Plant materials.

Seven sources of WN were used for evaluation of resistance to Phytophthora and graft compatibility with Persian walnut (Table 1). The seeds were collected from trees in September after they matured and became dry. Seed lots of NCB and PH were obtained from commercial nurseries in the fall. All of the seeds were air-dried, soaked overnight in clean tap water, rinsed thoroughly, and stratified for 3 to 4 months at 4 °C in moist sphagnum peat. The stratified seeds were planted in the spring in Jiffy Pots (8-cm diameter, 8-cm height) (Jiffy Products of America Inc., Lorain, OH) each filled with 300 mL UC Mix (Matkin and Chandler, 1957) and grown in a greenhouse (air temperature 18 to 33 °C, photoperiod 16 h). The plants were watered daily with deionized water and fertilized weekly with Grow More fertilizer (11-7-26, with micronutrients; Gardena, CA) until they were transplanted as described subsequently.

Table 1.Half-sib seed family selections of Chinese wingnut (Pterocarya stenoptera) used in the study.

View Fullscreen

Inocula.

Inocula of P. cinnamomi and P. citricola were grown on a V8 juice–oat–vermiculite substrate for 4 to 6 weeks as described previously (Matheron and Mircetich, 1985a). Three isolates collected from diseased walnut trees in different regions of California were used for each species of Phytophthora to represent pathogenic diversity. The isolates were grown individually in separate jars of substrate and were mixed in equal proportions just before introduction to soil. Sterile V8 juice–oat–vermiculite substrate was used as a control. Before mixing with soil, the inocula containing a Phytophthora sp. as well as the control were rinsed in sterile deionized water to remove unassimilated nutrients.

Evaluations of resistance to Phytophthora.

Three experiments were conducted. Expt. 1 used seed lots collected in 1996, and Expts. 2 and 3 used seed lots collected in 1999. At 2 to 3 months of age, each seedling was transplanted into a 1-L pot of UC mix soil amended with 45 mL of sterile V8 juice–oat–vermiculite substrate (the control) or 45 mL of the substrate colonized by P. cinnamomi or P. citricola. In Expt. 1, five replicate plants were used per combination of soil inoculum treatment and seedling type (i.e., PH, NCB, and the multiple selections of WN). In Expts. 2 and 3, six replicate plants were used per treatment combination, except that there were 12 replicate plants for PH seedling.

A split plot design was used for all three experiments. The inoculum treatments were allocated to main plots, which were randomized in complete blocks. The subplots, consisting of the individual potted plants, were randomized within the main plots. Each seedling type was represented once in each main plot, except that two seedlings of PH (each in a separate pot) were included in main plots of Expts. 2 and 3.

To facilitate infection by the pathogens, soil flooding was administered 7 to 10 d after transplanting. Then and every 2 weeks thereafter, all pots in the control and inoculated treatments were placed in bowls, and cool tap water was added to the soil to maintain the water surface at ≈1 cm above the soil surface for 48 h. After each flooding, the pots were removed from the bowls and allowed to drain freely. Between floods, the soil was watered daily using deionized water.

Three months after transplanting, resistance to the pathogens was assessed according to the severity of root and crown rot and root fresh weights. The root systems were washed free from soil, weighed, and rated for the percentage of total root length and total crown length exhibiting necrosis. The percentage of root length rotted was estimated visually. Root length was considered necrotic when the cortex and stele tissues were black to dark brown and healthy when the same tissues were white. The percentage of crown length rotted was measured with a ruler. For rating purposes, the crown was considered to be the stem and central root system axis starting from the point of main root convergence and extending to 3 cm above the soil surface. On the crown, a knife was used to reveal boundaries of healthy bark (white under the outer corky layer) and necrotic bark (black to brown).

