Incidence of resistance to ALS and ACCase inhibitors in Echinochloa species and soil microbial composition in Northern Italy

The increasing amount of weeds surviving herbicide represents a very serious problem for crop management. The interaction between microbial community of soil and herbicide resistance, along with the potential evolutive consequences, are still poorly known and need to be investigated to better understand the impact on agricultural management. In our study, we analyzed the microbial composition of soils in 32 farms, located in the Northern Italy rice-growing area (Lombardy) with the aim to evaluate the relationship between the microbial composition and the incidence of resistance to acetolactate synthase (ALS) and acetyl-CoA carboxylase (ACCase) inhibiting herbicides in Echinochloa species. We observed that the coverage of weeds survived herbicide treatment was higher than 60% in paddy fields with a low microbial biodiversity and less than 5% in those with a high microbial biodiversity. Fungal communities showed a greater reduction in richness than Bacteria. In soils with a reduced microbial diversity, a significant increase of some bacterial and fungal orders (i.e. Lactobacillales, Malasseziales and Diaporthales) was observed. Interestingly, we identified two different microbial profiles linked to the two conditions: high incidence of herbicide resistance (H-HeR) and low incidence of herbicide resistance (L-HeR). Overall, the results we obtained allow us to make hypotheses on the greater or lesser probability of herbicide resistance occurrence based on the composition of the soil microbiome and especially on the degree of biodiversity of the microbial communities.


