Grafted Coffee Increases Yield and Grafting is a viable solution for mitigating damage caused by root- Survivability knot nematodes in coffee orchards. Cleft grafts are created with scions of arabica coffee seedlings at the hypo- 1 2 3 Roxana Myers , Andrea Kawabata , Alyssa Cho , cotyledon stage and a coffee species and Stuart T. Nakamoto4 conferring resistance or tolerance to plant-parasitic nematodes as root- stocks (Reyna, 1966). Promising cof- ADDITIONAL INDEX WORDS. Coffea arabica, Coffea liberica, coffee decline, grafting, fee species that can be used as kona coffee, Meloidogyne konaensis, root-knot nematode rootstock typically have more vigor-
SUMMARY. Kona coffee root-knot nematodes (Meloidogyne konaensis) cause severe ous root development, even in the declines in ‘Kona Typica’ arabica coffee (Coffea arabica) trees in Hawaii. Defoliation presence of plant-parasitic nematodes and destruction of the root system result in significant yield losses and can kill the (Cabos et al., 2010). host. Grafting with other coffee (Coffea) species that exhibit tolerance to kona coffee Robusta coffee (Coffea canephora) root-knot nematodes is a viable solution for mitigating damage in the field. An is widely used as rootstock in Central infested field was established in 2006 with ‘Kona Typica’ scions grafted on seven America and South America. The accessions of promising rootstock and nongrafted ‘Kona Typica’ as the control. Four nematode-resistant ‘Nemaya’ was de- grafted trees of each accession were planted per plot with four repetitions. Yield data veloped through a series of controlled were assessed for the 2016–17, 2017–18, and 2018–19 seasons. Three liberica coffee crosses between promising cultivars (Coffea liberica) accessions [‘Arnoldiana’ (‘Arnoldiana’ 1 and ‘Arnoldiana’ 2), ‘Dewevrei’, and ‘Fukunaga’ 1], demonstrated higher yields of coffee cherry compared that showed resistance or tolerance to with nongrafted ‘Kona Typica’ in the 2016–17 season. In the 2017–18 and 2018–19 root-knot nematodes and lesion nem- seasons, five accessions of liberica and ‘Nemaya’ robusta coffee (Coffea canephora) atodes [Pratylenchus sp. (Bertrand exhibited higher cherry yields than ‘Kona Typica’. Plant vigor was greater in trees et al., 2000)]. ‘Nemaya’ expressed grafted on ‘Arnoldiana’ and ‘Fukunaga’ compared with other accessions and non- 64% resistance to southern root-knot grafted ‘Kona Typica’, with taller trees, higher vertical branches, thicker trunk cir- nematodes (Meloidogyne incognita) cumferences, and overall better health. After 13 years in the field, nongrafted ‘Kona and 78% resistance to a root-knot Typica’ showed the highest mortality, with 81% of trees lost. Liberica rootstocks nematode species from El Salvador. performed consistently well in the presence of kona coffee root-knot nematodes, with When inoculated with kona coffee the healthiest trees, highest yields, and least mortality of the coffee species evaluated. root-knot nematodes, ‘Nemaya’ dem- onstrated increased growth, healthier oot-knot nematodes (Meloido- root systems, and less nematode re- gyne sp.) adversely affect yields production compared with ‘Kona Received for publication 2 Dec. 2019. Accepted for of arabica coffee (Coffea arab- Typica’(Cabosetal.,2010). publication 12 Apr. 2020. R ica) in many coffee-growing regions. In Brazilian coffee plantations Published online 11 May 2020. In Kona, HI, kona coffee root-knot heavily infested with root-knot nem- 1United States Department of Agriculture, Agricul- tural Research Service, Daniel K. Inouye U.S. Pacific nematodes [Meloidogyne konaensis atodes, ‘Apoata’ robusta coffee (IAC Basin Agricultural Research Center, 64 Nowelo (Eisenback et al., 1994)] are ex- 2258) is used as a rootstock (Campos Street, Hilo, HI 96720 tremely pathogenic to ‘Kona Typica’ and Villain, 2005). Even with high 2University of Hawaii at Manoa, Department of arabica trees, causing a disease known initial populations of paran a coffee Tropical Plant and Soil Sciences, 79-7381 Hawaii Belt Road, Kealakekua, HI 96750 as coffee