Agroforestry Systems 29: 321-331, 1995. 9 1995 Kluwer Academic Publishers. Printed in the Netherlands. Black walnut curculio: patterns of nut damage in a plantation environment* M. J. LINIT and S. NECIBI Entomology Department, University of Missouri, Columbia, MO 65211, USA Key words: alleycropping, black walnut curculio, Conotrachelus retentus, eastern black walnut, Juglans nigra, nut mortality Abstract. Data on curculio-caused nut mortality have been collected since 1985 to assess the impact of this nut Predator on black walnut trees grown in a plantation environment. These data have been analyzed to investigate spatial and temporal patterns of curculio damage. Curculio damage was independent of nut cluster height or cardinal direction within the crowns of young, nut-bearing black walnut trees, and was greater in parts of a walnut planting that bordered a stand of native hardwoods than in a part remote from the stand. Nut losses caused by the curculio were consistently higher in an upland versus a bottomland planting, although the differences were not significant during most years of the study. Nut mortality caused by the curculio differed among years of the study and was negatively correlated with annual nut abundance. Introduction Agroforestry systems, including alleycropping, are an environmentally rational alternative to conventional farm practices. Eastern black walnut, Juglans nigra L., is especially well suited for use in agroforestry systems because of its high value wood and its nut crop. In an agroforestry system, nut production has the potential to provide income on an annual basis. Excellent markets exist for the fruit of eastern black walnut, however, strong year-to-year variation in the availability has prevented full development of market potential. It is estimated that the current market would accommodate in excess of 91 million kilograms annually. Unfortunately, with most of the present nut crop coming from wild trees, availability is often as low as a few million kilograms. Black walnut planted under agroforestry management has the potential to produce a large, stable nut crop which would enable the expansion of the current markets and development of new ones [Garrett et al., 1991 ]. The nut production potential of 13,000 trees grown under agroforestry management is currently under investigation on the Sho-Neff Plantation owned and operated by Hammons Products Company, Stockton, Missouri. The study of insect impact on nut production has been a part of this program since 1985. To date, we have concentrated on gathering data to assess the impact of the black walnut curculio, Conotrachelus retentus (Say), on nut survivorship. * Contribution from the Missouri Agricultural Experiment Station, Journal Series No 12, 156.
322 The curculio overwinters as an adult in the litter beneath black walnut trees [Blair, 1978]. Adults emerge from the soil in April and mate during the next few weeks. Oviposition begins shortly after pistillate flower fertilization and continues through June. Each egg is deposited in a developing nut through a sickle-shaped wound. A larva hatches from the egg and feeds within the nut. This disrupts normal nut development and results in 'June nut drop'. The larva crawls out of the nut to pupate in the soil. Newly formed adults emerge from the soil in mid-july through August, feed on the walnut foliage until leaf drop occurs, then seek overwintering sites. A life table approach has been used to follow and evaluate nut survivorship from pistillate flower formation through nut maturity. This approach has proven to be effective and we now have a data set containing several years of nut survivorship data. From this data set we can quantify nut survivorship and partition the mortality attributable to the curculio on plantation-grown black walnut trees. Additionally, we can determine temporal and spatial patterns of nut mortality. Long term monitoring of cu?culio-caused nut loss is necessary to determine the bounds of nut loss impact. Additionally, several consecutive years of nut mortality data are necessary to evaluate the relationship between nut crop abundance and curculio impact. Finally, these data document the establishment of the black walnut curculio as the plantation matures and the annual nut crop increases. The goal of our research project is to quantify the impact of the black walnut curculio on nut production in a plantation environment. We are also interested in understanding temporal and spatial patterns of curculio-caused nut losses. Materials and methods Nut damage caused by the black walnut curculio was monitored annually at the Sho-Neff Plantation near Stockton, Missouri from 1985 through 1992. The Sho-Neff Plantation consists of numerous black walnut plantings designated as 'areas'. Trees within each area were planted in north-south rows at 3.1 or 6.2 m spacing within rows and 12.3 m alleys between rows. Annual crops such as soybean, milo and wheat were grown in the alleys during plantation establishment. Alleys were switched to perennial ~ grass production at about year 10 when tree crown expansion reduced solar radiation needed for crop growth. Most work was conducted in two plantings: one on an upland site (Area 16), the other on a bottomland site (Area 2). During 1985 and 1986, a second upland planting (Area 23) also was utilized for study. Trees in all three areas were planted during 1975 and 1976. Commercial harvest of nuts started in 1985 on the upland sites and in 1986 on the bottomland site. Alleys were switched from crop production to grass production about the same time.