Isolations were conducted from the root systems of all plants to confirm that those grown in non-infested soil remained free from Phytophthora spp. and that those grown in soil infested with P. cinnamomi or P. citricola had been exposed to the appropriate pathogen. Ten 1-cm-long segments of roots ≈1 mm in diameter were removed from each root system, cultured in PARP medium (Kannwischer and Mitchell, 1978), incubated in the dark at ≈18 °C, and observed after 5 d for the presence of mycelium of a Phytophthora sp. Species of Phytophthora were identified according to their morphological characteristics (Erwin and Ribero, 1996).

Evaluations of graft compatibility.

Two field trials were conducted near Davis, CA, to examine graft compatibility of Persian walnut cultivars Chandler, Hartley, Serr, Tulare, and Vina on rootstocks of the seven WN selections and NCB and PH standards. In preparation for these trials, each scion–rootstock combination was propagated at two commercial walnut nurseries, one located near Oakdale, CA, (Nursery 1) and the other located near Yuba City, CA (Nursery 2). At each nursery, a split plot design was used with the rootstock selections allocated to main plots and scion cultivars allocated to subplots. At Nursery 1, there were four replicate main plots per rootstock (except for PH, which was propagated under identical conditions in adjacent rows) with a total of 16 to 20 trees (four to five per subplot) per rootstock–scion combination. At Nursery 2, there were five replicate main plots per rootstock with a total of 12 to 53 trees (two to 12 per subplot) per rootstock–scion combination. At each nursery, the main plots were randomized in complete blocks, and the subplots were randomized completely within main plots. At Nursery 1, trees were spaced 25 cm apart in the rows and 1.4 m apart between rows, whereas at Nursery 2, trees were 23 cm apart in the rows and 1.8 m apart between rows.

Briefly, propagation and grafting procedures at the nurseries were as follows: at Nursery 1, seeds collected in 1996 were stratified, grown into seedlings in 8-cm pots as described previously, and then transplanted into nursery plots in May 1997. At Nursery 2, seeds collected in 1997 were used. The PH and NCB seeds were planted directly in the nursery plots in Nov. 1997, whereas the WN selections, stratified and grown in pots as described previously, were transplanted into the nursery plots in May 1998. Grafting of the scion–rootstock combinations at Nursery 1 was first attempted in mid-Mar. 1998 using the whip graft method (Hartmann et al., 1997) at ≈35 cm above the soil line. Nearly all of the grafts on WN selections failed. The grafts were removed by cutting back to the rootstock. The ungrafted WN and NCB seedlings of similar size were transplanted to new plots in four randomized complete blocks at the same nursery in early Feb. 1999. There were five trees per rootstock–scion combination arranged in a split plot design as described for the first planting. Paradox hybrid seedlings were grown and grafted in adjacent rows under the same conditions. In mid-Apr. 1999, all of the seedlings were whip-grafted at ≈35 cm above the soil line with the same walnut cultivars as used previously (‘Chandler’, ‘Hartley’, ‘Serr’, ‘Tulare’, and ‘Vina’). Grafting of the scion–rootstock combinations at Nursery 2 also was completed in mid-Apr. 1999. The rootstocks were whip-grafted at ≈20 cm above the soil line. At both nurseries, the trees were dug during winter dormancy in early 2000 using conventional nursery equipment. Healthy grafted trees of each scion–rootstock combination were trimmed conventionally and stored with their roots covered in bins of moist sawdust. The trees were stored at 4 °C until transplanting to the orchard plots as described subsequently.

