Lost and rediscovered cider varieties

A cider variety exists only as living wood. There is no seed bank for it, because a seedling from its pips is a different individual, and there is no way to reconstruct it from a description. It persists as long as somebody keeps grafting it, and it stops existing when the last tree dies with no scion taken.

That makes the loss mechanism straightforward. Old orchards cleared for other uses, unreplaced trees falling in orchards that stopped being managed, and hedgerow and field-corner trees removed in field enlargement each took varieties out permanently and without record. The point of maximum loss coincides exactly with the period of maximum orchard clearance in the middle of the twentieth century.

The commercial narrowing did the rest. Twentieth-century planting concentrated on a short list of varieties selected for agronomic reliability and processing suitability, and there was no economic reason to propagate anything else. A variety that is not planted commercially and not held in a collection depends entirely on old trees, and old trees are a wasting asset.

The historic record for English cider fruit names a great many more varieties than are now known. County pomonas, nursery catalogues, estate planting records and the descriptive literature between the seventeenth and nineteenth centuries carry names that have no known tree behind them today.

Several different things sit behind such a name, and separating them matters. Some are genuinely gone. Some are alive under another name, because the same clone was independently named in more than one district. Some were never a distinct clone at all, but a local population of seedlings sharing a name and a rough type. And some are still growing in a hedge that nobody has surveyed.

Counts of lost varieties therefore need to be read carefully. A figure derived by subtracting the varieties now known from the names in the historic literature will overstate the loss, because it counts synonyms and populations as clones; a figure derived only from collection holdings will understate what survives in the field.

The method used by county orchard groups and cultivar networks is patient and unglamorous. Old orchards, farm boundaries and hedgerows are surveyed, generally with the landowner’s cooperation and often prompted by a local report of an unfamiliar tree. Fruit is collected in season, with several specimens from a well-cropped tree, and its position recorded.

The fruit is then compared against the descriptive literature and against reference material — the plates and text of the pomonas, and where possible fruit from collection trees of the candidate variety in the same season. Local knowledge is used as evidence in its own right: what the tree was called on the farm, when it was planted if anybody knows, and what the neighbouring parish called the same fruit.

If the identification is promising, wood is taken and grafted, and the resulting trees are grown on so that the variety exists in more than one place before the original tree fails. That step is the one that actually secures anything, and it has to happen whether or not the identification is ever settled. Networks in the Welsh Marches and the western counties have worked this way since the 1990s, and the perry pear equivalent is run by the specialist bodies for that fruit.

The popular framing of this work is a lost variety found, which suggests a closed case. What has usually happened is more provisional: a tree has been found whose fruit matches a historic description well, and whose location and local naming are consistent with that description, and which does not match any other variety in the reference collections.

Closing such a case requires reference material from the original variety, and for most historic cider fruit there is none. A plate and a paragraph in a nineteenth-century pomona cannot be fingerprinted. Where a collection does hold an accession under the historic name, a molecular comparison can either confirm the match or show that the collection accession and the found tree are different things — which then raises the question of which of them, if either, is the variety the pomona described.

None of this makes the work less valuable. A well-supported identification is a real result, and the grafted trees are secured whether the name is right or not. It does mean that a rediscovery should be reported as an identification with a stated basis, and CiderHQ records disputed identities as disputed rather than picking a side.

The genuine gains of the last three decades are concrete. Named varieties that existed on one or two ageing trees now exist as young grafted trees in several places; collection holdings have been audited and corrected; county groups hold survey records that did not exist before; and a body of people competent to identify fruit has been trained, which is itself a fragile resource.

The limits are equally concrete. A collection tree is a specimen, not a population, and a variety kept alive only in collections is one accident away from loss. Traditional orchards continue to age faster than they are replanted. And perry pears remain the hardest case of all, because a pear takes decades to bear and a rediscovered variety cannot be assessed as a perry fruit within the working life of the person who found it.

The honest summary is that the searching has slowed a loss rather than reversed one. What determines whether these varieties exist in a century is not identification work but whether anyone is planting them, and that depends on there being a reason to grow fruit that is difficult, low-yielding and known only to a few people.