasite to uptake water and nutrients from the host conductive system (Delavault, 2015). Then, broomrape begins to derive phloemic flow acting as a strong nutrient sink and a nutrient storage organ called tubercle develops quickly from which an underground shoot develops (Krupp et al., 2019). The shoots emerge from the soil, flower, and produce new seeds that contribute to the expansion of the seed bank (Molinero-Ruiz et al., 2015).

In contrast to most parasitic systems involving Orobanche or the closely related Phelipanche species, the interaction between sunflower broomrape and sunflower is primarily characterized by a gene-for-gene relationship. In this interaction, a dominant resistance gene in the sunflower host interacts with a corresponding dominant avirulence gene in the broomrape parasite (Rodríguez-Ojeda et al., 2013). This type of interaction, characterized by vertical or qualitative resistance mechanisms within the host, gives rise to distinct physiological races of the parasite (Fernández-Martínez et al., 2015). These races are designated with letters, starting with A, representing the initial race observed in Russia during the nineteenth century, and extending to G, which is currently the most virulent race prevalent in many regions (Cvejić et al., 2020). Nevertheless, it remains uncertain whether populations labeled with the same letter in various parts of the world truly represent the same physiological race, given the absence of a universally applicable set of differential lines for race classification (Velasco et al., 2016). For that reason, Martín-Sanz et al. (2016) proposed adding a subscript abbreviation to the letter of the race to indicate the country or region of origin.

Broomrape race evolution has been parallel in most of the areas where the parasite has been present for a long time, probably because the resistance genes deployed in commercial cultivars were the same. In these areas, race E was the most virulent race at the end of the 1970s (Vranceanu et al., 1980). Populations with increased virulence, named race F were detected in 1995 in Spain (Alonso et al., 1996) and shortly after in most areas of Eastern Europe (Škorić et al., 2010). New populations of race G overcoming race-F-resistant hybrids were first identified in Turkey (Kaya et al., 2004) and not long after in the main sunflower-producing areas of Europe and Asia (Kaya, 2014).

In Spain, where sunflower broomrape is not present in the wild flora, and the parasite only occurs in agricultural fields after its introduction from other areas, two genetically separated gene pools have been identified, one of them in the Guadalquivir Valley in southern Spain and another in Cuenca province in central Spain. Both gene pools exhibit shallow genetic diversity due to founder effects (Pineda-Martos et al., 2013). These authors observed for the first time the admixture of plants from both gene pools in a few sunflower fields. They also documented the existence of cross-fertilization between individuals of both gene pools in some of the populations, although genetic recombination did not affect virulence. A few years later, Martín-Sanz et al. (2016) identified populations in southern Spain derived from the genetic recombination of the Guadalquivir and Cuenca gene pools that possessed increased virulence compared to the original race-F populations. These populations overcame the genetic resistance in most race-F-resistant hybrids tested and were classified as race G_GV. The authors hypothesized that increased virulence resulted from the genetic recombination of avirulence genes from the two gene pools. It is important to recall that sunflower and sunflower broomrape generally follow a gene-for-gene interaction, in which a resistance gene in the crop interacts with an avirulence gene in the parasite (Vranceanu et al., 1980; Rodríguez-Ojeda et al., 2013).

The situation of the Cuenca gene pool has remained unexplored since the study of Pineda-Martos et al. (2013). Meanwhile, it was noticed that some populations of that area possessed the ability to parasitize on DEB2 sunflower line, which is resistant to race-G populations from eastern Europe and southern Spain (Martín-Sanz et al., 2016). Accordingly, this study aimed to evaluate the virulence and population structure of a set of sunflower broomrape populations collected in Cuenca Province, Spain.

Two gene pools of sunflower broomrape have been traditionally present in Spain, one in the Guadalquivir area of southern Spain and another in the Cuenca province in central Spain. The two gene pools are genetically distant and have shallow intrapopulation genetic diversity caused by a founder effect (Pineda-Martos et al., 2013). In that study, the introduction of populations from the Guadalquivir Valley into Cuenca and vice versa was detected for the first time. This fact has explained the appearance of a new race G in the Guadalquivir Valley, putatively caused by the genetic recombination of avirulence loci of both gene pools (Martín-Sanz et al., 2016). After detecting a new virulence of sunflower broomrape in Cuenca province, we investigated whether genetic recombination of avirulence genes could also be at the bottom of this new virulence. Notably, the new virulence observed in Cuenca is mainly characterized by the parasitization of the DEB2 sunflower line, which is resistant to all race-G populations evaluated so far (Martín-Sanz et al., 2016). Based on the results of the present study, the hypothesis of the genetic recombination of avirulence genes present must be discarded as the cause of DEB2 parasitization since the population with the highest degree of attack on DEB2, NCU1, showed full membership to the Cuenca gene pool, with no mixture with any external population. Accordingly, under no evidence of genetic recombination associated with the change in virulence, it can be hypothesized that a point mutation may have caused the virulence change in the Cuenca gene pool. A point mutation was also hypothesized as the genetic mechanism underlying the change from race E to race F virulence in the Guadalquivir Valley since the increased virulence was associated with dominant alleles at a single avirulence locus (Rodríguez-Ojeda et al., 2013) and no changes in population structure and diversity were observed between race-E and race-F populations (Pineda-Martos et al., 2013).