In Mar. 2000, the grafted rootstocks described previously were used to establish two graft compatibility trials in a field near Davis (long. 38°31.196′ N, lat. 121°45.579′ W). Expt. 1 included trees from Nursery 1, and Expt. 2 included trees from Nursery 2. The experiments were located adjacent to one another on Yolo loam and Yolo silty clay loam soil (fine-silty, mixed, superactive, nonacid, thermic Mollic Xerofluvents). The soil was not known to be infested with Phytophthora or phytopathogenic nematodes and did not have a history of walnut production. Each trial included five trees of each scion–rootstock combination in a randomized complete block design. The trees were spaced 4.6 m apart between and within rows. Scion and rootstock survival was tabulated 2 months after planting and at the end of the trial, ≈9 years after planting. Trunk circumference was measured at 30 cm above the graft union at the time of transplanting, every 1 to 2 years thereafter, and at the end of the trial. Scions that had died were assigned circumference values of zero. The intensity of shoot sprouting from the rootstocks was rated near the end of the experiments in May 2009. Until that time any sprouts had been trimmed back to the rootstock annually in February and in late summer. Rootstocks with no sprouts received a rating of 0, whereas those with one to five sprouts per tree, six to 10 sprouts per tree, and greater than 10 shoot sprouts per tree were assigned sprout ratings of 1, 2, and 3, respectively. Only rootstocks with living scions were rated.

Commercial orchard trial.

A third field trial was established in Apr. 2000 in a commercial walnut orchard near Linden, CA (long. 38°2.817′ N, lat. 120°59.818′ W) with a history of walnut tree death caused by P. cinnamomi. The soil was a Cogna loam (fine-silty, mixed, superactive, thermic Calcic Pachic Haploxerolls), and the trial was established in a rectangular area from which all walnut trees had been removed. The trial included only ‘Hartley’ grafted on the WN selections and PH, and the trees were produced at Nursery 2 along with those used in the graft compatibility trial at Davis. There were 15 to 20 trees per rootstock selection randomized in four complete blocks. Each block included a plot of three to five trees for each rootstock. Tree survival was monitored at 2 months after planting (June 2000) and at the end of the trial (July 2004). Tree circumference was measured at 1 and 3 years after planting, and dead scions were assigned trunk circumferences of zero. Trees that died and had symptoms of root and crown rot were subjected to diagnostic isolations as described previously (Matheron and Mircetich, 1985a).

Data analysis.

The MIXED procedure of SAS software Version 9.2 (SAS Institute Inc., Cary, NC) was used for analysis of variance (ANOVA) of the continuous variables, i.e., percentage of root length rotted, percentage of crown length rotted, increase in trunk circumference, and root fresh weight. The percentages were transformed to arcsine values before analysis. When there were multiple observations per plot, the plot means were the values used for ANOVA. Treatment factors of inoculum, rootstock, and scion were specified as fixed effects in the model statements of the MIXED procedure, whereas block was specified as a random effect in the random statements. In repeated experiments, the random statement specified block within the experiment. Additionally, in the greenhouse experiments, the random statement specified block × inoculum interaction. In cases in which treatment effects were significant (P ≤ 0.05) for continuous variables, 95% confidence intervals were either presented with the means (figure) or used to separate the means (tables). For clarity, non-transformed percentage means were used for presentation in tables. In the graft compatibility trials and the commercial orchard trial, Fisher's exact test (Agresti, 2007) was used to compare tree survival rates on the Paradox standard rootstock with survival rates on the other rootstocks; 2 × 2 contingency tables specifying counts of living and dead individuals for Paradox rootstock and each of the other rootstocks (in turn) were evaluated. Means and se of the means were calculated for the rootstock sprout severity scores in the graft compatibility trials.

Evaluations of resistance to Phytophthora.

In the greenhouse evaluations of resistance to P. cinnamomi and P. citricola, there was a significant experiment × inoculum × rootstock interaction (P < 0.0001, P = 0.009, and P < 0.0001 for percentage of root length rotted, percentage of crown length rotted, and root fresh weight, respectively). In Expt. 1, the NCB and PH rootstock standards were highly susceptible to P. cinnamomi and P. citricola and developed high levels of root and crown rot (68% to 100%; Table 2) and significant reductions in root fresh weight in soil with the pathogens (P = 0.05). In contrast, the WN selections were relatively resistant, sustaining negligible root and crown rot in soil infested with P. citricola and relatively low levels of root and crown rot in soil infested with P. cinnamomi. Neither pathogen significantly reduced root fresh weights of any WN selection.

