Frederick V. Coville and the History of North American Highbush Blueberry Culture

Abstract

November 15, 2010 marks the 100th anniversary of Dr. Frederick V. Coville's, U.S. Department of Agriculture, Bulletin 193, Experiments in Blueberry Culture. Prior to these studies, most attempts to move plants from the wild and maintain them in managed culture had been unsuccessful. In the short span of time from 1906 to 1910, Coville determined that blueberries required a moist but not wet soil with low pH, had a low nutrient requirement, and required winter chilling. He developed propagation procedures by cuttings, grafting, and budding. In 1909, he recognized that self-sterility could be an issue after few seeds and no plants resulted from selfing ‘Brooks’, an otherwise outstanding wild plant found near Greenfield, N.H. Elizabeth C. White, a commercial cranberry grower in Whitesbog, N.J., contacted Dr. Coville after reading Experiments in Blueberry Culture and offered encouragement and assistance. Cooperation in selecting wild breeding material, growing of seedlings, and cultivar selection continued for the next 26 years. In 1920, ‘Pioneer’ was the first cultivar released from their breeding effort, however, ‘Rubel’, a selection from the wild that is still grown today, was released in 1912. A total of 15 cultivars were released before Coville's death in 1937. Another 14 of his crosses were released after his death. These 29 cultivars still accounted for 75% of the commercial U.S. acreage in 1992. This early cultural work and variety development provided the basis for commercial plantings in New Jersey. Interest spread to other states with the first planting in Michigan in 1924, North Carolina in 1928, and Washington in the 1930s. By 1949, there were 1,082 ha (2,674 acres) in New Jersey, 701 ha (1,731 acres) in Michigan, 230 ha (568 acres) in North Carolina, and 84 ha (207 acres) in Washington, according to the U.S. Bureau of Census. Evaluation of rabbiteye (V. ashei) wild selections was begun in Georgia in 1939 followed by controlled crosses and the first releases, ‘Callaway’ and ‘Coastal’ in 1950 and ‘Tifblue’ and ‘Homebell’ in 1960. Development of highbush type bushes for warm climates was begun in Florida in 1948. The wild blueberry species V. darrowi was used as a parent to provide the low chilling requirement, while attempting to maintain the desirable fruiting characteristics of the northern highbush cultivars developed by Dr. Coville. The first of these cultivars, referred to as “southern highbush,” ‘Sharpeblue’ and ‘Floridablue’ were released in 1976. More recently, crosses of highbush and lowbush (V. angustifolium) have given bushes of reduced stature (“half-highs”) that are protected by snow cover. The first of these cultivars were released in Michigan, ‘Northland’, 1968 and Minnesota, ‘Northblue’ and ‘Northsky’, 1983. Could Fredrick Coville have envisioned an industry of 25,500 ha (63,000 acres) that returned over $500 million to growers in the U.S. in 2009?

Keywords:

• Frederick Coville
• Elizabeth White
• growth of blueberry industry

BLUEBERRY DOMESTICATION

Most of the important fruits of the world have been cultivated for hundreds, if not thousands, of years. DeCandolle (1964) traces grape cultivation in Egypt for more than 4,000 years. Only occasionally can a person's name be associated with the original domestication. Strawberries are one of the few examples. Strawberries from the wild were grown in French gardens as early as 1324 according to Hyams (1953). Domestication began in 1715 when the French Naval Officer Frezier took Fragaria chiloensis Duch. plants from Chile back to France. There the botanist Duchesne successfully made the cross with F. virginiana Duch. that led to the modern, large-fruited cultivated strawberry.

