Evolution of fruit and seed traits during almond naturalization

The domestication of crops has been ongoing since the beginning Holocene, determining characteristics cultivated plant species. transition towards domesticated breeds involved a series changes that resulted in phenotypes could be easily harvested (Allaby, 2014; Larson et al., Milla 2015). For example, fruit crops, there have relative size edible parts, which increased and become more accessible (Browicz & Zohary, 1996; Zohary Spiegel-Roy, 1975). Crop documented with detail at archaeological, botanical genetic levels. This abundance information shown (a) it is not linear process ‘ideotype’, (b) distinct event but rather gradient an indefinite no clear endpoint (c) between wild strains unidirectional phenotypic can reversible (Cornille 2012; Dickmann 1994; Richardson Rejmánek, 2011). Therefore, only fully comprehended considering its inverse process, de-domestication or naturalization breeds, role played evolution varieties (i.e. cultigens; Gross Wu 2021). However, available about crop very limited. Scientists studied small number cases, such as weeds emerged from specific cereal cultigens (Gering 2019; There also interest studying gene flow produce are difficult to eradicate due their similarity facilitate propagation genetically modified (GM) plants (Ellstrand, 2018). Nevertheless, cases naturalized do pose management challenges almost entirely unstudied, even though they highly valuable for breeding purposes Although incomplete, evidence obtained suggests involves some degree reversion phenotype, if does necessarily lead same genic regions. Research extremely rapid diversity domestic species tends significantly reduced direct ancestors Zhang 2017). As consequence, selection must draw comparatively restricted pool during feralization, providing intriguing case study evolutionary ecology. feral (wild) state conditioned by on pre-existing traits, will ultimately determine populations (Pannell Since response under feralization depends standing crop, expected proportional harboured pool, highest heterozygous polymorphic self-incompatible trees (Miller Gross, In addition limitations, ecological filters outcome events. Blackburn al. (2011) defined part continuum multiple phases, each characterized selective pressures environmental factors. Accordingly, introduced effectively colonize new area sufficient seed dispersal, germination, development seedlings juveniles (Blackburn 2011; 2000). other words, successful dispersal establishment constitute first prerequisites feralization. traits implicated germination likely naturalization. almond tree described undergoing SE Spain (Balaguer-Romano 2021; Homet-Gutiérrez Even closely related (Prunus subgen. Amygdalus) found this region, spontaneous recruitment individuals near farms frequent phenomenon (Homet-Gutiérrez 2015; Ruiz de la Torre, 2006). these P. dulcis result groves endo-feral, sensu Ellstrand 2010). Their presence inside orchards traditionally accepted extent, serve pollen donors mostly cultigens, source rootstocks identifiable boundary markers land olive (Ibancos Nuñez Rodriguez Franco, 2010; Rubio-Cabetas, 2016; spite long history tolerance occasional semi-wild individuals, dense, self-sustaining recently 2015) dynamics associated emergence remain poorly understood. Taking into account morphological almond, mechanisms gravity (barochory) synzoochory. latter strategy mediated granivores actively transport seeds store them (frequently underground) later consumption (Gómez 2019). Previous results our research team confirmed vertebrate vectors synzoochory almonds important note animal dispersers act primarily predators. Thus, dispersed ensure survival seed, might protective features hard shells targeted toxicity (García 2005; Struempf 1999; Vander Wall Beck, 2012). Seed considered trait increasing face potential predation (Freeland Janzen, 1974). Seeds Prunus spp. contain amygdalin, toxic substance mammals, including humans. Amygdalin cyanogenic diglucoside synthesized systemically prunasin accumulates cotyledons seeds. Once ingested, hydrolysed, releasing glucose, benzaldehyde (responsible bitter taste) hydrogen cyanide, harmful inhibits cellular respiration (Sánchez-Pérez (P. dulcis) genus ‘sweet’ species, amygdalin production appears Mendelian trait, controlled Sweet kernel (Sk) gene. sweet phenotype produced dominant form locus homozygous (SkSk) heterozygotes (Sksk) present characteristic (Heppner, 1923, 1926; Sánchez-Pérez relatively quantities (as