agronomy Article Viability of Deficit Irrigation Pre-Exposure in Adapting Robusta Coffee to Drought Stress Godfrey Sseremba 1,2,* , Pangirayi Bernard Tongoona 2, Pascal Musoli 1, John Saviour Yaw Eleblu 2 , Leander Dede Melomey 2, Daphne Nyachaki Bitalo 1, Evans Atwijukire 1, Joseph Mulindwa 1, Naome Aryatwijuka 1,3, Edgar Muhumuza 1, Judith Kobusinge 1, Betty Magambo 1, Godfrey Hubby Kagezi 1, Eric Yirenkyi Danquah 2 , Elizabeth Balyejusa Kizito 3, Gerald Kyalo 4, Emmanuel Iyamulemye 4 and Geofrey Arinaitwe 1 1 National Coffee Research Institute, National Agricultural Research Organisation, Mukono P.O. Box 185, Uganda 2 West Africa Centre for Crop Improvement, University of Ghana Legon, Accra PMB LG30, Ghana 3 Department of Agriculture, Faculty of Agricultural Science, Uganda Christian University, Mukono P.O. Box 4, Uganda 4 Uganda Coffee Development Authority, Kampala P.O. Box 7267, Uganda * Correspondence: gsseremba16@gmail.com Abstract: Coffea canephora has high but inadequately exploited genetic diversity. This diversity, if well exploited, can sustain coffee productivity amidst climate change effects. Drought and heat stress are major global threats to coffee productivity, quality, and tradable volumes. It is not well understood if there is a selectable variation for drought stress tolerance in Robusta coffee half-sibs as a result of watering deficit pre-exposure at the germination stage. Half-sib seeds from selected commercial clones (KR5, KR6, KR7) and a pipeline clone X1 were primed with deficit watering at two growth stages followed by recovery and later evaluated for tolerance to watering deficit stress in Citation: Sseremba, G.; Tongoona, three different temperature environments by estimation of plant growth and wilt parameters. Overall, P.B.; Musoli, P.; Eleblu, J.S.Y.; the KR7 family performed the best in terms of the number of individuals excelling for tolerance to Melomey, L.D.; Bitalo, D.N.; deficit watering. In order of decreasing tolerance, the 10 most promising individuals for drought and Atwijukire, E.; Mulindwa, J.; heat tolerance were identified as: 14.KR7.2, 25.X1.1, 35.KR5.5, 36.KR5.6, 41.KR7.5, 46.KR6.4, 47.KR6.5, Aryatwijuka, N.; Muhumuza, E.; et al. 291.X1.3, 318.X1.3, and 15.KR7.3. This is the first prospect into the potential of C. canephora half-sibs’ Viability of Deficit Irrigation Pre-Exposure in Adapting Robusta diversity as an unbound source of genetic variation for abiotic stress tolerance breeding. Coffee to Drought Stress. Agronomy 2023, 13, 674. https://doi.org/ Keywords: climate change adaptation; Coffea canephora; drought tolerance; drought stress recovery; 10.3390/agronomy13030674 heat stress adaptation; priming by deficit watering Academic Editors: Fábio Luiz Partelli, José Domingos Cochicho Ramalho and Douglas Silva Domingues 1. Introduction Received: 24 January 2023 Robusta coffee (Coffea canephora Pierre ex A. Froehner) and Arabica coffee are the Revised: 21 February 2023 most globally traded coffee species. C. canephora constitutes about 40% of the world’s total Accepted: 22 February 2023 coffee exports of 117.1 million bags [1]. There are other coffee species gaining attention, Published: 25 February 2023 such as C. liberica var. excelsa [2], C. stenophylla, and C. eugnoides, especially amidst climate change effects [3] and the emergence of specialty markets [4,5]. In the outcrossing diploid (2n = 2x = 22) C. canephora, the currently underutilized species may vitally be responsible Copyright: © 2023 by the authors. for the crop’s diversity and landscape adaptation [6] because of ‘unabated’ interspecific Licensee MDPI, Basel, Switzerland. fertilization among all known coffee species. Of the most commercial species, C. canephora is This article is an open access article believed to embody high genetic diversity partly attributed to wide geographic adaption [7] distributed under the terms and and at the plot level, because of the crop’s outcrossing behavior [8,9]. Similarly, in some conditions of the Creative Commons other crops, it is understood that pollination biology has a strong bearing on gene flow and Attribution (CC BY) license (https:// subsequent genetic diversity, e.g., in clonal Bromelia hieronymi Mez [10], Musa acuminata creativecommons.org/licenses/by/ Colla [11], and coffee-analogous tree species, Frangula alnus Mill. [12]. Kiwuka et al. [13] 4.0/). also alluded to similarities in gene pools among C. canephora populations in wild, feral, Agronomy 2023, 13, 674. https://doi.org/10.3390/agronomy13030674 https://www.mdpi.com/journal/agronomy Agronomy 2023, 13, 674 2 of 14 and cultivated landscapes. In fact, some genotypes of C. canephora in Uganda possess morphological architecture (e.g., tree vigor, plant height, leaf color, size, and texture) like that of C. liberica as trait similarities enabled by outcrossing (G. Sseremba, pers. comm.). Eight distinct genetic groups of Robusta coffee are known to exist [6,9,10,13–19], including a Ugandan group [13,18]. To further demonstrate the high genetic diversity in C. canephora, [13] reported distinct subgroups in Uganda; namely the southcentral (SC) and northwest (NW) clusters. These clusters are quite explained by temperature and drought gradients along the sampled locations [6,13]. In this case, the NW cluster is suggested to have a relatively higher tolerance to drought and high temperature than the SC cluster, whereas geographical location accounts for the wide genetic diversity of C. canephora is generally elucidated and planned for utilization [20,21], outcrossing behavior though appreciated [9], is not exploited. Outcrossing or exclusive cross-pollination of the C. canephora provides for the random constitution of offspring known as half-sibs of unknown pollen donors but known maternal parentage. In contrast, C. arabica withstands inbreeding, thereby supporting pure line selection and reproductively produced commercial seed. There is a need to develop innovative breeding approaches for Robusta coffee with a view to relieving farmers from long reproductive cycles and slow clonal growth progress before access to a commercial variety of guaranteed descriptor composition. The exploitation of natural half-sibs may be among the most important strategies to shorten a variety’s turn-around time in pursuit of meeting market demands arising from climate change effects and rapid changes in customer preferences. Drought and temperature [22,23] stresses among major global coffee productivity and sustainability constraints were selected for piloting the exploration of half-sibs-based genetic diversity in Robusta coffee. According to predictions by [22], an excess in minimum/maximum temperature of 1 ◦C away from an average of 20.5 ◦C (16.2–24 ◦C optimal range) results in at least 14% yield loss of Robusta coffee. Relatedly, drought stress, commonly mimicked under controlled conditions by deficit watering, causes loss of plant turgor, growth retardation, wilting, and ultimate plant death if the stress is unalleviated. Under extreme drought and temperature, yield and bean quality can be impacted to an excess of 80% loss [23]. As demonstrated by [2,13], geographically associated genetic variation exists in rela- tion to the response of C. canephora accessions to drought stress. At a variety level, more than one C. canephora clone is required to be