Incidence of resistance to ALS and ACCase inhibitors in Echinochloa species and soil microbial composition in Northern Italy
Carlo Maria Cusaro 1,2 , Enrica Capelli 1,2 , Anna Maria Picco 1,2 & Maura Brusoni 1,2* The increasing amount of weeds surviving herbicide represents a very serious problem for crop management.The interaction between microbial community of soil and herbicide resistance, along with the potential evolutive consequences, are still poorly known and need to be investigated to better understand the impact on agricultural management.In our study, we analyzed the microbial composition of soils in 32 farms, located in the Northern Italy rice-growing area (Lombardy) with the aim to evaluate the relationship between the microbial composition and the incidence of resistance to acetolactate synthase (ALS) and acetyl-CoA carboxylase (ACCase) inhibiting herbicides in Echinochloa species.We observed that the coverage of weeds survived herbicide treatment was higher than 60% in paddy fields with a low microbial biodiversity and less than 5% in those with a high microbial biodiversity.Fungal communities showed a greater reduction in richness than Bacteria.In soils with a reduced microbial diversity, a significant increase of some bacterial and fungal orders (i.e.Lactobacillales, Malasseziales and Diaporthales) was observed.Interestingly, we identified two different microbial profiles linked to the two conditions: high incidence of herbicide resistance (H-HeR) and low incidence of herbicide resistance (L-HeR).Overall, the results we obtained allow us to make hypotheses on the greater or lesser probability of herbicide resistance occurrence based on the composition of the soil microbiome and especially on the degree of biodiversity of the microbial communities.
Weed management is one of the most critical aspects in agriculture.Each year considerable worldwide yield losses are estimated, due to the negative impact of weeds on crop production 1 .The annual worldwide cost of crop losses caused by weed infestation is estimated around 32 billions of USD 2 .To date, chemical control has represented the most efficient tool for managing weeds.However, herbicide resistance (HeR) represents an upand-coming phytosanitary threat to worldwide agricultural systems and is an example of the adaptive evolution of weeds in response to human selective pressures 3 .The increasing amount of weeds surviving herbicide represents a very serious problem particularly for those territories traditionally suited to rice cultivation such as Po Valley in Northern Italy.In this area, a survey on the spread of herbicide resistance has been underway for a long time through monitoring studies and laboratory testing with the aim of highlighting the causes favoring the phenomenon.It was seen that resistance can persist for several years due to the seed stock in the soil 4 .
In addition, artificial selection of agronomic traits in rice (Oryza sativa L.), which are useful to humans, has unintentionally promoted the evolution of crop-like weed biotypes.As a result, the weeds can evade chemical control and eradication from fields, allowing them to spread throughout the agroecosystem.In fact, weeds has evolved an adaptive phenomenon (Vavilovian mimicry) which allowed them to resemble domesticated crops at specific stages in their life history.This phenotypic adaptation results in a morphological similarity of weeds to crops, making difficult for farmers to distinguish them 5,6 .
A lot of studies have demonstrated that herbicide resistance can be ascribed either to a DNA missense mutation in genes expressing specific proteins targeted by herbicides (target site resistance -TSR), or to herbicides detoxification processes (non-target site resistance -NTSR) 7 .Moreover, the increasing repeated field applications of an increasingly narrow range of herbicides, as a consequence of the withdrawal of many plant protection products (PPPs) from the EU market due to strict regulation (Reg EC/1107/2009) 8 , and the lack of herbicides with new modes of action (MoAs), determine a continuous rise and spread of the resistance [9][10][11][12] .Furthermore, other factors may influence herbicide resistance, such as epigenetic mechanisms (i.e.miRNAs) regulating the expression of genes encoding for enzymes involved in xenobiotic detoxification 13 .Environmental traits may also be related to the occurrence of herbicide resistance, such as the rise of temperature 14 which could affect chemical control efficacy.New information are emerging on the effects of herbicides on soil microbial communities.However, the interaction between microbial community and herbicide application, along with the potential consequences, are still poorly known and need to be investigated to better understand the possible relapses on agricultural management.Paddy soil microbiota was discovered to be critical to the wellness and fitness of rice, but also of herbicide resistant weeds.It has been discovered that different biomass allocation and root traits resulted for rice in the presence of herbicide-resistant Echinochloa crus-galli (L.) P. Beauv (barnyardgrass), since this weed recruited a distinct microbial consortium in the rhizosphere soil through its exudates 15 .It is known that root systems releases various plant metabolites through root exudation.Weeds and crops exudates may be selective for specific soil microbial groups, stimulating their development in the soil.This may favor symbiotic and associative interactions between plants and microorganisms, which could be important to plants in the uptake of nutrients and water, but also against phytopathogens 16 .
As a matter of facts, even if herbicides are thought to be specific to plants, it has been recently reported that they can affect other types of organisms by targeting evolutionarily conserved pathways.For instance, the growth of Stenotrophomonas maltophilia (Xanthomonadales order), an aerobic non-fermentative gram-negative bacillus, sometimes present in rice fields, resulted negatively affected by herbicides 17 .Moreover, an unsuspected impact on diversity and composition of microbial communities has been revealed by glyphosate, a common broad-spectrum herbicide, since its targets enzyme (shikimate pathway, 5-enolpyruvylshikimate-3-phosphate synthase -EPSPS) is present in many prokaryotes and fungi 18 .
Hence, herbicide exposure can change microbial communities because microbes differ in their intrinsic susceptibility.As a consequence, differences in sensitivity can lead to changes in the abundance of certain microbes under herbicide exposure.Moreover, many microbes can metabolize herbicides and some can use them as sources of nutrients 19 .Therefore, herbicide residues may increase the abundance of herbicide-metabolizing microbes with a cascade effect on the community.Thus, changes due to herbicide exposures can impact microbial composition inducing changes on community functionality.Further, the physical-chemical characteristics of the soils and climatic conditions influence soil microbiome, which results consequently variable 19,20 .
The aim of the present study was to evaluate the relationship between the soil microbial (Bacteria, Archaea and Fungi) composition of paddy fields in the Po Valley, an area traditionally vocated to rice production, and the incidence of herbicide-resistant weeds.E. crus-galli and Echinochloa oryzicola (Vasinger) Vasinger (late-watergrass) are the most problematic weeds infesting Italian rice fields due to their high ability to develop resistance to the most commonly used herbicides 21 .Among the most widespread and noxious weeds infesting rise cultivation, the species of the genus Echinochloa P. Beauv., tribe Paniceae R. Br. subfamily Panicoideae A. Br., family Gramineae Juss (= Poaceae Barnh.), are the worst due to their wide ecological success and ability to mimic the crops 5,6,22 .
Soil microorganisms have an important role in the metabolization of macronutrients and in the production of bioactive molecules that play a pivotal role at the rhizosphere level and in the interaction between plants.A recent study reported that the relationships between microbial communities and plants are controlled by the type of plants and the microorganisms involved.Plant microbiota could be beneficial or pathogenic to the plant 23 .Weeds can mitigate soil microbiota that colonize the rhizosphere and acquire nutrient sources unavailable to the crop, thereby creating soil conditions that favor weeds increasing their competitive ability [24][25][26] .Moreover, the composition of soil microbiota is impacted by herbicide treatments [27][28][29] .
Systemic herbicides released through weed roots may select microorganisms that mediate detrimental processes such as nutrient immobilization or serve as opportunistic pathogens.Kremer 16 highlighted the implications of herbicide resistance on soil biology and ecology, pointing out that herbicides may compound effects of weeds on soil microorganisms.Some studies have reported that repeated herbicide exposure during weed control favors increased herbicide tolerance in Bacteria 30 .Herbicide tolerance can be achieved via genetic changes in the herbicide target gene 31 or non-target genes linked with generalized stress tolerance 32,33 .However, the relationship among weeds, soil properties and soil microbial communities has not been fully explored.Soil organisms are an integral component of ecosystems, but their activities receive little recognition in agricultural management strategies.Soil organisms have the potential to enhance ecosystem service delivery and soil biodiversity promotes multiple ecosystem functions simultaneously (i.e.ecosystem multifunctionality).
The data obtained could be useful to implement more effective weed control strategies and to apply more sustainable methods of intervention in farming practices.

