J^-|p^;;., f

\h7^ K *

1 ’■ fsSftmt

ZTSin

, ( 8|f .t% y

m ^ Vm *

amoldia

Volume 62 • Number 2 • 2003

Amoldia (ISSN 004-2633; USPS 866-100) is
published quarterly by the Arnold Arboretum of
Harvard University. Periodicals postage paid at
Boston, Massachusetts.

Subscriptions are $20.00 per calendar year domestic,
$25.00 foreign, payable in advance. Single copies of
most issues are $5.00; the exceptions are 58/4-59/1
(Metasequoia After Fifty Years) and 54/4 (A Source-
book of Cultivar Names), which are $10.00. Remit-
tances may be made in U.S. dollars, by check drawn
on a U.S. bank; by international money order,- or
by Visa or Mastercard. Send orders, remittances,
change-of-address notices, and all other subscription-
related communications to: Circulation Manager,
Amoldia, The Arnold Arboretum, 125 Arborway,
Jamaica Plain, Massachusetts 02130-3500.

Telephone 617.524.1718; facsimile 617.524.1418;
e-mail arnoldia@arnarb.harvard.edu.

Postmaster: Send address changes to
Arnoldia Circulation Manager
The Arnold Arboretum
125 Arborway

Jamaica Plain, MA 02130-3500

Karen Madsen, Editor
Mary Jane Kaplan, Copyeditor
Andy Winther, Designer

Editorial Committee
Phyllis Andersen
Robert E. Cook
Peter Del Tredici
Jianhua Li

Stephen A. Spongberg
Kim E. Tripp

Copyright © 2003. The President and Fellows of
Harvard College

Page

2 Shrubs and Vines at the Arnold Arboretum:
A History
Sheila Connor

1 6 Ordering and Terracing in the Leventritt
Garden

Douglas P. Reed &) Gary Hilderbrand

20 Shrubs and Vines for the Leventritt Garden
Peter Del Tredici, Michael Dosmann, Tom
Ward &) Julie Coop

27 Genetic Piracy: A Newly Discovered
Marvel of the Plant World
Richard B. Primack

31 Book Note
Karen Madsen

32 Arnold Arboretum Weather Station
Data— 2002

Front Cover: Clockwise from the upper left, Lonicera
sempervirens (trumpet honeysuckle); Smilax spp.
(briar), both photographed by Peter Del Tredici;
Hamamelis mollis (Chinese witch hazel); and
Viburnum furcatum (forked viburnum), both by
Robert G. Mayer. The background, Ilex cornuta, is
also by Peter Del Tredici.

Inside front cover: Malus prunifolia var. xanthocarpa
on the left; M. 'Wyema' on the right.

Inside back cover: Inside a crabapple canopy looking
out: it is a very good year for crabapples.

Back cover: The center greensward of the M. Victor
and Frances Leventritt Garden, photographed by
Karen Madsen, as were the inside covers.

Erratum: In volume 62, number 1, our last issue, the
photographs of the Leventritt Garden on both front
and back covers were not credited; they are the work
of Alan Ward, author of American Designed Land-
scapes: A Photographic Interpretation.

MAR 0 9 2017

Shrubs and Vines at the Arnold Arboretum:

A History

Sheila Connor

In the Arboretum every effort has been made to foster the natural beauty of
the landscapes and add to them wherever possible, but it has not been able to
overcome the disadvantage of being a public garden. In the arrangement of the
collection of shrubs the aesthetic has had to give way to the practical. Even in
the matter of place there was no option since a considerable flat area of land
was necessary and there is but one in the Arboretum.

— E. H. Wilson, America's Greatest Garden (1925)

For almost one hundred years, gardeners,
homeowners, and students who wanted to
learn about shrubs made their way to a
gently sloping plot of land just inside and to the
right of the Arboretum's Forest Hills gate.
Bounded by the Arborway and Forest Hills Road
on the north and east and by the ponds, natural
woods, and maple collection on the south and
west, the linear arrangement of the three-acre
shrub collection made it easy to identify. Today
the site hosts the Bradley Collection of
Rosaceous Plants, but — as directors' reports and
archival records document — before this trans-
formation began in 1981, it had been home to
the oldest continually maintained collection in
the Arboretum's landscape.