few couple dozen) lethal (Ladizinsky, 1999), cultivation inextricably exclusively linked non-toxic (‘sweet’) genotypes Delplancke 2013; Ladizinsky, frequency allele low, point never individual (a puzzling problem was termed ‘riddle domestication’ 1999) perhaps because strongly favoured predator-mediated natural selection. Beyond toxicity, resistance predator deterrence involve harder endocarps increase handling costs promote caching than immediate deter animals solely predators Wall, 2001, Simultaneously, less attractive kernels, smaller sizes and/or lower nutritional value (Gómez, 2004). argued large nutritious desirable stimulate (Vander 2010) existence potentially conflicting pressures. Both (endocarp) heritable (Dicenta 1993; Spiegel-Roy Kochba, 1981). fact, both monogenic; endocarp proposed D-Q 2007) while recent genome-wide studies identified single marker significant association weight (Pavan Given high heritability three (seed weight) (Socias i Company Felipe, 1992), sufficiently strong cause shift means, over low generations. study, we attempted whether divergence indicative purpose, examined implies Iberian Peninsula; quantifiable differences populations, particularly synzoochorous interactions, what extent attributed namely resistance, weight. data were 2017 2018 total 39 (15 24 populations) southern (see Supplementary Methods details). We approximated age comparing photogrammetric images (Plan Nacional Ortofotografía Aérea, 1956–2017, National Geographic Institute Spain, IGN) setting baseline earliest date when previous coetaneous observed use compatible woody vegetation occupied roads; Table 1; Methods; Figure S1). visited period (which harvest season) September October. population, randomly sampled 10 12 reproductive least m apart. enough and, took samples all trees. From tree, collected 25–30 fruits around perimeter crown (or had <25 fruits). 3–5 leaves stored silica gel DNA extraction (Table 1). 37°01?27.0?N 3°57?45.2?W 37°01?24.7?N 3°58?13.5?W 37°13?07.5?N 3°33?44.4?W 37°13?05.5?N 3°33?48.7?W 36°46?27.8?N 4°56?16.6?W 37°47?31.5?N 4°39?21.4?W 36°48?57.4?N 4°57?5.7? W 36°46?28.8?N 4°56?23.0?W 37°01?18.1?N 4°56?29.9?W 37°40?10.0?N 2°48?21.6?W 37°10?44.6?N 4°40?56.9?W 37°11?38.3?N 3°34?56.4?W 37°35?06.9?N 2°15?25.2?W 37°16?37.4?N 3°39?46.5?W 37°19?18.1?N 3°34?07.1?W 37°19?55.2?N 3°33?21.7?W 37°54?06.8?N 2°25?41.5?W 37°23?28.2?N 3°01?07.2?W 37°23?24.9?N 3°01?08.2?W 36°45?25.7?N 5°22?0.2?W 37°19?50.5?N 3°03?12.4?W 37°19?47.1?N 3°03?19.8?W 36°48?38.8?N 3°17?46.7?W 37°42?58.3?N 2°27?39.1?W 36°47?51.8?N 3°17?33.3?W 37°09?44.4?N 3°43?28.0?W 36°52?55.1?N 4°57?19.9?W 37°42?05.6?N 2°11?03.0?W 37°00?15.1?N 4°33?32.2?W 37°00?14.4?N 4°33?45.6?W 37°00?12.8?N 4°33?37.4?W 37°07?41.2?N 3°21?42.4?W analysed structure using neutral (SSRs) determined (SNP Sk, kernel). extracted 17 16 used colorimetric quantification cyanide content (5–6 per population; 1) FavorPrep Plant Genomic Extraction Mini kit (FAVORGEN, Taiwan). carried out SSR genotyping CRAG (Center Agrigenomic Research, CSIC-IRTA-UAB-UB) (2007; Genotyping via SNPs Sk CEBAS-CSIC following protocol (2019). About 50 ng amplify PCR primers bHLH2FBamHI (CACCGCGGATCCGAATGGAAGAGATCATAGCCTCAT) bHLH2RXhoI (GATCCACTCGAGCTAGTTGTACCACCTTTTTATAAT) Phusion High-Fidelity polymerase (Thermo Scientific) conditions: 2 min 98°C, 35 cycles 20 s 60°C 1 72°C, one cycle 5 72°C. SNP detection performed 3500 Genetic Analyzer, Applied Biosystems, sequencing primer bHLH2_763F (AAGAGGGTGATACAAAAGAAGC, Tm 60°C), CAID (University Murcia). To seeds, kernels (volume, in-shell [endocarp + seed] mass mass), mechanical protection (thickness, density endocarps) (amygdalin/cyanide seeds). measurements 2707 fruits, pooled populations. measure 348 subset (6 feral, 1), 4–5 population individual. content, 134 30 (13 feral), utilizing possible measurements. When (e.g. destroyed cracking open measurements), ensuring every consistent level (at 2–3 individual). dried aliquot derived four five (total n = 2707) 50°C 72 h. dehydrated, entire (in-shell) (a: length, b: width, c: height M: mass). Afterward, them. estimate volume almonds, shape ellipsoid (V 4/3?abc). then estimated dividing difference unshelled corresponding [Min-shell ? Mseed]/[Vin-shell Vseed]). thickness halving without shell ([cin-shell cseed]/2). Based results, chose subgroup covered total, subjected endocarp-crushing experiments hydraulic press (S.A.E. IBERTEST model 1BTH-2730) quantify load. spectrophotometer assessment picric acid indicator reagent (Oshima 2003; Methods). quantified six