Transplanting Success of Balled-and-Burlapped Versus Bare-Root Trees in the Urban Landscape

Plant Material

Ten each of the following three species were harvested balled and burlapped (B&B) on October 23, 1997, from the Chenango gravelly loam of Schichtel’s Nursery, Springville, New York, United States: hackberry, American hophornbeam, and swamp white oak. In addition to the B&B trees, ten each of the same species were harvested BR on October 29, 1997. In spring 1998, hackberry and hophornbeam were harvested in the same manner and in the same numbers, but an extra ten B&B and ten BR swamp white oak were dug for use in the related experiment comparing effects of drought stress on B&B and BR trees. Due to weather and work scheduling factors, the spring B&B hackberry trees were harvested at different times ranging from March 16 to April 3, 1998. All spring B&B hophornbeam were harvested on March 11, 1998, and all spring B&B swamp white oak were harvested on March 6, 1998. Spring BR trees were harvested on dates as follows: hackberry, April 13, 1998; hophornbeam, April 12, 1998; swamp white oak, April 11, 1998. All hophornbeam in the study were dug from the same nursery block, as were all swamp white oak. Among hackberry, all but seven spring B&B trees were dug from the same nursery block. All trees were dormant when dug, with the exception of spring hophornbeam that, due to unusually warm weather, started to break bud just prior to harvest.

Fall BR trees were dug, dipped, and transported on October 29, 1997. B&B trees were also transported on October 29. Due to unusually warm weather, spring BR trees were dug four to six days before they were dipped and picked up on April 17, 1999. During the holding period before pickup, they were stored in a cool, shaded garage and watered every other day. Spring B&B trees were delivered to the Ithaca municipal nursery on April 13, 1998.

Trees of 40 mm (1.5 in.) caliper were dug according to ANSI Z60.1-1996 standards (American Association of Nurserymen 1996). B&B trees were dug with a Vermeer tree spade, and root balls were put into wire baskets lined with natural burlap. Root-ball diameters met or exceeded the standard of 51 cm (20 in.) for a 40-mm-caliper (1.5 in.) tree. Hophornbeam and hackberry root balls were 51 cm (20 in.) in diameter and swamp white oak root balls were 61 cm (24 in.) in diameter. BR trees were dug with a U-blade and exceeded the ANSI minimum root spread of 56 cm (22 in.) for 40-mm-caliper trees. All trees conformed to ANSI standards for appropriate height-to-caliper relationship.

Dipping Procedure for BR Trees

All BR trees were dipped in a hydrogel slurry at the nursery immediately prior to transport. Using a method modified from Haug (1996), approximately 15 oz (445 mL) of Soil Moist Fines™ (particles sized 700 μ or less) hydrogel was added for every 25 gal (95 L) of water, mixed in a 100-gal (380-L) plastic container, then left to hydrate for ten to fifteen minutes. Tree roots were dipped in the slurry and immediately slipped into large, pleated plastic bags. Bags were knotted around the trunk to hold in moisture, and trees were stacked gently in the bed of a dump truck. The truck bed was securely tarped for transport and upon arrival, BR trees were stored in a cool, shaded shed until planted two to five days later.

Planting Procedure

Seventy sites throughout the city of Ithaca were selected for pairing a total of 140 B&B and BR trees in tree lawns. At each site, one B&B and one BR tree of the same species and transplanting season were planted. B&B trees and their BR counterparts were planted in close proximity—usually, 4.5 to 6.0 m (15 to 20 ft)—thereby ensuring similar soil type and microclimate (Figure 1). Tree lawn width varied from 1.5 m (5 ft) in downtown areas to greater than 4.5 m (15 ft) in residential areas.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Paired B&B and bare-root Ostrya virginiana on city street representing typical spacing in this study. Tree closest to viewer is B&B; tree farther back is bare root.

Fall trees were planted October 31, 1997, and November 3 and 4, 1997. Spring trees were planted April 20 through 22, 1998. BR planting holes were dug with shovels; B&B holes were dug with a backhoe. Regardless of digging equipment, holes were comparable in width, and glazing of the sides of the planting hole was not observed. Turf was removed and holes were made at least as wide as the spread of the root system. Trees were planted so that the beginning of the root flare was just visible at grade. All trees were watered in immediately after planting and received 5 to 10 cm (2 to 4 in.) of wood-chip mulch in a 46-cm (18-in.) radius from the trunk. With only two exceptions, trees were not staked. Starting in the spring of 1998, all trees in the main study received 20-gal (75-L) Treegator^® drip irrigation bags that were filled at the discretion of the city of Ithaca forestry crew, usually once a week during dry stretches.

All 20 swamp white oak trees in the irrigation-withheld study were lined out in B&B-BR pairs along one street to minimize microclimatic and soil differences between sites. They were watered only once at planting and thereafter received no irrigation.