Visually, it was perhaps also the most distinc-
tive. Completely different from the rest of the
Arboretum, where the picturesque reigns
supreme despite the collection's taxonomic
arrangement, the shrub collection's rows of for-
mal, parallel beds probably confounded most
visitors. The very existence of a separate shrub
collection is a reminder of the difficulty of
accommodating in a single design the disparate
requirements of a scientific collection and a
public park. It also reminds us of the sometimes
opposing views held by the two men who

accomplished that task, Charles Sprague
Sargent and Frederick Law Olmsted.

In 1874, when he first approached Olmsted
with the notion that an arboretum would be a
valuable addition to the plan for the Boston park
system, Sargent was 33 and although he had
only "a modest reputation as gentleman land-
scape gardener," his professional standing was
impressive and growing. In May of 1872 he
was appointed professor of horticulture at the
Bussey Institution (founded 1869; closed 1940s),
a position first held by his friend, neighbor, and
mentor Francis Parkman. (Sargent resigned this
position at Harvard's school of scientific agricul-
ture, which abutted on the Arboretum's land, in
1877.) In November of 1873 he became director
of Harvard's botanic garden in Cambridge,
working under Asa Gray, the father of American
botany, and at the same time, he was named
director of Harvard's newly established Arnold
Arboretum in Jamaica Plain; it was in hopes
of linking the new arboretum to the Boston
park system that he addressed his proposal
to Olmsted.

Olmsted, then 55 and still working in New
York — on Central Park, among other projects —
had a national reputation in landscape architec-
ture and had been a proponent of arboreta since

The Arboretum’s shrub and vine collection as photographed from a helicoptor by Massachusetts State
Department of Public Works aerial photographers in fune 1969, the year horticulturist Donald Wyman retired.
The trellis, overwhelmed in places by vines, can be seen at right. From the Archives of the Arnold Arboretum.

4 Arnoldia 62/2

the beginning of his career. Sargent's proposal
may have appealed to him because an arbore-
tum had been included, though never realized,
in the Olmsted and Vaux 1858 plan for
Central Park.

SARGENT, OLMSTED, AND THE
SHRUB GARDEN

The collaboration between the novice botanist
and well-established landscape designer
worked. Under Asa Gray's tutelage, Sargent had
learned the importance of the individual ele-
ments within a landscape, and he valued the
ability to recognize, identify, and compare each
plant to all others. Olmsted, by contrast, looked
beyond individual plants; for him, they were
subordinate to the overall design and composi-
tion. But the two men agreed that the design
for the Arboretum's landscape would have to
satisfy both scientific and aesthetic require-
ments. And as Ida Hay writes in her preface to
Science in the Pleasure Ground (1995) "the
dynamic tension between science and aesthet-
ics turned out to be one of the most interesting
themes of Arboretum history."

Early in the planning process, Sargent insisted
on accommodating "a working or experimental
collection" within the otherwise wooded and
naturalistic design. The idea may in fact have
begun as a practical method for dealing with the
abundance of shrubby plant material already on
hand soon after the Arboretum's founding.
Months before Olmsted's involvement and
before any work had been done on the design, the
future site of the shrub collection served as a
nursery and a holding area for the thousands of
plants intended for the rest of the grounds. Estab-
lishing a nursery and making an inventory of
plants on hand were among the few projects that
Sargent could undertake with "the very limited
means at the disposal of the Director" during
1873, the first year of the Arboretum's existence.

The nursery site was possibly one of the most
fertile areas of what an elderly Sargent much
later recalled as the "worn-out farmland" on
which the Arboretum was established.
Described in 1692 as upland and meadow by its
first recorded owner, Samuel Gore, the land
became part of the farm of Joseph Weld in 1718
and remained in the Weld family until pur-
chased by Benjamin Bussey in
1810. It was also conveniently
adjacent to the Bussey Institu-
tion, where Sargent had estab-
lished an office for himself
and commandeered space in
the greenhouses that were
now to be "devoted to the
raising of forest and ornamen-
tal trees and shrubs for future
plantations" at the Arbore-
tum. Seed for the new arbore-
tum was pouring in from
other institutions through-
out the world, and Jackson
Dawson, the Arboretum's
first propagator (originally
the Bussey's), was making
ambitious collections of
native plants. Within just
three years, Sargent would
report that "to relieve the
overcrowded nurseries, 3,181
forest trees have been
planted out."

