Effect of Eight Container Types and Root Pruning During Nursery Production on Root Architecture of Acer rubrum

DISCUSSION

Differences in trunk diameter and tree height growth among container types were small, measured over the course of the three-year study; this was typical of red maple and other taxa in previous studies (Marshall and Gilman 1998, in Florida, U.S.; Owen and Stoven 2008, in Oregon, U.S.; Neal 2009, in New Hampshire, U.S.; Amoroso et al. 2010, in Italy; Stoven O’Conner et al. 2013, in Colorado, U.S.). It appears fair to conclude, based on these studies, that trees grow about the same trunk diameter and height in most container types given appropriate water and nutrient management for the taxa and region, even with a 40% difference in substrate volume among types (Table 1). Although results could have been different under a different irrigation regime, the cited studies were performed under varying cultural regimes in different climates.

When roots on this same set of trees as the current study were washed and measured from #3 containers, all container types showed less circling root length (by about one-third) compared to SS; however, circling root length still represented about half of the total root length (roots >2 mm diameter) growing on the root ball periphery, with the remaining growing in other directions (Gilman et al. 2010a). Similarly, trees in all #15 container types not shaved (except RT) had a smaller percentage of trunk circled with roots at the #3 position than those in SS (Figure 1), indicating roots remained in the deflected position and grew larger, thus forming an imprint on the root system. In contrast, only trees in half the #45 containers (AP, FN, JP, and SP) had a smaller percentage of trunk circled than SS at the #3 position. It appears that roots smaller than the threshold measured (3 mm diameter) continued to grow in diameter in the circled fashion at the #3 position as trees were shifted to larger containers, thus resulting in an increase in percentage of trunk circled with time for most container types (Table 8; Figure 4).

It was difficult to isolate one container type that consistently generated roots in all three sizes tested [#3 (Gilman et al. 2010a), #15, and #45] that would be considered the most desirable, although there were trends. Finished trees in #15 FN and JP were always in the group of containers that had all of the following attributes: the least percentage of trees with roots >5 mm diameter growing over the root collar, the smallest diameter roots at the #3 position, and a relatively small number of ascending roots at the #3 position (Table 3). Trees in JP and SP #15 containers had more radial roots growing away from the trunk than all other containers not shaved (Figure 2); abundant radial roots have been associated with good anchorage (Ortega et al. 2006). Less deflection may have allowed roots to grow freely outward in their natural, shallow position (Lyford and Wilson 1964) into the substrate of the larger container, thus reaching the container wall sooner as shown for Swietenia mahogani (Gilman and Paz 2014). Roots contacting walls sooner should grow to a larger size than those striking the wall later. It is not clear why the container with the smallest volume (JP) was among those with the least amount of root defects, other than roots appeared trapped by the fabric walls instead of deflected. This suggests that container wall attributes may impact root system architecture more so than container volume. There were no correlations between container volume and any measured root attribute. Whereas large roots on the periphery have been thought of as a negative (pers. comm. with horticulturists and arborists), a root system resembling the natural root system may be more likely to develop under this condition, provided these roots are removed by shaving the root ball when shifting to larger containers or planting into the landscape. Presence of only small diameter roots at the periphery of the container to be planted has been associated with more deflection—and therefore a larger imprint of woody roots on the interior of the root ball, resulting from the previous container—and less anchorage after planting (Gilman and Weise 2012; Gilman and Harchick 2014).

FN was the only container type to produce larger diameter roots than SS containers on both the north and south sides of the #45 root ball periphery (Table 6). This probably occurred because many roots had grown through the fabric of both the #3 and #15 container walls and therefore had to be cut for the trees to be removed from all replicates, resulting in a mild form of root-ball shaving. None of the other containers required this treatment.

Absent root pruning, no #45 container type generated a root system much different in number of radial roots than any other (Figure 3). Of the non-shaved containers, JP had the least percentage of trunk circled with roots at the #3 position, but was in the group of containers with the largest percent circled in #15 containers (Table 8). Trees in AP, RT, and SP #45 containers stood out because they had the least percentage of trunk circled with roots at the #15 position (Table 8). Arnold and McDonald (2009) also showed that SP dramatically reduced (by a factor of five) the amount of rose (Rosa spp.) roots at the root ball periphery, compared to SS containers. Marler and Willis (1996) said (no data were presented) that there were fewer circling roots on trees in RB than SS for two tropical species. Moore (2001) found that many Australian tree taxa grown in 20 cm diameter AP had far fewer circling roots than SS containers eight months after shifting to large containers. Owen and Stoven (2008) noted circling roots in all container types tested, including some in the current study. Red maple in AP for 14 months (Gilman et al. 2010b) and in other container types for 15 months (Marshall and Gilman 1998) eventually developed circling and descending roots, even prior to trees growing larger than standard (Anonymous 2014) size. Roots of younger trees (#15 containers) grew primarily down at the position of the #3 container; whereas roots of older trees primarily circled in #45 containers at the position of the #15.