grown on a farm to maximize clonal diversity, thereby increasing chances of inter-clone pollinations and subsequent fertilization, seed set, and berry development [24]. In addition, studies in other crops have suggested the potential of seed and/or seedling priming in either remodeling the genetic architecture (e.g., through DNA damage repair and new mitochondrial formation) of plants or activating immunity (say, through enzyme activation and protein synthesis) against field constraints so that future stresses would be tolerated or resisted [25–29]. As such, we hypothesized that genetic diversity due to outcrossing behavior might be genotype- and trait-specific depending on the crop seeds’ responsiveness to stress adaptation when primed. Plants can respond to drought stress through any or several of the four mechanisms: escape, avoidance, tolerance, and recovery [30–33], but there is no information relating to the impact of deficit watering priming on drought tolerance of Robusta coffee. This study explores the ability of C. canephora to utilize priming effects in half-sib families for protection against future drought and heat stress effects. Through exploration of both drought tolerance and recovery mechanisms under high-temperature environments, we specifically aimed to: (i) ascertain if priming influences tolerance to watering deficit stress; (ii) identify half-sib families that excel under watering deficit stress; and (iii) identify potentially drought tolerant individuals among half-sib families. The intention is to inform on the potential of utilizing genetic diversity created by the outcrossing nature of Robusta coffee in breeding for resilience to drought and high temperature and desired market traits. Agronomy 2023, 13, x FOR PEER REVIEW 3 of 14 environments, we specifically aimed to: (i) ascertain if priming influences tolerance to wa- tering deficit stress; (ii) identify half-sib families that excel under watering deficit stress; and (iii) identify potentially drought tolerant individuals among half-sib families. The in- tention is to inform on the potential of utilizing genetic diversity created by the outcross- ing nature of Robusta coffee in breeding for resilience to drought and high temperature and desired market traits. 2. Materials and Methods 2.1. Study Site The study was carried out at the National Coffee Research Institute (NaCORI), Kituza, Mukono, in central Uganda. Kituza is located (latitude −1.406, longitude 34.453) Agronomy 2023, 13, 674 37 km east of Kampala and 15 km off Kampala-Jinja Road towards the Katosi landing3 osfi1te4 of Lake Victoria. Mukono district experiences a tropical climate and lies in an agroecolog- ical zone of the L. Victoria crescent, characterized by nearly continuous rainfall with two p2.eMakast peruinalcstuaantdedM beyt hmooddserate dry spells in February and July. Mukono is among low- m2.1e.dSiutumdy alStiitteude areas at about 1200 m.a.s,l, experiences an ambient temperature range of 16.7–T2h7.e8 s°tCud, yanwda iss csaurirtiaebdleo ufot ra tRtohbeuNstaat icoonfafleCe opfrfoeeduRcetsieoanr.c ThhIen sotivteurtlea(pN ianC sOeaRsIo),nKali tpuazta-, tMerunkso onfo t,hine caerneatr alelaUdgs atnod uan. rKeiltiaubzlae isflolowcaerteindg( lpaatitttuedrnes−, t1h.o40u6g,hl ownigtiht updeeak3s4 .4(u5s3u) a3l7lyk min FeaesbtruoafrKy/aMmapraclha and J1u5lyk/mAuogfuf sKt)a. mMpaailna -aJinndja flRyo cardopto hwaarrvdestt hseaKsoantos sui sluanaldlyin ogcsciuter oinf NLaokveeVmibcteorr ian. dM uMkaoyn,o rdeissptericcttiveexlpye r(ihetntpcess://antorrodpiicapl pclriomaacthe.naon/dcolifefesei-ncalnenadgaror-e2c0o2lo2g/ iacac-l czeosnsedo of nth 1e4 FLe. bVriucatoryri 2a0c2r3e)s. cAen te,xcphearrimacetenrti wzeads sbeyt unpe ainrl cyocnotrnotlilneudo cuosndraitiinofnasll owf ait ghretweno- phoeauksse pcounnscitsutiantge dofb tywmo coodnetrraatsetidnrgy tesmpepllesraintuFreb rreugaimryesa nadndJ ual rya. iMnouukt osnhoeltisera. mThoins gstluodwy- wmaesd iiummplaelmtiteundt eda rferoams a Ft eabrouuatr1y2 2002m0 .tao. sM,l,aeyx p2e0r2i2e.n Wceisthainn aKmitbuiezna,t themrepe etreamtupreeraratunrgee- boaf s1e6d.7 e–n2v7i.8ro◦nCm, eantds wisesruei tcarbealetefdo ranRdo buusestda acto tfefsete lopcraotdiouncsti, onna.mTehlye goevneerrlapl girneesne ahsounsael p(GatHteGrn),s gorefetnhheoaurseea clheadmsbteor (uGnHreCli)a,b alnedfl opweenr irnaginpoauttt eshrneslt,etrh (oOuRgSh) w(Sicthepmeea k1s). (usually in FeTbhruea trhyr/eMe vaarcryhinang dteJmulpye/rAatugreu setn).vMiroaninmaendts flinycclurodpedh a(ri)v GesHt sGe acsoomnps ousseuda lolyf ao csciduer mineNtaollvice mchbaeirn-alnindkM neatyt,inregs apnedct iUvVel-ytr(ehatttepds :c/a/rnbornd picoalpypetrhoeanceh .snhoe/ect orfofoeefi-ncga lteon dcoanr-s2ti0t2u2te/ tahcec emssiedd-doany 1m4oFdeebrrautea rtyem20p2e3r)a.tuArne enxvpierroinmmenent tw oaf s2s6e–t3u5 p°Cin, (ciio) nGtrHoClle edreccotned iitniosindseo tfhae GgrHeeGn hto ucosensctointustiest tihneg mofidtw-doacyo hnitgrahs-ttienmgpteermatpuerrea teunrveirroegnimeensta onfd 32a–r4a3in °oCu, tasnhde l(tieiir). OThRiSs wstuhdicyhw waassi manp loepmeenn ftieedldf r‘ormooFf-eobnrluya’ rrya2in0o2u0tt oshMelatyer2 0w2i2t.hWouitt hsiindeK intuetztain, gth arenedt etrmanpsepraatruernet- ibraosne dsheenevtsir toon cmoennsttsituwteer neocrrmeaatel dfiealndd teumsepderaatttuerset leoncvaitrioonnms, ennatm aet lKy igtuenzaer faolrg mreiedn-dhaoyu soef (2G2–H2G7 )°,Cg.r eenhouse chamber (GHC), and open rainout shelter (ORS) (Scheme 1). Sccheemee 11.. SSoomee ooff tthee ssttudyy ccooffffeeee haallff--ssiibb pllaanttss undeerr tteempeerraatturree eenvviirroonmeenttss ooff ggeeneerraall greenhouse (GHG), greenhouse chamber (GHC)),, and open raiinoutt sshelltterr ((ORS)).. 2.2. PTlahnet tMhraeteerviaalrsy ing temperature environments included (i) GHG composed of a side metaFlluicllcyh raiipne- lcinhkernrye tbtienagrianngd hUalVf--stirbe asteeedd cwarabso hnaprvoelysteetdh efnroemsh tehertereo coofimngmteorccioanl s◦ c tliotnuetes othf ethmei dN-adtaioynmalo Adegrraictue ltteumrapl eRraetsuearercehn vOirrgoannmiseanttioonf 2(N6–A3R5 OC) ,K(iiit)uGzaH RCoebruescttae d(KinRs) isdeeritehse, nGaHmGeltyo KcoRn5s,t iKtuRt6e, tahnedm KidR-7d,a aynhdi gthhe-t efomuprethra pturoremeinsivnigro cnlmoneen t of 32–43 ◦C, and (iii) ORS which was an open field ‘roof-only’ rainout shelter without sidcoedneedtt iXn1g; aanlld otfr wanhsipcahr eanret kirnoonwsnh efeotrs tthoeicro rnesstiisttuatnecne otorm caolfffieee lwd itlet mdipseeraasteu, rheigehn vyiireolndm anendt gaotoKdi tpuhzyasifcoarl manidd- dcuapy of 22–27 ◦C. 2.2. Plant Materials Fully ripe cherry bearing half-sib seed was harvested from three commercial clones of the National Agricultural Research Organisation (NARO) Kituza Robusta (KR) series, namely KR5, KR6, and KR7, and the fourth promising clone coded X1; all of which are known for their resistance to coffee wilt disease, high yield and good physical and cup qualities [34,35]. The cherry seed was then de-pulped to obtain parchment from the coffee cherry. Based on slight modifications of standard protocols by Uganda Coffee Development Authority (UCDA) [36] and World Coffee Research (WCR) [37], freshly de-pulped parchment was sown directly into a substrate consisting of topsoil, sand, and manure in a ratio of 5:2:1 and placed in wooden boxes (Scheme 2). After the deficit watering (priming) treatment at either germination