Taken from the hillside just above the lilac and forsythia collections, this is one
of the earliest known images of the Arboretum. The upland in the middle
distance became the site of the shrub collection, and the treeline represents the
future route of the Arborway, constructed in the early 1890s. Putting Olmsted’s
design into effect transformed the standing water of the low-lying, wet meadow
into three distinct ponds, two of which were bisected by Meadow Road. This
image is from the Lantern Slide Collection at the Harvard Design School, an
especially comprehensive assemblage of images of early American landscape
design. They were taken between 1850 and 1920 and digitized in 1996 through
a Library of Congress-sponsored competition. The 2,800 lantern slides can be
accessed through the American Memory Page of the Library of Congress (http://
memory.loc.gov/ammem/award97/mhsdhtml/aladhome.html).

Shrub and Vine History 5

Published in the October 1898 issue of New England Magazine, this image shows an early version of the
collection’s layout when the rows of “young shrubs ” extended right up to the edge of the roadbed. Each plant’s
identity was painted on a wooden stake as the signage had not evolved to the “boldly lettered stand labels”
that became the norm for the collection. Taken just inside the Forest Hills gate, the long view shows the
remnant flora of “Woodland Hill, ” Benjamin Bussey’s estate. Several of the mature trees at the crest of the
hill have the characteristic shape of American elms, while the conifers on the right appear to be white pines.
The group of trees in the middle distance occupies the present site of the ash collection just above the lilacs.

By 1878, when Frederick Law Olmsted began
to work on a planting arrangement for the Arbo-
retum, the nursery area had been under cultiva-
tion for five years and occupied much of the
open land near the present-day ponds. Although
one of his early studies suggests that Olmsted
had considered a formal "greeting," or prom-
enade, for this location, it is only in the last
design, dated August 29, 1885, and signed by
John Charles Olmsted, nephew and stepson of
F. L. Olmsted, that the first reference to a plan
for the area appears.

In the Bentham and Hooker plant classifica-
tion sequence, which was used to arrange plant
families within the Arboretum's landscape, the
rose family should have been placed where it is
today, near the ponds. Surprisingly, however,
the plan calls for orchard-like rows of plants at

that location. A likely explanation for this
anomaly is that not only a nursery, but a collec-
tion of shrubs — arranged in rows — already occu-
pied the site.

While at the botanic garden in Cambridge,
Sargent had completed a successful and hand-
some renovation of the systematic beds. He had
weeded duplicate specimens, increased the
number of species, reordered the collection to
correspond with the organization of specimens
in the herbarium, and added grass paths to pro-
vide easy and attractive access to the plants. In
Sargent's estimation he had "brought order out
of Chaos." Faced with a "large and rich collec-
tion of shrubs located in the nursery" at the
Arboretum, it is no wonder that he proposed in
1884 "to arrange this collection ... in system-
atic order . . . until the completion of roads by

6 Amoldia 62j2

the city on the east side of the grounds make it
possible to permanently group the different
shrubs upon a [more] picturesque and natural
plan than can now be adopted."

Although doing so required draining the
ground below the nursery and relocating all the
shrubs, within a single year he could claim that
the "provisional or tentative arrangement of the
shrub collections referred to in my last report
has been completed. These now occupy thirty-
seven parallel beds, each ten feet wide and three
hundred feet long. This collection now contains
about eleven hundred species and varieties
arranged in botanical sequence, with provisions
for a considerable further increase." Thus came
into being the shrub collection whose design
would simultaneously please botanists and hor-
ticulturists, irritate directors and superinten-
dents, challenge landscape architects and
students, and mystify many visitors throughout
its century-long lifetime.

That year, 1885, was in general a momentous
one for Sargent, with several projects begun dur-
ing the previous decade now coming to culmi-
nation. In 1879, on the recommendation of Asa
Gray, Sargent had been appointed to manage the
nation's tenth forest census, and the results of
that study had just been published as The
Report of the Forests of North America. At the
same time, an exhibit of over 500 samples of
wood, amassed by Sargent for Morris K. Jesup,
was about to open at the American Museum of
Natural History in New York; the exhibit,
called "The Woods of the United States," dis-
played Sargent's notes on each species' struc-
ture, qualities, and uses. Lobbying for the
Arboretum to be included in the Boston park
system had succeeded, and the indenture
between the city and Harvard had been signed
three years earlier. And, most significant of all
for the Arboretum, the city had finally built
enough roads so that the systematic planting of
the "permanent" collection could begin.