Identification of farms affected by herbicide resistance phenomena
For the present research we have considered 32 rice farms, located in the provinces of Pavia and Milano (Lombardy region, Italy) and managed according to the Directive 2009/128 EC 34 that establishes a framework to achieve a sustainable use of pesticides promoting the use of integrated pest management (IPM) that aims to keep the use of pesticides and other forms of intervention only to levels that are economically and ecologically justified and which reduce or minimize risk to human health and the environment.The principles on which it is based concern seed selection, fertilization, the use of plant protection products and strategies to reduce herbicide resistance.IPM is mandatory since 1 January 2014.
www.nature.com/scientificreports/In the Figure 1 is displayed the rice cropping area of the Lombardy region (Figure 1a) within the provinces of Pavia and Milan where Echinochloa spp.specimens survived after the application of acetyl-CoA carboxylase (ACCase -HRAC group: A) and acetolactate synthase (ALS -HRAC group: B) inhibitors were identified (Figure 1b).The distribution of farms where herbicide resistant Echinochloa were observed are indicated with red marks.The map obtained following our survey indicates that this phoenomenon is widely spread-out within the Lombardy region, including areas where the presence of Echinochloa HeR was not reported (GIRE) 4 .

Echinochloa spp. herbicide resistance (HeR) incidence
Two species of Echinochloa were identified in the rice fields investigated for this study: E. crus-galli was recorded in 15 out of 32 of the surveyed farms, while E. oryzicola in 21 out of 32 of the farms.Both species were found only in 4 farms (Table 1).
The incidence of specimens surviving herbicide treatment (HeR incidence) was assessed as percent coverage.Based on the values in Table 1, it is possible to differentiate farms with a low incidence (coverage less than 5%), farms with a medium incidence (coverage between 5 and 60%) and farms with a high incidence (coverage higher than 60%).
Values of low HeR incidence were reported in 9 rice farms (FR.

Microbial biodiversity analysis
The sequencing of amplicons on the Illumina MiSeq Platform produced a total of 1,989,664 raw reads for Bacteria and Archaea (on average 62,177.00per sample) and a total of 3,000,941 raw reads for Fungi (on average 93,779.41per sample).
After quality filtering, 362,811.00sequences remained (on average 11,337.84per sample) for Bacteria and Archaea and 699,877.00sequences (on average 21,871.16per sample) for Fungi.
The Greengenes 16S and the UNITE ITS reference datasets were used for determining operational taxonomic units (OTUs) at the 99% level.A total of 51 phyla, 141 orders, 238 genera and 127 species was identified for Bacteria and Archaea.A total of 4 phyla, 16 orders, 16 genera and 15 species was identified for Fungi.Of these, the amount of unassigned and unidentified OTUs was 81% and 68% respectively.In Supplementary Table 1 are listed the identified bacterial, archaeal and fungal orders.Only taxa with a prevalence at phylum level equal or higher than 5% were considered for the analysis.
In the soil samples analyzed, the bacterial and archaeal communities showed higher average values of richness, diversity and evenness and lower average values of dominance compared to the fungal ones.
As concerns bacterial and archaeal communities, the highest values of the Margalef index were observed in samples FR.9, FR.15, FR.18, FR.22, FR.26 and FR.32 while the lowest were observed in FR.17 Vol:.( 1234567890

Herbicide resistance and microbial composition of soils
We searched for a relationship between the microbial composition of soils and the incidence of herbicide resistance.Figure 2 shows the heatmap and dendrogram obtained from the bootstrap-based hierarchical clustering of the paddy soils considering both the incidence of resistant Echinochloa spp.(HeR) and the prevalence of microbial orders.
Two main clusters were identified.Cluster A grouped the soils in which a higher incidence of resistant Echinochloa spp.and a lower microbial diversity were recorded.Otherwise, cluster B included the soils in which a lower incidence of resistant Echinochloa spp.and a higher microbial diversity were observed.It was also noted Table 1.Values of abundance (% coverage of the plant-area projection on the paddy area) of the E. crus-galli and E. oryzicola specimens survived herbicide administration in the surveyed rice farms.Total HeR incidence: midpoint of cover range (according to Braun-Blanquet scale) of the sum of E. crus-galli and E. oryzicola % coverage.www.nature.com/scientificreports/ that some bacterial, archaeal and fungal orders possessed higher prevalence in cluster A than in cluster B, and vice versa.For example, Rickettsiales, Enterobacteriales, Lactobacillales, Neisseriales, Malasseziales and Diaporthales were more abundant in soils of cluster A, while Rhizobiales, Methanosarcinales, Pseudanabaenales, Agaricales, Sordariales, Pezizales and Mortirellales were more represented in cluster B. These results suggested a possible relationship between certain bacterial and fungal orders and the presence of resistant Echinochloa spp.Envfit analysis revealed that Actinomycetales, Rhizobiales, Methanosarcinales, Gaiellales, Solibacterales, Desulfurococcales, Pseudanabaenales, Solirubrobacterales, Lactobacillales, Malasseziales, Agaricales, Sordariales, Diaporthales and Venturiales were the bacterial, archaeal and fungal orders which significantly contribute to the classification of farms within the two clusters A and B (Supplementary Table 2).