decline (Serracin et al., 1999). root-knot nematodes (Meloidogyne paranaensis), ‘Apoata’ maintains its 3University of Hawaii at Manoa, Department of This cultivar is known commercially Tropical Plant and Soil Sciences, 875 Komohana as Kona Typica, although it has been strong level of resistance (Andreazi Street, Hilo, HI 96720 referred to as Guatemalan in Hawaii et al., 2015). Grafting onto ‘Apoata’ 4University of Hawaii at Manoa, Department of for the last 100 years. Nematode- rootstock in nematode-infested fields Human Nutrition, Food and Animal Sciences, 1955 East-West Road, Honolulu, HI 96822 infested trees show symptoms similar resulted in 55% higher yields com- pared with yields from susceptible We thank Cathy Mello, Jen Burt, Marc Meisner, Nick to nutrient and water deficiencies, Yamauchi, Matt Miyahira, and the coffee harvesting such as leaf yellowing and flagging, cultivars (Barbosa et al., 2014). crew for their assistance. We also thank the University followed by defoliation and even tree Although liberica coffee (Coffea of Hawaii, College of Tropical Agriculture, and Human Resources staff who established and main- death. Underground, root galling liberica) is not as widely used, using tained the field plot over the 13-year period, and Dr. and necrosis occur. The subsequent ‘Fukunaga’ (‘Dewevrei’) as a root- Scot Nelson and Virginia Easton-Smith, who initiated lack of feeder roots and overall de- stock in Hawaii has been an effective the trial. struction of the root system result in management technique for control- Mention of trade names or commercial products in this publication is solely for the purpose of providing significant yield losses. ling kona coffee root-knot nematodes specific information and does not imply recommen- dation or endorsement by the U.S. Department of Agriculture. USDA is an equal opportunity provider and employer. Units R.M. is the corresponding author. E-mail: roxana. To convert U.S. to SI, To convert SI to U.S., [email protected]. multiply by U.S. unit SI unit multiply by This is an open access article distributed under the CC 0.3048 ft m 3.2808 BY-NC-ND license (https://creativecommons.org/ 2.54 inch(es) cm 0.3937 licenses/by-nc-nd/4.0/). 0.4536 lb kg 2.2046 https://doi.org/10.21273/HORTTECH04550-20 28.3495 oz g 0.0353
428 • June 2020 30(3) (Schmitt et al., 2001). The rootstock long-term effects of these rootstocks knot nematode eggs were recovered was named ‘Fukunaga’ by University on coffee yield and tree health under in an average density of 3029 eggs per of Hawaii researchers in honor of field conditions. gram of tomato root. Edward Fukunaga, a previous super- YIELD DATA. The yield per tree of intendent of the Kona Research Sta- Materials and methods ripe, pulpable berries was collected at tion (Kainaliu, HI) (Bittenbender 19- to 21-d intervals during the har- et al., 2001). A 1.4-fold to 2.6-fold FIELD ESTABLISHMENT. For the vest season in 2016–17, 2017–18, reduction in the kona coffee root- original investigation, the field was and 2018–19. Berries were harvested, knot nematode population was established in 2006 at the Kona Re- bagged, and weighed (kilograms) for detected when the ‘Fukunaga’ root- search Station (Kainaliu, HI). Potted each tree. Harvests were then aggre- stock was used (Zhang and Schmitt, ‘Orange Pixie’ tomato (Solanum lyco- gated by rootstock accession for fur- 1995). Higher yields and more ro- persicum) plants were inoculated with ther quality analyses. At the end of the bust root systems were also observed kona coffee root-knot nematode eggs season, the field was strip-picked to in infested field plantings of ‘Kona collected from coffee roots in an follow the recommended integrated Typica’ grafted on ‘Fukunaga’ com- adjacent field. Thirty days after in- pest management practice of sanitiz- pared with nongrafted controls oculation, the roots were removed, ing the field to prevent continued (Schmitt et al., 2001). The accessions finely chopped, and distributed uni- coffee berry borer development and were imported in the 1950s from the