Ecology

Insects and plantsLarvae of fall armyworm (FAW, Spodoptera frugiperda) were cultured at the Department of Entomology at the Max Planck Institute for Chemical Ecology, and reared on a semi-artificial diet based on pinto bean59, and maintained under controlled light and temperature conditions (12:12 h light/dark, 21 °C).Feeding experiments3rd–4th instar FAW larvae were utilized for all experiments. Insects were starved overnight prior to feeding experiments. The following day insects were fed with a semi-artificial, pinto bean-based diet or put on maize leaves in small plastic cups and allowed to feed on the respective diets for a day. Insects were dissected in cold phosphate buffered saline (PBS, pH = 7.4) to harvest larval tissues (guts, Malphigian tubules, fat bodies, cuticle), which were stored at − 80 °C until further use. For droplet feeding, 12.5 mM DIMBOA was prepared by dissolving the compound in DMSO. This DIMBOA solution was further diluted in 10% aqueous sucrose solution. The larvae were stimulated with forceps to encourage regurgitation, and 2 μL DIMBOA-sucrose solution was administered directly to the larval mouthparts. Insects were then fed on semi-artificial diet for up to 6 h; following which gut tissue was dissected using cold phosphate buffer and the tissue samples were stored at − 80 °C until further use.Insect cell culturesSpodoptera frugiperda Sf9 cells and Trichoplusia ni Hi5 cells were cultured in Sf-900 II serum-free medium (Gibco) and ExpressFive serum-free medium (Gibco), respectively. Adherent cultures were maintained at 27 °C, and sub-cultured every 3–4 days.Cell treatmentsInsect cells were seeded in 6 well culture plates (Corning) and left at 27 °C overnight. For transcript stability tests, a fresh cycloheximide (CHX) stock (50 mg/mL) was prepared in ethanol and added to the cultured cells at a concentration of 50 µg/mL. Incubations with CHX were performed up to 6 h. For testing substrate specificity, cells were then treated with the following compounds for 1 h—DIMBOA (25–100 μM), indole (50–100 μM), quercetin (50–100 μM), and esculetin (50–100 μM). All the stocks were prepared in DMSO and cells treated with the corresponding volume of pure DMSO served as a control. The range of concentrations used for the substrates was based on previous work38.RNA extraction, reverse transcription and real time-PCR analysisTissue samples from the larvae were homogenized and total RNA extracted using the innuPREP RNA Mini Kit (Analytik Jena). Cell cultures used for RNA extraction were obtained during sub-culturing at full confluency, and centrifuged at 500×g for 5 min. The culture medium was discarded, and the fresh pellets were directly used for RNA extraction. RNA concentrations were measured with the NanoDrop 2000 UV–Vis Spectrophotometer (Thermo Scientific). First strand cDNA was synthesized from 1 μg total RNA using SuperScript III Reverse Transcriptase and OligodT primers from Invitrogen. Sequences were successfully amplified using Phusion High Fidelity DNA Polymerase (New England Biolabs) (PCR protocol: 30 s at 98 °C; 35 cycles of 10 s at 98 °C, 20 s at 55 °C, 45 s at 72 °C; and 5 min at 72 °C). The PCR products were purified with a PCR cleanup kit (Qiagen) and cloned into pCR-Blunt II-TOPO vector (Life Technologies) and transformed into NEB cells (Life Technologies), which were plated on selective LB agar medium containing 100 μg/mL ampicillin and incubated overnight at 37 °C. Positive colonies were identified by PCR using vector-specific M13 primers. Positive clones were confirmed by sequencing. Real time PCR analyses were carried out using Brilliant III SYBR Master Mix, employing SYBR Green chemistry. Relative quantification of the transcript levels was done using the 2−∆∆Ct method60. SfRPL10 was used as reference gene for all analyses. The primer pairs used for distinguishing between the variants are listed in Supplementary Table 1. As the expression of full-length and variants of SfUGT33F28 differed according to the strains, tissues, and treatments being analyzed, variant expression is reported as ratios relative to the canonical transcript to facilitate comparisons.Preparation of minigenes for alternative splicing studiesGenomic DNA was isolated from S. frugiperda larvae using the cetyl trimethyl ammonium bromide (CTAB) protocol61. DNA concentration was measured with the NanoDrop 2000 UV–Vis Spectrophotometer (Thermo Scientific). The minigene was amplified using Phusion High Fidelity DNA Polymerase (New England Biolabs) (PCR protocol: 30 s at 98 °C; 35 cycles of 10 s at 98 °C, 30 s at 55–60 °C, 1 min 30 s at 72 °C; and 10 min at 72 °C), cloned into a pCR-Blunt II-TOPO vector (Life Technologies) and sequenced using M13 primers. The confirmed sequence was eventually cloned into a pIB/V5-His-TOPOvector (Life Technologies) and transformed into NEB cells (Life Technologies). Positive colonies were identified by colony PCR using vector-specific OpIE2 primers, sub-cultured overnight at 37 °C in liquid LB medium containing 100 μg/mL ampicillin and used for plasmid DNA purification with the NucleoSpin Plasmid kit (Macherey-Nagel). Concentration and purity of the obtained construct was assessed by the NanoDrop 2000 UV–Vis Spectrophotometer (Thermo Scientific) and the correct orientation of the PCR products was confirmed by DNA sequencing.Nuclear protein isolationNuclear proteins were isolated from insect cells62 using the protocol originally described with few modifications. Cells grown to concentrations of up to 1 × 106 cells/well were harvested and washed with PBS (pH 7.4). The extracts were centrifuged at 12,000×g for 10 min and pellets were re-suspended in 400 μL cell lysis buffer (10 mM HEPES, pH 7.5, 10 mM KCl, 0.1 mM EDTA pH 8.0, 1 mM DTT, 0.5% Nonidet-40 and 10 μL protease inhibitor cocktail). Cells were allowed to swell on ice for 20 min with intermittent mixing. Suspensions were vortexed to disrupt the cell membranes and then centrifuged at 12,000×g for 10 min at 4 °C. Pelleted nuclei were washed thrice with cell lysis buffer, re-suspended in 50 μL nuclear extraction buffer (20 mM HEPES pH 7.5, 400 mM KCl, 1 mM EDTA pH 8.0, 1 mM DTT, 10% glycerol and protease inhibitor) and incubated on ice for 30 min. Nuclear fractions were collected by centrifugation at 12,000g for 15 min at 4 °C. Protein concentrations were measured by Bradford and extracts were stored at − 80 °C until further use.Electrophoretic mobility shift assay (EMSA)EMSA was performed using the LightShift Chemiluminescent EMSA kit (Thermo Scientific) following the manufacturer’s instructions. Genomic DNA fragments of 20–25 bp corresponding to the 5′ flanking region of UGT33F28 exon 1 (with and without AhR-ARNT motif deletion) were synthesized with covalently linked biotin (Sigma). The DNA probes used in the experiment are listed in Supplementary Table 6. EMSA was performed in 20 µL reactions containing 20 fmol biotinylated DNA probe with 3.5–4 µg nuclear protein extracted from insect cells, according to manufacturer’s instructions. A reaction comprising the above along with the excess of unlabeled canonical DNA probe (200 molar excess) was further employed as a control. The reaction was assembled at room temperature and incubated for 30 min. The reactions were separated on a 5% TBE gel in 0.5X TBE at 100 V for 60 min. The samples were then transferred to a positively charged nylon membrane (Hybond N+, Amersham Bioscience) using semi-dry transfer at 15 V for 30 min. The membrane was cross-linked for 1 min using the auto cross-link function on the UV cross-linker (Stratagene). The biotinylated DNA–protein complex was detected by the streptavidin–horseradish peroxidase conjugated antibody provided in the kit. The membrane was washed and incubated with the chemiluminescence substrate for 5 min and the signals were developed by exposing the membrane to an X-ray film for 1 min.Streptavidin affinity purificationStreptavidin agarose (Sigma-Aldrich) was employed for protein purification. Briefly, 50–100 μL of agarose was packed into a 1.5 mL Eppendorf tube for each sample. The agarose was allowed to settle with a short centrifugation (500×g, 5 min) and the supernatant was discarded. The agarose was washed 4–5 times with binding buffer (PBS containing 1 mM EDTA, 1 mM DTT, 4 µg poly dI. dC as non-specific competitor DNA and protease inhibitor). Simultaneously, the binding reaction with the nuclear protein fraction and the DNA probe was assembled as described above. A 100 μg amount of total nuclear protein was incubated with 4 μg of biotinylated DNA probe at room temperature for 20 min. The reaction was loaded onto the streptavidin column equilibrated with the binding buffer and incubated for another 1 h at room temperature with gentle shaking. Subsequently, the agarose was washed 4–5 times with the binding buffer. After the final wash, the supernatant was aspirated and 10 μL was left above the beads. For protein separation, 20–30 μL pf the SDS loading buffer was added onto the agarose, boiled at 95 °C for 5 min and the sample thus obtained was utilized for electrophoresis.Deletion mutagenesisFor deletion mutagenesis, a pair of primers flanking the sequence to be deleted (non-overlapping) was designed. The pCR-Blunt II-TOPO vector (Life Technologies) clone for the SfUGT33F28 exon 1–2 minigene was utilized as a template. Sequence was successfully amplified using Phusion High Fidelity DNA Polymerase (New England Biolabs) (PCR protocol: 30 s at 98 °C; 20 cycles of 10 s at 98 °C, 30 s at 55–60 °C, 4 min at 72 °C; and 10 min at 72 °C). A DpnI digest was performed to remove the background DNA, followed by ligation and transformation into fresh cells. The sequence of the mutant TOPO clone was then confirmed and utilized as a template for cloning into pIB/V5-His-TOPO vector (Life Technologies) for transfection into insect cells.Cloning and heterologous expression of SfUGTsSequences were amplified from S. frugiperda gut cDNA samples using Phusion High Fidelity DNA Polymerase (New England Biolabs) (PCR protocol: 30 s at 98 °C; 35 cycles of 10 s at 98 °C, 20 s at 55–60 °C, 45 s at 72 °C; and 5 min at 72 °C). The resulting amplified products were purified with a PCR cleanup kit (Qiagen) and incubated with GoTaq DNA polymerase (Promega) for 15 min at 72 °C in order to add A overhangs. The products were cloned into the pIB/V5-His-TOPO vector (Life Technologies) and transformed into NEB cells (Life Technologies), which were plated on selective LB agar medium containing 100 μg/mL ampicillin and incubated overnight at 37 °C. Positive colonies were identified by PCR using vector-specific OpIE2 primers, sub-cultured overnight at 37 °C in liquid LB medium containing 100 μg/mL ampicillin and used for plasmid DNA purification with the NucleoSpin Plasmid kit (Macherey-Nagel). Concentration and purity of the obtained constructs were assessed by NanoDrop 2000 UV–Vis Spectrophotometer (Thermo Scientific) and the correct orientation of the PCR products was confirmed by DNA sequencing.Insect cell transfectionFor transfection, Sf9 cells and Hi5 cells were sub-cultured at full confluency in a 6-well plate in a 1:3 dilution and left overnight to adhere to the flask surface. The medium was replaced, and transfections were carried out using FuGENE HD Transfection Reagent (Promega) in a 1:3 plasmid/lipid ratio (1.7 μg plasmid and 5.0 μL lipid for 3 mL medium). Cells were incubated for 48–72 h at 27 °C and re-suspended in fresh medium containing 50 μg/mL blasticidin for 2 weeks. Stable cell cultures were subsequently maintained at 10 μg/mL blasticidin.Cell lysate preparationCells were obtained from cultures 2 weeks post transfection growing stably on 50 μg/mL blasticidin. A 1 mL quantity of cells was harvested for each construct and re-suspended into 100 µL buffer. Protein concentrations were measured using the Bradford reagent, and 1–2 μg of the cell lysate was used for enzyme assays.Microsome preparationFor microsome extraction, confluent, stably transfected cells from five T-75 flasks (10 mL culture) per recombinant plasmid were harvested by scraping the cells off the bottom using a sterile cell scraper (Sarstedt AG, Nuembrecht, Germany). The obtained cell suspensions were combined into a 50 mL falcon tube and centrifuged at 1000×g for 15 min at 4 °C (AvantiTM J-20 XP Centrifuge, Beckman Coulter, Krefeld, Germany). The supernatant was discarded, the cells were washed twice with ice-cold PBS buffer (pH 7.4) and centrifuged at 1000×g for 15 min. The resulting cell pellet was re-suspended in 10 mL hypotonic buffer (20 mM Tris, 5 mM EDTA, 1 mM DTT, 20% glycerol, pH 7.5), containing 0.1% BenzonaseR nuclease and 100 μL Protease Inhibitor Cocktail (Serva) followed by incubation on ice for 30 min. After cell lysis, the cells were homogenized by 20–30 strokes in a Potter–Elvehjem tissue grinder (Kontes Glass Co., Vineland, USA) and were subsequently mixed with an equal volume of sucrose buffer (20 mM Tris, 5 mM EDTA, 1 mM DTT, 500 mM sucrose, 20% glycerol, pH 7.5). The homogenate was centrifuged at 1200×g and 4 °C for 10 min (AvantiTM J-20 XP Centrifuge, Beckman Coulter), and the supernatant was transferred into Beckman polycarbonate ultracentrifugation bottles (25 × 89 mm) (Beckman Coulter) and centrifuged at 100,000×g and 4 °C for 1.5 h in a fixed angle Type 70 Ti rotor (OptimaTM L-90K Ultracentrifuge, Beckman Coulter). After ultracentrifugation, the clear supernatant, containing the cytosolic fraction, was aliquoted into 1.5 mL Eppendorf tubes. The pellet, containing the microsomal fractions, was re-suspended in 1 mL of phosphate buffer (100 mM K2HPO4, pH 7.0), containing 10 μL Protease Inhibitor Cocktail (Serva) and stored at − 80 °C until further use. Typically, 5–10 μg of the microsome fraction so obtained was utilized for the enzyme assays.Cross-linking assaysCross-linking assays were performed using dimethyl suberimidate (DMS) as the cross-linking agent. A fresh stock of DMS (5 mg/mL) was prepared in 0.2 M triethanolamine (pH 8.0) at the start of each assay. DMS was added to a final concentration of 2.5 mg/mL to insect cell microsomes with gentle shaking up to 3 h, and samples were subsequently stored at − 20 °C until further use. All protein samples were electrophoresed using a 12% Mini-PROTEAN tris glycine gel, blotted onto PVDF membrane using wet transfer at 70 V for 30–45 min, followed by detection using the V5-HRP conjugate.V5-based affinity purificationAnti-V5 agarose affinity gel (Sigma-Aldrich) was employed for protein purification. Briefly, 50–75 μL of the agarose was packed into a 1.5 mL Eppendorf tube for each sample. The agarose was allowed to settle with a short centrifugation and the supernatant was discarded. The agarose was washed 4–5 times with PBS (pH 7.4). Samples to be purified were incubated with 5% digitonin on ice for 20 min and subject to centrifugation at 16,000×g for 30 min. Clarified cell lysate or microsomal extract was added onto the resin (up to 200 μL, volume adjusted by addition of PBS) and incubated for 1.5 h on a shaker. Subsequently, the agarose was washed 4–5 times with PBS. After the final wash, the supernatant was aspirated and 10 μL was left above the beads. This fraction was used for both protein electrophoresis and enzyme assays (separate purifications). For SDS-PAGE, 20–30 μL pf the SDS loading buffer was added onto the agarose, boiled at 95 °C for 5 min and sample thus obtained was utilized for electrophoresis.LC–MS/MS peptide analysisProtein bands of Coomassie Brilliant blue R250 stained gels were cut from the gel matrix and tryptic digestion was carried out63. For LC–MS/MS analysis of the resulting peptides, samples were reconstituted in 20 μL aqueous 1% formic acid, and 1 μL was injected onto an UPLC M-class system (Waters, Manchester, UK) coupled to a Synapt G2-si mass spectrometer (Waters, Manchester, UK). Samples were first pre-concentrated and desalted using a Symmetry C18 trap column (100 Å, 180 µm × 20 mm, 5 µm particle size) at a flow rate of 15 µL/min (0.1% aqueous formic acid). Peptides were eluted onto a ACQUITY UPLC HSS T3 analytical column (100 Å, 75 µm × 200 mm, 1.8 µm particle size) at a flow rate of 350 nL/min with the following gradient: 3–15% over 3 min, 15–20% B over 7 min, 20–40% B over 30 min, 40–50% B over 5 min, 50–70% B over 5 min, 70–95% B over 3 min, isocratic at 95% B for 1 min, and a return to 1% B over 1 min. Phases A and B were composed of 0.1% formic acid and 100% acetonitrile in 0.1% formic acid, respectively). The analytical column was re-equilibrated for 10 min prior to the next injection. The eluted peptides were transferred into the mass spectrometer operated in V-mode with a resolving power of at least 20,000 full width at half height FWHM. All analyses were performed in a positive ESI mode. A 100 fmol/μL sample of human Glu-Fibrinopeptide B in 0.1% formic acid/acetonitrile (1:1 v/v) was infused at a flow rate of 1 μL/min through the reference sprayer every 45 s to compensate for mass shifts in MS and MS/MS fragmentation mode. Data were acquired using data-dependent acquisition (DDA). The acquisition cycle for DDA analysis consisted of a survey scan covering the range of m/z 400–1800 Da followed by MS/MS fragmentation of the ten most intense precursor ions collected at 0.5 s intervals in the range of 50–2000 m/z. Dynamic exclusion was applied to minimize multiple fragmentations for the same precursor ions. MS data were collected using MassLynx v4.1 software (Waters, Manchester, UK).Data processing and protein identificationDDA raw data were processed and searched against a sub-database containing common contaminants (human keratins and trypsin) using ProteinLynx Global Server (PLGS) version 2.5.2 (Waters, Manchester, UK). Spectra remaining unmatched by database searching were interpreted de novo to yield peptide sequences and subjected to homology-based searching using the MS BLAST program64 installed on a local server. MS BLAST searches were performed against a Spodoptera frugiperda database obtained by in silico translation of the S. frugiperda transcriptome37 and against arthropoda database (NCBI). PKL-files of MS/MS spectra were generated and searched against Spodoptera frugiperda database combined with NCBI nr (downloaded on May 24, 2020) using MASCOT software version 2.6.2. The following searching parameters were applied: fixed precursor ion mass tolerance of 15 ppm for the survey peptide, fragment ion mass tolerance of 0.1 Da, 1 missed cleavage, fixed carbamidomethylation of cysteines and possible oxidation of methionine.Enzymatic assaysFor UGT assays, samples from insect cell cultures (transient or stable) were prepared in phosphate buffer (pH 7.0, 100 mM). Typical enzyme reactions included 5–10 µg cell microsomal extracts, 2 μL of 12.5 mM DIMBOA in DMSO (25 nmol), 4 μL of 12.5 mM UDP-glucose in water (50 nmol), and phosphate buffer (pH 7.0, 100 mM) to give an assay volume of 50 μL. Controls containing either boiled enzymatic preparation, or only the protein suspension and buffer were included. After incubation at 30 °C for 60 min, the enzyme reactions were interrupted by adding 50 μL of 1:1 (v:v) methanol/formic acid solution. For enzyme assays involving resin purified microsomal extracts, equal amounts of extracts were employed for resin purification and the enzyme assay (buffer + substrate) was pipetted directly onto the resin. Post incubation, samples were centrifuged, supernatant was collected, and reaction was stopped by addition of methanol/formic acid solution. Assays were centrifuged at 5000g for 5 min and the obtained supernatant was collected and analyzed by LC–MS/MS.Chromatographic methodsFor all analytical chromatography procedures, formic acid (0.05%) in water and acetonitrile were used as mobile phases A and B, respectively, and the column temperature was maintained at 25 °C. Analyses of enzymatic assays and plant samples used a Zorbax Eclipse XDB-C18 column (50 × 4.6 mm, 1.8 μm, Agilent Technologies) with a flow rate of 1.1 mL/min and with the following elution profile: 0–0.5 min, 95% A; 0.5–6 min, 95–67.5% A; 6.02–7 min, 100% B; 7.1–9.5 min, 95% A. LC–MS/MS analyses were performed on an Agilent 1200 HPLC system (Agilent Technologies) coupled to an API 6500 tandem spectrometer (AB Sciex) equipped with a turbospray ion source operating in negative ionization mode. Multiple reaction monitoring (MRM) was used to monitor analyte parent ion to product ion conversion with parameters from the literature for DIMBOA65 and DIMBOA-Glc16. Analyst (version 1.6.3, Applied Biosystems) software was used for data acquisition and processing.Statistical analysisAll statistical analyses were carried out using SigmaPlot 12.0 and R studio (version 3.6.3). Data were tested for homogeneity of variance and normality and were appropriately transformed to meet these criteria where required. The specific statistical method used for each data set is described in the figure legends. More