Farm
From the Principal Coordinates Analysis (PCoA), three distinct groups of soils were identified in relation to the incidence of herbicide resistance and the hosted microbial communities (Figure 3): soils with low incidence (L-HeR -red ellipse), soils with medium incidence (M-HeR -yellow ellipse) and soils with high incidence (H-HeR -blue ellipse).The analysis revealed that in the H-HeR group of soils, Actinomycetales, Lactobacillales, Diaporthales and Malasseziales are the prevalent orders.Otherwise, in the L-HeR and M-HeR groups, Rhizobiales, Methanosarcinales, Gaiellales, Solibacterales, Desulfurococcales, Pseudanabaenales, Solirubrobacterales, Agaricales, Sordariales and Venturiales are the mainly present orders.
The contribution of microbial communities composition in the variations of the incidence of resistant E. crus-galli and E. oryzicola was tested by redundancy analysis (RDA).
Comparing paddies with a low resistance incidence (L-HeR) with those recording a high resistance incidence (H-HeR), it was observed that the microbial composition is different both from a quantitative and qualitative point of view (Supplementary Tables 4 and 5).
Considering Fungi, the comparison between the two groups showed that Malasseziales and Diaporthales were the most prevalent in H-HeR paddies (on average, 80.04% and 11.00% respectively), otherwise a high prevalence of Pezizales, Agaricales, Mortierellales and Sordariales (on average 8.52%, 7.62%, 3.83% and 2.70% respectively) was recorded in L-HeR soils (Supplementary Table 1).
At genus/species level, Malassezia, represented by the species M. restricta and M. globosa, recorded the highest prevalence values in H-HeR soils (Supplementary Table 5).
According to European and Mediterranean Plant Protection Organization (EPPO) Global Database 35 , providing basic information for species of interest to agriculture, forestry and plant protection and detailed information for pest species that are of regulatory interest (EPPO and EU listed pests, as well as pests regulated in other parts of the world), out of the micro-organisms identified and listed in Supplementary Tables 4 and 5, only Rhodococcus fascians, order Actinomycetales, is of phytosanitary relevance but EU regulated as no quarantine pest.

Physical-chemical properties of soils
The physical-chemical parameters of the different paddy soils are listed in the Supplementary Table 7.All the soils analyzed possessed a sandy/silty texture.Most of the soils resulted acidic and slightly acidic and with a low or medium nitrogen content (N).Considering C/N ratio, 8 soils (FR.6, FR.8, FR. 16   ).The majority of soils (25/32) resulted with a high or very high content of assimilable phosphorus (P) and the remaining soils with a low or very low content.Only one of the soils studied (FR.7) showed physical and chemical characters different from all others, with the highest content of silt and clay, calcium (Ca), magnesium (Mg) and with a high nitrogen (N) content.

Physical-chemical properties of soils and microbial composition
We also wanted to verify whether soil microbial biodiversity and composition could be related to soil physicalchemical properties.From the redundancy analysis (RDA) none of the analyzed physical-chemical characteristics recorded significant effect (anova.cca,P > 0.05) on the composition of bacterial, archaeal and fungal communities (Supplementary Table 8).

Herbicide resistance and physical-chemical properties of soils
The relationship between the incidence of resistant Echinochloa spp.(HeR) and the physical-chemical properties of soils was analyzed.The obtained heatmap and bootstrap-based dendrogram are reported in Figure 6.Two main clusters were identified: cluster A characterized by soils with thick texture (coarse and fine sand), low content of organic matter, organic carbon and macronutrients, cluster B including soils with fine texture (silt and clay), high content of organic matter, organic carbon and macronutrients.From the heatmap and hierarchical clustering results, it can be noticed that soils with high incidence of herbicide resistance (intense green color) are equally distributed in both clusters.www.nature.com/scientificreports/Envfit analysis revealed that soil texture, pH, organic matter content, CaCO 3 , N, Ca, Mg, K and Na content, cationic exchange capacity, base saturation, were the physical-chemical edaphic properties which significantly contribute to the classification of farms within the two clusters A and B (Supplementary Table 9).
The contribution of edaphic physical-chemical characteristics in the variations of the incidence of resistant Echinochloa spp.(HeR) was tested by redundancy analysis (RDA) that revealed no significant association (anova.cca, P > 0.05) (Supplementary Table 10).