Root pruning (shaving) had no impact on trunk growth measured in #3, #15, or #45 containers during the study, although trees may have become stunted from root pruning had irrigation been applied differently. No trees showed symptoms of stress, such as defoliation or tip dieback, and no trees died. Other researchers also reported that the shoots and trunk of trees grew similarly after cutting (by slicing the root ball top-to-bottom) and teasing roots away from the periphery of container root balls (Malieke and Hummel 1990; Arnold and Young 1991). Direct comparisons are difficult to make among the studies, but shaving the root ball likely removed more of the root system than slicing and teasing because shaving severs every root that reaches the periphery of the root ball. This demonstrates, as others have shown (Gilman et al. 2009; Gilman et al. 2010b), that root architecture can be improved by root-ball shaving container-grown shade trees receiving regular irrigation in the nursery without slowing their growth.

Shaving the outer portion of the #3 root ball when shifting to #15 appears to result in a substantial decrease in attributes associated with poor root systems (e.g., percentage of trunk circled with roots, Figure 1; percentage of trees graded as culls in #15 containers, and percentage of the five largest roots circling at the #3 position, Table 4). The reduction, by about half, in diameter of roots measured at the #3 position as a result of shaving (Table 4) shows the dramatic change in diameter of roots measured a few centimeters from the trunk. Instead of a few roots growing to a large diameter and deflecting in a circling, descending, or ascending manner at the #3 position, shaving nearly eliminated deflection (Figure 1; Figure 4). The increase in diameter of roots growing on the periphery of the finished #15 container (Table 4) showed that shaving increased the size of roots reaching the edge of the larger container. Although not directly measured, shaving appeared to shift the root mass farther from the trunk, as in mahogany (Gilman and Harchick 2014). Root deflection at the periphery was nearly eliminated by shaving the #15 when shifting into #45 containers (Figure 3).

The impact on anchorage by changing the root system architecture from a few large deflected roots to a larger number of non-deflected, smaller, radially-oriented roots (in seven of eight #15 containers tested, Figure 2) has not been extensively tested. Other researchers showed that young (four- to seven-year-old) Quercus virginiana Mill. trees with fewer deflected roots and more straight roots were either slightly more stable or equal in stability (depending on measurement date) to trees with many deflected roots inside the root ball (Gilman and Weise 2012). Moreover, reducing root deflections on top of the root ball by root remediation when planting into the landscape reduced occurrence of circling roots and roots growing over the root collar in Acer and Ulmus five years after planting (Gilman et al. 2015a). Although not shown empirically, this may reduce negative impacts on health by reducing the likelihood of stem-girdling root formation. The one container type (JP) may not have responded to shaving because the number of radial roots, even without shaving, was greater than all but one other type (SP, Figure 2). This desirable root attribute in the container with the smallest volume provides further evidence that the container wall configuration influenced root deflection more so than container volume.

Root pruning by shaving the #3 and #15 root ball peripheries, as trees were shifted, substantially reduced attributes associated with poor root systems in finished trees in all eight types of #45 containers (Table 7). This included a dramatic improvement in root system quality by reducing the percentage of trees considered culls from 95% to 42% at the #3, and from 50% to 2.5% at the #15 positions. Moreover, there was a substantial reduction in percent trunk circumference circled by roots at the #3 and #15 positions from root pruning (Table 8), and a reduced percentage of trees with roots growing over the root collar (Table 7). Root-ball shaving also increased the diameter of the five largest roots at the periphery of the #45 on the north and south sides of the ball, while decreasing the diameter of the five largest roots at the #15 position (Table 7). Like trees finished in #15 containers, this represented a shift in woody root mass away from the trunk. Shaving increased the number (except for RT, Figure 3) of radial roots in finished #45 containers compared to not shaving, similar to what occurred in #15 containers when #3 containers were shaved (Table 4). With more roots growing radially instead of downward or around, forming a cage or imprint of deflected roots, the likelihood of forming a natural root system (Lyford and Wilson 1964; Danjon et al. 2005), with some deep and some shallow roots more or less straight, may improve long-term root system quality and anchorage. The radially oriented root system at the #3 and #15 positions allowed more roots to reach the #45 container side wall (Figure 3), resulting in an increase in percent trunk circumference circled with roots at the #45 periphery for four of the eight container types, compared to not shaving (Table 8). However, these can be removed by shaving the root ball periphery when planting into the landscape (Gilman and Wiese 2012). However, it should be noted that shaving the periphery will not correct defects present on the root ball interior, which were severe unless trees were shaved during previous shifts. Landscape contractors should check the interior of the root ball and correct these defects at planting because this taxon appears very prone to root defects. The long-term (more than about seven years) impact on health and anchorage of the root architecture imposed on trees during field and container nursery practices has not been evaluated on any taxa.

Fewer and smaller roots grew on the south side of all #15 and #45 containers (Table 5), as Gilman et al. (2010a) observed (not measured) on the same trees as the current study finished in #3 containers and measured in other studies (Martin and Ingram 1988; Ruter 1993). However, there appeared (not measured) to be many live white roots 1 to 2 mm diameter on the south side positioned about 3 cm behind the container wall; this could result in more roots growing into landscape soil from the south side once planted into landscape soil. High substrate temperatures are known to cause root death, especially on the sunnier, hotter container side (i.e., south and west side, in Northern Hemisphere, Ruter and Ingram 1992; Owen and Stoven 2008). Results may have been different if containers were turned in random directions as they were shifted to larger containers and not maintained in the same cardinal direction as was done throughout the current study. It is not clear if the root asymmetry generated by the current study’s conditions is a desirable or undesirable occurrence regarding anchorage. Also unclear is the potential different response had this study been performed on trees grafted or budded onto seedling root stock.