Wink, M. Evolution of secondary metabolites from an ecological and molecular phylogenetic perspective. Phytochemistry 64, 3–19 (2003).CAS 
PubMed 
Article 

Google Scholar 
Khare, S. et al. Plant secondary metabolites synthesis and their regulations under biotic and abiotic constraints. J. Plant Biol. 63, 203–216 (2020).CAS 
Article 

Google Scholar 
Gajger, I. T. & Dar, S. A. Plant allelochemicals as sources of insecticides. Insects 12, 189 (2021).Article 

Google Scholar 
Vig, A. P., Rampal, G., Thind, T. S. & Arora, S. Bio-protective effects of glucosinolates: A review. LWT Food Sci. Technol. 42, 1561–1572 (2009).CAS 
Article 

Google Scholar 
Sikorska-Zimny, K. & Beneduce, L. The glucosinolates and their bioactive derivatives in Brassica: A review on classification, biosynthesis and content in plant tissues, fate during and after processing, effect on the human organism and interaction with the gut microbiota. Crit. Rev. Food Sci. Nutr. 61, 2544–2571. https://doi.org/10.1080/10408398.2020.1780193 (2020).CAS 
Article 
PubMed 

Google Scholar 
Radojčić Redovniković, I., Glivetić, T., Delonga, K. & Vorkapić-Furač, J. Glucosinolates and their potential role in plant. Period. Biol. 110, 297–309 (2008).
Google Scholar 
Wittstock, U., Kliebenstein, D. J., Lambrix, V., Reichelt, M. & Gershenzon, J. Chapter five glucosinolate hydrolysis and its impact on generalist and specialist insect herbivores. Recent Adv. Phytochem. 37, 101–125 (2003).CAS 
Article 

Google Scholar 
Noret, N. et al. Palatability of Thlaspi caerulescens for snails: Influence of zinc and glucosinolates. New Phytol. 165, 763–772 (2005).CAS 
PubMed 
Article 

Google Scholar 
Hopkins, R. J., Van Dam, N. M. & Van Loon, J. J. A. Role of glucosinolates in Insect-plant relationships and multitrophic interactions. Annu. Rev. Entomol. 54, 57–83 (2008).Article 
CAS 

Google Scholar 
Guleria, S. & Tiku, A. K. Botanicals in pest management: Current status and future perspectives. Integr. Pest Manag. 1, 317–329 (2009).
Google Scholar 
Clay, N. K., Adio, A. M., Denoux, C., Jander, G. & Ausubel, F. M. Glucosinolate metabolites required for an Arabidopsis innate immune response. Science 323, 95–101 (2009).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Yang, B. et al. Inhibitory effect of allyl and benzyl isothiocyanates on ochratoxin a producing fungi in grape and maize. Food Microbiol. 100, 103865 (2021).CAS 
PubMed 
Article 