323 During 1985 and 1986, the within-tree spatial dispersion of curculio-caused nut loss and the temporal pattern of flower and nut loss through the growing season were monitored. Intensive monitoring of curculio-caused nut loss was begun in 1987. All statistical analyses were conducted using the Statistical Analysis System [1985]. Temporal pattern of flower and nut mortality In 1985, twelve trees on the two upland sites were selected to determine the spatial dispersion of curculio-caused nut loss within the crowns of black walnut trees. The crown of each tree was divided horizontally into upper and lower halves, and each half was divided vertically into four sections. The octants were high southwest, high southeast, high northeast, high northwest, low southwest, low southeast, low northeast and low northwest. One cluster of pistillate flowers was chosen at random from each octant 1 week after appearance of pistillate flowers (9 May) and was checked weekly through 6 August, and then every 2 weeks until the nuts were mature (5 September). On each sample date, the number of nuts in each cluster was recorded along with evidence of curculio oviposition. Nut were considered mature in mid- September. A nut was considered destroyed by the curculio if it received an oviposition puncture and subsequently aborted. In 1986, 15 trees, including trees in the bottomland planting, were selected for use in the nut mortality study. Cluster selection was modified based upon analysis of 1985 data. Four clusters within reach from the ground (3 m) were chosen at random from each tree. Pistillate flowers appeared on 6 May and were checked weekly through 15 July, and then every 2 weeks until the nuts reached maturity. Data from both years were used to follow nut survivorship from the pistillate flower Stage through nut maturity. Nut mortality was partitioned through time to identify periods of flower/nut loss and to quantify the impact of the curculio on nut loss. Spatial patterns of curculio-caused nut mortality The spatial pattern of curculio-caused nut mortality within the crowns of black walnut trees was analyzed using the 1985 data described above. Analysis of variance was used to test the differences in curculio-caused nut mortality between heights and among directions within the crowns of study trees. The spatial dispersion of curculio-caused nut mortality across the landscape was investigated within one of the upland plantings (Area 16). Nut clusters on trees grouped in different parts of the planting were evaluated for curculio impact. Three sections of the planting were sampled; two sections (B and C) were adjacent to a stand of native hardwoods that contained wild black walnut trees in a mixture with other species (Fig. 1). The third section (A) was not adjacent to any hardwoods. This section was adjacent to a road and another
324 N Fig. 1. Diagrammatic representation of the upland planting (Area 16) at the Sho-Neff Plantation, Stockton, Missouri. Nut clusters evaluated for the occurrence of black walnut curculio oviposition were selected from trees in three sections of the planting. Section A was not adjacent to native trees, while sections B and C were adjacent to hardwood stands that contained native black walnut trees. The planting is bordered on the south by a road and another walnut planting; on the north and west by native hardwoods (shaded area). The pattern in the middle of the planting is a representation of the spacing and orientation of black walnut trees within the planting. walnut planting on the south and an open lawn-like area on the east. In 1991, 520 nut clusters were observed, while 300 clusters were observed during 1992. Each year, the number of curculio oviposition sites per nut cluster was determined. Analysis of variance was used to determine if the number of oviposition sites per cluster differed among the three sections within this planting. Curculio-caused nut mortality: 1987-92 Each year since 1987, nut clusters in the upland (Area 16) and bottomland (Area 2) sites were monitored to estimate annual curculio-caused nut ~ losses. In 1987, five pistillate flower clusters were selected randomly on 30 trees in the bottomland planting and 40 trees on the upland site. In 1988 and 1989, five clusters were selected on 50 trees in both plantings. Pistillate flower production was poor in 1990, making cluster selection within each planting difficult. A total of 151 and 136 clusters were selected for study in Areas 2 and
