Pollination of English Walnuts: Practices and Problems

Home , Juglans hindsii, Walnuts in California

William H. Krueger

ADDITIONAL INDEX WORDS. Juglans regia, Juglans nigra, pollination, monoecious, dichogamous, wind-pollinated, pistillate flower abortion.

SUMMARY. English walnut (Juglans regia, L.) is a monoecious species bearing staminate and pistillate flowers separately on the same tree. Walnuts are generally self-fruitful, cross-compatible and dichogamous, having incomplete overlap of pollen shed and female receptivity. It is this characteristic which led to the recommendation that about 10% of the trees in a commercial planting be a cultivar with a pollen shed period overlapping pistillate flower receptivity of the main cultivar. Excessive pollen load has been implicated in the ‘Serr’ cultivar in pistillate flower abortion (PFA), the loss of the female flowers early in the season before fruit drop due to lack of pollination. PFA can be reduced and yield improved in ‘Serr’ orchards by reducing pollen load. This can be accomplished by pollinizer removal, or catkin removal at the begin- ning of pollen shed by mechanical shaking. In years of significant bloom overlap between staminate and pistillate bloom, PFA can be further reduced and yield improved by removing ‘Serr’ catkins. PFA occurs to a lesser extent in other cultivars such as ‘Chico’, ‘Chandler’, ‘Vina’ and ‘Howard’. This information has led to the reevaluation of pollinizer recommendations. Research focused on optimum pollinizer levels in ‘Chandler’, a cultivar of increasing importance to the California walnut industry, has been inconclusive. Lack of pollinizers may impact yields to a greater extent in the in the northern San Joaquin Valley and Sacramento Valley than in the southern San Joaquin Valley. In any case the previously recommended 10% appears to be excessive. Two to three percent is probably adequate to limit losses due to lack of pollination without resulting in excessive PFA, and is currently being recommended by extension farm advisors and specialists. Factors to consider when determining the number of pollinators to plant include: cultivar susceptibility to PFA, walnut pollen load in the area and local pollination and fruit set experiences.

alifornia accounts for 99% of the U.S. English walnut production producing 220,000 tons (199,600 t) of inshell walnuts on C193,000 bearing acres (78,106 ha) in 1998 (California Agri- cultural Statistics Service, 1998). About 90% of this production is located in the Central Valley (comprised of the Sacramento Valley to the north and the San Joaquin Valley to the south). Of this, about 50% is located in the San Joaquin Valley and 40% is located in the Sacramento Valley. Walnuts are monoecious, (Polito, 1996) bearing staminate and pistillate flowers separately on the same tree. They are generally self fruitful and cross compatible. Typically the period of pollen shedding does not completely overlap the period of female receptivity. This is known as dichogamy. It is this characteristic which resulted in the recommendation that pollinizers with a pollen shedding period overlapping the pistillate flower receptivity of the main cultivar be included in commercial plantings (Forde and Griggs, 1975).

University of California Cooperative Extension, P.O. Box 697, Orland, CA 95963. The cost of publishing this paper was defrayed in part by the payment of page charges. Under postal regulation, this paper therefore must be hereby marked advertisement solely to indicate this fact.