Google Scholar 
Agrawal, A. A. & Kurashige, N. S. A role for isothiocyanates in plant resistance against the specialist herbivore Pieris rapae. J. Chem. Ecol. 296(29), 1403–1415 (2003).Article 

Google Scholar 
Müller, C. et al. The role of the glucosinolate-myrosinase system in mediating greater resistance of Barbarea verna than B. vulgaris to Mamestra brassicae Larvae. J. Chem. Ecol. 44, 1190–1205 (2018).PubMed 
Article 
CAS 

Google Scholar 
Kegley, S. E., Hill, B. R., Orme, S. & Choi, A. H. PAN Pesticide Database (Pesticide Action Network, 2000).
Google Scholar 
Worfel, R. C., Schneider, K. S. & Yang, T. C. S. Suppressive effect of allyl isothiocyanate on populations of stored grain insect pests. J. Food Process. Preserv. 21, 9–19 (1997).CAS 
Article 

Google Scholar 
Wu, H., Zhang, G. A., Zeng, S. & Lin, K. C. Extraction of allyl isothiocyanate from horseradish (Armoracia rusticana) and its fumigant insecticidal activity on four stored-product pests of paddy. Pest Manag. Sci. 65, 1003–1008 (2009).CAS 
PubMed 
Article 

Google Scholar 
Bhushan, S., Gupta, S., Kaur Sohal, S., Arora, S. & Saroj Arora, C. Assessment of insecticidal action of 3-Isothiocyanato-1-propene on the growth and development of Spodoptera litura (Fab.) (Lepidoptera: Noctuidae). J. Entomol. Zool. Stud. 4, 1068–1073 (2016).
Google Scholar 
Dhillon, M. K., Naresh, J. S., Singh, R. & Sharma, N. K. Reaction of different bitter gourd (Momordica charantia L.) genotypes to melon fruit fly, Bactrocera cucurbitae (Coquillett). Int. J. Plant Prot. 33, 55–59 (2005).
Google Scholar 
Ekesi, S., Nderitu, P. W. & Chang, C. L. Adaptation to and small-scale rearing of invasive fruit fly Bactrocera invadens (Diptera: Tephritidae) on artificial diet. Ann. Entomol. Soc. Am. 100, 562–567 (2007).Article 

Google Scholar 
Jakhar, S. et al. Estimation losses due to fruit fly, Bactrocera cucurbitae (Coquillett) on long melon in semi-arid region of Rajasthan. J. Entomol. Zool. Stud. 8, 632–635 (2020).MathSciNet 

Google Scholar 
Ladania, M. S. Physiological Disorders and Their Management. Citrus Fruit: Biology, Technology and Evaluation 451–463 (Academic press, 2008).
Google Scholar 
Du, Y., Grodowitz, M. J. & Chen, J. Insecticidal and enzyme inhibitory activities of isothiocyanates against red imported fire ants, Solenopsis invicta. Biomolecules 10, 716 (2020).CAS 
PubMed Central 
Article 

Google Scholar 
Tsao, R., Reuber, M., Johnson, L. & Coats, J. R. Insecticidal toxicities of glucosinolate· containing extracts from crambe seeds. J. Agric. Urban Entomol. 13, 109–120 (1996).CAS 

Google Scholar 
Li, Q., Eigenbrode, S. D., Stringam, G. R. & Thiagarajah, M. R. Feeding and growth of Plutella xylostella and Spodoptera eridania on Brassica juncea with varying glucosinolate concentrations and myrosinase activities. J. Chem. Ecol. 26, 2401–2419 (2000).CAS 
Article 

Google Scholar 
Noble, R. R., Harvey, S. G. & Sams, C. E. Toxicity of Indian mustard and allyl isothiocyanate to masked chafer beetle larvae. Plant Health Prog. 3, 9 (2002).Article 

Google Scholar 
Sousa, A. H., Faroni, L. R. A., Pimentel, M. A. G. & Freitas, R. S. Relative toxicity of mustard essential oil to insect-pests of stored products. Rev. Caatinga 27, 222–226 (2014).
Google Scholar 
de Souza, L. P., Faroni, L. R. D. A., Lopes, L. M., de Sousa, A. H. & Prates, L. H. F. Toxicity and sublethal effects of allyl isothiocyanate to Sitophilus zeamais on population development and walking behavior. J. Pest Sci. 91, 761–770 (2018).Article 

Google Scholar 
Freitas, R. C. P., Faroni, L. R. D. A., Haddi, K., Jumbo, L. O. V. & Oliveira, E. E. Allyl isothiocyanate actions on populations of Sitophilus zeamais resistant to phosphine: Toxicity, emergence inhibition and repellency. J. Stored Prod. Res. 69, 257–264 (2016).Article 

Google Scholar 
Jabeen, A., Zaitoon, A., Lim, L. T. & Scott-Dupree, C. Toxicity of five plant volatiles to adult and egg stages of Drosophila suzukii matsumura (Diptera: Drosophilidae), the spotted-wing Drosophila. J. Agric. Food Chem. 69, 9511–9519 (2021).CAS 
PubMed 
Article 

Google Scholar 
Wu, H., Liu, X. R., Yu, D. D., Zhang, X. & Feng, J. T. Effect of allyl isothiocyanate on ultra-structure and the activities of four enzymes in adult Sitophilus zeamais. Pestic. Biochem. Physiol. 109, 12–17 (2014).CAS 
PubMed 
Article 

Google Scholar 
Zhang, C., Wu, H., Zhao, Y., Ma, Z. & Zhang, X. Comparative studies on mitochondrial electron transport chain complexes of Sitophilus zeamais treated with allyl isothiocyanate and calcium phosphide. Pestic. Biochem. Physiol. 126, 70–75 (2016).CAS 
PubMed 
Article 

Google Scholar 
Jeschke, V. et al. How glucosinolates affect generalist lepidopteran larvae: Growth, development and glucosinolate metabolism. Front Plant Sci. 8, 1995. https://doi.org/10.3389/fpls.2017.01995 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Agnihotri, A. R., Hulagabali, C. V., Adhav, A. S. & Joshi, R. S. Mechanistic insight in potential dual role of sinigrin against Helicoverpa armigera. Phytochemistry 145, 121–127. https://doi.org/10.1016/j.phytochem.2017.10.014 (2018).CAS 
Article 
PubMed 

Google Scholar 
Jeschke, V. et al. So much for glucosinolates: A generalist does survive and develop on Brassicas, but at what cost?. Plants 10, 962. https://doi.org/10.3390/plants10050962 (2021).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Benrey, B. & Denno, R. F. The slow-growth-high-mortality hypothesis: A test using the cabbage butterfly. Ecology 78, 987–999 (1997).
Google Scholar 
Shroff, R., Vergara, F., Muck, A., Svatoš, A. & Gershenzon, J. Nonuniform distribution of glucosinolates in Arabidopsis thaliana leaves has important consequences for plant defense. Proc. Natl. Acad. Sci 05, 6196–6201 (2008).ADS 
Article 
CAS 

Google Scholar 
Bai, P. P. et al. Inhibition of phenoloxidase activity delays development in Bactrocera dorsalis (Diptera: Tephritidae). Fla. Entomol. 97, 477–485. https://doi.org/10.1653/024.097.0218 (2014).Article 

Google Scholar 
Datta, R., Kaur, A., Saraf, I., Singh, I. P. & Kaur, S. Effect of ethyl acetate extract and purified compounds of Alpinia galanga (L.) on Immune Response of a Polyphagous Lepidopteran pest, Spodoptera litura (Fabricius). Asian J. Adv. Basic Sci. 6, 16–21 (2018).CAS 

Google Scholar 
Hartzer, K. L., Zhu, K. Y. & Baker, J. E. Phenoloxidase in larvae of Plodia interpunctella (Lepidoptera: Pyralidae): Molecular cloning of the proenzyme cDNA and enzyme activity in larvae paralyzed and parasitized by Habrobracon hebetor (Hymenoptera: Braconidae). Arch. Insect Biochem. Physiol. 59, 67–79 (2005).CAS 
PubMed 
Article 

Google Scholar 
Silva, C. J. M. et al. Immune response triggered by the ingestion of polyethylene microplastics in the dipteran larvae Chironomus riparius. J. Hazard. Mater. 414, 125401. https://doi.org/10.1016/j.jhazmat.2021.125401 (2021).CAS 
Article 
PubMed 

Google Scholar 
Aucoin, R. R., Philogène, B. J. R. & Arnason, J. T. Antioxidant enzymes as biochemical defenses against phototoxin induced oxidative stress in three species of herbivorous Lepidoptera. Arch. Insect Biochem. Physiol. 16, 139–152 (1991).CAS 
Article 

Google Scholar 
Wang, Y., Branicky, R., Noë, A. & Hekimi, S. Superoxide dismutases: Dual roles in controlling ROS damage and regulating ROS signaling. Int. J. Cell Biol. 217, 1915–1928. https://doi.org/10.1083/jcb.201708007 (2018).CAS 
Article 

Google Scholar 
Zhang, C., Ma, Z., Zhang, X. & Wu, H. Transcriptomic alterations in Sitophilus zeamais in response to allyl isothiocyanate fumigation. Pest. Biochem. Physiol. 137, 62–70. https://doi.org/10.1016/j.pestbp.2016.10.001 (2017).CAS 
Article 

Google Scholar 
Felton, G. W. & Summers, C. B. Antioxidant systems in insects. Arch. Insect Biochem. Physiol. 29, 187–197. https://doi.org/10.1002/arch.940290208 (1995).CAS 
Article 
PubMed 

Google Scholar 
Cadenas, E. Mechanisms of oxygen activation and reactive oxygen species detoxification. In Oxidative Stress and Antioxidant Defenses in Biology (ed. Ahmad, S.) 1–46 (Chapman & Hall, 1995). https://doi.org/10.1007/978-1-4615-9689-9_1.Chapter 

Google Scholar 
Schramm, K., Vassão, D. G., Reichelt, M., Gershenzon, J. & Wittstock, U. Metabolism of glucosinolate- derived isothiocyanates to glutathione conjugates in generalist lepidopteran herbivores. Insect Biochem. Mol. Biol. 42, 174–182. https://doi.org/10.1016/j.ibmb.2011.12.002 (2012).CAS 
Article 
PubMed 

Google Scholar 
Falk, K. L. et al. The role of glucosinolates and the jasmonic acid pathway in resistance of Arabidopsis thaliana against molluscan herbivores. Mol. Ecol. 23, 1188–1203. https://doi.org/10.1111/mec.12610 (2014).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Gloss, A. D. et al. Evolution in an ancient detoxification pathway is coupled with a transition to herbivory in the Drosophilidae. Mol. Biol. Evol. 31, 2441–3245. https://doi.org/10.1093/molbev/msu201 (2014).CAS 
Article 
PubMed 
PubMed Central 

Google Scholar 
Bhatt, P., Zhou, X., Huang, Y., Zhang, W. & Chen, S. Characterization of the role of esterases in the biodegradation of organophosphate, carbamate, and pyrethroid pesticides. J. Hazard. Mater. 1, 125026. https://doi.org/10.1016/j.jhazmat.2020.125026 (2021).CAS 
Article 

Google Scholar 
Murfadunnisa, S. et al. Larvicidal and enzyme inhibition of essential oil from Spheranthus amaranthroids (Burm.) against lepidopteran pest Spodoptera litura (Fab.) and their impact on non-target earthworms. Biocatal. Agric. Biotechnol. 21, 101324. https://doi.org/10.1016/j.bcab.2019.101324 (2019).Article 

Google Scholar 
Sengottayan, S. N. Physiological and biochemical effect of neem and other Meliaceae plants secondary metabolites against Lepidopteran insects. Front. Physiol. 4, 359. https://doi.org/10.3389/fphys.2013.00359 (2013).Article 