325 16, respectively. The number of clusters selected per tree varied. In 1991, 250 nut clusters in each planting were monitored. Five nut clusters were selected randomly from each of 50 trees in Area 16, 10 nut clusters on each of 25 trees in Area 2. In 1992, 10 nut clusters were selected randomly from each of 25 trees on both sites. All nut clusters sampled since 1987 were located within six meters of the ground. Clusters were observed on a 10 or 11 day interval from pistillate flower formation through the end of curculio-caused nut drop in mid-july. This interval, based upon the timing of curculio oviposition and subsequent nut abortion observed during the 1985 and 1986 seasons, allowed sampling at a regular interval without missing evidence of curculio activity [Schreiber and Linit, 1991]. The final nut survey was made at the time of nut maturity. The number of fruits (flowers or nuts) in each cluster was recorded on each sample date. The impact of the black walnut curculio on nut mortality was calculated for each tree by dividing the number of curculio-destroyed nuts in each cluster by the number of nuts present at the time that curculio oviposition activity began. A nut was considered curculio-destroyed if an oviposition site was found and the nut subsequently aborted. Determination of the nut crop present at the time of curculio oviposition initiation was difficult because of the between-tree variation in fruit phenology. The nut crop present at the time of first oviposition at each planting area was used to calculate curculio-caused nut loss for each year. Some clusters had lost all their fruits by this time and were dropped from the analysis. T-tests were used to compare curculio-caused nut losses among sites for each year of the study. Analysis of variance was used to examine differences in curculio-caused nut losses among years. Annual nut abundance data were collected in Area 16 and Area 2 by the Hammons Products Co. (J. Jones, pers. comm.). These values represent the number of nuts that reached maturity and were harvested at the end of the growing season. The total number of nuts per area was divided by the total number of trees per area and related to the percentage of nuts destroyed by the black walnut curculio. Correlation analysis was conducted to determine if the pattern of curculio-caused nut loss among years was related to annual nut abundance. It was assumed that the harvest data were an accurate reflection of the relative abundance of nuts available for oviposition by the curcuiio. Results and discussion Temporai pattern of flower and nut mortality The 1985 nut cohort initially contained 219 pistillate flowers. Pistillate flower abscission was heavy during the first three weeks following pistillate flower formation (47.4%). Nut mortality was lower during the next three weeks; total
326 nut loss for this period was 17.4%. Mortality totaled 12.4% for the remaining weeks of nut development (Fig. 2). The pattern of overall nut mortality in 1986 was similar to that observed in 1985. The 1986 cohort began with 428 flowers. Pistillate flower abscision was high during the first three weeks (54.7%), and was particularly severe during week 1 (39%). Mortality was less severe during the next three weeks (7.7%). Mortality for the remaining weeks, 4.7%, was low. In 1985, black walnut curculio oviposition was first noted on 9 May and last recorded on 6 June. Sixty-seven nuts had oviposition punctures. Fiftyseven of these aborted due to black walnut curculio larval development within the nut. Peak black walnut curculio-caused nut loss occurred during weeks four, five, and six. During that time, 67% of the infested nuts aborted. In 1986, black walnut curculio oviposition was first noted on 7 May and last recorded on 3 June. Forty-three nuts had oviposition punctures. Sixteen of these failed to reach maturity. Peak black walnut curculio-caused nut loss occurred during weeks five, six, and seven. During both years nut mortality occurred in three distinct phases (Fig. 2). Heavy loss of pistillate flowers occurred during the weeks immediately following flower formation. Pistillate flower abscission could be caused by poor fertilization or other physiological processes. We found no evidence of insect involvement. A period of moderate nut loss followed, caused primarily by "7 o n- O o Z O V-- 8 LU z LU (.9 ILl o > b- 5 o O I I I I I I I I I I 1985 Curculio- I Late Season Nut ModaSty Caused ] Pistillate Nut Lo s ~ ~ 1986 Flower I ~ - ----"---'---- Abscission /.1/1/"-/I [ - - i _~ 1985 ]... s-'"... 2~J-" j JJ'TYJ~"............ 1986 I ' I I I I I I I I 0 2 4 6 8 10 12 14 16 18 WEEKS AFTER FLOWERING Fig. 2. Cumulative pistillate flower and nut mortality through the 1985 and 1986 seasons; Sho-Neff Plantation, Stockton, Missouri. Overall mortality (all causes) is represented by solid lines; curculio-caused mortality by dashed lines. The vertical lines partition mortality into three phases.