Data Collection

In August 1998, when terminal buds had set, four growth responses were measured on all trees. First, shoot extension was measured to the nearest half centimeter on three mid-canopy, full-sun terminal shoots per tree, and the mean was calculated. Second, dieback was also measured on three mid-canopy terminal shoots per tree, and the mean was calculated. We devised a third visual index called “leaf canopy rating” to give an approximation of the fullness of the canopy relative to its branch density Leaf canopy complemented the other measurements by taking into account the extent of dieback and bud break failure within the canopy. For leaf canopy rating, a percentage scale in increments of ten was used, with 100% corresponding to a tree that had fully leafed out and had no measurable dieback within the canopy. Figure 2 gives examples of this rating on hackberry. Finally, average area per leaf was calculated from a sample of half of the trees. Ten mid-canopy, representative leaves per tree were measured with a LI-COR LI-3100 leaf area meter. The mean area in square centimeters per leaf was then calculated. The same measurements were repeated in August 1999 to give second-year growth data. Soil samples weighing at least 50 g (1.8 oz) were taken from the majority of the sites by extracting soil 15 cm (6 in.) below turf roots with a shovel. Samples were dried and analyzed for pH and gravel, and for sand, silt and clay fractions.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

A visual “leaf canopy rating” was devised to capture the fullness of the canopy relative to its branch structure. The hackberry on the left rated 60%; the hackberry on the right rated 90%.

Statistical Analysis

The statistical software package SPSS 9.0 was used to analyze the data. Means of paired B&B and BR trees planted within the same season were analyzed using paired t-tests. Means for both root types between seasons were analyzed using independent samples t-tests. Because leaf canopy ratings were recorded as percentages and their mean distribution was not normal, leaf canopy means were analyzed using the nonparametric Wilcoxon signed rank test (for comparing within season) and the Mann-Whitney U-test (for comparing between season). Analysis of covariance was used to look at the potential influence of soil texture, specifically the sand fraction, on the results.

Comparing Production Methods

Comparing Season of Transplanting

Comparing Irrigation-Withheld Trees

Ten B&B and ten BR swamp white oak trees were planted in pairs in spring. Trees were watered in at planting but received no subsequent irrigation. In year one, mean leaf canopy and leaf area were significantly greater on the B&B trees (7% and 37%, respectively). In year two, there were no significant differences between B&B and BR trees (Table 8).

Soil Sampling Results

Results showed a surprisingly uniform soil texture and pH across sites. Most sites were characterized by loam, gravelly loam, silt loam, or gravelly silt loam soils; two-thirds of the sites had more than 50% combined sand and gravel fraction. The majority of the sites were in the 6.6 to 7.5 pH range. The covariate analysis on sand fraction addressed the concern that some sites might be drier than others. The analysis revealed no significant effects of this variable on the growth responses for any of the trees in either year (data not shown).

Hackberry

In this study, fall-planted BR hackberry grew just as well as fall-planted B&B hackberry. Furthermore, fall-planted BR trees grew better than spring-planted ones. The success of fall-planted BR hackberry in this study confirms some industry opinions and contradicts others. Hightshoe (1988) recommends BR fall planting of hackberry if done “with care.” Berrang and Karnosky (1983) say that hackberry can be planted in fall or spring. Bailey nurseries (1999) suggests avoiding planting hackberry in late fall, and Princeton nurseries (1999) lists hackberry as “very risky” to transplant, BR or B&B, in fall.

Fall planting comes with myriad potential advantages from which the BR hackberry in this study may have benefited. A previous study established that root growth in Ithaca ceases around the end of October (Harris and Bassuk 1994). Trees in this study were planted in late October and early November; therefore, fall root growth was unlikely to be one of those advantages. However, the physiological processes in roots that precede root growth might get underway in fall, giving fall-planted trees an advantage in the next growing season (Harris et al. 1996). Another benefit of fall planting is reduced water stress. Transpirational demand of leaves and shoots is lower in fall than spring because ambient temperatures are cooler, days are shorter, shoot extension has ceased, and plant cells have lignified (Good and Corell 1982). Roots of fall-transplanted trees are in place longer before new spring shoot growth begins, and root-to-soil contact is improved as a result. The roots of spring-harvested trees, by contrast, are severed shortly before shoot growth begins in spring. Hinesley (1986) suggests that fall-planted trees may do better because spring harvests interfere with the production of root-produced hormones necessary for good shoot extension.

Spring-planted B&B hackberry grew better than their spring-planted BR counterparts. This finding conforms to the conventional thinking that B&B transplanting is less stressful and leads to faster establishment than BR planting (Pirone et al. 1988). This outcome validates the recommendation by Hightshoe (1988) that hackberry be planted B&B in spring. However, the result in this study differs from that of Magley and Struve (1983) who found that spring-planted pin oak performed equally well when harvested by tree spade (equivalent in most respects to B&B) or BR. The result in this study also differs from Vanstone and Ronald (1981) who found that spring-dug BR hackberry actually surpassed tree spade hackberry in growth by the second year. A four-day holding period before dipping might have put the spring BR hackberry in this study at a disadvantage relative to the B&B trees. Whereas fall BR hackberry were dipped and transported the same day they were dug, spring trees were dipped and transported four days after they were dug. A shorter holding period might have resulted in less root desiccation and a growth response more comparable to that of the B&B trees.