or 4-leaf (seedling) stages, individual plants were transferred to 50-liter high-density (HD) polythene pots for full recovery from priming Agronomy 2023, 13, x FOR PEER REVIEW 4 of 14 qualities [34,35]. The cherry seed was then de-pulped to obtain parchment from the coffee cherry. Based on slight modifications of standard protocols by Uganda Coffee Develop- ment Authority (UCDA) [36] and World Coffee Research (WCR) [37], freshly de-pulped parchment was sown directly into a substrate consisting of topsoil, sand, and manure in a ratio of 5:2:1 and placed in wooden boxes (Scheme 2). After the deficit watering (prim- Agronomy 2023, 13, 674 ing) treatment at either germination or 4-leaf (seedling) stages, individual plants 4wofe1r4e transferred to 50-liter high-density (HD) polythene pots for full recovery from priming stress before re-imposition of deficit watering for experimental pots. Control plants were sktreepsts wbeeflol-rweartee-rimedp othsritoiuonghoofudte tfihceit ewxpateerriimngenftoarl epxepreioridm ienn taalll pthoets t.eCmopnetrroatluprlea netnsvwireorne- kmepent wts.e ll-watered throughout the experimental period in all the temperature environments. SScchheemee 22.. CCooffffeeee ppllaannttss uunnddeerr wooooddeenn bbooxxeess bbeeiinngg rreeccoovveerreedd tthhrroouugghh weellll--waatteerriinngg aafftteerr hhaallff ooff tthheemm hhaaddb beeeenne exxppoosseeddt tood deefifcicititw waatteerrininggo off2 255%%fi feieldldc caappaaccitiytyd duurrininggt htheeg geerrmmininaatitoionnp peeriroiodd. . 22..33.. Expeerriimeenttaall Seett Upp 22..33..11.. DDeessiiggnn AA nneesstteedd ‘‘wwiitthhiinn tteemmppeerraattuurree rereggimimee’ ’exepxpereirmimenetnatla sleste utpu wp aws aasrraarnrgaendg.e Tdh. reTeh wreae- wteartienrgin dgefdiceifit cpirtep-erex-peoxspuorseu trreeatrtemaetmntes nwtserwe earpepalpiepdl ioend soenpaseraptaer saetetss (eotnse( osnete fsoert efaocrhe pacreh- perxep-eoxsuproes utrreeattrmeaetnmt)e notf) ofofufor uhrahlfa-lsfi-bssib osfo cfocmommmerecricaila lclcolonnese s(N(NAARROO KKRR55,, NNAARROO KKRR66,, NNAARROO KKRR77,, aanndd aa ccaannddididaatet ecocmommmerecricaila clloclnoen Xe1X) 1o)f RofobRuosbtua sctoaffceoef fiene eainche aocf hthorefet htermee- tpemerpateurraetu erneveinrovnirmonemntesn. tEsa. cEha cfhamfaimlyi lwy aws arserperperseesnetnetde dbyb ysisxix toto 1122 ppllaannttss ppeerr iirrrriiggaattiioonn ttrreeaattmmeenntt ppeerr eennvviirroonnmmeenntt.. 2.3.2. Application of Deficit Watering as Priming Method 2.3.2. Application of Deficit Watering as Priming Method One of the half-sibs’ sets was treated with deficit watering at the germination stage. SimilaOrnlye, tohfe thseec hoanldf-ssiebtso’ fseptrsi mweads tprleaanttesdw waisthra diseefdiciut nwdaetrerwinegll -awt athteer ignegrmunintialttihone 4s-tlaegaef. (Sseimedillainrlgyo, trheea srleycovnedg esteatt iovfe p) rsitmageed, aptlawnhtsic whaths eradiseefidc iut nwdaetre rwinegll-swtreastesrwinags uimntpilo tsheed 4. -Tlehaef s(eseededlilninggs o(1rs etaarnlyd v2engdetsaettiv) eu)n sdtaegrew, aatt ewrhdicehfi ctihtes dtreefsiscipt rwimatienrgin(g2 5st%refisse wldacsa ipmapciotsye)dw. Terhee lsaeteedr lrienlgiesv (e1dst oafntdh e2nsdtr essest) autnndinere wmaotenrt hdsefoifciat gsetrebsys wpreilml-winagt e(2ri5n%g ,fiaenldd cthapisacpietyri)o wdeorfe wlaetlelr- wrealtieevriendg olfa sthteed stfroeusrs mat onninthes m, aotnwthhsi cohf aaggee bthy ewseelel-dwliantgesrihnagd, afnudll ythries cpoevreioredd o.fA wfteellr- cwomateprlienteg rleacsotevde rfyo,utrh me pornitmhse,d ast ewedhliicnhg asgwe etrheet hseeendelixnpgoss hedadt ofuwllayt erreicnogvedreefidc.i At sfttreers csofmor- opblseeter vraetcioovneoryf ,m thoerp phroimloegdic asel eredslpinognss we oenrea thweene kelxypboasesids .to watering deficit stress for ob- servaAtitohni rodf (mcoonrptrhool)losgeticoafl trheesphoanlfs-es iobns aw waseemkalyin btaaisnise.d under well-watering treatment throuAgh tohuirtdth (ecoenxtpreorli)m seetn toafl tpheer ihoadlf.-Fsiibelsd wcaaps amciatiyn(tFaCin)eodf uangdreorw wthelslu-wbsattreartienogf t5retaotpmsoeinl:t 2thsraonudg: h1omuta tnhuer eexwpaesriemsteinmtaalt epderbiyodv.o Fluiemlde cuaspinacgitayp (FroCc)e odfu ar egdroewsctrhib seudbisntr[a3t8e– o4f0 5] .toAptstohiel: s2t asratnodf:t 1h emeaxnpuerreim weanst ,etswtiemnatyte-fidv beyp veorcluenmte( 2u5s%in)go af FpCrowceadsuursee ddefsocrrtihbeedp rinim [3in8g–4tr0e]a. tAmt etnhte asntadrt7 o5%f thfoer etxhpeewrimelel-nwt,a ttwereinngtyt-rfeivaetm peenrcte, nfot l(l2o5w%e)d obf yFCde wmaasn uds-eddri vfoern tahde jupsritmienngt st,rei.aet.,- bmaseendt oanndth 7e5a%g efoorf tphlea nwtse,ldl-awilayteterminpge trraetautrme,eannt,d froelllaotwiveedh buym didemitya.nDdu-drirnigvethne aedxjupsetrmimeenntst,, wi.ee.,o bsaesrevde donth tahtew aagteer ionfg paltan10ts0,% daFiClyw taems qpueirtaetuexrece, sasnivde ,rie.lea.,tiivteb rhoumghidt iatbyo. uDtuflroinogd itnhge conditions in growth pots) and so, 75% was observed to be suited for the controls (well- watered plants). Adjustments were always based on routine visual observation for wilting symptoms and soil moisture content (SMC) to prevent extreme stress exposure and possible plant death. The SMC was monitored using a digital soil moisture meter (Model MO750 Extech® Instruments Corporation, Waltham, MA, USA) [38]. Other management practices such as weeding, fertilizer application, and pest/disease control were blanketly implemented Agronomy 2023, 13, 674 5 of 14 throughout the experimental boxes and pots depending on need as per standard recom- mendations for coffee nurseries and greenhouse practices [36,37]. 2.4. Data Collection Data were collected on a fortnightly basis starting from 28 February 2022 to 11 April 2022, every Monday of the week. Ten morphological growth and drought or water deficit response variables were measured following a standard descriptor manual for coffee [41] and applied by [42] with modifications as described in Table 1. Table 1. Description of variables measured on study individuals from different clonal coffee families. Variable Abbr. Variable Name Unit of Measure Measurement Procedure ILP Internode length on centimeters (cm) Obtained by dividing length of longest/sampled primaryprimary branch by the number of nodes on the primary ILS Internode length on stem cm Obtained by dividing plant height by the number of nodeson stem Leaf blade length was measured from base to apex of a LBL Leaf blade length cm sampled leaf on the first most fully open leaf pair from the primary’s growing tip LBW Leaf blade width cm Measured at broadest part of the leaf measured forleaf length LPP Number of leaves count Counted the number of green leaves per plant Measured length of a visually longest primary from its LLP Length of primary branch cm node (point of attachment to the stem) to the farthest lateral growth tip away from the stem. PLH Plant height cm Tape measure was used to record plant height from thecollar region to the apical tip of the coffee stem Measured girth of the stem at collar region of the plant STG Stem girth cm using digital vernier calliper at collar region of the potted coffee plant NOP Number of primaries count Counted the number of healthy primary branches on a plant WL Proportion of wilted leaves percentage (%) (No. of wilted leaves/No. of leaves on a plant) ∗ 100 WP Proportion of primary % (No. of primaries with at least