formly throughout the field in a water reproduction (Kawabata et al., 2017). Instituto Agronomico do Campinas slurry. The treatments included seven GROWTH DATA. Plant growth (Campinas, Brazil) and screened at rootstocks that had the potential for data were collected in 2016 and the University of Hawaii’s College resistance or tolerance to kona coffee 2017. The circumference and diame- of Tropical Agriculture and Human root-knot nematodes and a control ter of the trunk (centimeters) were Resources Kona Research Station in (nongrafted ‘Kona Typica’) (an measured for each tree. Overall tree Kainaliu, HI (Bittenbender et al., eighth rootstock, ‘Panamanian Dew- height (meters) and the height of the 2001). ‘Arnoldiana’ was also imported evrei’ liberica coffee, failed for un- tallest vertical (centimeters) were at that time, along with six other coffee known reasons and was subsequently evaluated. A health rating based on species. replaced by placeholders). The root- a scale of 0 to 5 was assigned for each Unfortunately, open pollination stocks included two accessions of tree, with 0 being morbid and 5 of robusta and liberica coffee can ‘Fukunaga’, two accessions of ‘Arnoldi- representing optimal health. significantly lower nematode resis- ana’, and one accession each of ‘Nem- tance levels when progeny segregate aya’, ‘Apoata’, and ‘Purpurascens’ Results for nematode susceptibility (Bertrand arabica. Rootstocks were propagated Over the 4-year pruning cycle, et al., 2000). A survey of coffee by seed. ‘Kona Typica’ scions were both accessions of ‘Arnoldiana’ pro- plantations in Esp ırito Santo, Brazil, grafted onto the rootstocks in 2005. duced the greatest amount of coffee found no differences in parasitism by Nongrafted ‘Kona Typica’ was also cherry of all the rootstocks evaluated. root-knot nematodes of arabica and propagated by seed. The field was Five of the grafted rootstocks (the robusta coffee in the field (Barros planted in a randomized complete liberica accessions and ‘Nemaya’) et al., 2014). Although it was pre- block design with four replications outperformed nongrafted ‘Kona sumed by growers that all robusta and four trees per replication (16 trees Typica’ in the last two harvest seasons coffee cultivars are resistant to root- pertreatment).Atthetimeofthe (Table 1). The four liberica accessions knot nematodes, some cultivars dem- current study, two trees or scions had and ‘Nemaya’ had yields that were onstrated susceptibility to root-knot died, leaving 15 trees for two of the higher than those of ‘Apoata’, ‘Pur- nematodes. treatments. Supplemental irrigation purascens’, and ‘Kona Typica’ in the In addition to outcrossing, levels was provided only until the trees were 2018–19 harvest season (P £ 0.05). of resistance to different root-knot established. The entire plot was treated ‘Fukunaga’ 1, ‘Arnoldiana’ 1, species can vary among coffee culti- uniformly for fertilization, pest control, ‘Arnoldiana’ 2, and ‘Nemaya’ had vars, and there is the potential for and other crop management activities. greater yields than ‘Apoata’, ‘Purpur- root-knot nematode populations to All trees were stumped 46 to 61 ascens’, and ‘Kona Typica’ in the overcome single resistance genes over cm from the ground in 2015. Emerg- 2017–18 harvest season (P £ 0.05). time. Continued evaluation of addi- ing new verticals were reduced to four In the 2016–17 harvest season, tional sources of host plant resistance or fewer per stump later that year by ‘Fukunaga’ 1, ‘Arnoldiana’ 1, and is recommended to ensure long-term the station staff. All remaining pro- ‘Arnoldiana’ 2 were the only grafted management of root-knot nematodes ductive verticals were of similar age at accessions that differed from the non- in Hawaii (Serracin and Schmitt, the initiation of data collection in grafted ‘Kona Typica’ (P £ 0.05), 2002). An investigation was con- 2016. although this is thought to be more ducted at a field plot infested with Field populations of kona coffee a result of the pruning style instead of kona coffee root-knot nematodes and root-knot