● January–March 2000 10(1) • POLLINATION WORKSHOP PROCEEDINGS 127 When pollen shed occurs before basal portion of the pistil is the ovary, tion to fertilization takes about 1 week. female flower receptivity they are said which eventually produces the nut. The Numerous pollen grains germinate and to be protandrous. This is the most pistil contains a relatively large two pollen tubes grow into the style, but common situation for California culti- branched stigma which is connected only one can penetrate and fertilize the vars. When female receptivity occurs to the ovary by a short style. egg. Unpollinated flowers will con- before pollen shed, the cultivars are Catkins begin to elongate in the tinue to grow for several weeks before said to be protogynous. Figure 1 shows spring. As they elongate the pollen is they drop from the tree. the relationship between pollen shed- released and dispersed by the wind. As ding period and peak pistillate bloom the pistillate flower develops, the stig- Pistillate flower abortion for the most common California culti- mas begin to separate. Cells on the Pistillate flower abortion is the vars (Hendricks et al., 1996). stigma surface secrete a sticky exudate loss of pistillate flowers early in the Staminate flowers develop on the which catches the pollen grains and season, about 2 to 3 weeks after bloom previous season’s growth and are small, acts as a substrate for germination. when the flowers have reached a size of inconspicuous and grouped together Pistillate flowers are receptive for up to 3/16 inch (4 to 5 mm). PFA flowers in hanging clusters known as catkins. 7 d under ideal conditions. Once the emerge and grow normally in the The individual flowers lack petals and stigma lobes open to about 45° they spring. After the stigmas expand, consist of a whorl of green sepals sur- stop secreting exudate and are no growth stops and they soon become rounding 40 pollen bearing stamens. longer receptive. When the pollen grain necrotic and drop. PFA occurs 2 to 3 Each stamen terminates in an anther reaches the stigma (pollination), it ger- weeks before flower drop due to lack which has numerous pollen grains. minates and the pollen tube penetrates of pollination. Losses from PFA can be Pistillate flowers are formed most between the cells on the stigma sur- particularly high in cultivars such as often in pairs and develop from the face, grows through the style into the ‘Serr’ where losses have reached 90% terminal bud on all cultivars (termi- ovary, into the ovule and to the egg (Polito, 1996.) nally fruitful) and from lateral buds on sac. It then releases two sperm. One Shortly after the first commercial certain cultivars (laterally fruitful). Pis- fuses with the egg effecting fertiliza- plantings of ‘Serr’ came into produc- tillate flowers begin to differentiate in tion and forming a zygote. The second tion in the mid 1970s, severe drop of the late spring or early summer of the sperm fuses with a two-celled nuclei in year before bloom. The final stages of the egg sac to form the endosperm floral bud initiation occur in the weeks which provides nutrition for the devel- Fig. 1. Relationship of pollen- before bloom. The basal portion of the oping embryo. It is consumed by the shedding period to time of peak flower is enclosed in a hairy involucre time the nut is mature. The zygote pistillate bloom. These data describe walnut trees growing in an orchard which is formed by the fusion of flower divides 7 to 10 d after fertilization to at the University of California, Davis. parts including bracts and sepals (Fig. begin the period of development which The number of days cited is an 2). This tissue eventually develops into ultimately results in the development average based on findings for 5 years the husk of the walnut. The enlarged of the embryo (the kernel). Pollina- (1986–90) (Hendricks et al., 1996).