Although genetic recombination seems not to be the cause for DEB2 parasitization, the results of this study showed that genetic recombination between local populations and populations introduced from the Guadalquivir Valley is occurring in Cuenca nowadays. This is more clearly seen in populations NCU2, NCU4, and NCU8, where F₁ individuals between the Cuenca gene pool plants and the Guadalquivir Valley gene pool have been identified. The case of NCU2 population deserves additional discussion. From 21 individuals analysed, 10 belonged to the Cuenca gene pool, 9 belonged to the Guadalquivir Valley gene pool, and two plants were F₁. This suggests that the introduction of plants from the Guadalquivir Valley into this population has been very recent, and genetic recombination is at the earliest stage. The case of population NCU10, where all the individuals belong to the Guadalquivir Valley gene pool, suggests that broomrape is being introduced from the Guadalquivir Valley even in areas where populations of the Cuenca gene pool were not present. Broomrape seeds are mainly dispersed through agricultural machinery and tools and together with the host seeds (Habimana et al., 2014). Although how and when the broomrape seeds of the Guadalquivir Valley gene pool have reached Cuenca province cannot be ascertained, it is important to note that populations of this gene pool have been also found in distant places such as northern Spain (Malek et al., 2017) and Morocco (Nabloussi et al., 2023).

This is the first time that complex sunflower broomrape populations have been characterized in detail. Their individuals could be unequivocally classified according to the gene pool of provenance, including identifying F₁ hybrid plants. This opens the possibility of extending the methodology used in this study to other areas, assuming that identifying the original gene pools is still possible. What we could not separate in the present study were the individuals of the Cuenca gene pool with classical race-F virulence, not attacking DEB2 sunflower population, from those with increased virulence that parasitize on DEB2. Assuming the point mutation hypothesis exposed above, separating both groups would require the availability of markers closely linked to the avirulence locus, which are not currently available.

The nomenclature for sunflower broomrape races has been traditionally based on the use of letters, with A indicating the initial population attacking sunflowers in Russia at the end of the nineteenth century, B the population that overcame resistance to race A, and so on (Cvejić et al., 2020). This led to using the same letter to designate the virulence of populations in different geographic areas with different virulence profiles, which caused great confusion. For that reason, Martín-Sanz et al. (2016) proposed adding a subscript indicating the geographical area of the population. In this case, considering that the most virulent race in the area so far was race F, the new populations attacking DEB2 sunflower line should be named race G. However, it must be considered that race G is being used in other areas to designate populations that overcome the resistance of race-F-resistant hybrids (Škorić et al., 2021), which is the case of population NCU9 but not the other populations. Consistent with this nomenclature, we propose to designate race G_CU only to the populations that show a number of shoots significantly higher than zero on Kiara, and race F⁺ _CU to the populations that show a number of shoots significantly higher than zero on DEB2 but not on Kiara. Any other population parasitizing on NR5 but not parasitizing significantly on Kiara or DEB2 should be considered race F_CU. Accordingly, NCU5, NCU7, NCU8, and NCU10 are currently race F_CU, NCU1 to NCU4 and NCU6 are race F⁺ _CU, and population NCU9 is race G_CU. This is the current situation, although some populations already show a low, non-significant number of shoots on Kiara and/or DEB2 and probably will evolve soon to higher levels of virulence. Thus, it cannot be discarded that some populations, such as NCU3, become virulent on race-F resistant hybrids in addition to their virulence on DEB2. In that case, we propose that they are classified as (F⁺G)_CU. Concerning race G_CU (population NCU9), it can be argued that it was probably introduced from the Guadalquivir Valley and could be considered a race G_GV. However, with the information available, we cannot discard that this population might have evolved locally in a similar way as populations G_GV evolved in the Guadalquivir Valley, as hybridization between both gene pools in Cuenca has been documented in the present research as it was documented previously in the Guadalquivir Valley (Pineda-Martos et al., 2013).

It is generally accepted that resistance to broomrape in sunflower is cumulative, i.e., resistance to a new race also provides resistance to the previous ones (Fernández-Martínez et al., 2015). In this study, we show that sunflower germplasm can be resistant to one race but not to another one, e.g., DEB2 is resistant to G_CU but not to F+_CU, whereas Kiara shows the opposite response. This observation emphasizes the need for a detailed characterization of the existing races of sunflower broomrape and the genetic structure of the populations to have a better picture of how the populations are evolving and the mechanisms driving race evolution. Also, our results highlight the importance of pyramiding resistance genes as a strategy to develop durable resistance to sunflower broomrape. Pyramiding strategies should include genes with a major effect, such as Or₇ (Duriez et al., 2019) and Or_Deb2 (Fernández-Aparicio et al., 2022), but also genes determining partial or posthaustorial resistance (Martín-Sanz et al., 2020), or even genes involved in polygenic resistance (Pérez-Vich et al., 2013; Louarn et al., 2016; Imerovski et al., 2019).

The race evolution of sunflower broomrape is primarily driven by mutations, consistent with the gene-for-gene host-parasite interaction between this parasite and the sunflower crop (Rodríguez-Ojeda et al., 2013; Škorić et al., 2021). Genetic recombination between avirulence loci has also been proposed as an additional mechanism for race evolution in this species (Martín-Sanz et al., 2016). In our present research, we identified sunflower broomrape populations in which both genetic mechanisms have modulated virulence. These findings provide valuable insights into the mechanisms of race evolution in sunflower broomrape and can aid in developing effective control strategies.

The original contributions presented in the study are included in the article/ Supplementary Material , further inquiries can be directed to the corresponding author.

LV, AM-S, and BP-V conceived the work and planned and supervised the research. LV and AM-S collected the broomrape populations and evaluated virulence. BF-M, LM, and BP-V carried out plant genotyping. BF-M, BP-V, and LV performed data analyses. BF-M and LV wrote the draft of the manuscript. All authors contributed to the article and approved the submitted version.