labelled trees. This emphasizes the fragility of these old collections, and reminds us of our need to complete our investigations soon. (NOTE: Several of the bulldozed trees were earlier propagated in Corwin Davis's collection for evaluation, and hence are not really lost, should they prove of any value).

The Blandy Experimental Farm was the first collection I ever investigated, back in 1979. It was the easiest of all to find, it never having changed hands and being well known around Winchester, VA. The Blandy pawpaw accessions date to the period 1926-1955, from seed that various individuals sent in from around the country, and from Zimmerman, the bulk of whose collection was donated to Blandy by his widow (Flory 1958). Around 60 trees, which I judge to be original, remain growing in two rows in the backwoods, and four more growing near the Quarters. All of these have been labelled. Some of the most interesting material we have discovered is at Blandy. An instance is BEF-53, which has large fruit (up to 14 oz.), very few seeds (4%), and an inferior flavor (unfortunately). We have been collecting seed from Blandy for several years.

The Hershey collection of Dowington, PA, dates from roughly 1925 to 1960. John Hershey was a fruit and nut grower, a commercial nurseryman, and a friend of Zimmerman's through the Pennsylvania Nut Growers Association. He acquired his stock from Zimmerman (Thomson 1974). At present, fewer than two dozen trees remained on the property. All are judged to be chance seedlings, except the old tree near the main house. We labelled the three best trees. Tree JH-2, which grows in a clump of ten near the "packing house," especially pleased us. Its fruit tastes exceptionally fine — very "cool," not too rich or cloying, and well balanced in flavor. We collected seed in 1984, and will later propagate it.

A fifth collection we have gathered from, in 1982 and 1983, belonged to Dr. H. A. Allard, of Arlington, VA, a scientist with the Department of Agriculture. Allard's pawpaws are of interest because he bred them for several generations, and because they do not derive from the Buckman-Zimmerman lines, and hence broaden our genetic base. They descend from some trees of superior flavor which Allard found on a camping trip in the Blue Ridge Mountains in the '20s or '30s. When Allard's land was subdivided and built upon in the 1970s, most of his trees were destroyed. However, one lot owner, Mrs. Ross, liked and kept her trees, which is quite fortunate because in that grove of pawpaws is a stem with one of the finest flavors I have ever tasted, equal or better than JH-2. I have labelled it HAA-1, and have collected seed of it. However, HAA-1 itself has not been propagated; the property has been sold, and it is uncertain if we will be allowed access.

Our nursery now contains almost 2,000 young pawpaws. Most of the seed comes from those five historic collections. The remainder came from trees along the C & O Canal near Washington, DC, where pawpaws grow extensively, from a couple of trees belonging to Schlaanstines of West Chester, PA, (beautiful vigorous trees, heavy fruiters, seedlings from John Hershey), and from seed that other pawpaw growers have sent us of named varieties, namely Overleese, Sunflower, Mitchell, Davis, Taytwo, etc. Over 700 of the trees are in their 3rd or 4th leaf, and are growing in a permanent nursery at the Wye Research and Education Center of the University of Maryland, on the Eastern Shore of Maryland. These may flower

75

in 1989. Another 700 are in their 2nd and 3rd leaf, and were just planted in March to a permanent nursery site at the Sharpsburg Research and Education Center in western Maryland. We do not expect these to flower before 1992. The remaining 450 are at the University greenhouses, where 150 are in their 2nd and 3rd leaf, and 300 just germinated this July. These are being grown in coarse well-drained medium in 2" × 2" × 12" cardboard tube-containers that are arranged 6 by 6 square in heavy duty plastic milk crates. The tubes are open at the bottom and the crates raised 2" off the ground to promote air root-pruning. The seedlings make good growth in these containers, and if germinated in the greenhouse in mid-March can make two years' growth in one year. (NOTE: Seeds sown in mid-February will germinate in mid-March at soil temperatures near 75°F. If germinated much earlier, when the day length is shorter, the seedlings will often set a terminal bud, thus halting growth). We like to set the seedlings in the nursery when they are about 12" high, which generally equals two years growth in containers. More than two years in these containers tends to stunt them.

Growth in the field has been quite good. In only a few progeny lines have we noticed stress from being grown in full sunlight. We are seeing considerable variation in growth rates both within and between different progeny lines. Those from the C & O Canal are noticeably the most uniformly vigorous (possibly because they were not selected for superior fruiting quality). Irrigation is provided as needed. Weeds are controlled with Surflan herbicide and hand cultivation. Nitrogen is applied annually at the rate of 50 lb/acre. Of the roughly 800 plants set out in the spring of 1984, only a couple dozen plants have been lost.

Another part of our collection consists of the southern species of Asimina, the dwarf pawpaws: parviflora, incarna (sometimes called speciosa), reticulata, obovata, longifolia, and pygmaea. We have several specimens of each, and are only missing tetramera, a rare and endangered species. All of our specimens were raised from seed collected in north central Florida, and are doubtfully hardy here in Maryland, even on the more temperate Eastern Shore. As an experiment, in the summer of 1984 we planted 63 plants of our six southern species at Wye, and kept another 36 at the University. Of the 63, only 31 made it through the winter of '85 (which was not abnormally cold) whereas 35 of the 36 survived that were overwintered in the walk-in refrigerators. Moreover, with the exception of A. parviflora, which is known to be hardier, the survivors at Wye were all topkilled. We conclude that the southern species, with the exception of A. parviflora, are not hardy here and require winter protection.

Raising the southern species in the artificial environment of pots and greenhouses is proving to be a major problem, and a painful learning experience. Container culture of dwarf pawpaws has never been tried before, so we are the first to make mistakes. And we have not been as attentive to the problems of culture as we might have. Every summer we lose some potted plants, and every winter in the refrigerators we lose some more. The major cause of death appears to be root rot, which reflects improper soil conditions, namely inadequate drainage, and possibly inappropriate pH and nutrient levels. The losses can be costly, as for example this summer, 98% of 2002-year old obovata seedlings were killed by overwatering, because some greenhouse staff moved the crates and placed them directly on the

76