327 the oviposition and feeding of the curculio. Little nut loss occurred during the last half of the growing season. Similar mortality patterns have been reported in pecans, Carya illinoensis (Wang) K. Koch [Woodroof et al., 1928; Sparks and Health, 1972], and in English walnuts, J. regia L. [Catlin and Ramos, 1985]. Spatial patterns of curculio-caused nut mortality The within-tree distribution of nut losses was examined for three different time periods in 1985 corresponding to the three distinct phases of nut mortality described above. At the end of week 3, nut mortality was independent of cluster height (F = 1.67; df = 1,83; p = 0.20) and direction (F = 0.32; df= 3,83; p = 0.81). Similar relationships were found at week six, the period immediately following curculio oviposition (F -- 0.41; df= 1,83; p = 0.52 and F = 1,11; df-- 3,83; p = 0.35 respectively). It appears, therefore, that early mortality factors, including the curculio, affected nuts in all sections of the crown in the same manner. At the end of the growing season, nut mortality was still independent of cluster height (F = 0.82; df= 1,83; p -- 0.37), but there was a significant difference in nut mortality among directions (F = 2.90; df = 3,83; p --- 0.03). Significantly more nuts had aborted from the southwest octants than from the northeast octants. The prevailing wind direction is from the southwest [Anonymous, 1977], thus, nuts on that side of the tree may have been subjected to detrimental effects of wind, leading to early nut drop. Many of these nuts, however, were probably fully mature when dropped. Significant differences were found in the mean number of curculio oviposition sites per nut cluster between sections of Area 16 in both 1991 (F = 3.73; df= 2,517; p -- 0.025) and 1992 (F = 16.49; df= 2,297; p = 0.0001). In the first year of the study, curculio oviposition was significantly lower on nut clusters in the nonadjacent section (A) than in one of the adjacent sections but not the other (Table 1). The following year, curculio oviposition was lower Table 1. Mean number of black walnut curculio oviposition scars per nut cluster on trees adjacent to and nonadjacent to native hardwood stands in the upland planting (Area 16) at the Sho-Neff Plantation during 1991 and 1992. Year Section Type Mean a SD 1991 A Nonadjacent 0.69 a 1.23 B Adjacent 0.69 ab 0.92 C Adjacent 1.07 b 1.15 1992 A Nonadjacent 0.82 a 1.32 B Adjacent 2.20 c 2.04 C Adjacent 1.40 b 1.69 a Means within the same year followed by the same letter did not differ according to a test of least squares means test (p = 0.05).