choosing nine 134; liquid nitrogen grind fine, homogeneous powder ?20°C. Then, digested citric tartaric acid. amount liberated paper strips stained washed ethanol h later, optical fluid (proportional content) Tecan NanoQuant Infinite M200 spectrophotometer. parallel, calibration method, 82 (50 cultivated, 32 feral; through ultra-performance chromatography (UPLC) coupled spectrogrametry established Arrazola (2013) Waters Acquity UPLC system interfaced triple quadrupole spectrometer Xevo TQS (Waters). analyses University Granada's Center Scientific Instrumentation (CIC-UGR). among types analysis partition molecular variance (AMOVA) (SSR) sequenced orchards. compute overall fixation index (Fst) construct phylogenetic neighbour joining Nei's distances (Nei, 1972). Population packages ‘hierfstat’, ‘poppr’ ‘vegan’ (Goudet, Kamvar Oksanen Additionally, functional (differences allele) cyanogenesis modelled effect type GLM fitted binomial negative distribution, genotype factor variable. distribution this, generated contingency table showing against evaluated correlation two variables chi-square test. affected type, sets traits. First, across almonds. similar paired included limiting sampling sister OTUs adjacent tree. last compared focus proxy did positively correlated (R ? 0.7; p-value 2.2 × 10?16). analyses, models predictors: ‘type’ (feral cultivated) origin. case, fit predictors, carrying transformations satisfy criteria parametric modelling necessary. Furthermore, mixed random factor, function link best suited variable functions ‘descdist’ ‘fitdist’ packet ‘fitdistrplus’ R (Delignette-Muller Dutang, preliminary explorations residual deviance degrees freedom chi-squared test selected lowest AIC value. comparable values (six units less) simplest, parameterized model, simplest uni-factorial (the applied indicated S3). Analyses Molecular Variance showed within type. Also, explained differed locations interaction term S2). Differences accounted much proportion (R2 0.02 vs. 0.29). Local variation influence term) explanatory power 0.08). These seem indicate stochastic variability cultivation. component most 0.60) congruent intrapopulational corroborate limited differentiation Of pairs analysed, recovered (Figure contrast pattern markers, (SNPs clearly diverged different alleles varied (X2[2, 95] 10.671, p < 0.01). dominants Sk/Sk (associated phenotype) recessive 2). homozygotes (sk/sk Sk/Sk) Sk/sk Sk/Sk, 0.0001 cases), (p 0.063, frequencies concomitant toxicity. All young. Only (Don Fadrique Vereda Estrella), unable ascertain after 1956. Most (17 appeared 40 years (after 1980) eight 25 1998; short time span hindered differentiation, volume, density, comparisons. characterizations. differ groups S3; 3). therefore, should disregarded, trends apparent. Feral naked higher resistance. Toxicity significant. average 3; Pst versus Fst comparisons effects directional 0.128 (95% CI 0.114 – 0.140). any c/h2 > 0.2 (endocarp toxicity). Moreover, 0 4). additive <20% ? Fst. confidence intervals hypothesis c h2 bootstrapping especially (PstEndocarp Resistance 0.668 [95% 0.587–0.793]; PstSeed Weight 0.948 0.943–0.956]; PstCyanide Content 0.463 0.403–0.727]; According Conversely, differentiation. Results demonstrate several shifts may taken place (likely <4) genomic supports acting certain stands marked (functional) types. Neutral inter- intra-populational levels, (self-incompatible insect pollinated; Hamrick Godt, Tamura majority variance. On hand, types, albeit significant, lower. allelic Orchards putatively genotypes; Heppner, 2008) counterparts. seems (Messer 2016). reckoned data, 24) exist ago. Consequently, occurred generations, probably <4 (assuming fast trees; Petit Hampe, It unclear why scientifically until recently, fact Iberia resource A explanation, supported young so common historically represent novel feature. coincidental conditions occur presently facilitating surface spreading decades (from ~564,000 ha 1980 >657,000 2018; MAPAMA, comprises concurrent phenomena. New plantations tend located productive, irrigated areas, mountain marginal (almond soils hillslopes; van Wesemael 2006) abandoned steep pace, still many (MAPAMA, combination propagule pools areas colonization created perfect spread communities. lack origin demonstrated Some plastic, driven practices like pruning fertilization. patterns depend local conditions, come narrow basis. Variation density/resistance 2007, Selection led wide range (hardness), paper-thin stone-hard (Fornés-Comas Our variability; measured crac