Google Scholar 
Augustyniak, M., Gladysz, M. & Dziewięcka, M. The Comet assay in insects: Status, prospects and benefits for science. Mutat. Res. Rev. Mutat. Res. 767, 67–76. https://doi.org/10.1016/j.mrrev.2015.09.001Get (2016).CAS 
Article 
PubMed 

Google Scholar 
Foster, E. R. & Downs, J. A. Histone H2A phosphorylation in DNA double strand break repair. FEBS J. 272, 3231–3240. https://doi.org/10.1111/j.1742-4658.2005.04741.x (2005).CAS 
Article 
PubMed 

Google Scholar 
Porichha, S. K., Sarangi, P. K. & Prasad, R. Genotoxic effect of chlorpyrifosin Channa punctatus. Cytol. Genet. 9, 631–638 (1998).
Google Scholar 
Kalita, M. K., Haloi, K. & Devi, D. Larval exposure to chlorpyrifos affects nutritional physiology and induces genotoxicity in silkworm Philosamia ricini (Lepidoptera: Saturniidae). Front. physiol. 7, 1–14. https://doi.org/10.3389/fphys.2016.00535 (2016).Article 

Google Scholar 
Datta, R. et al. Assessment of genotoxic and biochemical effects of purified compounds of Alpinia galanga on a polyphagous lepidopteran pest Spodoptera litura (Fabricius). Phytoparasitica 48, 501–511. https://doi.org/10.1007/s12600-020-00813-8 (2020).CAS 
Article 

Google Scholar 
Afify, A. & Negm, A. A. K. H. Genotoxic effect of insect growth regulators on different stages of peach fruit fly, Bactrocera zonata (Saunders)(Diptera: Tephritidae). Afr. Entomol. 26, 154–161 (2018).Article 

Google Scholar 
Gupta, J. N., Verma, A. N. & Kashyap, R. K. An improved method for mass rearing for melon fruit fly Dacus cucurbitae Coquillett. Indian J. Entomol. 40, 470–471 (1978).
Google Scholar 
Srivastava, B. G. A chemically defined diet for Dacus cucurbitae (Coq.) larvae under aseptic conditions. Entomol. News Lett. 5, 24 (1975).
Google Scholar 
Kumar, A., Sood, S., Mehta, V., Nadda, G. & Shanker, A. Biology of Thysanoplusia orichalcea (Fab.) in relation to host preference and suitability for insect culture and bioefficacy. Indian J. Appl. Entomol. 18, 16–21 (2004).
Google Scholar 
Martinez, S. S. & Emden, H. F. V. Growth disruption, abnormalities and mortality of Spodoptera littoralis (Boisduval) (Lepidoptera: Noctuidae) caused by azadirachtin. Neotrop. Entomol. 30, 113–125 (2001).CAS 
Article 

Google Scholar 
Khan, Z. R. & Saxena, R. C. Behavioural and physiological responses of Sogatella furcifera (Homoptera: Delphacidae) to selected resistant and susceptible rice cultivars. J. Econ. Entomol. 78, 1280–1286 (1985).Article 

Google Scholar 
Zimmer, M. Phenol oxidation. In Methods to Study Litter Decomposition (eds Graça, M. A. et al.) (Springer, 2005).
Google Scholar 
Kono, Y. Generation of superoxide radical during auto-oxidation of hydroxylamine and an assay for superoxide dismutase. Arch. Biochem. Biophys. 186, 189–195. https://doi.org/10.1016/0003-9861(78)90479-4 (1978).CAS 
Article 
PubMed 

Google Scholar 
Bergmeyer, H. U. Reagents for enzymatic analysis. In Methods of Enzymatic Analysis (eds Bergmeyer, H. U. & Gawehn, K.) 438 (Verlag Chemie, 1974).
Google Scholar 
Chien, C. & Dauterman, W. C. Studies on glutathione S-transferases in Helicoverpa (=Heliothis) zea. Insect Biochem. 21, 857–864. https://doi.org/10.1016/0020-1790(91)90092-S (1991).CAS 
Article 

Google Scholar 
Katzenellenbogen, B. & Kafatos, F. C. General esterases of silk worm moth moulting fluid: Preliminary characterization. J. Insect Physiol. 17, 1139–1151. https://doi.org/10.1016/0022-1910(71)90016-3 (1971).CAS 
Article 

Google Scholar 
Mac Intyre, R. J. A method for measuring activities of acid phosphatases separated by acrylamide gel electrophoresis. Biochem. Genet. 5, 45–56 (1971).CAS 
Article 

Google Scholar 
Singh, N. P., McCoy, M. T., Tice, R. R. & Schneider, E. L. A simple technique for quantitation of low levels of DNA damage in individual cells. Exp. Cell Res. 175, 184–191 (1988).CAS 
PubMed 
Article 

Google Scholar  More

Infection properties of clade A and clade B T7-like cyanophagesWe set out to test the hypothesis that the phylogenetic separation of T7-like cyanophages into two major clades reflects differences in their infection physiology. To do this we investigated a suite of infection properties of three pairs of clade A and B phages, each pair infecting the same Synechococcus host (Table 1) to allow us to control for variability in host genetics and physiology. These six cyanophages are representatives of 3 clade A and 2 clade B cyanophage subclades (SI Appendix, Table S1).Table 1 Summary of infection physiology of three pairs of clade A and clade B cyanophages infecting the same Synechococcus hosts.Full size tableWe began by investigating adsorption kinetics and the length of time taken to produce new phages in the infection cycle, the latent period, from phage growth curve experiments. In all three pairs of phages, adsorption was 7–15-fold more rapid in the clade A phage versus the clade B phage (Fig. 1, Table 1). Furthermore, the clade A phage had a faster infection cycle with a latent period that was 3-5-fold shorter than the clade B phage on the same host (Fig. 1a–c) (Table 1). To determine how representative these findings are for a greater diversity of T7-like cyanophages we report the latent period of nine additional non-paired phages that infect a variety of hosts and span the diversity of this cyanophage genus, measured here and taken from the literature (SI Appendix, Table S1). These phages showed the same pattern as observed between phage pairs, although one clade A phage had a relatively long latent period (see SI Appendix, Table S1). Overall, the 5 clade A phages representative of 5 subclades had a significantly shorter latent period (3.3 ± 3.6 h, n = 5 phages (mean ± SD) than the 10 clade B phages from 7 subclades (7.7 ± 2.0 h, n = 10 phages) (Kruskal-Wallis: χ2 = 4.72, df = 1; p = 0.029, n = 15). No significant differences in the length of the latent period were found for clade B phages that infected Synechococcus and Prochlorococcus (Kruskal-Wallis: χ2 = 1.13, df = 1; p = 0.29, n = 10).Fig. 1: Comparison of the infection physiology between pairs of clade A and clade B T7-like cyanophage infecting the same Synechococcus host.a–c Cyanophage growth curves, d–f burst sizes, g–i virulence as the percentage of lysed host cells, j–l decay as loss of infectivity, m–o plaque sizes. a, d, g, j, m Clade A Syn5 phage and clade B S-TIP37 phage infecting WH8109. b, e, h, k, n Clade A S-CBP42 phage and clade B S-RIP2 phage infecting WH7803. c, f, i, l, o Clade A S-TIP28 phage and clade B S-TIP67 phage infecting CC9605. The host strain is shown at the right of the panels. Red and blue lines or bars show results for clade A and clade B phages, respectively. a–c, g–I Error bars indicate standard deviations. d–f Burst size results are for single cells. j–l The solid line shows the fitted multi-level linear model. m–o The time after infection at which plaques were photographed appears above the images. *p value  More