At the end of the year, a confident Sargent
described his vision for a public arboretum in
the 1884-1885 Annual Report of the Director of
the Arnold Arboretum to the President and
Fellows of Harvard College; he had formally
accepted the Olmsted firm's final plan for the
planting arrangement and declared the first

of the Arboretum's collections — the shrubs —
completed. By this time he had decided that
two distinct collections should be maintained
within the Arboretum: the first, or "permanent
collection for display, [would consist] of a selec-
tion of species intended to illustrate . . . the
most important types of arborescent vegeta-
tion"; while the second, a "collection for
investigation, which need not necessarily be
permanent . . . should be arranged in a manner
to permit the admission of . . . new forms and
the removal of others which have served their
purpose. To this second collection would natu-
rally be joined all minor collections like that of
shrubs and other plants of less enduring charac-
ter than trees."

Olmsted's feelings about the displacement
of Rosaceae by Sargent's shrub collection are
unknown. Little correspondence between the
two men survives — perhaps because, as Sargent
testified many years later at a Park Department
hearing, the two "spoke daily." The conversa-
tions between Olmsted and Sargent about the
placement of shrubs must have been interest-
ing. One letter, written by Sargent to Olmsted
on April 22, 1888, when the tree groups were
being set out on the hillside above the ponds,
suggests that discussions about plant placement
continued for years after Olmsted's final plans
had been accepted. Of today's much beloved
lilac collection and its bright yellow compan-
ions and fellow harbingers of spring, Sargent
wrote, "Isn't it a mistake to plant forsythia,
syringa [lilacs] and other showy flowered
garden shrubs on the Arboretum Hill? I should
be afraid that they would not harmonize with
the general scheme of planting . . . How will a
mass of bright colored garden flowers look rising
above the softer first tints seen everywhere else
in the Arboretum?"

Four months later the debate about the use of
"showy" exotic shrubbery spilled over into the
pages of Sargent's newly founded Garden and
Forest, a fournal of Horticulture, Landscape
Art, and Forestry. The first sally in the debate,
appearing in the issue of August 1, 1888, was
an unsigned note placed among the editorial
articles. The sentiments may have come from
the journal's knowledgeable New York editor,
Williams Stiles, but they are as likely to have

Shrub and Vine History 7

originated with the journal's "conductor,"
Charles Sprague Sargent:

It is not easy to explain why certain plants look
distinctly in place in certain situations and why
other plants look as distinctly out of place in the
same situations. This is a matter which nature
perhaps has settled for us. It is certain at any rate
that combinations of plants other than those
which nature makes or adopts inevitably possess
inharmonious elements which no amount of
familiarity can ever quite reconcile to the edu-
cated eye. Examples of what we wish to explain
abound in all our public parks, and especially in
Prospect Park in Brooklyn . . . where along the
borders of some of the natural woods and in con-
nection with native shrubbery great masses of
garden shrubs, Diervillas, Philadelphus, Deut-
zias, Forsythias and Lilacs, have been inserted.
These are all beautiful plants. They never seem
out of place in a garden; but the moment they are
placed in contact with our wild plants growing
naturally as they do, fortunately, in the Brooklyn
park, they look not only out of place, but are a
positive injury to the scene.

As the designers of Prospect Park had been
Olmsted and Calvert Vaux, it is not surprising
that the editorial elicited a prompt reply from
Mr. Olmsted.

To the Editor of GARDEN AND FOREST:

Sir.— In GARDEN AND FOREST of August
1st, page 266, the law seems to me to have been
laid down that the introduction of foreign plants
in our scenery is destructive of landscape repose
and harmony. No exception was suggested, and
the word harmony was used, if I am not mistaken,
as it commonly is in criticism of landscape paint-
ing, not of matters of scientific interest.

The question, as we understand it, is essen-
tially this: Would all of the trees and bushes that
had come of a foreign ancestry be noted before
any of the old native stock?

It appears to us . . . that the American
Chionanthus, Angelica, Cercis, Ptelia, Sumachs,
Flowering Dogwood, Pipevine and Rhododen-
drons would be placed before some of the foreign
Barberries, Privets, Spireas, Loniceras, For-
sythias, Diervillas or even Lilacs. We doubt if the
stranger, seeing some of these latter bushes
forming groups spontaneously with the natives,
would suspect them to be of foreign origin . . .