Dates and a name are much easier to document for highbush blueberry domestication. The date was 1906 and the name was Frederick Vernon Coville, born March 23, 1867 at Preston, N.Y. He attended Oxford Academy in Oxford, N.Y. and went on to Cornell where he was described as the leading scholar in his class and also its leading athlete. Baseball remained a lifelong interest. Graduation from Cornell was in 1887 (Brown and Palmer, 1967). On the 100th anniversary of his birth, Brown and Palmer (1967) described Coville in the years leading up to and concurrent with the blueberry studies. In 1988, he entered the Department of Agriculture and became the botanist on the Geological Survey of Arkansas. A List of the Plants of Arkansas, accepted as a standard work to this day, was published by Coville and John C. Branner in 1891. Also in 1891, he accompanied the Death Valley expedition as the botanist. As a result, there was published in 1893 Botany of the Death Valley Expedition, described as one of the classic studies of desert vegetation. Other investigations in the far west included useful plants of the American Indians and grazing management. Grazing-management policies of the Forest Service became, in large measure, the outcome of these investigations. He was credited with the efforts that lead the Carnegie Institution to establish its Desert Botanical Laboratory at Tucson, Arizona in 1903. Collaboration with Lester F. Ward brought about the National Herbarium in 1894 that combined the collections of the Department of Agriculture and the Smithsonian Institution. Coville was appointed the first curator and continued in the post until his death. In collaboration with Frederick Law Olmsted and H. P. Kelsey he edited Standardized Plant Names, the indispensable text that assigned English names to plants previously known only in Latin or Greek.

The National Arboretum will probably be Frederick V. Coville's most remembered accomplishment. Both George Washington and city designer, Pierre L'Enfat, had envisioned such an institution, but by 1897 only a few small gardens and some plantings in the Mall and on Capitol Hill had materialized. The need was recognized for a scientific institution that might be likened to Kew Gardens in England, with research and educational capabilities, including a herbarium, botanical library, and photographic files. Disputes over location and jurisdiction and by lethargy in high places had kept such an institution from being established. Coville's fellow botanists and scientists gave their blessing for him to attempt to overcome the impasses. Brown and Palmer (1967), describe the effort in these words: “Launching the undertaking proved a task formidable enough and consumed more than 30 years of unremitting effort, punctuated with recurring setbacks and disappointments.” After enlisting sponsors, convincing Congress, examining hundreds of acres of land on foot, arranging soil surveys, and eventually drafting the legislation, the enabling Act was passed in 1927. In 1929, he became the unsalaried acting director.

The National Geographic Society was a life-long interest. He served on the Board of Directors for at least 6 years before becoming chairman of the Research Committee in 1920, a position he maintained until his death. Close ties were established with Gilbert Hovey Grosvenor, who was director of the National Geographic Society from 1899 to 1919 and president from 1920 to 1954. In addition, Grosvenor was editor-in-chief of National Geographic Magazine from 1903 to 1954. Coville is pictured with Grosvenor, President Theodore Roosevelt, and Admiral Robert E. Peary, the arctic explorer, in National Geographic Magazine, October 1963, page 542. This issue commemorated the 75th anniversary of the magazine. The picture was taken during a dinner in 1914 to honor Roosevelt before his National Geographic lecture about South American explorations. Interesting genealogy is associated with the Coville and Grosvenor families. Gilbert Grosvenor's wife, Elsie May Bell Grosvenor, was the daughter of inventor Alexander Graham Bell. Frederick Coville's son, Cabot, married Lilian Waters Grosvenor, daughter of Gilbert and Elsie May. The children of Cabot Coville, Gilbert Grosvenor Coville born in 1928 and Cabot Coville, Jr. born in 1932, have Alexander Graham Bell as a great-grandfather and Fredrick V. Coville as a grandfather.

Many honors were bestowed on Frederick V. Coville, but among the most important were the honorary degree of Doctor of Science conferred by George Washington University in 1921, Fellow of the American Association for the Advancement of Science and the George Robert White Medal of Honor from the Massachusetts Horticultural Society for his outstanding work with blueberries.

BLUEBERRY INTEREST BY COINCIDENCE

Coville was worried that his four children (Stanley, 11; Katherine, 9; Cabot, 3; Fredrick, 1) growing up in Washington D.C. would never learn the rural skills that he had acquired in his childhood in central New York state. A geologist friend in Washington, Arthur Keith, told him about a farm for sale in Greenfield, New Hampshire. He bought 40 acres with an abandoned house and barn for $400 on May 2, 1905. Blueberries, both highbush and lowbush, flourished in the fields no longer grazed by cattle around Greenfield. The highbush were picked by hand or shaken and the lowbush were raked for sale in Boston and local cities. Hand-picked berries averaged about $.11/quart and raked or shaken berries averaged about $.02/quart. In 1906, less than a year after coming to Greenfield, his interest was attracted to the subject of blueberry culture. He began looking at the diverse blueberry plants in the area. Convinced that it should be possible to cultivate blueberries instead of simply harvesting wild ones, Coville arranged for George W. Oliver, a colleague at the Bureau of Plant Industry, to germinate some blueberry seed in the fall of 1906. This was the first experiment conducted on blueberries.