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
and JavaScript. More

Six biological replicates each were sampled from four fruit species (blueberries, cherries, raspberries, and strawberries) at four developmental stages. Developmental stages were based on fruit pigmentation ranging from unripe (green) to fully ripe (red/purple/navy; Fig. S1) throughout June to September in 2018. Ten fruits (except blueberries N = 20) were collected for each species per replicate, and this was replicated six times for each ripening stage for each fruit at different sites.Quantitative analysis of fungal communitiesMetabarcoding analysis is generally not quantitative, but the addition of 265 P. cucumerina cells to sub-samples prior to DNA extraction served as an internal standard to attempt an estimation of the size of fungal populations. One replicate spiked with the internal standard of the strawberry stage 3 samples was removed due to poor sequence quality leaving 96 non-spiked and 95 spiked samples which produced a total of 38,445,395 reads that clustered into 1712  > 97% identity Amplicon Sequence Variants (ASV), which from here-in we call phylotypes (Table S1). Blast searches across all phylotypes for matches to the P. cucumerina internal standard’s ITS sequence generated from Sanger sequencing revealed one phylotype that matched with 100% identity. Plectosphaerella cucumerina was naturally present in 21 of the 95 non-spiked samples and comprised of a total of 444 reads. Cherry was the only fruit where the internal standard was reliably recovered: 23 of 24 spiked samples and only one of 24 non-spiked samples contained the internal standard phylotype. After internal standard DNA read normalisation, the mean (± SE) number of fungal cells from each of the useable 23 pairs of cherry replicates was 307,323 (± 39,090) cells. The range of phylotype cell abundance across all cherry samples was 3.9 million for an Aureobasidium phylotype to 3 cells for a phylotype taxonomically assigned no higher level than kingdom. There was no significant change in total fungal cell numbers across cherry maturation stage (Kruskal–Wallis, chi-squared = 2.63, P = 0.45; Fig. S2), but fruit surface areas also increased significantly (Kruskal–Wallis, chi-squared = 19.70, P = 0.0002, Fig. S2). When cell numbers were normalised for surface area this revealed that absolute fungal population sizes remained static across cherry maturation stages (Kruskal–Wallis, chi-squared = 2.49, P = 0.48; Fig. 1A). However, there was a significant change in absolute Saccharomycetales cell numbers when normalised for cherry surface area across maturation (Kruskal–Wallis, chi-squared = 15.30, P = 0.002): stage 1 had significantly greater absolute Saccharomycetales cell numbers than stage 4 (P = 0.0007; Fig. 1B). Six individual Saccharomycetales yeast phylotypes from the genera Debaryomyces, Saccharomyces, Kodamaea, one from the family Pichiaceae, and phylotypes with  > 97% homology to M. pulcherrima and Metschnikowia gruessii, had significantly greater abundances on ripening stage 1 than 4 (P values span 0.045 to 0.006).Figure 1Absolute fungal cell abundances on cherry epicarp. Number of total fungal (A) and Saccharomycetales yeasts (B) cells per mm2 of cherry epicarp (N = 6 except, stage 3 and 4, N = 5) at four ripening stages (1, unripe/green fruit; 2, de-greening fruit; 3, ripening fruit; and 4, fully ripe/harvest fruit) estimated from DNA read abundances normalised to DNA abundances from the deliberate addition of 265 live Plectosphaerella cucumerina cells prior to DNA extraction. Different lower-case letters above bars show significant differences between ripening stages at P  > 0.05, Dunn’s comparisons post-hoc with Benjamini–Hochberg multiple comparison correction.Full size imageOverview of fungal diversity across all fruit samplesThe P. cucumerina internal standard phylotype was removed from all samples, and the sequence data normalised and analysed. A total of 1712 fungal phylotypes was revealed, comprising seven phyla, 25 classes, 96 orders, 197 families, and 280 genera. The most abundant and diverse phylum was Ascomycota, comprising 92.2% of reads and 57.3% of phylotypes, followed by Basidiomycota (7.7% reads and 33.6% phylotypes), Zygomycota (0.1% and 1.1%), Chytridiomycota ( > 0.1% and 0.7%), Mucoromycota ( > 0.1% and 0.3%), Glomeromycota and Rozellomycota (both  > 0.1% and 0.1%; Fig. S3A). A phylotype from the Cladosporium genus was the most common phylotype across all samples, comprising 60.8% of reads. A total of 87 phylotypes from the order Saccharomycetales (budding yeasts) was detected, comprising 1,792,782 DNA reads (4.7% of the total) spanning 10 families and 25 genera. Metschnikowia was the most abundant Saccharomycetales genus (40.0% of Saccharomycetales reads), followed by Hanseniaspora (38.2%), then Pichia (5.2%), with the remaining genera contributing fewer than 3% each. Candida was the most diverse genus within the order Saccharomycetales accounting for 21.8% of phylotypes, despite only comprising 2.4% of reads, followed by Metschnikowia (11.5%), Hanseniaspora (8.0%) and Pichia (6.9%), with each of the remaining genera contributing fewer than 3.5% of phylotypes each (Fig. S3B). The most common Saccharomycetales yeast across all samples was a phylotype from the genus Hanseniaspora with  > 97% homology to H. uvarum and comprised 38.2% of the total Saccharomycetales reads (Fig. S3B).The effect of fruit species and ripening stage on epicarp fungal communitiesWe analysed differences in three biodiversity metrics to evaluate the effect of fruit species and maturation stage on fungal communities: differences in the absolute numbers of phylotypes (richness); differences in the types of phylotypes (i.e. presences/absences); and differences in the relative abundances of phylotypes (community composition) following Morrison-Whittle et al.14 and Morrison‐Whittle and Goddard37.
Fungal phylotype richnessPhylotype richness was not normally distributed (Shapiro-Wilks, P = 0.008) but square root transformation allowed the data to conform to the assumptions of ANOVA. There was a significant effect of both fruit type and ripening stage on the number of fungal phylotypes, including an interaction between the two (F3,175 = 18.58, P = 1.65 × 10–10; F3,175 = 5.00, P = 0.002 and F9,175 = 6.69, P = 3.25 × 10–8 respectively). Comparisons of effect sizes revealed fruit type (ω2 = 0.30) had a 4.4 times greater effect than ripening stage (ω2 = 0.068) on fungal phylotype richness. Disregarding ripening stage, cherry (mean ± SE number of phylotypes = 98 ± 4.1) had significantly more fungal phylotypes than blueberry (68 ± 3.7), raspberry (72 ± 2.9) and strawberry (76 ± 3.2) (Tukey’s HSD, P  0.05) and there was a significant effect of ripening stage on the number of fungal phylotypes for cherry, raspberry, and strawberry (one-way ANOVA: F3,44 = 4.33, P = 0.0093; F3,44 = 13.56, P = 2.11 × 10–6 and F3,44 = 13.86, P = 1.84 × 10–6, respectively, Fig. 2), but not blueberry (F3,44 = 2.27, P = 0.055). On cherries phylotype numbers increased during ripening, but raspberry and strawberry had greater numbers at intermediate stages of fruit maturation (Fig. 2).Figure 2Number of observed phylotypes across fruit types and maturation stages. Number of fungal phylotypes across four ripening stages (1, unripe/green fruit; 2, de-greening fruit; 3, ripening fruit; and 4, fully ripe/harvest fruit) for blueberry, cherry, raspberry and strawberry (N = 12 except N = 11 for strawberry stage 3). Numbers of fungal phylotypes differ across ripening stages for cherry, raspberry and strawberry but not blueberry (ANOVA, P values shown). Where significant, different lowercase letters represent significant differences in phylotype numbers within each fruit (P  97% homology to Metschnikowia kunwiensis and H. uvarum on raspberry; and phylotypes with  > 97% homology to Kalmanozyma fusiformata (Ustilaginaceae smut fungi) and Podosphaera aphanis on strawberry.Twenty-four of the 195 indicator phylotypes belonged to the Saccharomycetales budding yeasts (Table S13). There were no Saccharomycetales indicator phylotypes for cherry, and just one for blueberry, a fungal phylotype with  > 97% homology to Metschnikowia koreensis. Raspberry had 15 Saccharomycetales indicator phylotypes: three with  > 97% homology to the Metschnikowia and, Candida genera, two Pichia and Schwanniomyces, and one each from Hanseniaspora, Barnettozyma, Debaryomyces, Candida, Geotrichum and Martiniozyma. There were eight indicator phylotypes for strawberry; two Candida and one from each of the Metschnikowia, Starmerella, Kodamaea and Hyphopichia genera and the Pichiaceae family, and a phylotype assigned to the no higher level than fungal kingdom (with  > 97% homology to deposit from Candida genus). The dynamics of Saccharomycetales yeast indicator phylotypes abundances across maturation for raspberry and strawberry is shown in Fig. 6.Figure 6Dynamics of changes in the proportion of budding yeast indicator phylotypes. Mean proportion of reads for the Saccharomycetales budding yeast indicator phylotypes that are significantly overrepresented on (A) raspberry and (B) strawberry (P  97% homology identified by manual Blast searches.Full size imageDifferences of yeast known to be attractive to D. suzukii
Yeast from the Hanseniaspora, Pichia, Saccharomyces, Candida and Metschnikowia genera and their combinations are attractive to D. suzukii27,28,30,31, and phylotypes belonging to these genera were recovered here. The combined relative read abundances of all phylotypes assigned to these genera were significantly different between fruit types and ripening stages (Kruskal–Wallis chi-squared = 60.54, P = 4.51 × 10–13; chi-squared = 10.11, P = 0.018, respectively). Raspberry had the highest relative abundance of yeast genera known to be attractive to D. suzukii (mean ± SE = 21,539 ± 4339) and this was significantly greater than on the other fruits (P  97% homology to H. uvarum as over-represented on raspberry generally, and especially at later stages (Fig. 6A).Differences of Botrytis cinerea, known to be repulsive to D. suzukii
The relative read abundances of B. cinerea were significantly different between fruit types and ripening stages (Kruskal–Wallis chi-squared = 73.45, P = 7.80 × 10–16; Kruskal–Wallis chi-squared = 23.81, P = 2.74 × 10–5, respectively). Raspberry had the lowest relative abundance of B. cinerea (mean ± SE = 800 ± 136) and this was significantly lower than strawberry (1994 ± 292) and blueberry (5990 ± 1305) (P  More

The conceptual flow chart of the process is provided in Fig. 1. We used seven reanalysis SM data (Table 2) masked with soil temperature (ST) and soil freeze/thaw status to calculate water table depth, i.e. the input of TOPMODEL, given the obvious disagreements between the input datasets. The diagnostic algorithms based on TOPMODEL were used following Stocker et al. (ref. 20) and Xi et al. (ref. 25), where the optimized parameters were calibrated with long-term maximum wetland areas from four observation-based wetland datasets (Table 1). Details about these datasets and computational processing are shown as follows.Fig. 1Diagram of workflow for parameter calibration and the simulation of global wetland dynamics.Full size imageTable 2 Key characteristics of seven global soil moisture reanalysis data used in this study.Full size tableReanalysis soil moisture datasetsSeven long-term reanalysis SM datasets used in this study include NCEP-DOE (National Centers for Environmental Prediction-the Department of Energy)26, MERRA-Land (the Modern-Era Retrospective Analysis for Research and Applications)27, MERRA-2 (ref. 28), GLDAS-Noah v2.0 (the Global Land Data Assimilation System)29, GLDAS-Noah v2.1 (ref. 29), ERA5 (European Environment Agency)30,31, and ERA5-Land30,31. Key characteristics of the seven SM data are listed in Table 2. The datasets differ by their spatial and temporal resolutions, the time-period they cover, as well as the definition of the soil layers. More details are provided for each dataset below.

NCEP-DOE
NCEP-DOE is an updated version of the National Centers for Environmental Prediction/National Center for Atmospheric Research (NCEP/NCAR) Reanalysis 1 project, which uses a state-of-the-art analysis/forecast system to perform data assimilation with past data from 1948 to the present32. NCEP-DOE features the newer physics and observed SM forcing and also eliminates several previous errors, such as oceanic albedo and snowmelt term during the entire period, and snow cover analysis error from 1974 to 1994 (ref. 26). With a spatial resolution of about 210 km, there are two vertical soil layers in NCEP-DOE for both SM and ST: 0–0.1 and 0.1–2 m.

MERRA-Land and MERRA-2
MERRA-Land soil moisture is generated by driving the Goddard Earth Observing System model version 5.7.2 (GEOS-5.7.2) with meteorological forcing from the MERRA reanalysis product27. The precipitation forcing in MERRA-Land merges MERRA precipitation with a gauge-based data product from the National Oceanic and Atmospheric Administration (NOAA) Climate Prediction Center, and the Catchment land surface model used in MERRA-Land is updated to the “Fortuna-2.5” version27. MERRA-2 intends to replace the original MERRA reanalysis and ingests important new data types28. The Catchment model in MERRA-2 has been updated with rainfall interception and snow model parameters of MERRA-Land, and the precipitation correction is a refined version of MERRA-Land. For MERRA-Land and MERRA-2, there is only one layer for SM from the surface to the bedrock, with “depth-to-rock” depending on local conditions. ST is computed on six vertical soil layers: 0–0.10, 0.10–0.29, 0.29–0.68, 0.68–1.44, 1.44–2.95, and 2.95–12.95 m.

ERA5 and ERA5-Land
ERA5 is the fifth generation ECMWF (European Centre for Medium-Range Weather Forecasts) reanalysis of global climate and weather, replacing ERA-Interim30,31. Based on a decade of developments in model dynamics and data assimilation, there is a significantly enhanced horizontal resolution (31 km), temporal resolution (hourly) and uncertainty estimation. ERA5 covers 1979–2020 and continues to be updated in near-real-time. ERA5-Land is produced with a finer horizontal resolution of 9 km by running the land component of the ERA5 climate reanalysis but without data assimilation. By March of 2021, the ERA5-Land outputs are only available since 1981. SM and ST are computed on four vertical soil layers (0–0.07, 0.07–0.28, 0.28–1, and 1–2.89 m) for both ERA5 and ERA5-Land.

GLDAS-Noah v2.0 and GLDAS-Noah v2.1

GLDAS is a global, moderate-resolution (0.25° × 0.25°) offline terrestrial modeling system developed by NASA Goddard Space Flight Center (GSFC) and the NOAA National Centers for Environmental Prediction29, thus similar to ERA5. To produce optimal fields of land surface variables in near-real-time, it incorporates satellite- and ground-based observations. GLDAS-Noah drives the Noah land surface model and has two components: one forced with the Princeton meteorological forcing data (i.e. GLDAS-Noah v2.0) and the other forced with a combination of model and observation (i.e. GLDAS-Noah v2.1). GLDAS-Noah v2.0 covers the period 1948–2014, while GLDAS-Noah v2.1 is available from 2000 to the present. There are four vertical layers in the Noah land surface model for both ST and SM: 0–0.1, 0.1–0.4, 0.4–1, and 1–2 m.Observation-based wetland/flooded area dataIn terms of large uncertainties in current wetland datasets (Table 1) we selected four widely used and available satellite/satellite-based wetland/flooded area data including GIEMS-2 (ref. 14), RFW (the Regularly Flooded Wetland map)10, WAD2M (a global dataset of Wetland Area and Dynamics for Methane Modeling)33, and G2017 (the pantropical wetland extent from an expert system model)9 for parameter calibration. Among them, GIEMS-2 and WAD2M include monthly wetland dynamics, while RFW and G2017 are static. The comparison of the four wetland datasets is shown in Supplementary Fig. 1; details on each data are provided below.