When compared across seasons, B&B and BR hackberry grew equally well in fall and spring. This contradicts Dirr (1998), Princeton nurseries (1999), Bailey nurseries (1999), and others who specify spring planting or warn against fall planting.

Hophornbeam

In this study, fall-planted B&B and fall-planted BR hophornbeam grew equally well. Both fall-planted B&B and BR hophornbeam grew equally well as their spring-planted counterparts. Significant differences were present only when comparing spring-planted trees; then, B&B outperformed BR. It is possible that spring-planted BR hophornbeam would have performed just as well if not subject to early bud break and a five-day holding period before dipping. Nonetheless, based on the results of this study, if one must plant in spring, B&B appears to be the better choice. If fall-planting is an option, the results of this study suggest that BR is just as good. This finding contradicts many sources recommending spring planting for hophornbeam (Berrang and Karnosky 1983; Hightshoe 1988; Schein 1993; Watson and Himelick 1997; Dirr 1998) or that caution against fall planting (Bailey Nurseries 1999; Princeton Nurseries 1999). This contradiction may be explained by the fact that the BR trees in this study had the benefit of the hydrogel root-dip protocol.

Swamp White Oak

Comparison of paired B&B and BR swamp white oak in both fall and spring showed no advantage to either production method. Oaks generally are considered fall transplanting hazards (Bailey Nurseries 1999; Northern Nurseries 1999; Princeton Nurseries 1999) or are recommended for spring planting (Berrang and Karnosky 1983; Schein 1993; Watson and Himelick 1997). By year two of this study, growth of swamp white oak showed no significant difference between fall and spring planting, only a mild trend in favor of fall.

Comparing First and Second Growing Seasons

In the first year, there were significant differences in growth responses between many treatment combinations of season, species, and production method. By the end of the second year following transplanting, however, most of these differences disappeared. This finding is consistent with Vanstone and Ronald (1981) who found that the first-year effects of transplanting method on their species (including hackberry) did not persist into the second year.

The second year of this study was characterized by the driest April through July on record in Ithaca, New York. Whereas hackberry and hophornbeam put on more shoot extension in year two than in year one in spite of the drought, swamp white oak set less shoot growth in year two. Yet of the three species, swamp white oak had the heaviest leaf canopy and largest leaf area both years and had negligible dieback in both seasons. It appears that swamp white oak, while not incurring dieback, responded to the drought by reducing shoot extension though a full canopy of large leaves was maintained. In the first year, hackberry had the longest shoot extension of the three species, but also had the most dieback. Rapid growth was noted in the wet spring and early summer of 1998, growth that was then checked by dieback in the drought of mid- to late summer. In the second year, hackberry again exhibited the most growth of the three species, but this time had negligible dieback. Root growth that took place in the intervening period between firstand second-year measurements may have been sufficient to support the rapid second-year shoot growth without exhibiting any dieback. For all three species, leaf area approximately doubled from year one to year two in spite of the second-year drought.

Comparing Irrigation-Withheld Trees

For 40-mm-caliper (1.5-in.) swamp white oak—hydrogel-dipped if BR and normal handling if B&B—there does not appear to be any advantage to planting B&B in terms of ability to cope with drought stress. The second year of this experiment was the driest April through July on record; in that year, there was no difference in growth response between B&B and BR trees. This finding challenges the conventional thinking that under drought stress, B&B trees are at an advantage over BR trees.

It is important to look at second-year growth in this study. If looking at only first-year data, one could assume that because B&B trees had significantly more leaf canopy and leaf area than BR trees, B&B was superior. However, year two showed a leveling out of differences in the midst of record drought.

Research in Urban Settings

One concern about conducting research in urban settings was the role of lurking variables. For example, it was thought that soil texture might have a confounding influence on measurements between sites. In Ithaca, where the great majority of soils in this study were found to be gravelly loams, loams, or silt loams, the concern was not with wet soils but with excessive drainage and limited soil moisture during two dry summers. The covariate analysis on sand fraction addressed the concern that some sites might be drier than others. The analysis revealed no significant effects of this variable on the growth responses for any of the trees in either year. Soil pH was also tested and was found sufficiently uniform to prevent concern that it might confound results. Other lurking variables include whether a home owner watered trees or not, dog urination patterns, children shaking tree trunks, and the like. This variability was not possible to control or quantify; however, the study employed a large number of replicates to ensure that these variables would average out in the summary statistics.

It should be noted that the months of December through March in the winter of 1997-1998 and December through February in the winter of 1998-1999 were warmer than normal. These comparatively warmer winters may have benefited fall-planted trees in this study by making winter injury less likely