'Golden Delicious' on MM.111 rootstock and a 12" M.9 interstem in one of Roy Simons' trials at Illinois. Photo: Rich Marini, Penn State
Dwarfing rootstocks have been used in various situations for more than 2,000 years and the first apple rootstock breeding program is about 100 years old, but none of the dwarfing rootstocks currently available are without problems. As I have said before, "new ideas" seem to get recycled about every 25 years and I am old enough to remember the last time that interstems were widely tested. I think this might be a good time to review what we know about interstems.
In the 1920s and 1930s rootstock researchers at East Malling used interstem trees as a tool for studying the effects of scions and rootstocks on each other and learned that the scion influenced the morphology of seedling roots, but not Malling rootstocks.
One of the major reasons that dwarf rootstocks were not widely planted in North America until about 1995 was because their root systems are one-sided or brittle. Early experiences with dwarf rootstocks were disappointing because non-supported dwarf trees are poorly anchored, brake easily or fall over, and the central leaders lean or bend under the weight of a crop. Additionally, M.9 and M.26 were thought to lack adequate cold hardiness for many North American apple-growing regions. A potential solution to this problem was to use a hardy dwarf interstem with a vigorous rootstock to produce a free-standing, cold-hardy, semi-dwarf tree. The practice of double-grafting to produce interstem trees had been used to develop dwarf trees for many years, but interstem trees were more expensive and the system is more complex because it involves two graft unions.
Virginia Crab is winter hardy, tolerates collar rot, and was commonly used as a rootstock and as an interstem in the 1930s and 1940s in North America. Stark Brothers Nursery in Missouri sold a four-part tree with Virginia Crab root, 4"-long Clark Dwarf stem piece, Hibernal hardy stem and the cultivar, and some commercial orchards planted these trees with little research data to support their use. The rootstock known as Clark Dwarf was later identified as M.8. The plantings were generally not very profitable because trees were too small for their spacing and trees tended to lean, produced excessive root suckers and had virus problems. It turned out that interstem trees needed to be supported. By the 1950s there were reports of poor tree growth due to incompatibility with some scion varieties, possibly caused by a virus. Later it was learned that Virginia Crab was sensitive to stem pitting, caused by apple stem grooving virus, which is latent in most apple scion/rootstock combinations. Ironically, Virginia crab was later used as an indicator host for some viruses.
Following World War II, due to increasing labor costs there was renewed commercial interest in smaller trees and the use of interstems to produce hardy, precocious, semi-dwarf trees in America. In general, trees with dwarf interstems were smaller and more precocious than trees on seedling roots, but to a lesser extent than trees on dwarfing rootstocks. The degree of dwarfing increased as the length of the interstem increased regardless of the vigor of the rootstock. In Michigan, Carlson reported that the size of 12 year-old 'Delicious' and 'Jonathan' trees decreased as the length of M.8 interstem on Alnarp 2 rootstock increased from 4 to 10". Yield per tree declined linearly with increasing interstem length, but 4 and 8" long interstem trees had higher yields than trees on Alnarp 2. Yield estimates, based on tree spacing adjusted for canopy diameter of the 12-year-old trees, showed that yield per acre increased as interstem length increased. Researchers in Oregon placed interstems of varying lengths at 3 or 12" above ground and found that the lower placement had a greater effect on scion growth, but they found that increased flowering was not closely related to growth suppression.
As the desire for free-standing semi-dwarf trees increased, a number of rootstock trials were established on American research stations and in commercial orchards in the 1970s. In most trials, the rootstocks were seedling, MM.111 and MM.106, and M.26 was often used as the interstem. Early observations of young trees indicated that interstems often developed burr knots and had excessive suckering, but research by Simons in Illinois and Costante and Lord in Vermont and Massachusetts showed that suckering could be reduced by deep planting. Root suckering was a problem in other trials and the number of suckers per tree was positively associated with tree vigor in Ohio. Sometimes tree mortality was unacceptable because fire blight infected the susceptible interstem. Grower experiences with interstem trees were variable. The potential size of the trees was unknown and trees often did not fill their allotted space because trees were spaced too widely and young trees were allowed to crop too heavily. In general, interstem trees and trees on dwarfing rootstocks performed better when supported by a trellis or a tree stake.
In Ohio, Ferree found that freestanding 'Mutsu' trees on M.9/MM.106 or M.9 /MM.111 were larger but more productive than trees on MAC.9 (later named Mark) rootstock. In Virginia, Barden and Marini reported results from an 18-year experiment with 'Delicious' and 'Golden Delicious', where M.9/MM.106 interstem trees were similar in size or smaller than trees on M.9, whereas M.9/MM.111 trees were similar in size to M.26. Survival of trees on M.9/MM.106 was lower than for trees on M.9, M.26 or M.9/MM.111. Cumulative yield efficiency for both interstem trees was similar to M.9. In Massachusetts, Lord et al. compared M.9 and M.27 interstems on MM.111 and MM.106 rootstocks with trees on M.26, M.9 and M.27 (where trees on M.9 and M.27 were supported with stakes). After eight years all interstem trees were similar in size, and had cumulative yield and yield efficiency similar to trees on M.26.
In the late 1970s and early 1980s I saw a number of research and commercial interstem plantings in New England, New York, Illinois, and Virginia and most had trees that were leaning due to inadequate support and high tree mortality due to breakage at one of the unions, apple union necrosis (tomato ringspot virus) and fireblight. By the late 1980s, most researchers and commercial growers felt that interstem trees offered few advantages over dwarfing rootstocks.
In general the experiences with interstem trees in the 1970s were disappointing enough that researchers and growers lost interest in interstem trees. To my knowledge, newer dwarfing rootstocks, such as B.9, O.3, G.65 and other members of the CG series have not been tested as interstems. But based on past experiences, we know that interstem trees have their own set of problems and growers interested in these types of trees should proceed with caution.
Fifteen-year old 'Redchief Delicious' trees on M.26, M.9 and M.9/MM.106 interstem trees in Blacksburg, where the M.9/MM.106 graft union broke. Photo: Rich Marini, Penn State