128 ● January–March 2000 10(1) year, but was al- the tree. At this time pollen shed has ways greatest next begun, but most of the catkins are to the pollinizer about half elongated, have not begun row and declined to shed pollen and can be removed with distance from the tree without excessive force. from the Up to 80% of the catkins can be re- pollinizer, usually moved at this time by shaking. In the stabilizing 250 to case of an adjacent block of pollinizers, 300 ft (76 to 91 all trees within 150 ft (46 m) of the m) from the ‘‘Serrs’ should be shaken. pollinizer. Some- PFA IN OTHER CULTIVARS. PFA due times PFA was as to excess pollen load is known to oc- high as 80% adja- cur, although to a lesser degree, in cent to the cultivars other than ‘Serr’. (Catlin and pollinizer. Olsson, 1990). Fig 3 shows PFA rela- PFA reduces tive to distance from the pollinizer in yield. In ‘Serr’ or- representative orchards for ‘Serr’, chards yield was ‘Vina’, ‘Chandler’ and ‘Chico’ (Polito consistently lower et al., 1996a). PFA for ‘Chandler’ and Fig. 2. Longitudinal section of a next to the pollinizer row and in- ‘Vina’ adjacent to pollinizers typically pistillate flower (Polito, 1996). creased with distance from the was about 40% with average PFA for pollinizer. the orchards between 15 and 20%. the pistillate flowers shortly after bloom REDUCING POLLEN LOAD REDUCES ‘Chico’ typically showed PFA levels was noted (Polito, 1996). Research by PFA AND INCREASES YIELD. In experi- near the pollinizer about 20 to 25% University of California (UC) during ments where pollen load was reduced, with average PFA for the blocks being the 1980s ruled out pests and diseases, either by pollinizer removal or by shak- around 10%. Results from experiments nutrient deficiencies, tree age, shad- ing catkins from the pollinizer trees conducted with ‘Chandler’ cultivar ing, pruning practices, and incompat- before pollen dispersal, PFA was sig- designed to determine what levels of ible pollen or lack of pollen as possible nificantly reduced and yield was sig- pollinizers would result in adequate causes (Catlin et al., 1987). During nificantly increased compared to areas pollination without resulting in exces- this time period, it was observed that in the same orchards where catkins or sive PFA were variable and inconclu- PFA varied among sites from year to pollinizers were not removed. sive (Polito, et al., 1996b). In Tulare year and was not limited to the ‘Serr’ In experiments where ‘Tehama’ County in the southern San Joaquin cultivar. In the late 1980s and early pollinizers were removed, PFA was Valley, orchards without pollinizers 1990s, UC researchers began to inves- reduced and yield was improved by 20 did have a slightly reduced set due to tigate the role of excessive pollen load to 86% compared to where pollinizers lack of pollination, but losses were not in PFA. Early experiments with flow- were not removed. Response would as great as losses due to PFA in or- ers that received high doses of pollen be expected to be greatest where PFA chards with pollinizers. Pollinizer re- showed that PFA was correlated with has been chronic and or where high moval was correlated to reduced PFA pollen load. Walnut cultivars varied in percentages of pollinizers had been and increased yield. In the northern their response to high pollen load and used. San Joaquin Valley and the Sacramento PFA was unrelated to pollen source Pollen load can be reduced and Valley, the results were less clear. In (Catlin et al., 1990; McGranahan et yield improved by mechanically re- San Joaquin County in the northern al., 1991). moving catkins from pollinizer trees San Joaquin Valley no correlation was Research conducted by Polito et using the same shaker harvester used found between distance from the al. (1996a) during the 1990s on ‘Serr’ for walnut harvest. In years of signifi- pollinizer and PFA and no significant has shown that: PFA is positively cor- cant overlap of staminate and pistillate difference in yield resulted from pollen related with pollen load. Counts of flowers, yields can be increased further removal. In a Yolo County orchard in pollen grains on pistillate flowers show by shaking the catkins from the ‘Serr’ the southern Sacramento Valley, no that high PFA is always associated with trees. In experiments where catkins correlation between distance from the high numbers of pollen grains. Dose were removed from ‘Tehama’ pollinizer and PFA was measured. Post response curves indicated that 70 to pollinizers in ‘Serr’ orchards, PFA was PFA drop, most likely due to lack of 100 pollen grains per pistillate flower reduced and yields were increased 16 pollination was reduced by the pres- would induce PFA in 50 % of the to 26% compared to nonremoval of ence of pollinizers. In three other or- flowers. This relationship was consis- catkins. In years of extensive overlap chards in the Sacramento Valley, in tent for orchards from different parts between pollen shedding and pistillate Sutter and Butte Counties, PFA was of California. bloom, catkin removal from both ‘Serr’ not reduced by removing catkins from PFA is positively correlated with and ‘Tehama’ reduced PFA and im- pollinizers or with distance from the distance to a pollen source. The proved yields. This increased yield by pollinizers. In one Butte County or- ‘Tehama’ cultivar has been widely used 16% compared to an area where cat- chard, slight increases in yield were as a pollinizer for ‘Serr’ in California. kins were removed only from ‘Tehama’. detected with increasing distances from The percentage of PFA varied from For best results, catkins should be pollen sources, indicating that the orchard to orchard and from year to shaken when the first catkins fall from method of analysis may have been