BLUEBERRY CULTURAL REQUIREMENTS

A popular idea that blueberries could not be transplanted or cultivated existed when the studies were begun in 1906 (Coville, 1910). Experiment stations in Maine, Rhode Island, New York, and Michigan had attempted to grow the blueberry as a fruit but none had been commercially successful. However, blueberries from the wild were sold on the Boston market. Shipments began in early June from North Carolina, followed later in the month from Pennsylvania, New Jersey, and New York. In late June or early July, they came from Massachusetts and New Hampshire followed by Maine, Nova Scotia, and New Brunswick in August and September. If the berries were hand picked, the quality was generally good. Coville (1910) noted, “The blueberry withstands the rough treatment incident to shipment so much better than most other berries that with proper handling it should always reach the market in first-class condition.” The owners of land with blueberry plants commonly paid pickers two-thirds of the price they received from the sales. If the berries were harvested with rakes, which reduced the harvest cost about 75% compared with hand harvest, the fruit was often partially crushed, and the berries covered with soured juice. He did not consider raked lowbush blueberries suitable for sale on the fresh market. Records from a shipper in southern New Hampshire indicated that prices and demand for his hand-picked, wild highbush blueberries were good.

Coville was very aware of the economic potential that existed if larger berries could be produced, particularly if they could be harvested more rapidly from bushes growing in a cultivated plantation. To this end, in the fall of 1906, he began testing methods for germinating seed in cooperation with Mr. George W. Oliver of the Bureau of Plant Industry. The method developed was to remove the seeds from fully matured and slightly fermented berries, sow them in a shallow wooden flat with fibrous kalmia peat on the bottom (to insure good drainage) and then cover them with a germinating medium of peat, sand, and a small amount of loam soil. Germination of seeds sown in August occurred in about 5 weeks and, with proper handling in the greenhouse, robust plants were produced by the beginning of summer. Plants from these experiments were used for determining the soil requirements for growing blueberries.

One of the very definitive soil studies was begun on December 22, 1908. Rose cuttings, alfalfa seeds, and blueberry seedlings were planted in a rich garden soil or a peaty soil taken from a location where wild blueberries were thriving. Both the roses and alfalfa grew poorly in the peat while the blueberries grew well. In the rich garden soil, the roses and alfalfa grew well while the blueberries barely remained alive. The results of this experiment seem so incredibly obvious based on the present understanding of the requirements of acid loving plants. However, horticultural literature of the day made no mention of certain plants requiring an acid soil. The assumption seemed to have been that all plants would produce maximum growth and be most healthy when planted in a well limed, rich, garden-type soil. The pH meters we take for granted and use so routinely for selecting soils suitable for blueberries had not been developed. Coville first tried litmus paper to measure acidity, but found phenolphthalein indicators to be the most satisfactory for the weak acids of the blueberry soils. The finding that blueberries required an acid soil and the development of a means to accurately measure soil acidity were key to blueberry domestication. Additional studies provided much detailed and practical information about soil properties, nutrient requirements, and root characteristics. Blueberries were found to not thrive on a heavy clay soil. Aeration conditions satisfactory for blueberries were found in both sandy soils and soils consisting of drained fibrous peat, while soils that often became saturated with water were deemed not satisfactory. Heavily manured soil, favorable to many plants, was detrimental to blueberries, but occasional applications of small amounts of liquid from a manure pit during the growing season were beneficial. Blueberry roots were found to not have root hairs. Healthy roots were found to be inhabited by a fungus known as endo-tropic mycorrhiza. He speculated that the mycorrhiza was involved with nitrogen nutrition since available nitrogen levels were very low in the soils favorable for blueberries.