GIEMS-2
The GIEMS-1 is the first global estimate of monthly inundated areas, derived from passive microwave land surface emissivity34. With a 0.25° × 0.25° resolution, GIEMS-1 documents a mean annual maximum inundated area of 9.5 Mkm2 for 1993–2007 (including open water, wetlands, and rice paddies, but excluding large lakes), which shows good agreement with existing independent, static inventories as well as regional high-resolution synthetic aperture radar observations34. Based on similar retrieval principles with GIEMS-1, GIEMS-2 is developed to less depend on ancillary data with an updated microwave emissivity, and correct a known overestimation over low vegetated areas from GIEMS-1 (ref. 14). The period is extended to 1992–2015 for GIEMS-2 and can be updated with the availability of observations. Globally, the mean annual maximum and long-term maximum inundated extent after removing the rice paddies using the Monthly Irrigated and Rainfed Crop Areas dataset (MIRCA2000)35 for the period 1992–2015 are 6.7 and 10.9 million km2 (hereafter Mkm2; sum of mean annual maximum or long-term maximum inundated extent for each grid cell) respectively. The rice paddies are removed here as they are not natural wetlands and cannot be simulated with TOPMODEL.

RFW
RFW is a static, high-resolution map (15 arc-sec) of regularly flooded wetlands, developed by overlapping flooded areas (permanent wetlands and flooded vegetation classes) for 2008–2012 from the ESA-CCI land cover map36, mean annual maximum inundated areas (including wetlands, rivers, small lakes, and irrigated rice) for 1993–2004 from GIEMS-D15 global inundation extent (downscaled using GIEMS-1)37, and long-term maximum surface water areas for 1984–2015 from JRC global surface water bodies product13. The large permanent lakes and reservoirs are distinguished using the HydroLAKES database38. Globally, RFW covers 9.7% of the land surface area (~13.0 Mkm2) including wetlands, river channels, deltas, and flooded lake margins, but excluding large lakes10. Due to the mean annual maximum or long-term maximum inundation/surface water extent for 1984–2016 from the three wetland data is used, we treated RFW as the long-term maximum wetland extent in this study. Besides, given that GIEMS-D15 includes artificial rice paddies, we removed them with MIRCA2000 from RFW (~11.9 Mkm2 after removing rice paddies).

WAD2M
WAD2M dataset used in this study is an improved version of the SWAMPS v3.2 from Jensen et al. (ref. 15), covering the years 2000 to 2018. With a spatial resolution of 25 km × 25 km, this data was used as input wetland area data of phase 2 of the Global Methane Budget33. Given that the initial SWAMPS failed to detect wetlands lacking surface inundation and to differentiate between lakes, wetlands, and other surface water bodies, Zhang et al. (ref. 33) modified it using a series of independent static wetland distribution data7,9,39,40,41 in an attempt to include missing wetlands under dense canopies. Besides, they removed inland waters (lakes, rivers, and ponds) and rice agriculture with JRC and MIRCA2000, respectively. Globally, the mean annual maximum and long-term maximum wetland extent for the period 2000–2018 estimated by WAD2M are 8.1 Mkm2 and 13.2 Mkm2 (sum of mean annual maximum or long-term maximum inundated extent for each grid cell) respectively.

G2017

G2017 (ref. 9) is a static, pantropical wetland and peatland extent map (covering 60°S–40°N) at 232 m × 232 m resolution, derived from a hybrid expert model system. With three biophysical indices related to wetland and peat formation (long-term water supply exceeding atmospheric water demand, annually or seasonally waterlogged soils, and geomorphological position where water is supplied and retained), G2017 identifies not only permanently and seasonally wetland areas, but also soil wetness and topographic conditions that favor waterlogging in the absence of flooding for the end of the 20th century. Given the broad coverage of different types of wetlands, we also treated this map as long-term maximum wetland areas. This ‘pantropical’ data (60°S to 40°N) offers the advantage to include non-flooded wetland areas that are missed in satellite-based wetland products. However, note that not all detected wetlands or peatlands in G2017 have been observed. Rice agriculture was also removed with MIRCA2000 from G2017. The resulting wetland and peatland area for 60°S–40°N is 4.0 Mkm2.The TOPMODEL-based diagnostic modelTOPMODEL as improved by Stocker et al. (ref. 20) and Xi et al. (ref. 25) was used to calculate the inundated fraction from WTD at grid-scale in this study. Based on the assumptions that the local hydraulic gradient is approximated by the local topographic slope and the water table variations can be assimilated to a succession of steady states with uniform recharge, the classical TOPMODEL establishes an analytical relationship between the soil moisture deficit and the distributions of local topographic index within a catchment. At grid-scale, the analytical relationship can be represented as:$$CT{I}_{i}-overline{CT{I}_{x}}=mathrm{-M}left({{Gamma }}_{i}-overline{{{Gamma }}_{x}}right)$$
(1)
where CTI indicates the topographic index, defined as the log of the ratio of contributing area to the local slope. We used the CTI data at 500 m × 500 m resolution from Marthews et al. (ref. 22), where lakes, reservoirs, mountain glaciers, and ice caps are removed using the Global Lakes and Wetlands Database7. The (overline{CT{I}_{x}}) indicates the average of CTIi of all sub-grids (index i) within the grid cell x. M indicates a tunable parameter that describes the exponential decrease of soil transmissivity with depth21. Γi is the water table of the pixel i and (overline{{{Gamma }}_{{x}}}) is the mean water table of the grid x. When Γi is at the soil surface (i.e. Γi = 0), the threshold (CT{I}_{x}^{* }) above which all pixels are flooded for the grid x is derived:$$CT{I}_{x}^{* }=overline{CT{I}_{x}}+{rm{M}}cdot overline{{{Gamma }}_{x}}$$
(2)
The wetlands area is defined as the flooded areas (i.e. Γ ≤ 0), the flooded fraction in the grid x (fx) being the percentage of pixels with CTIi larger than a threshold (CT{I}_{x}^{* }):$${f}_{x}=frac{1}{{A}_{x}}{sum }_{i}{A}_{i}^{* }$$with$${A}_{i}^{* }=left{begin{array}{c}{A}_{i},if,CT{I}_{i}ge CT{I}_{x}^{* }\ 0,if,CT{I}_{i} < CT{I}_{x}^{* }end{array}right.$$ (3) To reduce the computational costs from the high-resolution CTI data for predicting long time series of wetland area, we used the asymmetric sigmoid function from Stocker et al. (ref. 20) to fit the “empirical” relationship (widehat{{Psi }}) between (widehat{f}) and Γ:$${{rm{psi }}}_{x}left({{Gamma }}_{x}right)={left(1+{v}_{x}cdot {e}^{-{k}_{x}left({{Gamma }}_{x}-{q}_{x}right)}right)}^{-1/{v}_{x}}$$ (4) where vx, kx, qx are three parameters of the function. Given a value of parameter M, the three parameters can be derived with a sequence of Γx spanning a plausible range of values (−1 m to 2 m) and corresponding fx from the initial TOPMODEL approach (Eq. (3)). Thus, the wetlands in our study are defined as the flooded area simulated by TOPMODEL. As for the range of parameter M, Stocker et al. (ref. 20) used a global uniform value for M (M = 8) after testing simulated wetland fraction for a range of M (7, 8, 9). Nevertheless, given that distinct topography, soil types, and other intrinsic characteristics in different regions, we considered M as a tunable, spatially heterogeneous, and grid-specific parameter, with a range of 1–15 following Xi et al. (ref. 25). Thus, for each grid cell x there are 15 choices for M, and then 15 sets of (vx, kx, qx). The optimized parameter combination of (vx, kx, qx) is determined by selecting minimum root-mean-square-error (RMSE) between simulated inundated fractions and observations:$$RMSE=sqrt{frac{{sum }_{i=1}^{n}{left({O}_{i}-{P}_{i}right)}^{2}}{n}}$$ (5) where Oi and Pi are observed and simulated wetland fraction, respectively. n represents the time-series length for wetland extent. For simulations calibrated with RFW and G2017, the RMSE was computed with the long-term maximum (hereafter called MAX) monthly wetland area because the two data sets are static and only record the MAX wetland extent. While for simulations calibrated with GIEMS-2 and WAD2M which include temporal variations of wetland area, we calibrated the parameters with all months, mean seasonal cycle, yearly maximum, and MAX wetland area, but only showed the optimal simulations calibrated with MAX wetland area in this work to keep consistency with RFW and G2017. Besides, to provide more choices for users, we combined all of the four wetland datasets (i.e. the union of long-term maximum wetland extent) to generate a new wetland map (hereafter called MAX_all), and then used the MAX_all to calibrate the parameters to produce seven sets of global wetland extent products with seven soil moisture datasets. The simulations calibrated with yearly maximum wetland area from GIEMS-2 and WAD2M and long-term maximum wetland area from MAX_all are also provided in our resulting products.Finally, to avoid unrealistically high wetland fraction output from the function, the simulated maximum wetland fraction fx is constrained by the observed MAX wetland area with a parameter ({f}_{x}^{max}) (Eq. (6)), which is different from Stocker et al. (ref. 20). The determination of ({f}_{x}^{max}) is analyzed in the supplemental material in detail (Supplementary Text 1). Once the value of (vx, kx, qx) are determined, the wetland fraction fx can be directly derived from the monthly water table Γx according to Eqs. (4) and (6).$${f}_{x}=minleft({{Psi }}_{x}left({{Gamma }}_{x}right),{f}_{x}^{max}right)$$ (6) Calculation of water table depthWater table depth is not computed by land surface models, given their coarse soil vertical discretization. We thus used the saturation deficit of soil moisture (θSD) as a surrogate of water table depth, θSD being defined as an index consisting of saturated volumetric water content and the “actual” soil depth modified by soil freeze/thaw status:$${theta }_{SD}={z}_{{l}_{0}}-{sum }_{l=1}^{{l}_{0}}{theta }_{l}cdot frac{Delta {z}_{l}}{{theta }_{S}}$$ (7) Subscript l represents the lth soil layer, l0 is the number of layers above the first frozen soil layer counted from the top (l = 1 at the soil surface), θl is the monthly volumetric water content in the lth soil layer (m3 m−3), (Delta {z}_{l}) is the thickness of the lth soil layer (m), θS is the saturated volumetric water content (in m3 m−3 units, uniform over depth).As formulated in Eq. (7), ({z}_{{l}_{0}}) is the thickness of all soil layers (or depth to bedrock) when there is no frozen soil layer. If there exists at least one frozen layer, ({z}_{{l}_{0}}) is set to the depth of the uppermost frozen soil layer. We excluded the frozen soil layers here given that some important wetland processes such as methane production and transport are insignificant when the soils are frozen. In high latitudes, the presence of frozen soil layers may lead to an overestimation of the wetland fraction due to relatively large θSD values even if there is little liquid soil water above the uppermost frozen soil layer. Hence, we used monthly soil temperature (ST) at 70 cm, the Global Record of Daily Landscape Freeze/Thaw Status data42, and the Köppen climate classification system43 to refine the frozen mask. When the monthly mean ST at 70 cm is below 0 °C, or soil freezing days are more than 5 in a month, or the grid is classified as the Hot desert (BWh) in the Köppen climate classification system, the wetland fraction for the grid is set to zero. However, it should be noted that the algorithm using the ST at 70 cm could omit some unfrozen soil layers above 70 cm, which could lead to bias in estimation of methane emissions from these unfrozen layers. We provided the global wetland maps in our resulting products, but the potential uncertainties in wetland estimation due to the omitted unfrozen layers should be considered, particularly at high latitudes. We used seven reanalysis SM products to compute θSD to provide the uncertainty in SM input (Table 2). All data are re-interpolated to 0.25° × 0.25° resolution.Evaluation against wetland calibration data and independent satellite productsAlthough we calibrated parameters of the TOPMODEL-based diagnostic model with the observation-based wetland data, to what extent the simulations can reproduce the spatial patterns and temporal dynamics of the calibration wetland data must be evaluated. For spatial patterns, we calculated the RMSE of wetland area between our simulations and corresponding wetland calibration data following Eq. (5), and evaluated the spatial patterns of simulated wetland extent in two wetland hotspots including Amazon basin and Western Siberia lowlands with three independent global/regional water products. For Amazon basin, we used the global surface water dataset from JRC13 (optical satellite images) and the wetland map produced using mosaics of Japanese Earth Resources Satellite (JERS-1) L-band SAR imagery from Hess et al. (ref. 44, hereafter H2015). For West Siberian lowlands, we used JRC and the Boreal–Arctic Wetland and Lake Dataset (BAWLD, only covers the north of ~55°N) produced using an expert assessment and extrapolated using random forest modelling from climate, topography, soils, permafrost conditions, vegetation, wetlands, and surface water extents and dynamics45. For temporal dynamics, since we only used the static wetland area (long-term maximum) from all of the four observation-based wetland products to calibrate parameters, the simulated temporal dynamics can be evaluated with the two dynamic wetland products (GIEMS-2 and WAD2M). Besides, we also used the terrestrial water storage (TWS) from the Gravity Recovery and Climate Experiment (GRACE), which retrieves relative change in TWS from the monthly anomalies of the Earth’s gravity field for 2003–2016 measured by the twin GRACE satellites46,47 to evaluate the simulated temporal dynamics. More