one wilted leaf/No.branches with wilted leaves primaries on a plant) ∗ 100 wilting score at scale 0–5 (adaptation of Banik et al., 2016: 0 = no leaf is wilted, 1 = 1–25% of leaves are wilted, WS Wilting Score 0+5− score of 0–5 2 = 26–50% of leaves are wilted, 3 = 51–75% of leaves are wilted, 4 = 76–100% of leaves are wilted, and 5 = 100% leaf plus stem wilting) 2.5. Statistical Analysis 2.5.1. Priming Effect on Drought Tolerance and Recovery For drought tolerance study, F-test in GenStat 12th edition for a general linear model providing for nesting of replications (R) within temperature environments (E/R), as well as an interaction of growth stage (weeks, W), environment (E), priming stage (P), and family (F), was used as follows: Yijkl = µ + E/Ri +Wj + Pk + Fl + EWij + EPik + WPjk + EFik + WFjl + PFkl + EWPijk + EWFijl + EPFikl (1) +WPFjkl + EWPFijkl + εijkl where µ is the grand mean, Yijkl stands for the measured response at the ith replication nested within the environment, jth growth stage, kth priming stage, lth family and all interactions, while ε is the overall random error term. The decision for discriminability value of a morphological character among families, other factors such as environment and any level of interactions were made at α = 0.05. For analysis of the priming effect on recovery from deficit watering stress, F-test at 5% error margin was also used: in this case, considering temperature environment (E), replication within the environment (E/R), priming (P), and family (F) treatments. The same levels for each of the study factors as in the drought tolerance analysis were applied for the recovery analysis. In both cases, mean squares for the different sources of variation Agronomy 2023, 13, 674 6 of 14 and mean values of measured variables per family for control and experimental plants for demonstrating the impact of stress memory are reported. 2.5.2. Identification of Potentially Drought-Tolerant Plants For both drought tolerance and recovery assessments, the K-means cluster analysis method based on R packages tidyverse for data manipulation [43], cluster for clustering algorithms [44,45], and factoextra for clustering algorithms and visualization (https:// cran.r-project.org/web/packages/factoextra/ accessed on 23 December 2022). K-means cluster analysis is appropriate for continuous data [46,47], and it is the case for this study. In general, clustering algorithms are tailored to minimize intra-cluster and maximize inter- cluster variations [45,46]. Of several options for determining the optimum number of clusters based on the location of a bend along a ‘total within-cluster sums of squares and the number of clusters (k)’ function, the elbow method [45,47] was applied in this study. While analyzing cluster membership, the focus was put on resilience to both water deficit stress (wilting score, WS) and high temperature (growth environment of either GHC or GHG), excluding the ORS at this stage being a relatively low-temperature environment. 3. Results 3.1. Deficit Watering Effect on Growth and Wilting Response 3.1.1. Family-Level Response The temperature environment had a significant (p < 0.05) effect on all 10 measured variables (Table S1). Plants in open rainout shelter (ORS) had the longest internodes length on primary branches (ILP) at 13.14 cm followed by greenhouse chamber (GHC) at 10.62 cm and general greenhouse (GHG) at 10.22 cm. Priming stage significantly (p < 0.05) affected eight of the 10 measured variables, with non-significant ones being the length of internodes on primary branches (ILP) and plant height (PLH). Significant P × F interactions were exhibited for seven of the measured variables except for ILP, the proportion of wilted primaries (WP), and wilting score (WS). The proportion of wilted leaves (WL) for KR5 was lower for priming at germination (14.4%) than at the 4-leaf stage (18.1%) (Table 2). The WL of KR6 and KR7 was also lower for priming at the germination stage than at the 4-leaf stage. It is notable that the KR7 family had the lowest values across priming stages than the rest of the families. These observed trends also hold for WP of KR5 and KR7. However, a reverse trend (compared to that for KR5, KR6, and KR7) for the X1 family was obtained for WL, WP, and WS in that priming at the 4-leaf stage was observed with lower values of wilting than at the germination stage. Table 2. Mean values for growth and wilting traits measured on four different half-sib families of Robusta coffee. Variable Priming Stage KR5 KR6 KR7 X1 s.e.d l.s.d c.v. Control 11.004 10.987 10.738 11.105 ILP (cm) Germination 11.401 10.354 11.246 11.601 0.3952 0.7754 30.3 4-leaf stage 11.041 10.924 10.550 11.638 Control 5.662 5.586 5.97 5.691 ILS (cm) Germination 6.399 6.093 5.976 6.099 0.1328 0.2605 19.0 4-leaf stage 5.886 5.379 6.173 6.174 Control 20.994 21.735 20.74 19.807 LBL (cm) Germination 21.685 22.125 20.151 21.542 0.4530 0.8887 18.1 4-leaf stage 21.809 21.038 20.848 22.664 Control 9.707 10.253 9.374 8.766 Leaf blade Germination 10.07 10.307 8.927 9.817 width (cm) 0.2180 0.4276 19.14-leaf stage 9.777 9.795 9.140 10.243 Agronomy 2023, 13, 674 7 of 14 Table 2. Cont. Variable Priming Stage KR5 KR6 KR7 X1 s.e.d l.s.d c.v. Control 23.59 20.00 25.55 24.07 LLP (cm) Germination 24.82 20.30 29.81 19.17 1.330 2.609 49.2 4-leaf stage 17.58 21.14 26.64 22.30 Control 73.36 73.46 76.85 73.43 PLH (cm) Germination 74.77 73.39 73.53 70.80 1.590 3.120 18.3 4-leaf stage 74.07 71.16 73.00 76.25 Control 1.2028 1.134 1.2015 1.1578 STG (cm) Germination 1.1268 1.0929 1.1839 1.0456 0.0225 0.0441 16.9 4-leaf stage 1.0617 1.1215 1.124 1.1059 Control 0.01 −0.09 −0.03 −0.01 WL (%) Germination 14.35 13.48 11.30 18.99 2.127 4.173 - 4-leaf stage 18.11 18.51 14.13 13.74 Control 0.02 0.50 0.27 0.43 WP (%) Germination 32.62 33.40 31.62 35.34 2.273 4.459 84.6 4-leaf stage 35.98 32.88 36.56 34.00 Control 0.004 0.009 0.005 −0.004 WS (0+5−) Germination 1.344 1.214 1.141 1.399 0.0979 0.1920 98.5 4-leaf stage 1.255 1.339 1.122 1.287 s.e.d., standard error of differences of means for priming stage by family (P × F); l.s.d., least significant differences of means for P × F at α = 0.05; c.v., percentage coefficient of variation; ILP, internode length on primary; ILS, internode length on stem; LBL, leaf blade length; LBW, leaf blade width; LLP, length of primary branch; PLH, plant height; STG, stem girth; WL, proportion of wilted leaves; proportion of primary branches with wilted leaves; WS, wilting score. 3.1.2. Identity of Tolerant Half-Sibs Either four or eight clusters (Figure 1) were observed to be feasible based on the visible bends in the figure. Subsequent analyses were based on eight clusters (Figure 2). Cluster 7 of size 19 (Table S2) is composed of plants having a relatively low WS of 1.5 and 74% of the individuals located under GHG. The majority (42%) of the cluster 7 members belong to the KR7 family, 26% KR5, while KR6 and XI families are each represented at 16% only. The cluster 7 members include 9.KR6.3, 14.KR7.2, 15.KR7.3, 18.KR7.6, 207.KR5.3, 209.KR5.5, Agronomy 2201230, 1.K3, xR F5OR.6 P,E2ER9 R1E.VXIE1W.3 , 296.KR7.2, 318.X1.3, 320.KR7.2, 25.X1.1, 35.KR5.5, 36.KR5.68 ,o4f 114. KR7.5, 46.KR6.4, 47.KR6.5, 185.KR7.5, and 243.KR7.3. Figure 1. VariationFigoufreto 1t.a Vlawriaittihonin o-fc tloutaslt ewritshuinm-cluosftesr qsuumar oefs swquiatrhesn wuitmh bnuemr boefr colfu cslutesters (k) indicating optimum k is either 4 or 8 for grouping of Robusta coffee half-sibs grown underr sco(nktr)aisntindgi cteamti-ng opti- mum k is either 4 poerra8tufroe rengvriorounpmienngts oanfdR doebficuits wtaatceorinffge. e half-sibs grown under contrasting temperature environments and deficit watering. Figure 2. Cluster plot for the first two principal components of variation among Robusta coffee half- sibs studied for tolerance to deficit watering. 