328 in the nonadjacent section than in both of the adjacent sections. These data suggest that black walnut curculios move into the plantation from overwintering sites beneath the litter layer of hardwood stands. Trees in the proximity of the hardwood stand are more likely to receive curculio oviposition than those in more remote parts of the planting. We know little about the dispersal capabilities of this insect, however, closely related weevils such as the plum curculio, Conotrachelus nenuphar (Herbst), and the cotton boll weevil, Anthonomus grandis Boheman, are known to overwinter in areas adjacent to orchards or fields and move onto their host plant after emergence from overwintering sites [Slosser and Boring, 1980; Le Blanc et al., 1984]. Curculio-caused nut mortality: 1987-92 Curculio-caused nut losses were consistently higher at the upland site than the bottomland site. These differences, however, were only significant in 1989 and 1992 (Table 2). The bottomland planting was far from any native hardwood stands and was subject to moist soil conditions during the spring prior to curculio emergence from overwintering sites within the soil. These factors may have contributed to lower curculio-caused nut mortality values in the bottomland planting. Annual nut losses due to the black walnut curculio differed significantly among years (F = 57.95; df= 5,1715; p = 0.0001). Mean nut losses ranged from 1.9% of the available nut crop in 1988 to 32.0% in 1990 (Table 3). Kessler [1978] monitored trees in Wisconsin from bloom to nut maturity and estimated that 8 and 11% of all developing nuts were destroyed by black walnut curculio in 1975 and 1976 respectively. Blair and Kearby [1978] Table 2. Mean black walnut curculio-caused nut mortality (%) at the upland planting (Area 16) and the bottomland planting (Area 2); Sho-Neff Plantation, 1987 to 1992. Year Planting n Mean" SD p > t 1987 Upland 158 3.16 15.75 0.110 Bottomland 126 0.79 8.91 1988 Upland 204 1.96 12.98 0.993 Bottomland 194 1.89 11.72 1989 Upland 106 18.40 36.73 < 0.001 Bottomland 117 4.27 18.07 1990 Upland 136 34.19 40.13 0.340 Bottomland 149 29.98 37.57 1991 Upland 117 25.36 39.31 0.468 Bottomland 247 22.23 34.21 1992 Upland 97 14.95 33.98 0.012 Bottomland 70 4.29 20.40 a Analyses were conducted on arcsin transformed data; raw means are reported above.
329 Table 3. Mean black walnut curculio-caused nut mortality (%); Sho-Neff Plantation, 1987 to 1992. Nut losses for the upland planting (Area 16) and the bottomland planting (Area 2) were combined. Year n Mean a SD 1987 284 2.11 d 13.20 1988 398 1.93 d 12.40 1989 223 10.99 c 29.30 1990 285 31.99 a 38.80 1991 364 23.24 b 35.90 1992 167 10.48 c 29.47 a Means within the same year followed by the same letter did not differ according to a Student- Newman-Keuls multiple range test (p = 0.05). Analysis was conducted on arcsin transformed data; raw means are reported above. reported that black walnut curculio destroyed 51% of Missouri's 1977 nut crop. Their estimate was based on the nut cohort present in June, after pistillate flower abscission was complete. Houkal (pers. comm., 1984) tagged 498 pistillate flowers from mature trees. Ninety percent of these failed to reach maturity; an estimated 25% of the loss was due to black walnut curculio. Our estimation of black walnut curculio nut loss was similar to Houkal's and Blair and Kearby's (once differences in calculation of nut mortality are considered). Brooks [1922] stated that black walnut curculio infestations were more severe between the latitudes of 35 and 40 degrees, perhaps explaining the low losses reported by Kessler (Wisconsin lies between 42 and 47 degrees of latitude). Nut abundance in Area 16 ranged from 68.6 nuts per tree in 1988 to virtually 0 nuts per tree in 1990 and 1991. Area 2 nut abundance varied from 129.5 in 1989 to virtually 0 in 1990 and 1991 (Fig. 3). The percentage of nuts destroyed by black walnut curculio was negatively correlated with the mean number of nuts per tree in both the upland and bottomland sites during the period 1987 through 1991 (r = -0.94; p -- 0.01 and r = -0.87; p = 0.05, respectively). The percentage of nut loss caused by the curculio was low during years of high nut abundance and was high during years of low nut abundance. A classic predator-prey cycle appears if one