Frederick Law Olmsted,
Brookline, September 1888

Perhaps the debate had ranged too widely — or
come too close — for at this point the editor
stepped in:

Mr. Olmsted's letter should be read with the
greatest care and attention. No man now living
has created so much and such admirable land-
scape, and no man is better equipped to discuss
all that relates to his art. The position which
GARDEN AND FOREST has taken upon the
question of composition in plantations made
with the view of landscape effect is embraced in
the following sentence, extracted from the
article to which Mr. Olmsted refers: "It is cer-
tain, at any rate, that combinations of plants,
other than those which nature makes or adopts,
inevitably possess inharmonious elements
which no amount of familiarity can ever quite
reconcile to the educated eye." This sentence
was written with special reference to the fact
that in Prospect Park, in Brooklyn, various
showy flowered garden-shrubs of foreign origin
had been massed among native shrubs growing
apparently spontaneously along the borders of a
natural wood in the most sylvan part of the park.
The effect which this combination produced
appeared to us inharmonious, and therefore less
pleasing than if the plantation had been confined
to such shrubs as may be found growing natu-
rally on Long Island in similar situations. How
far the idea of harmony in composition in
landscape is dependent upon association it is
hard to say.

The truth is that great masters of landscape
construction can combine material drawn from
many climates and many countries into one har-
monious whole, but the masters of the art are
not many, and the planter who is not sure of his
genius can wisely follow nature in her teachings
of harmony in composition. Had this reservation
been made in the article referred to, our state-
ment that "all attempts to force Nature, so to
speak, by bringing in alien elements from remote
continents and climates, must inevitably pro-
duce inharmonious results," would, perhaps,
have been less open to criticism. — ED.

But by then the Arboretum's Diervilla, Phila-
delphus, and Deutzia (though not the large col-
lections of Syringa and Forsythia) had been
relegated to the shrub collection, away from
the more "naturalistic" and "permanent" plan-
tations. But this disposition may only have
brought more trouble for Sargent. In almost
every succeeding annual report he showed his

8 Arnold ia 62/2

Although undated, this photograph from the J. Horace McFarland Collection was probably taken during the
winter of 1905-1906, shortly after the construction of the vine trellis that enclosed the shrub garden on three
sides. McFarland (1859-1948), a lifelong friend of Charles Sprague Sargent and a well-known author, publisher,
horticulturist, and rosarian, wrote as floridly on the Arboretum as he did on roses. In “A Tree Garden to Last
a Thousand Years, ” an article published in the first volume of “The Country Calendar” in 1 905, he emphasized
that “Even in the ‘Order, ' which is the name of the amphitheater of single specimens of shrubs, arranged in
botanical order for easy comparison and study, there is informal formality and continuous beauty. ”

Part of the Smithsonian Institution Archives of American Gardens, which hold approximately 80,000
photographic images and records documenting historic and contemporary American gardens (available
online at http://gardens.si.edu/horticulture/res_ed/AAG/coll-mcf.htm) the McFarland Collection includes
over 3,100 black-and-white mounted photographs and 445 glass lantern slides of gardens throughout the
United States and dates from 1900 to 1962.

obsession with the collection's ongoing needs
and with his pursuit of the perfect design. If he
was not "improving its arrangement," he was
"devoting more time to studying and improv-
ing" it, or had to "extend and rearrange" it by
"lifting every plant and rearranging the beds."
One design revision — presumably made to
accommodate more plants — changed the arrange-
ment from 37 seven-foot beds to 15 eleven-foot
beds, achieving an increase of only five feet
(from 8,325 to 8,330). The collection was clearly

"suffering from the want of proper space," and
he worried how it would be "adequately pro-
vided for in the future." By 1906, the collection
had undergone three complete rearrangements,
and Sargent admitted that although it was a
principal feature of the Arboretum and its most
complete collection, "the arrangement has not,
however, proved entirely satisfactory."

After one last attempt at redesign, this time
with the plants displayed in 19 beds, each ten
feet wide, and occupying a total of only 7,765

Shrub and Vine History 9

linear feet, Sargent conceded defeat. In 1907,
he reports that it has been necessary to plant
genera formerly accommodated in the shrub
collection along the drives and in other parts of
the arboretum. The "excluded" groups include
the rhododendrons, azaleas, kalmias, and vibur-
num and dogwood collections. Mention of the
shrub collection subsequently disappears from
his annual reports.

He does, however, introduce a new element
into the design. A ten-foot tall, 1,280-foot-long
vine trellis was erected to provide shelter for the
beds along the northern and eastern sides of the
collection. Made of concrete posts strung with

galvanized wire, it was the Arboretum's first
attempt to bring together a collection of vines
and other climbers. The trellis itself, not unlike
the layout of the shrub collection, would be
modified and rebuilt, but the two collections
would soon be thought of as one: the shrub and
vine collection.