nematodes were confirmed rootstock performance. In the 2018– planted with ‘Kona Typica’ trees to be increased in 2019 by planting 19 harvest season, ‘Arnoldiana’ 1 had grafted on seven cultivars of promis- ‘Orange Pixie’ tomato plants ran- the highest red cherry yields of 9.8 ing rootstock material compared with domly throughout the field next to kg/tree, whereas ‘Arnoldiana’ 2 had a control (nongrafted ‘Kona Typica’). grafted and nongrafted trees. Plants the greatest yields in 2017–18, at The objective of this study was to were removed after one or two nem- 10.9 kg/tree. In the first harvest revisit that field plot and evaluate the atode generations. Kona coffee root- season, trees grafted on ‘Arnoldiana’
• June 2020 30(3) 429 1 had the highest yields, followed by Verticals of liberica rootstocks ‘Nemaya’ rootstocks also had high ‘Fukunaga’ 1. Nongrafted ‘Kona were taller than nongrafted ‘Kona survival rates, with 94% of trees Typica’ was the lowest producer in Typica’, ‘Apoata’, and ‘Purpurascens’ remaining. ‘Apoata’ (81%) and ‘Pur- all harvest seasons, followed by ‘Pur- (P £ 0.05) in 2017 (Table 2). In purascens’ (69%) had lower survivabil- purascens’ and ‘Apoata’. 2016, the grafted rootstock plants ity than others, with the nongrafted The trunk circumference was dif- had taller verticals than the non- ‘Kona Typica’ having the lowest sur- ferent for all grafted accessions (26– grafted control (P £ 0.05). In 2016, vivability at 19% (Fig. 1). 30 cm) and the ungrafted control (19 ‘Arnoldiana’ 2 had the tallest mean Cupping quality was unaffected cm) in 2017 (P £ 0.05), with the verticals at 167 cm, which was greater by grafting on any of the rootstocks exception of ‘Apoata’ (26 cm), which than those of ‘Kona Typica (85 cm) examined in this study, with cupping was not significantly different from and ‘Apoata’ (128 cm) (P £ 0.05). scores similar to those of nongrafted the control. A similar trend was ob- Plant health of all grafted trees ‘Kona Typica’ (data not shown). served in 2016, with grafted trunks was greater than that of nongrafted Based on Specialty Coffee Association ranging from 24 to 27 cm compared ‘Kona Typica’ in 2017 and 2016 (P £ (SCA) protocols and a 100-point with nongrafted ‘Kona Typica’ at 16 0.05) (Table 2). The liberica acces- grading scale, blind cupping scores cm (P £ 0.05). Both accessions of sions were also healthier than of the various treatments ranged from ‘Arnoldiana’ had the largest trunk ‘Apoata’ and ‘Purpurascens’ (P £ 81.00 to 83.75, similar to those of circumference in 2017 and 2016. 0.05) in 2017. This was also true in nongrafted ‘Kona Typica’, which had In 2016, all accessions were taller 2016 (P £ 0.05), with the exception a score of 83.50. The final SCA (1.8–2.2 m) than ‘Kona Typica’ of ‘Arnoldiana’ 1. scoring rates these coffees as ‘‘Very (1.3 m) (P £ 0.05). ‘Arnoldiana’ 2 No mortality occurred in both Good Specialty’’ coffees. (2.2 m) was taller than ‘Purpurascens’ accessions of ‘Arnoldiana’ after 13 (1.8 m) and ‘Apoata’ (1.8 m). years (Table 2). ‘Dewevrei’ 1 and Discussion This study confirmed that kona coffee root-knot nematodes are a seri- Table 1. Coffee cherry yields for ‘Kona Typica’ arabica coffee grafted on root- ous pest of ‘Kona Typica’ because knot nematode-tolerant rootstocks of liberica, robusta, or arabica coffee and they severely limit survivability, lon- nongrafted ‘Kona Typica’ planted in a root-knot nematode-infested field 11–13 gevity, and yield potential in infested years prior. fields. During the 13 years since field