assumes the percentage of nuts lost to curculio oviposition within a given season is correlated with the population density of the curculio. Following years of high nut abundance, the percentage of curculio-destroyed nuts rises, contributing the low abundance of nuts in subsequent years. This in turn limits the availability of curculio oviposition sites and contributes to the decline in curculio population density. Observation of these processes over additional seasons is necessary to determine if this pattern will continue. Brooks [1992] made similar observations for black walnut curculio in West Virginia. Harris et al. [1986] stated that damage by the pecan weevil, Curculio caryae (Horn), and
330 w LU n" I--- t~ w Q.. oo I- z d z 60 40 20 0 120 80 I 1 I I I I.,,...A- ~ Upland Planting // --\\ Area 16 ///~ " // "A [ Z~ Curculio-causedMortality ^../ \I 9 No. Nuts Per Tree I I I I I I I ~ Bottomland Planting // \X\\ 30 o c 20 O c I" 5 lo o > c 0 m z c.-i 30 ~ 0 2o ~ r'- 40 // \ \\... -a_... J ~_ "a 0,.o I I I I I I 1987 1988 1989 1990 1991 1992 YEAR Fig. 3. Mean number of nuts per tree and curculio-caused nut mortality, 1987 through 1992 at the Sho-Neff Plantation, Stockton, Missouri. pecan nut casebearer, Acrobasis nuxvorella Neuenzig, were more severe during years of low nut densities. Little is known about the functional and numerical responses of the black walnut curculio to changes in the within- and between-generation abundance of nuts. As black walnut plantations mature and nut abundance increases, it will be important to determine if the curculio can respond to increased host availability. Our data suggest that black walnut curculio has the potential to significantly impact nut production in managed black walnut plantations. This impact will increase as the value of plantation-grown nuts increase. Further studies on the biology of the black walnut curculio and its impact on nut production in managed plantations is necessary to formulate management strategies. Acknowledgments We thank the following individuals who have participated in this project since 1985: Alan Schreiber, Jay Pershing, Stewart Humphry, Jim Warren, Bart Plotter, Terrell Stamps, and Xu Zhang. This research was funded in part by the Missouri Department of Conservation.
331 References Anonymous (1977) Climatic Atlas of the United States. Dept Commerce Natl Ocean Atmos Admin, 89 pp Blair LM (1978) The bionornics of the black walnut curculio. MS thesis. University of Missouri- Columbia, 145 pp Blair LM and Kearby WC (1978) The black walnut curculio and its impact on nut production. In: Walnut Insects and Diseases, pp 51-54. USDA Forest Serv, Gen Tech Rep NC-52, 100 PP Brooks FE (1922) Curculio that attack the young fruits and shoots of walnut and hickory. USDA Bull 1066, 16 pp Catlin DB and Ramos DE (1985) Pistillate flower abscission. In: Ramos DE (ed) Walnut Orchard Management, pp 87-90. University of California, Davis, 178 pp Garrett HE, Jones JE, Haines J and Slusher JP (1991) Black walnut nut production under alleycropping management: an old but new cash crop for the farm community. Proceedings 2nd Conference on Agroforestry in North America, Springfield, MO, August 1991 Harris MK, Culter BL and Ring (1986) Pecan nut loss from pollination to harvest. J Econ Entomol 79:1653-1657 Kessler KJ Jr (1978) Premature loss of developing black walnut fruit. In: Walnut Insects and Disease, pp 1-4. USDA Forest Serv, Gen Tech Rep NC-52, 100 pp Le Blanc JPR, Hill SB and Paradis RO (1984) Oviposition in scout-apples by plum curculio, Conotrachelus nenuphar (Herbst) (Coleoptera: Curculioidae), and its relationship to subsequent damage. Environ Entomol 13:286-291 Schreiber AA and Linit MJ (1991) Patterns of nut mortality in a black walnut plantation with particular reference to insects. J Kansas Entomological Society 64:38-44 Slosser JE and Boring EP III (1980) Shelterhelts and boll weevils: a control strategy based on management of overwintering habitat. Environ Entomol 9:1-6 Sparks D and Heath JL (1972) Pistillate flower and fruit drop of pecan as a function of time and shoot length. Hort Sci 7:402-404 Statistical Analysis System Institute (1985) SAS User's Guide: Statistics. SAS Institute, Cary, NC, 956 pp Woodroof JG, Woodroof NC and Bailey JE (1928) Unfruitfulness of the pecan. Ga Expt Sta Bull, 148 pp