As for those conspicuous plantings on Bussey
Hill, by the 1920s Sargent appears to have had a
change of heart. Writing in The Bulletin of
Popular Information, a seasonal guide to plants
made available to Arboretum visitors for a dol-
lar a year, he notes, "Forsythias are often badly
planted; they require space in which to spread

“The shrubs are brought together in their respective genera and natural orders and there is only a single row
of plants down the centre of the border. This allows each plant to stand on its own ground without interference
from its neighbor, and the convenience of the student is further aided by each specimen having a label on
either side. The general idea is somewhat similar to that of the arrangement of the herbaceous plants at Kew.
Considered as part of the landscape this system is not beautiful, especially in the early stages, when the plants
are not fully grown and bare spaces yet await their destined occupants. But for purposes of botanical study no
other arrangement is so convenient. ” So wrote W. J. Bean, Curator at the Royal Botanic Gardens, Kew, and
author of Trees and Shrubs Hardy in the British Isles, in the Bulletin of Miscellaneous Information of the Royal
Garden, Kew, in October 1910, after his visit to the Arnold Arboretum.

10 Ainoldia 62/2

their long, gracefully arching branches and are
not suitable for small gardens. To be most effec-
tive they should be planted as in the Arboretum,
in a great mass on a bank or hillside."

The lilacs were also given their due by
Sargent in the Bulletin, "The flowers of no other
plant bring so many visitors to the Arboretum."
Beginning in 1915, in what surely gave rise to
the now famous Sunday celebration, Sargent
began predicting when the lilacs would attain
their fullest bloom. In 1920 he forecast that

"Most of the varieties of the common lilac will
be in flower when this Bulletin reaches its
readers living near Boston." In May of 1920, he
wrote, "The plants will be in full bloom by
Saturday, the 29th," and when the spring of
1924 proved to be slow in coming, he com-
mented that "They are late blooming this year
but are now fast opening their flowers, and
it is possible that Sunday, the 25th, will bring
the largest number of visitors of the year to
the Arboretum."

The first trellis, constructed of ten-foot high posts of reinforced concrete set fifteen feet apart, provided seven
parallel strands of wire for the vines to climb. According to E. H. Wilson, with this layout, all that was needed
to keep the shrubs from interfering with their neighbors was the use of “a knife in late winter. ” Apparently
some did overstep their bounds. Beginning in the mid 1930s, many of the most “aggressive” were removed and
grown on the stone walls. While working on the rearrangement of the shrubs in 1 950, Farr and proposed letting
“delicate creepers” such as species clematis and single roses scramble over the rock walls or climb trees. She
thought that “loose growing crabapples, small cherries, or pears would please the clematis tribe” and asked
to be provided with a list of “tree-climbing creepers. ”

Shrub and Vine History 1 1

WILSON AND THE SHRUB GARDEN

When E. H. Wilson's book about the Arboretum,
America's Greatest Garden, was published in
1925, most of the chapters had fanciful titles
such as "Spring Pageantry," "Summer Luxu-
ries," "Cherry Blossom Festival," "Crabapple
Opulence," and "Azalea Carnival." By contrast,
the chapter appearing second to last had the
unadorned title "The Shrub Garden," followed
only by "What the Arboretum Does." The chap-
ter on the shrub collection opens with Wilson's
declaration that "A garden where the conve-
nience of the public has to be shown preference,
labors under disadvantages unknown to private
gardens where landscape effects alone have to be
considered." Acknowledging that the shrub gar-
den was established for the instruction of stu-
dents, landscape gardeners, and plant lovers, and
that its long, formal beds gave each plant ample
room to develop, he described its arrangement
as practical, convenient, and useful.

The design, he thought, did have some draw-
backs: although every shrub was identified by
name on two "boldly lettered" labels, one facing
the grass paths on each side of the plant, some
careless visitors made shortcuts across the beds
and "many a small plant has carelessly been
trodden to death as a consequence." As to its
location, he could not say enough. It was the
coldest spot in the Arboretum in the winter, and
the hottest in the summer; and because it was
low-lying, its air drainage was the poorest, so
that it suffered the first frosts in fall and the
last frosts of spring. If nothing else, it made an
splendid testing ground for plant hardiness — in
Wilson's estimation, if a plant could survive in
the Arboretum's shrub collection it would
"withstand the winter's cold and summer's
heat of any part of New England." Regardless
of its lack of "landscape beauty" and its less-
than-ideal site, he thought there was probably
"no more instructive a collection of shrubs
in existence."