z establishment, coffee decline, caused Yield (kg/tree) by kona coffee root-knot nematodes, Cultivar 2016–17 2017–18 2018–19 weakened ‘Kona Typica’ trees and Fukunaga 1 3.8 ay 9.6 a 7.7 a caused high mortality. Yields were Fukunaga 2 2.3 abc 7.1 ab 9.1 a reduced at even greater levels than Arnoldiana 1 3.5 ab 9.1 a 9.8 a the reported 45% observed after a 2- Arnoldiana 2 4.0 a 10.9 a 9.4 a year field trial (Schmitt et al., 2001) Nemaya 2.3 abc 8.4 a 8.3 a showing that damage from kona cof- Apoata 1.8 abc 3.1 bc 2.9 b fee root-knot nematodes is both cu- Purpurascens 0.7 bc 3.2 bc 2.5 b mulative and irreversible. With no Kona Typica 0.0 c 0.2 c 0.0 b chemical or cultural method for erad- z1 kg = 2.2046 lb. icating kona coffee root-knot nema- yTotals with the same letter were not significantly different according to the Tukey-Kramer highly significant > todes in infested fields, grafting has difference test (P 0.05). proven itself as the only viable cultural management practice for mitigating damage caused by nematodes. It is an Table 2. Mean vertical heights and plant health ratings of ‘Kona Typica’ arabica effective technology that is economi- coffee grafted on root-knot nematode-tolerant rootstocks of liberica, robusta, or cally viable, environmentally compat- arabica coffee and nongrafted ‘Kona Typica’ planted in a root-knot nematode- ible, and easily adopted by growers. infested field 12 years prior (2017). Percent survivability was recorded in 2018, Trees grafted on ‘Arnoldiana’ 13 years after planting. had the highest yields, health rating, Cultivar Vertical ht (cm)z Plant health rating (0–5 scale)y Alive trees (%) and plant vigor comparable to ‘Fuku- naga’, with no mortality after 13 Fukunaga 1 273 ax 4.15 a 94 a years. The greatest amount of coffee Fukunaga 2 264 a 3.63 a 88 a cherry from an individual tree in the Arnoldiana 1 255 a 3.81 a 100 a field was harvested from a tree grafted Arnoldiana 2 276 a 4.31 a 100 a on ‘Arnoldiana’. In addition, ‘Arnoldi- Nemaya 243 ab 3.03 ab 94 a ana’ rootstocks had the highest yields Apoata 205 b 1.91 bc 81 a of all the rootstocks tested when aver- Purpurascens 205 b 1.50 cd 69 a aged over the entire pruning cycle. Kona Typica 122 c 0.16 d 19 b This cultivar shows promise as a new z1 cm = 0.3937 inch. rootstock that can be incorporated into yPlant health rating based on a scale of 0 to 5, with 0 being morbid and 5 representing optimal health. xMeans with the same letter are not significantly different according to the Tukey-Kramer highly significant the long-term management of kona difference test (P > 0.05). coffee root-knot nematodes in Hawaii.
430 • June 2020 30(3) ‘Fukunaga’ is currently used by variation, as seen in a previous potted the trees sampled was not observed some growers in Hawaii and is rec- plant bioassay (Cabos et al., 2010). for any one cultivar, and no AMF ommended by the University of To maintain nematode tolerance, were found to be associated with Hawaii College of Tropical Agriculture clones should be made by selecting Kona Typica. The dataset was insuffi- and Human Resources (Bittenbender robusta trees that remain vigorous cient to draw any conclusions at this et al., 2001). It has been proven to be and productive in infested fields time, but future research of the role of a reliable rootstock for use in com- (Fatobene et al., 2018). Even varia- AMF in the tolerance of coffee culti- mercial orchards infested with kona tions among accessions of both liber- vars to kona coffee root-knot nema- coffee root-knot nematodes, sustain- ica cultivars require further study and todes should be explored. ing high yields, vigorous growth, and consideration when selecting material Due to its isolated location, low mortality over a 13-year period. for distribution to growers. Hawaii has few of the diseases and Necrosis was mild and few galls were The presence of naturally occur- pests plaguing most coffee-growing observed on roots of ‘Fukunaga’ in ring arbuscular mycorrhizal fungi regions. Kona coffee root-knot nem- contrast to severe necrosis and galling (AMF) in the field plot could also atodes are likely the most significant on ‘Kona Typica’ roots when inocu- explain some of the variation. The contributor to declining yields in lated with kona coffee root-knot AMF compete with nematodes for Kona orchards. With high nematode nematodes in a potted plant bioassay space and nutrients and improve plant populations in the field, symptoms of (Cabos et al., 2010). Root weights tolerance by increasing root vigor coffee decline, and no other observ- and plant growth were also greater