The effect of geotextile fabric bag volume and media on the growth of two tree species (Dodonaea viscosa 'Purpurea' and Populus x canadensis 'Tasman')

Abstract

Two species of trees grown in geotextile fabric bags indicate that bags may not be a suitable method for long term restriction of growth. This study shows how root restriction occurs in bags and rejects the comparison that bags act in a manner similar to the process of bonsai. Media in bags had a significant effect on trunk diameter, height, total growth, changes in bulk density and the number of root escapes.

Four experiments were carried out to determine the long term effect of geotextile bags, (in ground fabric containers), on restriction of tree growth. The main experiment was a randomised complete block design of five blocks each with two tree species, two media and four bag volumes plus control (no bag treatment equal to the largest bag volume). Trees were planted 3 metres apart in ten rows (2 rows per block), in the field in September 1993. Dodonaea viscosa 'Purpurea' was harvested after two years and Populus x canadensis 'Tasman' after three years growth. Measurements of height and trunk diameter were made during, and at the end of each growing season. Fresh and dry weights of above and below ground parts were also determined after the trees were harvested.

Results showed media had an effect mainly in the first season, but this effect carried over for the life of the experiment with trees in potting mix larger than trees in field soil for most measurements. Bag volume was generally not significant until the end of the second season's growth. As trees became progressively more restricted within bags, the root : shoot ratio in g/g dry weight reduced. Increased root volume within bags increased media bulk density, indicating extreme levels of compaction especially in the smallest bag sizes.

A second experiment measured the force required to topple trees in bags compared with trees not in bags. Dodonaea viscosa 'Purpurea' trees planted for experiment one were wrenched from the vertical axis after two season’s growth, just before the final harvest. Results showed trees in potting mix were more difficult to topple than trees in field soil. There was no difference in force required for trees in 139 l bags or control.

The third experiment assessed the potential for water loss through the wicking effect of the 50 mm of fabric that sits above ground level after planting. The purpose was to determine if the potential drying out of the fabric around the roots affected the process of restriction. Subsequent analysis showed no difference between fabric containers with 50 mm of bag above the surface, at surface level, or if there was no bag. The conclusion was that bag fabric does not dry the area around the roots.

The fourth study considered the response of the geotextile fabric bag when placed under lateral tension. Observations showed fabric samples with an average of 48 needle punched holes per cm² with hole diameters between 200 and 500 µm. Tests showed that when lateral tension was applied to the fabric the number of holes and the diameter of these holes reduced. This suggested roots may have increasing difficulty penetrating the fabric wall as internal tension within the bag increased due to root growth.

Excavations made, where surface cracking was observed, at the base of trees at the end of the third season showed large roots (root escapes), had occurred in Populus x canadensis 'Tasman'. Root escapes occurred by exploiting the 3rd dimension of the bag fabric. The restricted root inside the bag produced secondary roots able enter the wall of the fabric at an oblique angle and grow sufficient length, to exploit the fabric at it's weakest point, and finally break out of the bag through 2° thickening. Root escapes were confirmed and numbers recorded after three season’s growth when they were harvested.