Discussion
For the present study, we analyzed the microbial composition in the soils of 32 farms located in the Lombardy rice-growing area, all managed according to Directive 2009/128 EC 34 , in which specimens of E. crus-galli and E. oryzicola resistant to herbicide treatments were detected.
We examined the incidence of resistant Echinochloa spp.specimens in the paddies of the considered rice farms, highlighting that herbicide resistance phenomenon is particularly spread within the rice cropping territory of the Lombardy region, also in areas where it resulted underestimated.Various incidence values of herbicide resistance were observed, however this variability did not result related to the different types of herbicide used and their MoAs.Weed infestation is likely to be generated from a highly persistent soil seed bank, as a result of seed dormancy and resistance to conventional weed control strategies, as reported by previous studies 36 .
Microbial communities-diversity analysis highlighted a variability in the richness, diversity, dominance and evenness indexes assessed in the considered soils for Bacteria, Archaea and Fungi.Also in this case no relation with herbicide treatment and their modes of action was revealed.In fact, it has been noticed that bacterial and fungal communities characterized by very differing indexes were detected also in rice farms where the same www.nature.com/scientificreports/type of herbicide has been applied.For example, FR.17 and FR.18 paddies, treated both with clethodim (ACCase inhibitor), recorded respectively the lowest and the highest value in bacterial richness.Moreover, the highest richness and diversity of fungal communities were recorded in FR.2, FR.20, FR.23, FR.28, FR.30 and FR.31 paddies, which were treated with different herbicides.Concerning soil physical-chemical properties, our findings revealed that, according to the Italian standards (see Methods), the majority of the analyzed soils possess a prevalent sandy/silty texture, an acid or slightly acid pH and a medium or low content of macronutrient.Most of the considered soils possess a low or medium content of organic matter, a balanced mineralization degree (C/N), a medium nitrogen content, a low or medium cationic exchange capacity, a very low content of Ca, Mg, K and Na and a medium or high content of assimilable P. All soils are extremely poor in CaCO 3 .In general, all soils possess a quite similar chemistry, hence the variation in microbial composition cannot be related to physical-chemical parameters.These results were expected because all the rice fields are located in the Po Valley, a homogeneous area from a climatic and geological point of view and furthermore all the considered farms perform the same type of agronomic management.
The abundance of weeds that survived herbicides was greater in paddy fields where a low microbial biodiversity was recorded.Fungal communities appeared to be more involved in the phenomenon, showing a greater reduction in richness than Bacteria and Archaea.Together with a general reduction of microbial diversity, we observed a significant increase of some orders, particularly Malasseziales, Diaporthales and Lactobacillales.
From RDA analysis, Lactobacillales, Malasseziales and Diaporthales resulted in a positive relation with the high incidence of HeR, otherwise Rhizobiales and Solibacterales resulted in a positive relation with the low incidence of HeR.
Lactobacillales are an order of gram-positive, acid-tolerant bacteria, usually found in decomposing plants and milk products.They produce lactic acid as the major metabolic end-product of carbohydrate fermentation, giving them the common name lactic acid bacteria (LAB).Proteinaceous bacteriocins are produced by several LAB strains and provide an additional hurdle for spoilage and pathogenic microorganisms.Bacteriocin metabolites are toxic to microbes 37 .In addition, organic acids are the prominent secondary metabolites that exhibit antifungal www.nature.com/scientificreports/activity and preservative effects in fermented food and silage 38 .Recent studies reported LAB strains as promising candidates for sustainable agriculture, since they promote soil health and fertility 39 .Considering the results of our research, it would be useful to conduct further analysis to deepen the relation between high abundance of Lactobacillales in soils and high incidence of herbicide resistant weeds, that forces the farmer to perform more frequent treatments, in contrast to agricultural sustainability.
Malasseziales is a heterogeneous group of species, and several species comprise multiple genotypes associated with mammalian hosts, but using culture-independent techniques they were also retrieved from much widerspread habitats, including various terrestrial and marine ecosystems and even deep-sea sediments.Furthermore, Malassezia DNA was detected from soil nematodes in Central European forests, and it has been hypothesized that nematodes may serve as a vector for Malassezia species.M. restricta and M. globosa were associated with the nematode genus Malenchus spp., whereas another nematode, Tylolaimophorus typicus hosted only M. restricta 40,41 .
In our study we identified at species level M. restricta and M. globosa with the highest prevalence in H-HeR soils.Moreover, H-HeR paddies were characterized by the absence of some genera and species that were found in L-HeR ones (i.e.Lycoperdon pratense (Agaricales) recovered in 80% of soils and Bolbitius coprophilus (Agaricales) recovered in 40% of soils).Mortierella (Mortierellales) observed in both groups, but showed the lowest values in H-HeR soils.
The order Diaporthales includes plant pathogens, plant endophytes, saprobes, human-animal pathogens, and soil inhabitants, some of which have extensive host ranges and geographical distributions 42 .As endophytes, they live in medicinal plants and are used for studies that investigate antimicrobial activities, e.g., Diaporthe spp., which were isolated from the hosts Copaifera langsdorffii and C. pubiflora 43 .Antibacterial activity has been demonstrated using extracts of two unidentified Diaporthe spp.and D. miriciae 44 .
The high prevalence of Diaporthales and Malasseziales we observed in association with the high incidence of herbicide-resistant weeds and with the reduced microbial biodiversity, suggests a possible influence of environmental conditions (i.e.herbicide treatments) causing the selective growth of stress tolerant microorganisms, with a high capacity to regulate gene expression in response to changes in temperature, saltiness, nutrient content, acidity fluctuations as reported for different Bacteria and Fungi (i.e.Malasseziales) [45][46][47][48] .
Rhizobiales are well-known beneficial partners in plant-microbe interactions.They commonly exert beneficial functions for their hosts by providing various nutrients, phytohormons as well as precursors for essential plant metabolites [49][50][51] .The order contains many genera of nitrogen-fixing, methanotrophic, legume-nodulating and microsymbiotic bacteria 52,53 .
The order Solibacterales produces enzymes to metabolize complex organic compounds or macromolecules available in its environment for metabolism and participates in nitrate and nitrite reduction.Some studies have shown that Solibacterales thrives in soils rich in phosphorus.In fact, Solibacterales is a proficient P miner and a bioindicator of the organic P nutrient-use economy, as its relative abundance is strongly related to phosphatase activity 54 .This order might play a role in making phosphorous bioavailable in ecosystems with differing nutrient levels 55,56 .
One of the soils studied (FR.7) showed a microbial composition clearly differing from all the others, with increased abundance of Porphyromonas and Prevotella (Bacteroidales) and Neisseria (Neisseriales).Strains of the family Porphyromonadaceae were isolated from temperate deciduous forest soil exposed to pesticide 57 .Prevotella is a genus of bacteria that commonly associate with humans in various body sites and is spread in various environmental niches 58 .Neisseria is a large genus of human and animal commensal bacteria within the normal microbiota of the human and animal nasopharynx.It comprises opportunistic pathogens and only two strains are recognized as pathogenic (N.gonorreae and N. meningitidis).Neisseria was recovered also in environment from soil and water.The ability of Neisseria to degrade organic pollutants has been confirmed in different contexts.Neisseria have been also found in sites closely associated with humans 59 .Recent reports evidenced the presence of Neisseria in soil treated with herbicides 60,61 .This fact could explain the higher abundance of Neisseriales in a soil characterized by a high incidence of HeR.
The different microbial profiles that characterize H-HeR and L-HeR soils can be explained as a possible selection of certain microbes due to the capacity of weeds to release in soil, through root exudation, metabolites able to drive the composition of microbial communities in accordance with previously published studies 16 .
Plants can modify the rhizosphere microbiota to affect the growth of conspecific and interspecific plants, displaying a variety of effects on each other.Plant-plant interactions can shape soil bacterial and fungal communities via root exudates.The root exudates from resistant E. crus-galli are responsible for the assembly and establishment of the root microbial structure.In particular, resistant E. crus galli is able to assemble more Proteobacteria and Ascomycota to enhance plant stress tolerance 15 .
Another possible explanation is the selection, due to repeated weeding, of stress-tolerant microorganisms with a high capacity to regulate gene expression, as reported by some researchers.It is known that herbicide treatments induce modifications in the transcription activity (RNA metabolism) and in the carbohydrate metabolism of the bacteria present in the soil.For example, an increase in the production of the F0F1 ATP synthase subunit, along with increased expression of a cytochrome c-553-related transcript, was observed in bacteria exposed to glyphosate 29,33,48,62 .
In our condition, the differential microbial profiles observed cannot be related to the different type of management, because all farms are conducted with the same strategy, as mentioned above.Oryza sativa L. cover cannot be a factor influencing the variability of soil microbial composition since it is homogeneous in all the rice fields investigated as found in intensive crops.
The relationship that we highlighted between the incidence of herbicide resistance in E. crus-galli and E. oryzicola and the composition and biodiversity of the microbial communities of paddy soils represents an innovative result.Recent studies have demonstrated the ability of herbicide-resistant specimens of E. crus-galli to influence the microbial diversity and composition of paddy fields by means of root exudates 15  www.nature.com/scientificreports/how seed microbiome assembly is associated with the herbicide resistance evolution in barnyardgrass 63,64 .To date, no study had yet shown differences in soil microbial communities related with higher or lower incidence of herbicide resistant E. crus-galli and E. oryzicola specimens.Considering that plant-associated soil microbial communities have an important role not only in soil nutrient cycling, soil enzyme activity, plant growth promotion, disease suppression but also in abiotic stress tolerance, the assessment of soil microbial composition can represent an important indicator for predicting the development of herbicide resistance phenomena.