3.2. Recovery from Watering Deficit Stress 3.2.1. Family-Level Response Temperature environment significantly (p < 0.001) affected all 10 measured variables (Table S3), including length of primary (LLP), number of primaries (NOP), and WS. Plants of the longest primary branches were observed under GHC (45.3 cm), followed by GHG (41.2 cm) and ORS (21.3 cm). The highest NOP was recorded under GHC (~7), followed by GHG (~4) and ORS (~3). Aside from the internode length on the stem (ILS), priming significantly affected all measured variables, including LLP, NOP, and WS. There were significant (p < 0.05) priming × family (P × F) interactions for internode length on the stem (ILS), leaf blade width (LBW), number of leaves per plant (LPP), PLH, stem girth (STG), and WS. KR6 and KR7 families’ values of most of the variables (e.g., LBW, LPP, NOP, PLH, and STG) were higher for plants primed at the germination stage than those at the Agronomy 2023, 13, x FOR PEER REVIEW 8 of 14 Figure 1. Variation of total within-cluster sum of squares with number of clusters (k) indicating Agronomy 2023, 13, 674 8 of 14 optimum k is either 4 or 8 for grouping of Robusta coffee half-sibs grown under contrasting tem- perature environments and deficit watering. Fiigurree 22.. CClulussteter rpplolto ftofro trhteh feirfistr stwt tow poripnrciinpcailp caolmcopmonpeonntse notfs voafrivatairoinat aiomnoanmg oRnogbuRsotab ucosftfaeec ohfafelfe- hsiablsf- sstiubsdisetdu dfoierd tofolerrtaonlceera tnoc deetfoicdite wficaittewriantge.r ing. 3..2.. Reeccoveerry ffrrom Watteerriing Deefificciitt Sttrreessss 33..22..11.. FFaammiillyy--LLeevveell RReessppoonnssee TTeemmppeerraattuurree eennvviirroonnmmeenntt ssiiggnniifificcaannttllyy ((pp << 00..000011)) aaffffeecctteedd aallll 1100 mmeeaassuurreedd vvaarriiaabblleess ((TTaabbllee SS33)),, iinncclluuddiinngg lleennggtthh ooff pprriimmaarryy ((LLLLPP)),, nnuummbbeerr ooff pprriimmaarriieess ((NNOOPP)),, aanndd WWSS.. PPllaannttss ooff tthhee lloonnggeesstt pprriimmaarryy bbrraanncchheess wweerree oobbsseerrvveedd uunnddeerr GGHHCC ((4455..33 ccmm)),, ffoolllloowweedd bbyy GGHHGG ((4411..22 ccmm)) aanndd OORRSS ((2211..33 ccmm)).. TThhee hhiigghheesstt NNOOPP wwaass rreeccoorrddeedd uunnddeerr GGHHCC ((~~77)),, ffoolllloowweedd bbyy GGHHGG ((~~44)) aanndd OORRSS ((~~33)).. AAssiiddee ffrroomm tthhee iinntteerrnnooddee lleennggtthh oonn tthhee sstteemm ((IILLSS)),, pprriimmiinngg ssiiggnniiffiiccaannttllyy aaffffeecctteedd aallll mmeeaassuurreedd vvaarriiaabblleess,, iinncclluuddiinngg LLLLPP,, NNOOPP,, aanndd WWSS.. TThheerree wweerree ssiiggnniiffiiccaanntt ((pp << 00.0.055) )pprirmiminingg × ×famfaimlyi l(yP (×P F×) inFt)erinactetiroancsti ofonrs infoterrinnotedren loednegtlhe nognt hthoe nsttehme s(ItLemS), (lIeLaSf) b, llaedafe bwlaiddtehw (LidBtWh )(,L nBuWm)b, enru omf bleearvoefs lpeaevr epslapnetr (pLlPaPn)t, (PLLPHP,) ,sPteLmH g, isrttehm (SgTirGth), (aSnTdG W), aSn. dKWR6S .aKnRd 6KaRn7d fKaRm7ilfiaems’ ilviaeslu’ vesa loufe smoofsmt oosf ttohfet hvearviaarbilaebsl e(es.(ge..,g L.,BLWBW, L, LPPPP, ,NNOOPP,, PPLLHH,, aanndd SSTTGG)) wweerree hhiigghheerr ffoorr ppllaannttss pprriimmeedd aatt tthhee ggeerrmmiinnaattiioonn ssttaaggee tthhaann tthhoossee aatt tthhee 4-leaf stage (Table 3). Notably, WS for the K7 family was also lower under priming at the germination stage than at the 4-leaf stage. Overall, however, the lowest WS was obtained with the KR5 family when primed at the 4-leaf stage. Table 3. Mean values for recovery traits measured on four different half-sib families of Robusta coffee. Variable Priming Stage KR5 KR6 KR7 X1 s.e.d l.s.d c.v. Control 9.9 10.01 9.85 9.52 ILP (cm) Seedling 9.01 8.27 9.29 9.01 0.499 0.983 23.0 Vegetative 8.93 8.84 8.82 9.1 Control 6.236 6.237 6.752 6.459 ILS (cm) Seedling 6.911 6.665 6.395 6.245 0.2997 0.5901 19.9 Vegetative 6.25 5.767 6.28 6.563 Control 20.9 19.58 19.9 21.59 LBL (cm) Seedling 17.75 17.63 15.44 15.81 0.956 1.882 27.7 Vegetative 17.81 15.56 15.33 17.43 Control 8.752 8.617 8.268 8.662 LBW (cm) Seedling 7.443 8.012 6.488 6.492 0.4132 0.8135 23.2 Vegetative 7.417 6.487 6.356 7.728 Agronomy 2023, 13, 674 9 of 14 Table 3. Cont. Variable Priming Stage KR5 KR6 KR7 X1 s.e.d l.s.d c.v. Control 27.74 23.77 28.43 26.7 LLP (cm) Seedling 20.12 21.56 19 17.71 3.028 5.962 59.4 Vegetative 19.36 16.95 19.36 18.7 Control 53.6 48.3 53.6 54.2 LPP Seedling 20.5 41.9 34.2 31.1 5.76 11.35 65.7 Vegetative 29.6 22.8 27.7 29 Control 7.27 5.97 7.8 6.61 NOP Seedling 2.48 3.48 3.89 3.39 0.885 1.742 83.2 Vegetative 3.45 2.64 3.67 3.5 Control 84.89 85.2 88.98 86.76 PLH (cm) Seedling 78.54 83.26 81.87 75.67 3.544 3.489 18.3 Vegetative 82.52 76.68 76.25 84.09 Control 10.594 10.501 10.935 10.387 STG (cm) Seedling 9.973 10.222 10.644 9.844 0.411 0.810 17.0 Vegetative 10.064 9.691 9.339 10.734 Control 0.006 0.054 0.034 −0.046 WS (0+5−) Seedling 3.727 2.81 2.929 3.426 0.310 0.611 65.5 Vegetative 2.212 2.957 2.573 3.446 s.e.d., standard error of differences of means for priming stage by family (P × F); l.s.d., least significant differences of means for P × F at α = 0.05; c.v., percentage coefficient of variation; ILP, internode length on primary; ILS, internode length on stem; LBL, leaf blade length; LBW, leaf blade width; LLP, length of primary branch; LPP, number of leaves per plant; NOP, number of primaries; PLH, plant height; STG, stem girth; WS, wilting score. 3.2.2. Identity of Half-Sibs Recovering Four optimum clusters (Figures 3 and 4) were obtained. Cluster 3, having the lowest Agronomy 2023, 13, x FOR PEER REVIEW 10 of 14 WS of 0.9 and size 36 (Table S4), is composed of 72% of its members located under GHG, while the rest (28%) are located under GHC. The majority (33%) of cluster 3 members belong to the K1R857.KfRa7m.5,i l1y9,2.fKoRll5o.6w, e24d3.KbRy7.23,5 %252X.X1.,3,2 27%6.XK1.3R, 62,8a9.nX1d.1,1 92%91.Xb1e.3l,o n29g2.KtoR6t.1h, e KR5 family. The clu2s9t7e.KrR37.3m, 30e9m.KbR5e.r3s, 31in2.cXl1u.3d, aend4 3.X181.X.41.,3. 8 .KR6.2, 12.KR6.6, 13.KR7.1, 14.KR7.2, 15.KR7.3, 16.KR7.4,Sp1e8c.iKficR e7xamgenotypes that a.6p, in2a5ti.oXn1 o.f1 d, a3ta1 .sKetsR o5n pear in both sets and p.o1 W, S3 f2o.rK bRot5h. t2o,le3ra3n.Kce Ran5d.3 r,ec3o5v.eKryR re5v.e5a,ls3 160. ssess between 0 (no leaf is wilted) and 2 (only KR5.6, 37.KR7.1, 38.KR276.–25,04%1 .oKf Rle7a.v5e,s 4a4re.K wRil6te.d2),, 4a5s .KwRilt6in.g3 ,s4co6r.eKs Rin6c.l4u,de4 71.4K.KRR67..25, , 2458.X.K1.1R, 63.56.K, R151.15,. KR7.3, 169.X1.1, 185.KR376..5K,R15.962, 4.K1.KRR57..65,, 4264.K3R.K6.4R, 747.3.K,R265.52, .2X911..X31,.32, 73618.X.X11..33, ,a2nd8 91.5X.K1R.71.,3.2 T9h1e. X101 b.e3s,t 2g9en2-.KR6.1, 297.KR7.3, 309.KotRyp5e.3s ,w3e1re2 m.Xa1d.e3 f,roamn dhig3h1 8te.mXp1e.r3a.ture (291.X1.3 and 318.X1.3) and moderately high-temperature environments. Figure 3. VariationFigiunret o3.t Valarwiatiitohni inn- tcoltuals wteitrhisnu-cmlustoefr ssuqmu oafr seqsuawreist whinthu nmumbbeerr ooff clculsutesrtse (rks) i(nkd)iciantidngi cthaatti ng that optimum k = 4 for grouping Robusta coffee half-sibs based on drought recovery. optimum k = 4 for grouping Robusta coffee half-sibs based on drought recovery. Figure 4. Cluster plot for the first two principal components of variation among Robusta coffee half- sibs studied for recovery from watering deficit stress. 