Propagation tests went from the initial seed germination in 1906 to successful increases of desirable plants by grafting, budding, division, layering, twig cuttings, and root cuttings. Additional recommendations for propagation were published in 1921 (Coville, 1921). Also by 1910, the cycle of flower bud formation and flowering had been determined. The need for a winter chilling period for normal leafing and flowering in the spring was determined. Flower configuration was observed to almost completely prevent self pollination and the need for insect pollination was recognized.

The first field trial was made in early July 1908 near Greenfield, N.H. from plants grown from seed in the fall of 1907. Drought followed the planting but 97% survived and 7% flowered in 1909. In 1910, 70% flowered. Thus, by 1910, Coville's goal of being able to asexually propagate blueberries and grow them in a cultivated plantation was accomplished. The remaining step at this point was larger berries.

IN SEARCH OF BIGGER BERRIES

The first wild blueberry for breeding purposes was selected in July of 1908 in a pasture near Greenfield, N.H. The Coville family spent summers near Greenfield after buying the farm in 1905. Coville began casually observing the native blueberries in the area in 1905, but a diligent search was undertaken in the summer of 1908 (Coville, 1937). In July, the first superior bush was found on the farm of Fred Brooks, near the Coville farm. The selection was aptly named Brooks. Many berries on the bush were more than 1.27cm (1/2 in.) in diameter, unusually light blue, firm, juicy, and exceptionally flavored, with a good balance between sugars and acids. Yield on the bush that season was about 2,840 cc (3 qt). Coville (1937) says the following about the selection of ‘Brooks’: “I regard its selections of fundamental importance to the success of the Department's blueberry-breeding experiments. Every breeder of race horses or of milk cows understands that the choosing of the individuals to be interbred is of the highest importance. Plant breeders usually select carefully the species they intend to interbreed, but often make the mistake of paying too little attention to the choice of superior individual plants within the species.”

The first attempt at controlled breeding to improve the wild highbush blueberry was undertaken in 1909 with the self-pollination of flowers from ‘Brooks’. This technique that had proven so successful with field crops, such as corn and bean, produced no viable seeds. The pollinations were repeated with similar negative results in 1910. This was the first indication of self sterility that is common in Vaccinium species.

Failures also often occurred when crosses were attempted among species of similar size and habit. An example was the highbush blueberry, Vaccinium corymbosum L. and bigbush blueberry, V. atrococcum (A. Gray) Heller. These resembled each other so closely that Asa Gray considered one to be a variety of the other. Cytological studies later showed highbush blueberries to have 24 chromosomes and bigbush to have 12 (Coville, 1927). After many attempts at crosses of diverse Vaccinium species and plant habits, Coville concluded that species of the most diverse appearance and characteristics, but with the same number of chromosomes, hydridized most readily.

The second wild blueberry selected for breeding purposes also came from near Greenfield, N.H. It was named Russell, for Frank Russell who described it as the best lowbush blueberry on his farm. Selected in 1909, Russell berries reached a diameter of 1.43 cm (9/16 in.), were light blue, and most importantly, ripened earlier than ‘Brooks’. Coville recognized that the earlier berries brought the higher prices. The first crosses between ‘Brooks’ and ‘Russell’ were made in the spring of 1911. Plants resulting from this cross were cross-pollinated with each other in 1913. About 3,000 hybrids of the first and second generations were grown to maturity in the field, with remarkable results. A wide range of many traits were expressed in the seedlings.