Concluding remarks
Understanding complex interactions of weeds with soil microorganisms and improving the biodiversity of soil microbial communities could be strategic in developing more effective approaches to herbicide-resistant weed management and plant protection, with the final aim of optimizing precision weed management (PWM) technologies.
The relation between HeR incidence variability and different microbial communities allows us to make hypotheses on the greater or lesser probability of herbicide resistance occurrence, based on the composition and α-diversity of the soil microbiome.The type of weed affects the composition of soil microbiome, favoring the growth of some microbial groups to the detriment of others.The repeated treatments with the same type of herbicide can favor both the selection of resistant weeds and microbial groups capable of greater proliferation after chemical treatments.It is therefore necessary in the future, to investigate in all these directions to better understand the resistance phenomenon and therefore to implement more targeted control strategies.
Management practices that cause beneficial changes in soil community composition are likely to increase agricultural sustainability.Minimum tillage, cover cropping, organic fertilization and the use of bio-stimulants able to sustain benefit microorganisms can be useful tools for enhancing the biodiversity of soil microbial communities and keeping the problem of herbicide resistance under control.Furthermore, following the assessment of different microbial profiles in soils with different incidences of herbicide resistance, the enrichment of the soil with microorganisms inhibiting the spread of weeds could be of benefit in the context of weed management.
The selection of microorganisms suitable to be associated and the composition of an effective microbial consortium is not easy.It must be based on the understanding of the functionality of the different species and the potential interactions between the different components of the microbial consortium.
The methodology applied in this study and the results obtained can be a useful support for the development of predictive models of the possible evolution of herbicide resistance in relation to microbial composition/diversity in other types of crops, including the vineyard, where recent years have witnessed the evolution of weed populations resistant to the most commonly used herbicides.
Our findings can contribute to planning more targeted and sustainable weed management strategies considering the preservation and enhancement of soil biodiversity, through conservative and regenerative agriculture practices adoption, reducing chemical inputs, improving food health and protecting the environment and human health.In compliance with the Farm to Fork strategy, that provides for a reduction and more sustainable use of pesticides and inputs, regardless of the distribution tool used, the management of resistance phenomena, that can reduce the effectiveness of treatments, is a strategic aspect to be considered.