Liu, H. et al. Optimal nitrogen input for higher efficiency and lower environmental impacts of winter wheat production in China. Agr. Ecosyst. Environ. 224, 1–11 (2016).Article 

Google Scholar 
Guang-hao, L., Gui-gen, C., Wei-ping, L. & Da-lei, L. Differences of yield and nitrogen use efficiency under different applications of slow-release fertilizer in spring maize. J. Integr. Agric. 20(2), 554–564 (2021).Article 

Google Scholar 
Kumar, V. V. Role of Rhizospheric Microbes in Soil 377–398 (Springer, 2018).Book 

Google Scholar 
Ullah, A. et al. Factors affecting the adoption of organic farming in Peshawar-Pakistan. Agric. Sci. 6(06), 587–593 (2015).
Google Scholar 
Cui, Z. et al. Pursuing sustainable productivity with millions of smallholder farmers. Nature 555(7696), 363–366 (2018).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Zhang, X. et al. Managing nitrogen for sustainable development. Nature 528(7580), 51–59 (2015).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Alzaidi, A. A., Baig, M. B. & Elhag, E. A. An investigation into the farmers ’ attitudes towards organic farming in Riyadh Region–Kingdom of Saudi Arabia. Bulg. J. Agric. Sci. 19(3), 426–431 (2013).
Google Scholar 
Zhihui, W. et al. Combined applications of nitrogen and phosphorus fertilizers with manure increase maize yield and nutrient uptake via stimulating root growth in a long-term experiment. Pedosphere 26(1), 62–73 (2016).Article 
CAS 

Google Scholar 
Guang-hao, L., Gui-gen, C., Wei-ping, L. & Da-lei, L. Differences of yield and nitrogen use efficiency under different applications of slow release fertilizer in spring maize. J. Integr. Agric. 20(2), 554–564 (2020).
Google Scholar 
Zant, W. Is organic fertilizer going to be helpful in bringing a green revolution to sub-Saharan Africa? Economic explorations for Malawi agriculture (Working Paper). International House Hold Survey Network (2010).Barman, M., Paul, S., Choudhury, A. G., Roy, P. & Sen, J. Biofertilizer as prospective input for sustainable agriculture in India. Int. J. Curr. Microbiol. App. Sci. 6(11), 1177–1186 (2017).Article 

Google Scholar 
Kalhapure, A. H., Shete, B. T. & Dhonde, M. B. Integrated nutrient management in maize (Zea Mays L.) for increasing production with sustainability. Int. J. Agric. Food Sci. Technol. 4(3), 2249–3050 (2013).
Google Scholar 
Nazli, R. I., Kuşvuran, A., Inal, I., Demirbaş, A. & Tansi, V. Effects of different organic materials on forage yield and quality of silage maize (Zea mays L.). Turk. J. Agric. For. 38(1), 23–31 (2014).CAS 
Article 

Google Scholar 
Niu, Z. et al. Total factor productivity growth in china’s corn farming: an application of generalized productivity indicator. J. Bus. Econ. Manag. 22(5), 1189–1208 (2021).Article 

Google Scholar 
van Wesenbeeck, C. F. A., Keyzer, M. A., van Veen, W. C. M. & Qiu, H. Can China’s overuse of fertilizer be reduced without threatening food security and farm incomes?. Agric. Syst. 190, 103093 (2021).Article 

Google Scholar 
Ji, Y., Liu, H. & Shi, Y. Will China’s fertilizer use continue to decline? Evidence from LMDI analysis based on crops, regions and fertilizer types. PLoS ONE 15, e0237234 (2020).CAS 
PubMed 
PubMed Central 
Article 

Google Scholar 
Jiao, X. et al. Grain production versus resource and environmental costs: towards increasing sustainability of nutrient use in China. J. Exp. Bot. 67(17), 4935–4949 (2016).CAS 
PubMed 
Article 

Google Scholar 
Sher, A. et al. Response of maize grown under high plant density; performance, issues and management: a critical review. Adv. Crop Sci. Technol. 5(3), 1–8 (2017).Article 

Google Scholar 
De-yang, S. H. I. et al. Increased plant density and reduced N rate lead to more grain yield and higher resource utilization in summer maize. J. Integr. Agric. 15(11), 2515–2528 (2016).Article 

Google Scholar 
Du, X., Wang, Z., Lei, W. & Kong, L. Increased planting density combined with reduced nitrogen rate to achieve high yield in maize. Sci. Rep. 11(1), 1–12 (2021).CAS 
Article 

Google Scholar 
Li, T., Zhang, W., Yin, J., Chadwick, D., Norse, D., Lu, Y., Liu, X., Chen, X., Zhang, F., Powlson, D., & Dou, Z. Enhanced-efficiency fertilizers are not a panacea for resolving the nitrogen problem (2017).Adu-gyamfi, R. et al. One-time fertilizer briquettes application for maize production in savanna agroecologies of Ghana. Soil Fertil. Crop Prod. 111(6), 3339–3350 (2019).CAS 

Google Scholar 
Jiang, C. et al. Optimal nitrogen application rates of one-time root zone fertilization and the effect of reducing nitrogen application on summer maize. Sustainability 11, 2979 (2019).Article 

Google Scholar 
Jiang, C. et al. One-time root-zone N fertilization increases maize yield, NUE and reduces soil N losses in lime concretion black soil. Sci. Rep. 8(1), 1–10 (2018).ADS 

Google Scholar 
Li, G., Zhao, B., Dong, S., Liu, P. & Vyn, T. J. Impact of controlled release urea on maize yield and nitrogen use efficiency under different water conditions. PLoS ONE 12(7), 1–16 (2017).
Google Scholar 
Sikora, J. et al. Assessment of the efficiency of nitrogen slow-release fertilizers in integrated production of carrot depending on fertilization strategy. Sustainability (Switzerland) 12(5), 1–10 (2020).
Google Scholar 
Tian, C. et al. Effects of a controlled-release fertilizer on yield, nutrient uptake, and fertilizer usage efficiency in early ripening rapeseed (Brassica napus L.). J. Zhejian Univ. Sci. B (Biomed. Biotechnol.) 17(14), 775–786 (2016).CAS 
Article 

Google Scholar 
Tong, D. & Xu, R. Effects of urea and ( NH4)2SO4 on nitrification and acidification of Ultisols from Southern China. J. Environ. Sci. 24(4), 682–689 (2012).CAS 
Article 

Google Scholar 
El-rokiek, K. G., Ahmed, S. A. & Abd-elsamad, E. E. H. Effect of adding urea or ammonium sulphate on some herbicides efficiency in controlling weeds in onion plants. J. Am. Sci. 6(11), 536–543 (2010).
Google Scholar 
FAO. Guidelines for soil description. Enhanced Recovery After Surgery, (2006).Landon, J. Booker Tropical Soil manual: A Handbook for Soil Survey and Agriculture Land Evaluation in the Tropics and Subtropics (2013).Zhao, R. F. et al. Fertilization and nitrogen balance in a wheat-maize rotation system in North China. Agron. J. 98(4), 938–945 (2006).CAS 
Article 

Google Scholar 
Huang, S. et al. Estimation of nitrogen supply for summer maize production through a long-term field trial in china. Agronomy 11(7), 1358 (2021).CAS 
Article 

Google Scholar 
Dong, Y. J. et al. Effects of new coated release fertilizer on the growth of maize. J. Soil Sci. Plant Nutr. 16(3), 637–649 (2016).CAS 

Google Scholar 
Ngosong, C., Bongkisheri, V., Tanyi, C. B., Nanganoa, L. T. & Tening, A. S. Optimizing nitrogen fertilization regimes for sustainable maize (Zea mays L.) production on the volcanic soils of Buea Cameroon. Adv. Agric. 2019, 1–8 (2019).
Google Scholar 
Su, W., Ahmad, S., Ahmad, I. & Han, Q. Nitrogen fertilization affects maize grain yield through regulating nitrogen uptake, radiation and water use efficiency, photosynthesis and root distribution. PeerJ 8, 1–21 (2020).CAS 

Google Scholar 
Sainju, U. M, Ghimire, R., & Pradhan, G.P. Nitrogen Fertilization I: Impact on Crop, Soil, and Environment. IntechOpen https://doi.org/10.5772/intechopen.86028 (2020).Sha, Z. et al. Effect of N stabilizers on fertilizer-N fate in the soil-crop system: a meta- analysis. Agr. Ecosyst. Environ. 2020, 290 (2019).
Google Scholar 
Chen, K. & Vyn, T. J. Post-silking factor consequences for N efficiency changes over 38 years of commercial maize hybrids. Front. Plant Sci. https://doi.org/10.3389/fpls.2017.01737 (2017).Article 
PubMed 
PubMed Central 

Google Scholar 
Jia, X. P. et al. Farmer’s adoption of improved nitrogen management strategies in maize production in China: an experimental knowledge training. J. Integr. Agric. 12(2), 364–373 (2013).Article 

Google Scholar 
Amanullah,. Rate and timing of nitrogen application influence partial factor productivity and agronomic NUE of maize (Zea mays L.) planted at low and high densities on calcareous soil in northwest Pakistan. J. Plant Nutr. 39(5), 683–690 (2016).CAS 
Article 

Google Scholar 
Draman, A., Almas, L. K. Partial factor productivity, agronomic efficiency, and economic analyses of maize in wheat-maize cropping system in Pakistan. Southern Agricultural Economics Association Annual Meetings, 2009 (January 2009).Yan, P. et al. Interaction between plant density and nitrogen management strategy in improving maize grain yield and nitrogen use efficiency on the North China Plain. Agric. Sci. 154, 978–988 (2016).Article 

Google Scholar 
Oenema, O. Nitrogen use efficiency (NUE) an indicator for the utilization of nitrogen in food systems. EU Nitrogen Expert Panel, January 2017, 1–4 (2015).Venterea, R. T., Coulter, J. A. & Dolan, M. S. Evaluation of intensive “4R” strategies for decreasing nitrous oxide emissions and nitrogen surplus in rainfed corn. J. Environ. Qual. 45(4), 1186–1195 (2016).CAS 
PubMed 
Article 

Google Scholar 
Zhang, C., Ju, X., Powlson, D., Oenema, O. & Smith, P. Nitrogen surplus benchmarks for controlling N pollution in the main cropping systems of China. Environ. Sci. Technol. 53(12), 6678–6687 (2019).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Fernández, C., Koop, G. & Steel, M. F. J. Multiple-output production with undesirable outputs multiple-output production with undesirable outputs : an application to nitrogen surplus in agriculture. J. Am. Stat. Assoc. 97(458), 432–442 (2013).MATH 
Article 