in (Schouteden et al., 2015). When my- able biotic or abiotic factors, it is infested ‘Fukunaga’ plants in contrast corrhizal root exudates in the soil reasonable to conclude that the dif- to infested ‘Kona Typica’. A histology were encountered, root-knot nema- ferences in health and yield of the study showed fewer galls and poorly tode motility and root penetration grafted trees were due to the varying developed giant cells in liberica were reduced (Vos et al., 2012). In- levels of tolerance to kona coffee roots, suggesting a restriction in nem- oculation of coffee seedlings with root-knot nematodes by the atode development and reproduction AMF in a greenhouse bioassay dem- rootstocks. (Arango et al., 1982). Liberica coffee onstrated increased tolerance and re- There is a need to continue out- was also found to have a more robust duced damage from parasitism by reach efforts to educate growers re- tap root that penetrated deeper in the root-knot nematodes (Alban et al., garding the adverse effects of kona soil profile than arabica root systems 2013). Subsampling from the field coffee root-knot nematodes on coffee (Bertrand et al., 2001). The ability of plot revealed AMF colonizing 23% production. Lack of awareness of the liberica coffee to produce new roots of the tree roots. Colonization of all existence of this pest has resulted in even in the presence of high nema- tode population densities makes it an effective rootstock (Cabos et al., 2010). Although it performed well in the initial years of the trial, ‘Purpur- ascens’ fared poorly in this long-term study, with low yields, poor health, and 31% mortality. As observed with other arabica cultivars like Kona Typica, the inability to produce vig- orous roots in the presence of a nem- atode infestation makes it highly intolerant to kona coffee root-knot nematodes. Although ‘Apoata’ is successfully used as a rootstock against root-knot nematodes in Brazil (Campos and Villain, 2005), it did not perform well against kona coffee root-knot nema- todes in this study. This further jus- tifies the importance of screening multiple coffee species against local nematode isolates to find the optimal rootstock for that region. ‘Nemaya’ rootstocks demon- strated greater yields and less mortal- Fig. 1. Aerial view of a 13-year-old field containing ‘Kona Typica’ arabica coffee ity than nongrafted ‘Kona Typica’, grafted on root-knot nematode-tolerant rootstocks of liberica, robusta, and but they were not as productive or arabica coffee and nongrafted ‘Kona Typica’. Blocks consist of four grafted trees healthy as the liberica cultivars. High bordered on each side by nongrafted rootstock cultivars. Boxes highlight low variability among individual trees survivability in blocks of nongrafted ‘Kona Typica’ (yellow) and ‘Kona Typica’ suggests outcrossing and genetic grafted on ‘Purpurascens’ (purple).
• June 2020 30(3) 431 a continuing, but not well-docu- enxertia e do nematoide Meloidogyne exi- Fatobene, B.J.R., V.T. Andrade, W. mented, decrease in yields. Increased gua sobre o crescimento radicular e Goncxalves, and O.G. Filho. 2018. Coffee availability of rootstock material a produtividade de cafeeiros. Coffee Sci. canephora clones with multiple resistance needs to be addressed with careful 9:427–434. to Meloidogyne incognita and M. para- naensis. Exp. Agr. 55:443–451. selection and clonal propagation of Barros, A.F., R.D.L. Oliveira, I.M. Lima, high yielding and highly tolerant R.R. Coutinho, A.O. Ferreira, and A. Kawabata, A.M., S.T. Nakamoto, and accessions. Costa. 2014. Root-knot nematodes, R.T. Curtiss. 2017. Recommendations Grafting ‘Kona Typica’ onto a growing problem for Conilon coffee in for coffee berry borer integrated pest ‘Arnoldiana’ and ‘Fukunaga’ root- Esp ırito Santo state, Brazil. Crop Prot. management in Hawaii 2016. Univ. stocks when establishing or replant- 55:74–79. Hawaii Manoa, College Trop. Agr. Hu- ing coffee orchards in Hawaii can be man Resources, Insect Pests IP-41. 29 Bertrand, B., H. Etienne, and A. Eskes. Nov. 2019.
432 • June 2020 30(3)