Between Sargent's death in 1927 and his own
in 1930, Wilson, as "Keeper of the Arboretum,"
dutifully took up authorship of the Bulletin. Of
the shrub collection he wrote that it "is a never
failing source of interest to all visitors, filled as
it is with a general miscellany of shrubs." He

also liked to point out the attractiveness of
various Asiatic representatives, many of which
were his own introductions.

WYMAN AND THE SHRUB GARDEN
The collection's greatest champion since
Sargent, Donald Wyman, was appointed horti-
culturist, a newly minted position, in 1936. In
1970, a year after his retirement, Wyman
reflected on the state of horticulture at the
Arboretum when he first arrived there, shortly
after the deaths of both Sargent and Wilson,
describing it as at an "all time low," with a pro-
nounced "lack of staff interest in the living col-
lection." In his opinion, "there was much work
to be done."

Oakes Ames had replaced Sargent, though as
supervisor rather than director. During his
administration, from 1927 to 1935, specialists
in plant pathology, wood anatomy, and genetics
filled new staff positions. Elmer Drew Merrill,
who succeeded Ames in 1935 and led the Arbo-
retum until 1946, created the position of horti-
culturist and hired Wyman, who remained to
serve under two more directors, Karl Sax (1946-
1954) and Richard A. Howard (1954-1978).

Wyman valued shrubs, as well as
groundcovers, hedges, vines, small trees,
"choice evergreens that never grow tall," the
best of flowering crabapples and lilacs, and all
plants ornamental. As Sargent had championed
trees, Wyman carried the banner for shrubs — his
Shrubs and Vines for American Gardens came
out two years before its much slimmer compan-
ion volume, Trees for American Gardens. A pro-
lific author, his bibliography reflects his
leanings: generally, the more ornamental a
plant, the better, and even better yet if the plant
was a shrub. He wrote articles about shrubs on
the color, sequence of bloom, and seasonal
interest of early, late, and summer blooming
ones; he wrote about those not to be overlooked;
those that were dwarf, fruited, and rare,- and
those that could be used for hedges or in the
shade. He listed "Forty-five of the Best Shrubs
for Massachusetts Gardens" in Arnoldia (1951,
vol. 11, no. 1), suggesting that shrubs could "fill
your garden with color," but cautioning, in an
October 1950 issue of Horticulture magazine

12 Ainoldia 62/2

(vol. 28), "Be Careful with These Colored Foli-
age Shrubs." His advice in a November issue
(1947, vol. 25) was to "Look Around This Month
for Fruited Shrubs to Plant."

He introduced the practice of displaying "the
best, most ornamental" plants in the Arbore-
tum. From a horticultural standpoint, the
opportunity to compare plants growing side-by-
side — especially the many that were by now
mature specimens — was distinctly useful. So
valuable did he find the shrub collection, now
numbering over a thousand plants, that in
"How to Spend an Hour in the Arboretum,"
published in 1945 in Ainoldia (the renamed
Bulletin of Popular Information ), he suggested
that visitors with only a limited amount of time
could best spend it "examining the long rows of
shrubs many of which are commonly grown in
nurseries and hence are available to every home-
owner." He concluded the article with these
remarks: "This shrub collection is one of the
very few places in this country where so many
different kinds of woody plants can be seen and
closely examined in such a small area. An hour
spent here is well worth while to both amateur
gardeners and professional plantsmen."

An additional motivation for writing the
article may have been that Wyman now knew
each of the shrubs intimately, having just com-
pleted his own revamping of the collection. This
rearrangement was driven more by the necessity
to reduce the amount of hand labor needed to
manage the collection than to increase its num-
bers or to enhance its display. The institution
was on the cusp of recovering from war-related
labor shortages that had exacerbated the effects
of the hurricane of 1938 as well as ongoing sea-
sonal damage due to "snow and ice . . . storm
and fire." In Merrill's annual report for 1945-
1946 he decries the condition of the grounds,
admitting, "some of the plantings actually
approached a deplorable condition" and hoped
to "go far in repairing the ravages due to
neglect." One laborsaving cost reduction
Merrill could report on was that "through the
rearrangement of the very large shrub collec-
tion, permitting the use of mechanical equip-
ment, the actual maintenance cost has been
reduced to one-fifth of what it was."