Google Scholar 
Børsting, C. F., Kristensen, T., Misciattelli, L., Hvelplund, T. & Weisbjerg, M. R. Reducing nitrogen surplus from dairy farms. Effects of feeding and management. Livest. Prod. Sci. 83(2–3), 165–178 (2003).Article 

Google Scholar 
Liang, K. et al. Reducing nitrogen surplus and environmental losses by optimized nitrogen and water management in double rice cropping system of South China. Agric. Ecosyst. Environ. 286, 106680 (2019).CAS 
Article 

Google Scholar 
Klages, S. et al. Nitrogen surplus-a unified indicator for water pollution in Europe?. Water (Switzerland) 12(4), 1197 (2020).CAS 

Google Scholar 
Muratoglu, A. Grey water footprint of agricultural production: an assessment based on nitrogen surplus and high-resolution leaching runoff fractions in Turkey. Sci. Total Environ. 742, 140553 (2020).ADS 
CAS 
PubMed 
Article 

Google Scholar 
Niemiec, M. & Komorowska, M. The use of slow-release fertilizers as a part of optimization of celeriac production technology. Agric. Eng. 22(2), 59–68 (2018).
Google Scholar 
Ranum, P., Peña-Rosas, J. P. & Garcia-Casal, M. N. Global maize production, utilization, and consumption. Ann. N. Y. Acad. Sci. 1312(1), 105–112 (2014).ADS 
PubMed 
Article 

Google Scholar 
HLPE. Biofules and food security. High Level Panel of Experts on Food Security and Nutrition of the Committee on World Food Security, Rome (2013).Karp, A., Beale, M. H., Beaudoin, F. & Eastmond, P. J. Growing innovations for the bioeconomy. Nat. Plants https://doi.org/10.1038/nplants.2015.193 (2015).Article 
PubMed 

Google Scholar 
Chavarria, H., Trigo, E., Villarreal, F., Elverdin, P., & Piñeiro, V. Policy brief bioeconomy: a sustainable development strategy task force 10 sustainable energy, water, and food systems. T20, Saudi Arabia (2020). More

Here we have used existing surveillance to detect an emerging wildlife disease and appraise its impact by combining traditional host and vector screening with utilisation of national datasets generated by citizen scientists. Following the detection of USUV in the UK in 20207, whilst national surveillance identified no further cases of USUV infection in wild birds that year, we discovered a significant cluster of blackbird DIRs and an overlapping regional reduction in reported blackbird observations, possibly indicating disease-mediated population decline. Our investigation also identified mosquito vectors at the index site that were positive for USUV RNA, suggesting that ongoing virus transmission was likely.The most prevalent and notable histological changes in the blackbirds and house sparrow with confirmed USUV infection were those in the liver and spleen, consisting of necrosis and lymphohistiocytic inflammation along with moderate to abundant virus antigen labelling. Whilst neurotropism resulting in brain necrosis and lymphohistiocytic inflammation has been reported in studies which examined large numbers of wild blackbirds with USUV infection in continental Europe4,12, we found minimal evidence of neural lesions in the five wild birds examined in this study. Although, histopathological changes in other tissues were generally non-specific, immunolabelling demonstrated widespread virus antigen distribution in both bird species, which is similar to reports of USUV infection elsewhere4,13,14. Immunolabelling was disproportionately greater in the brain and heart in contrast to the minimal or absent histological changes observed in these organs: similar contrasting results of histological and immunohistochemical examinations of USUV-infected wild birds have previously been reported12. Although only brain and kidney samples were examined using USUV RT-PCR, our findings, together with earlier reports4,14, demonstrate that viral antigen can be detected in abundance in the heart and liver, suggesting that these organs could be useful for molecular diagnostic sampling.A differential for necrotising lesions in European passerines, and a comorbidity detected in blackbirds with USUV infection, is Plasmodium spp. infection4,8,15. DNA of the same Plasmodium spp. as detected in the tissues of USUV-positive blackbirds from the ZSL London Zoo site in 2020 was identified in Cx. pipiens s.l. that fed on blackbird hosts at this site previously in 2015, supporting endemic avian haemoparasite infection of this wild bird species at this location. In contrast to the results reported from USUV-positive blackbirds in the Netherlands4, no exo-erythrocytic stages of haemoprotozoa indicative of avian malaria were observed histologically in the two UK blackbirds positive for Plasmodium DNA. Since histological examination has limited sensitivity, in situ hybridisation could be used to further appraise the clinical significance of this co-infection in the future16.Zoological collections are ideally placed to form part of wildlife disease surveillance networks and have already contributed to flavivirus detection in mainland Europe10,13,17,18. The collection grounds at ZSL London Zoo are well monitored for evidence of morbidity or mortality in synanthropic wildlife; this unusually high level of vigilance is considered the likely explanation for detection of USUV at such a location. Recent import of infected captive birds can be excluded as a potential route of USUV introduction as the COVID-19 pandemic had led to suspension of animal movements into the zoological collection. Following USUV detection in synanthropic wildlife, preemptive management practices were employed to safeguard the health of captive animals (Supplementary Materials 1); there was no evidence of USUV-associated disease in the collection animals.The majority of mosquitoes trapped in 2020 were primarily ornithophagic Cx. pipiens s.l., a known vector for USUV1 and a common species in temperate urban habitats. This mosquito species was also the most frequently detected at the ZSL London Zoo site in 2015, during historical trapping sessions19 and at two zoological collections in northern England20. Bloodmeal analyses from mosquitoes at ZSL London Zoo in 2015 and 2020 demonstrate that this species feeds on both wild and collection birds, as would be expected for a generalist ornithophagic mosquito21. In addition, targeted mosquito surveillance in 2020 confirmed circulating USUV in multiple Cx. pipiens s.l. pools at the index site over a three-week period subsequent to the detection of USUV-associated wild bird mortality. This further demonstrates that local mosquito trapping combined with PCR screening is useful as part of an integrated surveillance programme22 and provides evidence that native vectors in the UK may facilitate the onward transmission of USUV to susceptible hosts following an emergence event.Wild bird flavivirus surveillance in Great Britain integrates submissions from three schemes, each with a different taxonomic focus. These convenience samples inevitably lead to skews in species coverage. Although a common garden bird, the number of blackbirds tested for USUV was modest at 2–8 per annum over the period 2012–2019. A communication programme to raise awareness of blackbirds as a sentinel species for USUV, involving a range of stakeholder communities (e.g. non-governmental organisations, wildlife rehabilitators and veterinary surgeons) could help to increase the volume of submissions and, by extension, the ability to rapidly identify the occurrence of USUV in this species. The potential value of target species as sentinels within wild bird surveillance networks has been highlighted for other pathogens, e.g. highly pathogenic H5N1 avian influenza and West Nile virus23,24. In addition to this passive surveillance focused on disease detection in avian hosts, active targeted serosurveys could be conducted to identify cryptic exposure of subclinically affected birds in the future. Given the logistical challenges around active serosurveys in wild birds, screening of archived samples from captive birds in the zoological collection may provide a means to further appraise the extent of USUV circulation, as has previously been undertaken at other collections in mainland Europe25,26.Local reductions of blackbird populations have been reported following USUV outbreaks in mainland Europe27,28,29, but numbers recorded by the BBS have been stable in the UK and Greater London since 2011 when USUV incursion would be predicted most likely to have occurred on the basis of spatio-temporal patterns of spread in mainland Europe3 until the latest data are available from 2019 (Supplementary Figure 5). Whilst our index site detection of USUV is unlikely to represent the incursion event, and earlier sporadic or localised USUV incidents prior to 2020 may have occurred7, based on historical blackbird population trends it seems plausible that the existing surveillance system enabled rapid detection of this emerging infectious disease.Significant clustering of blackbird DIRs was observed in the Greater London, South East and East of England regions in 2020. These results should be interpreted with care given the potential for biases with these opportunistic data and the absence of confirmed aetiology for the DIRs, however, these findings are consistent with a regional increase in blackbird morbidity and mortality in summer 2020 around the USUV index site. Consequently, it is likely that further blackbirds, in addition to those recovered for PME, were infected and died with USUV. Whilst no evidence of an increase in generalised ill health or neurological disease category blackbird DIRs was found in 2020, particular attention should be paid to early detection of clusters of DIRs of these categories as a potential signal of USUV occurrence in the future.One indicator, the dead bird ringed recovery dataset, did not support increased scale of blackbird mortality in Greater London; however, the dataset is small and vulnerable to variation in observer bias (e.g. related to COVID-19 induced lockdown and travel restrictions). In contrast, using the GBW dataset, we identified a substantial seasonal decline in the blackbird weekly reporting rate which was associated with a concomitant reduction in weekly count in gardens, but not in ecologically similar control species, which was contemporaneous with the period of detected USUV activity in Greater London. These population trends are consistent with a hypothesis of disease-mediated decline. Alternative explanations, such as variation in climate, food availability or bird movement need consideration and are discussed next.Exploration of climate data indicates that, whilst the spring and early summer of 2020 was noteworthy with a high daily temperature average and low rainfall, at the time of USUV detection and the decline in the blackbird reporting rate, these parameters were within historical ranges (Supplementary Table 7). Consequently, while the climate may have been permissive for USUV transmission, there is no evidence to support variation in the weather alone as an explanation for the seasonal pattern of blackbird reporting rate decline; nor were declines observed in the robin or starling data, the control species with similar soil invertebrate diet and therefore similar vulnerability to summer drought. Blackbird, robin and starling populations in the UK are partially migratory; however, birds from mainland Europe do not migrate to overwinter in England until mid-October (i.e. after the decline in blackbird reporting rate occurred): consequently international bird movement does not offer an explanation for the observed regional blackbird decline. During the late summer season, short-distance movement from garden to non-garden habitats typically occurs, during the period of moult; however, the extent of the decline in blackbird reporting rate in gardens that occurred in Greater London in 2020 markedly exceeds that of the historical trend (2011–2019 inclusive; BTO unpubl. data). In summary, despite the fact that surveillance did not confirm further cases of wild bird USUV infection in 2020, and whilst it is not possible to ascribe causality, or exclude the chance that other factors may have contributed to the observed population trend, it remains possible that large-scale blackbird mortality due to USUV occurred in Greater London in summer 2020.Our study and others30 illustrate the need to integrate disease surveillance and long-term population monitoring schemes to evaluate disease impact, and to use control species to explore potential confounding drivers of population change (e.g. climate, food availability). Since GBW reporting rates are generated online in real-time, and nationwide, they offer a tool to rapidly detect changes in species presence (i.e. reporting rate) or flock size in gardens (i.e. weekly maximum count) that can be used to strategically enhance surveillance effort for disease detection. As wild bird ring recovery reports are also submitted online, there is also the potential to develop a complementary system that monitors for trends in occurrence of dead birds that might signal a disease outbreak. The BBS survey provides the most robust available data on population trends to appraise disease impact, however there is a delay of some months until data from this scheme become available. Since repeated incursions have occurred in mainland Europe following first detection1,17,31, it is likely that USUV will emerge in the UK again, either through overwintering or repeat incursion(s). Integrated disease surveillance in combination with bird population monitoring using the various available datasets, as we have capitalised on here, is required to assess whether USUV re-occurs, or becomes endemic, in UK wild birds and to identify any associated population impacts.By combining a range of professional and citizen science datasets our study approach facilitates the rapid detection of an emerging disease in free-living wildlife and enables insights into its incipient impact. We believe this multidisciplinary approach presents a framework for the early detection of disease outbreaks and incursion, thus helping to safeguard animal and public health. Such early warning systems could facilitate prompt mitigation action, for example targeted biosecurity measures and enhanced vigilance by medical and veterinary authorities. In addition, there is opportunity to further develop collaboration with ornithologists through active surveillance of wild birds, as was recently employed to detect West Nile virus in a migratory bird in the Netherlands32. Whilst population monitoring schemes are most developed for wild birds, lessons learned may be applied for the surveillance of diseases affecting other taxa. More