Study area
The study area is located in the North-western Italy in the Lombardy region (Po Valley) in one of the most important rice-growing vocation territories, characterized by homogeneous climatic and geological conditions with debris, alluvial, fluviolacustrine and fluvioglacial deposits (Supplementary Figure 1).

Sampling
Soil samples and E. crus-galli and E. oryzicola specimens that survived herbicide treatments were collected from 32 rice farms managed according to the Directive 2009/128 EC 34 .Samples collection was conducted during summer 2023.
Soil collection was conducted according to the non-systematic X scheme (~ 1 kg of soil obtained from 12 equidistant collections of equal amount along the diagonals of the sampling unit) by Lambkin 65 .Soil samples were aseptically taken from 10 to 25 cm depth.Three different samples, recovered from the same experimental site were pooled and subdivided into aliquots of about 350 mg for metagenomic analysis and stored at − 30°C.Aliquots of about 1 kg for the analysis of physical-chemical characteristics were kept at 4°C.
The presence of E. crus-galli and E. oryzicola specimens survived herbicide application was recorded by means of cover/abundance values in accordance to Braun-Blanquet 66 at an equal and random defined sampling unit of about 100 m 2 extension within each paddy (% coverage of the plant-area projection on the paddy area).The total HeR incidence value was assigned as the midpoint of cover range (according to Braun-Blanquet scale) of the sum of E. crus-galli and E. oryzicola % coverage.
Sampling units were signed with an identification code.Herbicides applied, classified as mechanism of action (MoA) and following the Herbicide Resistance Action Committee (HRAC) 67 , and Echinochloa spp.survived chemical control are listed in Table 3. Herbicide resistance was tested according to Cusaro 13 : (i) controlled green-house growth trials were performed applying the same herbicides used in fields; (ii) plants were maintained in a growth chamber with a mean temperature of 20 °C, a relative humidity of 70% and a photoperiod of 14/10 h (day/night); (iii) three weeks after treatment, the sensitivity/resistance of plants to herbicides was tested following European and Mediterranean Plant Protection Organization (EPPO) standards 68