● January–March 2000 10(1) • POLLINATION WORKSHOP PROCEEDINGS 129 insufficient to provide a complete pic- of determining if pollen is infected. tance from the pollinizer row, but ture of PFA throughout the orchard Therefore, the practice is of question- drop due to lack of pollination was canopy. This may have underestimated able value and is not recommended. sometimes noted. PFA in ‘Chandler’ orchards. Pollen is required for crop set, but In general, recommendations for These results make it difficult to the necessity and density of pollinizers pollinizers have been reduced in recent generalize regarding the effect of is unclear. The long standing recom- years and are currently between 2 and pollinizers on PFA in ‘Chandler’. mendation had been that about 10% of 3% pollinizers depending on the factors However, there does appear to be the trees in a planting be of a cultivar listed above. greater need for pollinizers in the north- with a good overlap of the staminate ern San Joaquin and Sacramento Val- bloom with the pistillate bloom of the Literature cited leys than in the southern San Joaquin main cultivar (Forde and Griggs 1975.) Valley. The need for pollinizers is prob- For ease of harvest, the pollinizers California Agricultural Statistics Service. ably greater in areas of limited walnut were usually placed in complete rows 1998. California walnut acreage survey. plantings. starting with the first row on the up- Calif. Dept. Food Agr., Sacramento. wind side of the orchard. This level of Catlin, P.B. , D.E. Ramos, G.S. Sibbett, Current pollinizer pollinizers is almost certainly more and W.H. Olson. 1987. Pistillate flower recommendations than required to set the crop and may abscission of the persian walnut. Hort- Science 22:201–295 In the past some growers have contribute to crop loss from PFA. supplemented natural pollination with Factors that should be considered Catlin, P.B. and E.A. Olsson. 1990. Pistil- aerial application of supplemental pol- when considering pollination require- late flower abcission of walnut-‘Serr’, len. Pollen infected with the walnut ments for a new orchard follow. ‘Sunland’, ‘Howard’ and ‘Chandler’. Hort- strain of the cherry leaf roll virus has COINCIDENCE BETWEEN THE STAMI- Science 25:1391–1392. been shown to spread walnut blackline NATE AND PISTILLATE BLOOM ON THE CUL- Catlin, P.B., G. McGranahan and D. disease, a serious disease of english TIVAR BEING PLANTED. If all but earliest Voyatzis. 1990. Pistillate flower abscis- walnut in California which limits pro- or latest pistillate flowers coincide with sion. Walnut Mktg. Board Res. Rpt. Calif. duction in some areas and threatens to pollen release, it is unlikely that a Walnut Comm., Sacramento. pollinizer is necessary. increase in incidence in others Forde, H.I., and W.H. Griggs. 1975. Pol- CULTIVAR SUSCEPTIBILITY TO PFA. (Mircetich et al., 1996.) The virus lination and blooming habits of walnuts. enters the tree through infected pollen PFA susceptibility varies with cultivars. Univ. Calif. Div. Agr. Sci. Lflt. 2753. and travels slowly through the english For a cultivar such as ‘Serr’, which is scion, a symptomless carrier. Most highly susceptible, the risk of crop loss Hendricks, L.C., W.W. Coates, R.B. Elkins, due to PFA is probably greater than the G.H. McGranahan, H.A. Phillips, D.E. english walnuts in California are grown Ramos, W.O. Reil, and R.G. Snyder. 1996. either on northern california black potential loss due to lack of pollination. With a less susceptible cultivar, the pros Selection of varieties, p. 84–89. In: D.E. (Juglans hindsii Jepson) or Paradox Ramos (ed.). Walnut production manual. hybrid (Juglens regia x Juglans hindsii) and cons of a small number of pollinizers Univ. Calif. Div. Agr. Natural Resources rootstock. When the virus arrives at should be considered. Publ. 3373. the graft union, the first cells are killed, WALNUT DENSITY IN THE AREA. Wal- nut pollen is wind borne. In areas where Mircetich, S.M., A. Rowhani, E.L. Civerelo, forming a black line, girdling the tree and D.E. Ramos. 1996. Blackline disease, and leading to decline and eventual walnut plantings are common, there may be large amounts of pollen in the air p. 233-241. In: D.E. Ramos (ed.). Walnut tree death. Trees may be infected for production manual. Univ. Calif. Div. Agr. many years before the symptoms ap- during walnut bloom. This may be ad- Natural Resources Publ. 3373. pear, and there are no practical means equate or may require the addition of a small number of additional pollinizers McGranahan G., P. Catlin, V. Polito, S. to insure adequate pollination. Sibbett, K. Kelly, L. Hendricks, J. Grant, J. Fig. 3. Pistillate flower abortion Luza, and R. Synder. 1991. Pollen and (PFA) in representative orchards of Local experience with pistillate flower abscission, p. 65-71. In: ‘Serr’, ‘Vina’, ‘Chandler’, and ‘Chico’ 1991 Walnut Mktg. Board Res. Rpt. as a function of distance from the walnut pollination pollinizer (in parentheses) (Polito et As has been noted, the need for Polito, V.S. 1996. Floral biology: Struc- al., 1996a); 10 ft = 3.0 m. pollinizers may vary ture, development and pollination, depending on loca- p.-127-132. In: D.E. Ramos (ed.). Walnut tion. In the southern production manual. Univ. Calif. Div. Agr. Natural Resources Publ. 3373. San Joaquin Valley, ‘Chandler’ losses due Polito, V.S., G.S. Sibbett, J.A. Grant, K.M. to PFA were gener- Kelley, and P.B. Catlin. 1996a. Pistillate ally greater than were flower abortion and pollination manage- losses due to lack of ment, p. 133-138. In: D.E. Ramos (ed.). pollination. This gen- Walnut production manual. Univ. Calif. Div. Agr. Natural Resources Publ. 3373. erally was not the case in the northern San Polito, V.S., B. Coates, J. Grant, J. Hasey, Joaquin and Sacra- W. Micke, B. Olson, and K. Pinney. 1996b. mento Valleys, where Pollen, pistillate flower abortion/abscis- it was difficult to cor- sion, p. 77-87. In: 1996 Walnut Mktg. relate PFA with dis- Board Res. Rpt.

130 ● January–March 2000 10(1)