Table 2.Comparative susceptibility of open-pollinated seedlings from seven sources of Chinese wingnut (Pterocarya stenoptera) and standard walnut rootstock seedlings to Phytophthora cinnamomi and P. citricola in three greenhouse experiments.^z

View Fullscreen

In Expts. 2 and 3, results were similar to those of Expt. 1, except PH exhibited only low to moderate levels of root and crown rot in soil infested with P. citricola (11% to 53%; Table 2) and no root fresh weight reduction with either species of Phytophthora. The paradox hybrid still was significantly more susceptible than WN selections to both Phytophthora spp., except for percentage of crown length rotted in Expt. 3. All the selections of WN developed low levels of root or crown rot that were not significantly different from those of the controls (Table 2). In Expts. 2 and 3, P. cinnamomi significantly reduced root fresh weights of ‘WN 1.01’ and ‘WN Minicenter’ seedlings, respectively. In Expt. 3, P. citricola significantly reduced the root weights of ‘WN Minicenter’. Root weights of the other WN selections were unaffected significantly by the pathogen.

In each trial, root isolations from the experimental plants confirmed that Phytophthora was absent on root systems grown in non-infested control plots and that the expected species of Phytophthora was present in plots infested with one of the pathogens. The pathogens were isolated from all selections of WN as well as from NCB and PH in treatments infested with Phytophthora.

Evaluations of graft compatibility.

Tree survival and rootstock sprout results were similar in Expts. 1 and 2 and therefore were combined for presentation (Table 3). Most of the scion–rootstock combinations initially survived in high proportion after transplanting into the Davis graft compatibility trials (Table 3). Exceptions were that only six of the 10 trees of ‘Chandler’ on WN ’10.01’ and six of the 10 trees of cultivar Hartley on WN ‘Minicenter’ rootstock survived initially. Final scion and rootstock survival rates varied greatly among the scion-rootstock combinations (Table 3). All scion cultivars survived and grew adequately on the NCB and PH standards. In contrast, high proportions of ‘Chandler’ scions ultimately failed on all of the WN selections. A relatively low proportion of ‘Serr’ trees ultimately survived on WN ‘2.08’, ‘11.02’, ‘17.01’, and ‘Dairy Road’ selections, but most ‘Serr’ scions survived on WN ‘1.01’, ‘10.01’, and ‘Minicenter’ selections. ‘Tulare’ and ‘Vina’ survived in high proportion across all WN selections (Table 3).

Table 3.Comparative performance of Persian walnut scion cultivars on rootstocks of Northern California black walnut (Juglans hindsii), Paradox hybrid (typically J. hindsii × J. regia), and seedlings from seven selections of Chinese wingnut (Pterocarya stenoptera).^z

View Fullscreen

The scion–rootstock combinations that had high proportions of final survival tended to produce satisfactory scion growth (Fig. 1A–B). Trunk circumference increases were affected by a significant scion × rootstock × experiment interaction (P = 0.01). All of the scion cultivars grew relatively well on the standard NCB and PH rootstocks, but ‘Chandler’ grew poorly on all WN selections. Including dead scions, which were assigned trunk circumference increases of zero, mean trunk circumference increases for ‘Chandler’ on WN selections ranged from 4% to 47% of those on NCB and PH rootstocks (Fig. 1A–B). Including only the surviving ‘Chandler’ scions, mean trunk circumference increases during the experiment for ‘Chandler’ on WN selections ranged from 14% to 85% of the increases on NCB and Paradox rootstocks. A few trees of ‘Chandler’ on WN (i.e., one on WN ‘11.02’, two on WN ‘17.01’, and one on WN ‘Minicenter’) generated increases in trunk circumference that fell within the 95% confidence intervals for mean growth of ‘Chandler’ on NCB and PH. Surviving scions of ‘Serr’ and ‘Hartley’ on WN selections generally fared better than those of ‘Chandler’. Mean increases in trunk circumference of surviving trees of ‘Serr’ and ‘Hartley’ on WN selections ranged from 51% to 137% of the corresponding increases on NCB and PH.

