Field studies
In the summer of 1981, a sample of 254 ex vitro 'Thornless Evergreen' plants were moved to the field. 
Of these, 200 were planted at Urbana, Illinois and 54 were placed at Dixon Springs, Illinois. Plants were spaced 10 x 10 feet in the plots, grown as if they were seedlings, and irrigated as needed until well established. The only other cultural care was occasional weeding by mechanical cultivation or hand pulling; a single application of granular herbicide was made in late fall of 1981. In December of that year, a straw mulch was placed over rows of plants.
First fruiting season
Vigor and fruitfulness were evaluated in 1982 by inspection, measuring plant height, determining the number of primocanes per plant, and recording the number of flowering laterals per plant and fruit set. Data were recorded in the third week of July at the Dixon Springs plot, and two weeks later at the Urbana plot. Each plant was evaluated individually, and location effects within fields were not considered. Categories were created based on growth of primocanes.
Plant survival at both locations was excellent (only two plants died), despite the facts that 'Thornless Evergreen' is not recommended to be hardy in the midwestern United States and that the winter of 1981-82 included temperatures of -20°F and lower. Straw mulch and nearly continuous snow cover certainly contributed to good survival rates.
Under field conditions the growth habit of these blackberries was more variable than expected. Since segregation of the components of this periclinal chimera had been verified by screening root suckers, different growth types were expected. But the field plantings revealed a continuum of plant sizes. For convenience, three size designations were made (full=1.00 m or more in canopy diameter, intermediate= 0.50 to 0.99 m, and dwarf= less than 0.50 m). Intermediates and dwarfs were bent petiole types.
Average fruit set among plant sizes at the Urbana plot was low (0.7 to 27 percent); Dixon Springs plants had better fruit set (28 to 90%). All the first-test fruiting was considered preliminary and not statistically analyzed.
Second fruiting season
In the second fruiting season (1983), data were collected on ex vitro field-grown plants in Urbana and Dixon Springs. Evaluations were made in mid-July in Dixon Springs and late July to early August in Urbana. Primocane erectness, plant spread, number of primocanes, height of primocanes, number of flowers per inflorescence, and percentage of fruit set were recorded.
Primocane erectness was similar in all full and intermediate morphological types. Average erectness varied from 3.1 to 3.3, when judged on a scale of 1 = prostrate to 5 = erect (as Rubus idaeus 'Heritage' red raspberry). However, dwarf types were significantly less erect than other types, and had a mean erectness rating of 2.2.
There were considerable differences in plant spread among the morphological types. By the second test year, only 16 of the original 143 plants classified as dwarfs remained in that group. No bent petiole plants became full sized.
Urbana full-sized types and Dixon Springs intermediates were not significantly different in mean number of primocanes, but Dixon Springs intermediates did produce a greater number of primocanes than Urbana intermediates or dwarfs. Data showed Urbana dwarfs less vigorous than any other group.
A similar pattern was noted in height of primocanes with significant differences among all morphological types. Fruit set in the second-test was more similar than the first-test data. In the second-test year, fruit set on individual plants varied from 0% (Urbana dwarfs) to 90% (Urbana fulls). The full types produced the most fruit in the second-test year, with an average fruit set (85.7%) which was significantly greater than other types. One Urbana intermediate had 85% fruit set; this plant's plot number was 6-6.
Plant selections
Selections from a segregating population need to be based on desirable characters. Bramble breeding objectives typically include erect growth habit, vigor, productivity, and thornlessness. In our case, a desirable blackberry selection needed to be vigorous, thornless, and produce abundant, high-quality fruit.
A weight-free index selection scheme was described by Elston (1963) to select for economic characters. By establishing lower acceptable limits for traits under consideration, best variants can be selected for further studies. This type of selection was imposed on ex vitro 'Thornless Evergreen'.
In the Urbana plot, weight-free selection was imposed such that full-sized plants with 90% or greater fruit set, seven or more primocanes, and an erectness rating of 3.5 or greater were retained.