4. Discussion The significance of difference in tolerance to watering deficit stress with stress- primed plants exhibiting better tolerance and recovery than controls suggest a potential of breeding for resilience to drought stress in C. canephora using the half-sib selection Agronomy 2023, 13, x FOR PEER REVIEW 10 of 14 185.KR7.5, 192.KR5.6, 243.KR7.3, 252.X1.3, 276.X1.3, 289.X1.1, 291.X1.3, 292.KR6.1, 297.KR7.3, 309.KR5.3, 312.X1.3, and 318.X1.3. Specific examination of data sets on WS for both tolerance and recovery reveals 10 genotypes that appear in both sets and possess between 0 (no leaf is wilted) and 2 (only 26–50% of leaves are wilted), as wilting scores include 14.KR7.2, 25.X1.1, 35.KR5.5, 36.KR5.6, 41.KR7.5, 46.KR6.4, 47.KR6.5, 291.X1.3, 318.X1.3, and 15.KR7.3. The 10 best gen- otypes were made from high temperature (291.X1.3 and 318.X1.3) and moderately high- temperature environments. Agronomy 2023, 13, 674 10 of 14 Figure 3. Variation in total within-cluster sum of squares with number of clusters (k) indicating that optimum k = 4 for grouping Robusta coffee half-sibs based on drought recovery. Figure 4. CClulussteterr pplolot tfofro trhteh feirfistr stwt tow poripnrciinpcailp caolmcopmonpeonntes notfs voafrivaatiroinat aiomnoanmg oRnogbuRsotab ucostfafeeco hfafelfe- hsiablsf- sstibusdsietudd fioerd refocrovrecroyv ferroymf rwomatewriantge rdinegficdite fistcrietssst.r ess. 4. DiSspcuescsifiiocne xamination of data sets on WS for both tolerance and recovery reveals 10 geTnhoety spiegsnitfhicaatnacpep oefa rdiinffebroetnhces eitns atonlderpanocsese stos bweatwteerienng 0d(enfoicilte asftriessws iwltietdh) satnredss2- (pornimlye2d6 –p5la0n%tso efxlehaibvietisnagr ebewttieltre dto)l,earsanwciel tainngd srceocroevseirnyc ltuhdane 1c4o.nKtrRo7l.s2 s, u2g5.gXe1s.t1 a, 3p5o.KteRn5ti.a5l, 3o6f .KbrRe5e.d6i,n4g1 .fKorR 7re.5s,il4ie6n.KceR 6to.4 ,d4ro7u.KgRh6t .s5t,re2s9s1 .iXn1 .C3., c3a1n8e.pXh1o.r3a, aunsidng1 5t.hKeR h7a.3l.f-sTibh ese1l0ecbtieosnt genotypes were made from high temperature (291.X1.3 and 318.X1.3) and moderately high-temperature environments. 4. Discussion The significance of difference in tolerance to watering deficit stress with stress-primed plants exhibiting better tolerance and recovery than controls suggest a potential of breeding for resilience to drought stress in C. canephora using the half-sib selection approach. This study’s findings indicate that C. canephora populations differ in their ability to acquire tolerance to drought and high temperatures. Individuals, namely 14.KR7.2, 15.KR7.3, 41.KR7.5, and 243.KR7.3, belonging to the KR7 family, posed a relatively greater potential in tolerating simulated drought/water deficit and high temperature than the other families. Similarly, the greatest potential for recovery from drought stress was exhibited by individual plants of the KR7 family, i.e., 14.KR7.2, 15.KR7.3, 41.KR7.5, 185.KR7.5, and 243.KR7.3. For both tolerance and recovery, potential genotypes for drought tolerance are indicated as 14.KR7.2, 25.X1.1, 35.KR5.5, 36.KR5.6, 41.KR7.5, 46.KR6.4, 47.KR6.5, 291.X1.3, 318.X1.3, and 15.KR7.3. Although at the family average level, KR7 responded better to drought tolerance and recovery, indicating a possible role of maternal genetic factors [33,48]; the 10 best individual plants (genotypes) are a representation of each family studied. Of the best 10, three come from KR7 (14.KR7.2, 41.KR7.5 and 15.KR7.3), three from X1 (25.X1.1, 291.X1.3 and 318.X1.3), two from KR5 (35.KR5.5 and 36.KR5.6), and two from KR6 (46.KR6.4 and 47.KR6.5). Our view is that the observed distribution of best genotypes is a result of (i) the genetic background of mother clone KR7 being more drought tolerant than that of the rest of the studied materials, but (ii) there is also a random pollen grain movement due to an exclusive cross pollination tendency in Robusta coffee thereby increasing geneflow among the study half-sib offspring. The random mating facilitated by self-incompatibility of C. canephora [24,49] and pollination agents [24,50] creates unlimited genetic diversity that suites the crop to recur- ring climate change effects (biotic and abiotic) [2,22,23], including the studied watering deficit tolerance and recovery from the stress. Thus, some of the recombinant genetic compositions from outcrossing tendency could have occurred and conditioned differential Agronomy 2023, 13, 674 11 of 14 stress memory abilities, especially when exposure occurred at the germination stage. This resilience development technique is popularly known as priming, and positive results have been realized in other crops [25,28,29]. The ability of crop plants to adapt to abiotic stresses is often influenced by a combination of genetic and environmental factors. The significance of any of these factors, in turn, determines the heritability and durability of desired attributes in a genotype. Genetically, a plant may possess active tolerance factors outright; otherwise, activity has to be initiated early in the development stage by way of the plant’s innate immunity activation or enforcement of adaptation through artificial means such as crossbreeding and backcrossing, mutation breeding, and genetic engineering. The coffee sector is not yet ready for genetic engineering. Thus, an option of stimulation of remodeling in genetic architecture by stress pre-exposure is imminent, especially in the case of Robusta coffee breeding that is constrained by unpredictable flowering time, making synchronization for parents difficult [24,50], long reproductive cycles, and slow clonal propagation [51,52]. During artificial immunity activation or induction, a plant’s response may manifest and qualify as either being an escapee, avoidant, tolerant, or recoveree. The ability and level of response, as well as the acquired mechanism of drought tolerance, is understood to be both genotype and environment dependent [7,14,23,31]. Our findings generally indicate better watering deficit stress tolerance and recovery for plants primed at germination than at the 4-leaf stage. The views held based on the findings are subject to a follow-up validation study. The validation involving both mother clones and the promising half-sibs needs to be undertaken beyond the controlled conditions of the greenhouse. This can demonstrate the use of half-sibs’ unlimited genetic constitution and diversity for selecting varieties amenable to the increasing effects of climate change that risk on-farm sustainability. 5. Conclusions Watering deficit pre-exposure improves subsequent tolerance to the stress in Robusta coffee half-sibs; maternal origin or family of the half-sib influenced the response to priming stress. KR7 family was better than others in tolerance to deficit watering stress. Selections recommended for onward use in breeding for drought and high-temperature tolerance include 14.KR7.2, 25.X1.1, 35.KR5.5, 36.KR5.6, 41.KR7.5, 46.KR6.4, 47.KR6.5, 291.X1.3, 318.X1.3, and 15.KR7.3. Four of the selections (14.KR7.2, 291.X1.3, 318.X1.3, and 15.KR7.3) are products of stress pre-exposure at the germination stage. We observed that genetic diversity due to outcrossing behavior is genotype- and trait-specific depending on the crop seeds’ ability to respond to priming at early development stages. Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/agronomy13030674/s1, Table S1: Mean squares (MS) and sig- nificance levels (p) for rejecting a hypothesis of no difference in morphological response among priming stages and Robusta coffee families; Table S2: Cluster vector and attributes of Robusta coffee half-sibs under deficit watering conditions; Table S3: Mean squares (MS) and significance levels (p) for rejecting a hypothesis of no difference in recovery response among priming stages and Robusta coffee families; Table S4: Cluster vector and attributes of Robusta coffee half-sibs based on recovery from deficit watering stress. Author Contributions: Conceptualization, G.S., P.B.T., P.M., J.S.Y.E., G.H.K., E.B.K. and G.A.; Method- ology, G.S., P.M. and J.M.; Validation, L.D.M., D.N.B., G.H.K., E.Y.D. and G.K.; Formal analysis, G.S., L.D.M. and N.A.; Investigation, G.S., E.A., J.M., N.A., E.M., J.K., B.M. and G.H.K.; Resources, P.B.T., E.Y.D., E.B.K., G.K. and E.I.; Data curation, D.N.B., E.A., N.A. and E.M.; Writing—original draft, G.S.; Writing—review and editing, P.B.T., P.M., J.S.Y.E., L.D.M., D.N.B., E.A., J.M., J.K., B.M., E.B.K., G.K. and E.I.; Supervision, P.B.T., P.M., J.S.Y.E., E.Y.D., E.B.K. and G.A.; Funding acquisition, G.S., E.Y.D., G.K. and G.A. All authors have read and agreed to the published version of the manuscript. Agronomy 2023, 13, 674 12 of 14 Funding: This research was funded by TWAS/UNESCO and Sida grant number 20-354 RG/BIO/AF/ AC_G—FR3240314179 for research equipment, DAAD under the Postdoctoral Fellowships in Sub- Saharan Africa at DAAD supported Centres grant number 91817875 and Uganda Coffee Development Authority (UCDA). The APC was funded by UCDA. Data Availability Statement: The data used in article is available on request and we can upload it online for access as per data policies. Conflicts of Interest: The authors have no conflict of interest to declare. References 1. ICO. 2021 Coffee Development Report; International Coffee Organisation (ICO): London, UK, 2021. 2. Davis, A.P.; Kiwuka, C.; Faruk, A.; Walubiri, M.J.; Kalema, J. The re-emergence of Liberica coffee as a major crop plant. Nat. Plants 2022, 8, 1322–1328. [CrossRef] [PubMed] 3. Tournebize, R.; Borner, L.; Manel, S.; Meynard, C.N.; Vigouroux, Y.; Crouzillat, D.; Fournier, C.; Kassam, M.; Descombes, P.; Tranchant-Dubreuil, C.; et al. Ecological and genomic vulnerability to climate change across native populations of Robusta coffee (Coffea canephora). Glob. Chang. Biol. 2022, 28, 4124–4142. [CrossRef] [PubMed] 4. Ferreira, J.; Ferreira, C. Challenges and opportunities of new retail horizons in emerging markets: The case of a rising coffee culture in China. Bus. Horiz. 2018, 61, 783–796. [CrossRef] 5. Quintão, R.T.; Brito, E.P.Z.; Belk, R.W. The taste transformational ritual in the specialty coffee market. Rev. Adm. Empres 2017, 57, 483–494. [CrossRef] 6. Aquino, S.O.; Kiwuka, C.; Tournebize, R.; Gain, C.; Marraccini, P.; Mariac, C.; Bethune, K.; Couderc, M.; Cubry, P.; Andrade, A.C.; et al. Adaptive potential of Coffea canephora from Uganda in response to climate change. Mol. Ecol. 2022, 31, 1800–1819. [CrossRef] [PubMed] 7. Nunney, L. Adapting to a changing environment: Modeling the interaction of directional selection and plasticity. J. Hered. 2016, 107, 15–24. [CrossRef] 8. Blambert, L.; Mallet, B.; Humeau, L.; Pailler, T. Reproductive patterns, genetic diversity and inbreeding depression in two closely related Jumellea species with contrasting patterns of commonness and distribution. Ann. Bot. 2016, 118, 93–103. [CrossRef] 9. Depecker, J.; Verleysen, L.; Asimonyio, J.A.; Hatangi, Y.; Kambale, J.; Mwanga, I.M.; Ebele, T.; Dhed’a, B.; Bawin, Y.; Staelens, A.; et al. Genetic diversity and structure in wild Robusta coffee (Coffea canephora A. Froehner) populations in Yangambi (DR Congo) and their relation to forest disturbance. Heredity 2023, 1, 1–9. [CrossRef] 10. Godoy, F.M.d.R.; Lenzi, M.; Ferreira, B.H.D.S.; Silva, L.V.D.; Zanella, C.M.; Paggi, G.M. High genetic diversity and moderate genetic structure in the self-incompatible, clonal Bromelia hieronymi (Bromeliaceae). Bot. J. Linn. Soc. 2018, 187, 672–688. [CrossRef] 11. Kallow, S.; Panis, B.; Vu, D.T.; Vu, T.D.; Paofa, J.; Mertens, A.; Swennen, R.; Janssens, S.B. Maximizing genetic representation in seed collections from populations of self and cross-pollinated banana wild relatives. BMC Plant Biol. 2021, 21, 415. [CrossRef] 12. Finlay, C.M.V.; Bradley, C.R.; Preston, S.J.; Provan, J. Low genetic diversity and potential inbreeding in an isolated population of alder buckthorn (Frangula alnus) following a founder effect. Sci. Rep. 2017, 7, 3010. [CrossRef] 13. Kiwuka, C.; Goudsmit, E.; Tournebize, R.; de Aquino, S.O.; Douma, J.C.; Bellanger, L.; Crouzillat, D.; Stoffelen, P.; Sumirat, U.; Legnaté, H.; et al. Genetic diversity of native and cultivated Ugandan Robusta coffee (Coffea canephora Pierre ex A. Froehner): Climate influences, breeding potential and diversity conservation. PLoS ONE 2021, 16, e0245965. [CrossRef] [PubMed] 14. Cubry, P.; De Bellis, F.; Pot, D.; Musoli, P.; Leroy, T. Global analysis of Coffea canephora Pierre ex Froehner (Rubiaceae) from the Guineo-Congolese region reveals impacts from climatic refuges and migration effects. Genet Resour. Crop. Evol. 2013, 60, 483–501. [CrossRef] 15. Hendre, P.S.; Aggarwal, R.K. Development of Genic and Genomic SSR Markers of Robusta Coffee (Coffea canephora Pierre Ex A. Froehner). PLoS ONE 2014, 9, e113661. [CrossRef] 16. Leroy, T.; Cubry, P.; Durand, N.; Dufour, M.; De Bellis, F.; Jourdan, I.; Vieira, L.G.; Musoli, P.; Aluka, P.; Legnate, H.; et al. Coffea spp. and Coffea canephora diversity evaluated with microsatellites and SNPs. Lessons from comparative analysis. In Proceedings of the 21st International Conference on Coffee Science, Montpellier, France, 11–15 December 2006. 17. Leroy, T.; De Bellis, F.; Legnate, H.; Musoli, P.; Kalonji, A.; Loor Solórzano, R.G.; Cubry, P. Developing core collections to optimize the management and the exploitation of diversity of the coffee Coffea canephora. Genetica 2014, 142, 185–199. [CrossRef] 18. Musoli, P.; Cubry, P.; Aluka, P.; Billot, C.; Dufour, M.; De Bellis, F.; Pot, D.; Bieysse, D.; Charrier, A.; Leroy, T. Genetic differentiation of wild and cultivated populations: Diversity of Coffea canephora Pierre in Uganda. Genome 2009, 52, 634–646. [CrossRef] 19. Pot, D.; Bouchet, S.; Marraccini, P.; De Bellis, F.; Cubry, P.; Jourdan, I.; Pereira, L.F.P.; Vieira, L.G.E.; Musoli, C.P.; Legnate, H.; et al. Nucleotide diversity of genes involved in sucrose metabolism. Towards the identification of candidates genes controlling sucrose variability in Coffea sp. In Conilon Coffee; Capixaba Institute for Research, Technical Assistance and Rural Extension: Linhares, Brazil, 2019; pp. 679–686. 20. Dieters, M.; Hardner, C.; Lambrides, C. BPAT Assessment of NaCORI Coffee Programs; University of Queensland: Brisbane, Australia, 2020. 