View Figure

• Download as Powerpoint
• Download Figure

Among the trees that survived the initial transplanting on WN but failed later, it was possible to distinguish three arbitrary classes. In one category, the scions died back either partly or completely, starting with distal portions of scaffolds or shoots. In many cases, living basal portions of the affected scions initiated new growth for 1 or more years in the spring and summer, but the dieback recurred annually until the scions ultimately died. This category predominated in failed ‘Chandler’ scions, occurring on all WN selections (incidence 38% to 67% of trees that survived transplanting), but it also occurred in ‘Hartley’ (11% on WN ‘Dairy Rd.’), ‘Serr’ (12% and 22% on WN ‘10.01’ and ‘2.08’, respectively), ‘Tulare’ (10% on ‘10.01’, ‘17.01’, ‘Dairy Rd.’), and ‘Vina’ (11% on ‘10.01’, 10% on ‘17.01’). In a second category, tree failure involved sudden trunk breakage at the graft union. This situation did not occur with ‘Chandler’ but did so with ‘Hartley’ (incidence 22% on WN ‘17.01’), ‘Serr’ (10% to 40% among WN ‘1.01’, ‘11.2’, ‘17.01’, and ‘Dairy Rd.’), and ‘Vina’ (11% on ‘1.01’). A third category of tree failure involved death of both the scion and rootstock. Its incidence was greatest with ‘Chandler’ on WN (12% to 22% among WN ‘1.01’, ‘2.08’, ‘Dairy Rd.’), but it also occurred on ‘Hartley’ (10% on ‘2.08’ and ‘11.02’), ‘Serr’ (10% to 20% among ‘1.01’, ‘2.08’, ‘17.01’, and ‘Dairy Rd.’), ‘Tulare’ (10% on ‘17.01’ and ‘Dairy Rd.’), and ‘Vina’ (10% on ‘2.08’ and ‘Dairy Rd.’).

Among the trees that survived for the duration of the Davis trials, mean severity of rootstock shoot sprouts at the end of the trial period was affected by significant rootstock × scion interaction (P = 0.0001), but there was no significant experiment × rootstock × scion interaction (P = 0.16). Mean sprout severity scores were relatively low (i.e., 0.0 to 0.5) for NCB and PH rootstocks. Sprout severity scores were relatively high (i.e., 1.0 to 3.0) for most WN rootstocks grafted with ‘Chandler’, ‘Hartley’, and, in some cases, with ‘Serr’ (Table 3). The limited tree survival of ‘Serr’ on ‘17.01’ prevented estimation of its mean rootstock sprout score.

Commercial orchard trial.

Most trees of ‘Hartley’ grafted on the standard seedling PH rootstock survived the initial transplant and began growth in the commercial orchard known to be infested with P. cinnamomi, but few of the trees (two of 18) survived through the end of the trial (Table 4). In contrast, most trees of ‘Hartley’ grafted on WN rootstocks survived the initial transplanting and grew through the end of the trial. The mean increase in scion trunk circumference during the trial was significantly greater on all WN rootstocks except for WN ‘Dairy Road’ compared with the increase on PH (Table 4). The trees that died on PH exhibited symptoms of Phytophthora root and crown rot, and P. cinnamomi was isolated from them.

Table 4.Comparative performance of ‘Hartley’ walnut on seedling rootstocks of Paradox hybrid (typically Juglans hindsii × J. regia) and open-pollinated seed families of Chinese wingnut (Pterocarya stenoptera) in a commercial orchard infested with Phytophthora cinnamomi.^z

View Fullscreen