Shortened internode (intermediate and dwarf) plants were subjected to the same type of screening, with fruit set being considered, but not at such a high level. This was due to the poor fruiting of most of the shortened-internode plants.
Three full-sized, six intermediate, and three dwarfs were selected for further testing. Of the intermediate types, one plant (previously mentioned as plant 6-6) was the most desirable. Its growth in both fruiting seasons was vigorous, upright, and fruit set was high, especially in comparison to other shortened-internode types.
Propagation for further studies
Vegetative one-node cuttings (with leaves attached) were made from primocanes of these selections in early September, 1984 to propagate stock for advanced testing.
Cuttings were treated with a commercial rooting compound (Funginex) and inserted in 1:1 (v:v) peat/perlite and placed under intermittent mist on a 12-hour cycle of misting, 2 sec/6 minutes.
After about 6 weeks, rooted cuttings were potted, grown for a short time in the greenhouse, and placed in cold storage at about 2° to 4°C for a 2- to 3-month cold period.
Plants were sent in 1985 to Oregon State University for final testing at the North Willamete Experiment Station, Aurora. The plants were allowed to grow for several years. Observations and data were recorded over several fruiting seasons, with the final data collected in August 1989 and spring 1990. The plant formerly known as 6-6 continued to be an outstanding selection--this plant is the one described in the plant patent document.
For more complete information:
McPheeters, K. 1985. Stability of 'Thornless Evergreen' blackberry in vitro and ex vitro. Doctor of Philosophy Dissertation. 131 pages. University of Illinois, Urbana-Champaign.
McPheeters, K. and R. Skirvin. 1989. Somaclonal variation among ex vitro'Thornless Evergreen' trailing blackberries. Euphytica 42:155-162.
Riordan, T. 1996. Sweet blackberries without the thorns. New York Times, Patents column Feb. 5, 1996.
Link to the National Clonal Germplasm Repository in Corvallis, Oregon for more information about their germplasm collection, which includes Rubus.
Background Information
Plant Patent 9407
More Information
Of these, 200 were planted at Urbana, Illinois and 54 were placed at Dixon Springs, Illinois. Plants were spaced 10 x 10 feet in the plots, grown as if they were seedlings, and irrigated as needed until well established. The only other cultural care was occasional weeding by mechanical cultivation or hand pulling; a single application of granular herbicide was made in late fall of 1981. In December of that year, a straw mulch was placed over rows of plants.
Under field conditions the growth habit of these blackberries was more variable than expected. Since segregation of the components of this periclinal chimera had been verified by screening root suckers, different growth types were expected. But the field plantings revealed a continuum of plant sizes. For convenience, three size designations were made (full=1.00 m or more in canopy diameter, intermediate= 0.50 to 0.99 m, and dwarf= less than 0.50 m). Intermediates and dwarfs were bent petiole types.
There were considerable differences in plant spread among the morphological types. By the second test year, only 16 of the original 143 plants classified as dwarfs remained in that group. No bent petiole plants became full sized. 
A similar pattern was noted in height of primocanes with significant differences among all morphological types. Fruit set in the second-test was more similar than the first-test data. In the second-test year, fruit set on individual plants varied from 0% (Urbana dwarfs) to 90% (Urbana fulls). The full types produced the most fruit in the second-test year, with an average fruit set (85.7%) which was significantly greater than other types. One Urbana intermediate had 85% fruit set; this plant's plot number was 6-6.
Selections from a segregating population need to be based on desirable characters. Bramble breeding objectives typically include erect growth habit, vigor, productivity, and thornlessness. In our case, a desirable blackberry selection needed to be vigorous, thornless, and produce abundant, high-quality fruit.
Plants were sent in 1985 to Oregon State University for final testing at the North Willamete Experiment Station, Aurora. The plants were allowed to grow for several years. Observations and data were recorded over several fruiting seasons, with the final data collected in August 1989 and spring 1990. The plant formerly known as 6-6 continued to be an outstanding selection--this plant is the one described in the plant patent document.