21. Kawuki, R. Coffee Breeding Improvement Plan for NaCORI; National Coffee Research Institute: Mukono, Uganda, 2021. Agronomy 2023, 13, 674 13 of 14 22. Kath, J.; Mittahalli Byrareddy, V.; Mushtaq, S.; Craparo, A.; Porcel, M. Temperature and rainfall impacts on robusta coffee bean characteristics. Clim. Risk Manag. 2021, 32, 100281. [CrossRef] 23. Venancio, L.P.; Filgueiras, R.; Mantovani, E.C.; do Amaral, C.H.; da Cunha, F.F.; dos Santos Silva, F.C.; Althoff, D.; dos Santos, R.A.; Cavatte, P.C. Impact of drought associated with high temperatures on Coffea canephora plantations: A case study in Espírito Santo State, Brazil. Sci. Rep. 2020, 10, 19719. [CrossRef] [PubMed] 24. Campuzano-Duque, L.F.; Blair, M.W. Strategies for Robusta Coffee (Coffea canephora) Improvement as a New Crop in Colombia. Agriculture 2022, 12, 1576. [CrossRef] 25. Aswathi, K.P.R.; Kalaji, H.M.; Puthur, J.T. Seed priming of plants aiding in drought stress tolerance and faster recovery: A review. Plant Growth Regul. 2022, 97, 235–253. [CrossRef] 26. Devika, O.S.; Singh, S.; Sarkar, D.; Barnwal, P.; Suman, J.; Rakshit, A. Seed Priming: A Potential Supplement in Integrated Resource Management Under Fragile Intensive Ecosystems. Front. Sustain. Food Syst. 2021, 5, 654001. [CrossRef] 27. Hameed, A.; Farooq, T.; Hameed, A.; Sheikh, M.A. Silicon-Mediated Priming Induces Acclimation to Mild Water-Deficit Stress by Altering Physio-Biochemical Attributes in Wheat Plants. Front. Plant Sci. 2021, 12, 625541. [CrossRef] [PubMed] 28. Ling, Y.; Zhao, Y.; Cheng, B.; Tan, M.; Zhang, Y.; Li, Z. Seed priming with chitosan improves germination characteristics associated with alterations in antioxidant defense and dehydration-responsive pathway in white clover under water stress. Plants 2022, 11, 2015. [CrossRef] [PubMed] 29. Liu, X.; Quan, W.; Bartels, D. Stress memory responses and seed priming correlate with drought tolerance in plants: An overview. Planta 2022, 255, 45. [CrossRef] [PubMed] 30. Harfouche, A.; Meilan, R.; Altman, A. Molecular and physiological responses to abiotic stress in forest trees and their relevance to tree improvement. Tree Physiol. 2014, 34, 1181–1198. [CrossRef] [PubMed] 31. Khan, A.; Pan, X.; Najeeb, U.; Tan, D.K.Y.; Fahad, S.; Zahoor, R.; Luo, H. Coping with drought: Stress and adaptive mechanisms, and management through cultural and molecular alternatives in cotton as vital constituents for plant stress resilience and fitness. Biol. Res. 2018, 51, 47. [CrossRef] 32. Saharan, B.S.; Brar, B.; Duhan, J.S.; Kumar, R.; Marwaha, S.; Rajput, V.D.; Minkina, T. Molecular and Physiological Mechanisms to Mitigate Abiotic Stress Conditions in Plants. Life 2022, 12, 1634. [CrossRef] 33. Shanker, A.K.; Venkateswarlu, B. (Eds.) Abiotic Stress Response in Plants—Physiological, Biochemical and Genetic Perspectives; InTech: Rijeka, Croatia, 2011. 34. Musoli, C.P.; Pinard, F.; Charrier, A.; Kangire, A.; ten Hoopen, G.M.; Kabole, C.; Ogwang, J.; Bieysse, D.; Cilas, C. Spatial and temporal analysis of coffee wilt disease caused by Fusarium xylarioides in Coffea canephora. Eur. J. Plant Pathol. 2008, 122, 451–460. [CrossRef] 35. Musoli, P.C.; Kangire, A.; Leroy, T.; Nabaggala, A.; Nakendo, S.; Olal, S.; Ochugo, J.; Kabole, C.; Pande, J.; Cilas, C.; et al. Towards a variety resistant to coffee wilt disease (CWD): A case for Robusta coffee (Coffea canephora) in Uganda. In Conilon Coffee; Capixaba Institute for Research, Technical Assistance and Rural Extension: Linhares, Brazil, 2019; pp. 1472–1479. 36. UCDA. Robusta Coffee Handbook; Uganda Coffee Development Authority (UCDA): Kampala, Uganda, 2019. 37. WCR Good Practice Guide: Coffee Nursery Management. Available online: https://worldcoffeeresearch.org/resources/good- practice-guide-coffee-nursery-management (accessed on 23 December 2022). 38. Sseremba, G.; Tongoona, P.; Eleblu, J.S.Y.; Danquah, E.Y.; Kaweesi, T.; Baguma, Y.; Masanza, M.; Kizito, E.B. Stability of Solanum aethiopicum Shum accessions under varied water deficit stress levels and identification of pertinent breeding traits for resistance to water shortage. Euphytica 2018, 214, 11. [CrossRef] 39. Sseremba, G.; Tongoona, P.; Eleblu, J.; Danquah, E.Y.; Kizito, E.B. Heritability of drought resistance in Solanum aethiopicum Shum group and combining ability of genotypes for drought tolerance and recovery. Sci. Hortic. 2018, 240, 213–220. [CrossRef] 40. Nakanwagi, M.J.; Sseremba, G.; Kabod, N.P.; Masanza, M.; Kizito, E.B. Identification of growth stage-specific watering thresholds for drought screening in Solanum aethiopicum Shum. Sci. Rep. 2020, 10, 862. [CrossRef] 41. IPGRI. Descriptors for Coffee (Coffea spp. and Psilanthus spp.); International Plant Genetic Resources Institute IPGRI: Washington, DC, USA, 1996. 42. Rakocevic, M.; Baroni, D.F. Correlating Coffea Canephora 3D Architecture to Plant Photosynthesis at a Daily Scale and Vegetative Biomass Allocation. Tree Physiol. Available online: https://academic.oup.com/treephys/advance-article-abstract/doi/10.1093/ treephys/tpac138/6909003?login=false (accessed on 23 December 2022). 43. Wickham, H.; Averick, M.; Bryan, J.; Chang, W.; McGowan, L.; François, R.; Grolemund, G.; Hayes, A.; Henry, L.; Hester, J.; et al. Welcome to the Tidyverse. J. Open Source Softw. 2019, 4, 1686. [CrossRef] 44. Maechler, M.; Rousseeuw, P.; Struyf, A.; Hubert, M.; Hornik, K.; Studer, M.; Roudier, P.; Gonzalez, J.; Kozlowski, K.; Schubert, E. Package “Cluster”. Available online: https://svn.r-project.org/R-packages/trunk/cluster/ (accessed on 23 December 2022). 45. Martos, L.A.P.; García-Vico, A.M.; González, P.; Carmona, C.J. Clustering: An R library to facilitate the analysis and comparison of cluster algorithms. Prog. Artif. Intell. 2022, 1, 12. [CrossRef] 46. James, G.; Witten, D.; Hastie, T.; Tibshirani, R. An Introduction to Statistical Learning with Applications in R; James, G., Witten, D., Hastie, T., Tibshirani, R., Eds.; Springer Science and Business Media: New York, NY, USA, 2021; ISBN 978-1-4614-7137-7. 47. Morissette, L.; Chartier, S. The k-means clustering technique: General considerations and implementation in Mathematica. Tutor. Quant. Methods Psychol. 2013, 9, 15–24. [CrossRef] Agronomy 2023, 13, 674 14 of 14 48. Griffing, B. Concept of general and specific combining ability in relation to diallel crossing systems. Aust. J. Biol. Sci. 1956, 9, 463–493. [CrossRef] 49. Moraes, M.S.; Teixeira, A.L.; Ramalho, A.R.; EspAA ndula, M.C.; FerrAAo, M.A.G.; Rocha, R.B. Characterization of gametophytic self-incompatibility of superior clones of Coffea canephora. Genet. Mol. Res. 2018, 17, gmr16039876. [CrossRef] 50. Ngo, H.T.; Mojica, A.C.; Packer, L. Coffee plant—Pollinator interactions: A review. Can. J. Zool. 2011, 89, 647–660. [CrossRef] 51. Chichipe Oyarce, J.; Camacho, A.; Bobadilla, L.G.; Vigo, C.N.; Vásquez, H.V.; Silva Valqui, G. Clonal propagation of Coffea arabica with indole butyric acid and acclimatization conditions in Amazonas, Peru. Int. J. Agron. 2021, 214, 1–8. [CrossRef] 52. Etienne, H.; Breton, D.; Breitler, J.C.; Bertrand, B.; Déchamp, E.; Awada, R.; Marraccini, P.; Léran, S.; Alpizar, E.; Campa, C.; et al. Coffee somatic embryogenesis: How did research, experience gained and innovations promote the commercial propagation of elite clones from the two cultivated species? Front. Plant Sci. 2018, 9, 1630. [CrossRef] Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.