DNA extraction and metagenomic amplicon production
Total DNA was extracted from thawed soil samples using the NucleoSpin Soil kit (Macherey-Nagel, Düren, Germany) according to the manufacturer specifications.DNA was then quantified on a Qubit fluorometer (ThermoFisher Scientific, Waltham, MA).Metagenomic amplicons of bacterial and fungal communities were obtained following PCR amplification, using primers linked to Illumina adapters.To produce bacterial amplicons, the V3-V4 hypervariable region of the prokariotic 16S rRNA gene was targeted, using the primers designed by Takahashi et al. 71 (Pro341_Forward: 5′-TCG TCG GCA GCG TCA GAT GTG TAT AAG AGA CAG-CCT ACG GGNBGCASCAG-3′ and Pro805_Reverse: 5′-GTC TCG TGG GCT CGG AGA TGT GTA TAA GAG ACAG-GAC TAC NVGGG TAT CTA ATC C-3′).To obtain fungal amplicons the ribosomal ITS1 region was targeted, by using

Figure 1 .
Figure 1.(a) map of the land use in the Lombardy region (SIARL 2019; DUSAF 7.0, 2023).The rice cropping area, in the provinces of Pavia and Milano is bordered in red.(b) distribution of the rice farms (FR) where E. crus-galli and E. oryzicola survived herbicide were detected.The maps were generated through Q-GIS software, version 3.32 Lima (http:// qgis.osgeo.org).

Figure 2 .
Figure 2. Heatmap and bootstrap-based hierarchical clustering based on "Bray-Curtis" distance and "ward.D2" algorithm.The incidence of resistant Echinochloa spp.(HeR-green), the prevalence of bacterial and archaeal (blue) and of fungal (red) orders were considered.FR.: rice farm.N.I.: taxa not identified.Other: all orders with a prevalence value < 5%.au: approximately unbiased.bp: bootstrap probability.

Figure 3 .
Figure 3. Principal Coordinates Analysis (PCoA) of the 32 soil samples collected.Values on brackets represent the percent variation explained by coordinate 1 and coordinate 2 respectively.For microbial communities, the taxonomic rank of order was considered.Blue arrows: bacterial and archaeal orders.Red arrows: fungal orders.Red ellipse: L-HeR soils group.Yellow ellipse: M-HeR soils group.Blue ellipse: H-HeR soils group.Ellipses assumed a multivariate normal distribution.Confidence level: 0.99.

Figure 4 .
Figure 4. Redundancy analysis (RDA) ordination diagram of the first two axes for the incidence of herbicide resistance (HeR).Values on brackets represent the percent variation explained by axis 1 and axis 2 respectively.The constrained sets of bacterial, archaeal and fungal orders analyzed are indicated as vectors.Lactobacillales, Rhizobiales, Solibacterales, Malasseziales and Diaporthales resulted in a significant relation with HeR incidence in the RDA (anova.cca,P < 0.01**).
As concerns fungal communities, samples FR.2, FR.6, FR.20, FR.23, FR.28, FR.30 and FR.31 were characterized by the highest values of the Margalef richness index, while the soil communities sampled in farms FR.7, FR.8, FR.17 and FR.19 were the poorest.Shannon index showed in general low values except in FR.14 and FR.30 soil communities.Simpson index showed low values in samples FR.10, FR.11, FR.12, FR.14, FR.28, FR.30 and FR.32, communities characterized by high values of Pielou evenness index.The highest value of Simpson index and the lowest value of Pielou index were registered in FR.8 soil community, which is characterized by the dominance of the Malasseziales order (with 92%).
) Scientific Reports | (2024) 14:10544 | https://doi.org/10.1038/s41598-024-59856-0www.nature.com/scientificreports/values were low in all samples except for FR.11 and FR.19, that represent poor communities characterized by the dominance of one order (Actinomycetales with 32% in FR.11; Actinomycetales and Rickettsiales both with 35% in FR.19).The Pielou index, which ranges from 0 to 1, showed high values except in FR.11 and FR.19, which were characterized by communities with dominance of a single or few orders.The high evenness recorded in all samples is an indicator that communities draw on resources equally.

Table 2 .
Margalef richness, Shannon diversity, Simpson dominance and Pielou evenness indexes at the taxonomic rank of orders for Bacteria plus Archaea and for Fungi.ID: identification code. IDBacteria
. Others have highlighted

Table 3 .
List of herbicides applied and Echinochloa spp.survived chemical control in each surveyed paddy.Farm ID: identification code of each soil sample.MoA: herbicide classification in reference to mechanism of action.HRAC group: herbicide classification in reference to HRAC mode of action classification 2022 map.