Elizabeth C. White, the daughter of a cranberry grower near New Lisbon, N.J., read Coville's 1910 publication, Experiments in Blueberry Culture. She harvested and shipped wild blueberries from the area and was extremely fascinated by the possibility of growing blueberries in a plantation. She offered land and cooperation to Coville's blueberry improvement projects. Her offer was readily accepted and many seedlings were tested in this ideal blueberry environment. Of equal or greater importance was the help she and her pickers of wild berries provided in finding outstanding wild selections. Her pickers were given a “blueberry gage” that consisted of a plate with holes ranging from 0.7 cm (0.28 in.) to 2.5 cm (1.0 in.). They were asked to mark bushes that had large berries based on the gage measurement. Later, either Elizabeth White or Coville would be taken to observe the bush. ‘Sooy’, named for Ezekiel Sooy, in 1911 was the first New Jersey selection. Crosses with ‘Brooks’ in 1912 provided some of the best hybrids and included the cultivars ‘Pioneer’ and ‘Katharine’. ‘Rubel’, named for Rube Leek, was selected in 1912 and became a cultivar that is still grown today and used widely as a breeding parent. ‘Grover’, named for Russell Grover, and ‘Chatsworth’, named for the town of Chatsworth, were other important breeding parents. The cross of ‘Rubel’ by ‘Grover’ in 1916 begat ‘Jersey’, which is still a major cultivar today. The last Coville release was ‘Dixi’ in 1936. Dixi, the Latin word meaning I have spoken or I am through, was to punctuate his retirement at the end of 1936. However, in addition to the 15 cultivars he released, an additional 14 of his seedlings were released after his retirement at the end of 1936 and untimely death on January 9, 1937 (Table 1). His second objective, to develop blueberries that were large and could be rapidly harvested by hand, was certainly met. The size was obtained with constant vigilance to maintaining excellent flavor and other horticultural characteristics.

TABLE 1 Cultivars Released by F. V. Coville and His Crosses That Were Released as Cultivars after His Death in 1937

Cultivars released	Year	After 1937	Year
Rube	— ^z	Angola	1951
Pioneer	1920	Atlantic	1939
Katharine	1920	Berkeley	1949
Greenfield	1926	Bluecrop	1952
Rancocas	1926	Blueray	1955
Jersey	1928	Burlington	1939
Concord	1928	Collins	1959
June	1930	Coville	1949
Stanley	1930	Croatan	1954
Scammell	1931	Earliblue	1952
Redskin	1932	Ivanhoe	1951
Catawba	1932	Murphy	1950
Wareham	1936	Pemberton	1939
Weymouth	1936	Wolcott	1950
Dixi	1936
^zRubel was selected from the wild in 1912.

The value of Dr. Coville's contribution to understanding the cultural requirements of blueberries cannot be measured, but we can estimate the value of the fruit from bushes he bred. In 1992, Dr. James N. Moore of the University of Arkansas and one of the distinguished blueberry breeders with the United States Department of Agriculture that followed Dr. Coville, surveyed the blueberry cultivars being grown in the United States (Moore, 1993). The cultivars released or seedlings he produced that were released later are shown in their order of planted acreage (Table 2). No less than 75% of the acreage and undoubtedly a higher percentage of the production came from Coville's cultivars or seedlings.

TABLE 2 Ranking by planted hectares in the United States in 1992 of cultivars released by F. V. Coville and his seedlings that were released after his death in 1937 (figures from Moore, 1993)

Ranking	Cultivar	Hectares	Acres	U.S. acreage, (%)
1.	Bluecrop	6,173	15,253	35
2.	Jersey	3,030	7,487	17
3.	Weymouth	996	2,461	6
4.	Croatan	948	2,342	5
5.	Blueray	740	1,829	4
7.	Rubel	551	1,362	3
8.	Berkeley	420	1,038	2
11.	Earliblue	189	467	1
14.	Collins	106	262	1
15.	Murphy	95	235	0.5
16.	Coville	85	210	0.5
Total		13,333	32,945	75

A BILLION DOLLARS WORTH OF BLUEBERRIES

In 1992, 81,000 mt (180,000,000 lbs) of highbush blueberries were produced and sold from production in the United States. The Coville portion of 75% was 61,000 mt (135,000,000 lbs). At the 1992 price of $.95/kg ($.43/lb), this would make the value of Coville's portion of the blueberry crop to be $58,000,000. During the 10-year period, from 1983 to 1992, highbush production in the U.S. was 636,000 mt (1,400,000,000 lbs); if we again use the conservative estimate of 75% attributed to Coville's cultivars and a value of $.95/kg ($.43/lb), this makes a total value of $451,000,000. If the years before 1983 and the years after 1992 are included in the estimate of Coville's economic contribution, along with the production of cultivars grown abroad, the total value of his cultivars would undoubtedly exceed $1 billion.

Dr. Coville's success with domesticating the highbush blueberry and providing the basis for a major fruit industry can only be described as incredible. It is even more incredible to consider that he was a major force in establishing the National Arboretum in Washington and was Chairman of the National Geographic Society during this same period.