A NEW COLLABORATION: WYMAN
AND FARRAND

To help restore the collection, Beatrix Farrand
was "added to the staff, on a retainer basis, as a
consulting landscape gardener." According to
Donald Wyman, "Although many had tried to
obtain permission to become consulting land-
scape architects," she was the first landscape
architect since Olmsted to work at the Arbore-
tum. Farrand's association with what she con-
sidered her alma mater had begun in the 1890s,
when as a young student she had studied at
the Arboretum under Sargent's guidance. While
Farrand may have known the collections well,
in Ida Hay's estimation her "grasp of the
raison d’etre behind the Arboretum's original
layout and subsequent development was some-
what sketchy."

The Arboretum's archives include much of
Farrand's correspondence and other writings
regarding her plans for rejuvenating the collec-
tions. The notes she made after a visit to the
Arboretum in the spring of 1947 include her
first mention of possible changes to the shrub
collection. "The remodeling of the present
shrub collection should be done as soon as
financially possible. All the formal beds in
straight rows should be removed and the space
now occupied by their stiff lines made into a big
meadow, where moisture loving trees might be
planted ..." She pointed out that there was suf-
ficient space nearby "to exhibit a collection of
the best shrubs, choosing the most attractive
varieties and relegating the less good sorts to the
'study' nursery whether at Weston [the recently
acquired Case Estates] or elsewhere." She also
asked for a list of the shrubs (presumably from
Dr. Wyman) "with comments on their attrac-
tiveness and their size and condition . . . some
would be banished . . . and others, perhaps the
newer sorts, would be added." Wyman, just
having successfully laid the shrub collection
"to rest," must have wondered just exactly
where the role of the horticulturist intersected
with that of the consulting landscape gardener.
The Wyman-Farrand collaboration was to prove
every bit as challenging as that of the earlier
team of plantsman and landscape architect,
Sargent and Olmsted.

Shrub and Vine History 13

In 1946, in an effort to eliminate hand labor and reduce costs, a “judicious
arrangement” of the shrub collection’s beds coupled with replacement of
alternate grass walks with harrowed swaths and repositioned labels enabled
the use of mechanized equipment and weed killers. A Ford-Ferguson tractor is
seen here harrowing between the rows for weed control.

Donald Wyman experimented with a series of weed killers that ran the gamut
from using flame throwers to “burn off”’ poison ivy to spraying sodium
arsenate and sodium chlorate on annual weeds. He also introduced spent hops
(probably acquired from local breweries in Jamaica Plain) as a weed-
controlling mulch. In this photograph “Dowpon” (2,2-dichloropropionic acid)
sprayed on the grass is used to edge the beds.

Donald Wyman and Beatrix
Farrand did agree on the state
of the grounds. Faced with an
overcrowded and undermain-
tained landscape, they both saw
a need for better management.
While Wyman wanted to select
the most ornamental material
and weed out the rest. Farrand
wanted to eliminate "duplicate
and over-aged plants." But
she too wished to "display to
advantage the best and most
ornamental of the plants now
growing at the arboretum,"
according to her 1946 article
"Contemplated Landscape
Changes at the Arboretum."

Wyman did act on many
of Farrand's suggestions. Al-
though it bore no relationship
at all to the Bentham and
Hooker system, the azalea
border along Meadow Road,
arranged by color and designed
by Farrand to present a
sequence of bloom, was attrac-
tive and an instructional addi-
tion. The north slope of Peters
Hill was rehabilitated, as was
part of the top of Bussey Hill.
The dogwoods were relocated,
some along Bridal Path and
others on Bussey Hill. The
shrub collection itself, how-
ever, remained inviolate.

Between 1947 and 1952,
while her other suggestions
were being carried out,
Farrand's proposed renovation
of the collection was the sub-
ject of over fifteen letters
between her and Karl Sax,
the Arboretum's director. Sax
thought at first that there
"were some differences of opin-
ion regarding this project.
Wyman thinks that the present
arrangement is much less

14 Amoldia 62/2

Beatrix Fanand envisioned a delicate rose arbor to replace the trellis, but the
unanticipated relocation of the wisteria collection in 1951 required the
construction of a very long, substantial edifice of cedar posts. Farrand, ever
gracious, wrote to Sax inquiring about its design. “ The news you have finished
the arbor and planted the wisterias is thrilling. You are most secret as to what
the arbor is made of, how large and long it is, so I am in a twitter to hear more
about it. ” Sax’s reply was less than enthusiastic. “The new arbor in the shrub
collection looks a bit raw at present, but it will mellow with time, I hope. As is
the case in all our work we have to make some compromise between expense,
botanic garden objectives and landscape design. ”

expensive to maintain," and though Sax is clear
that he had "never liked the present arrange-
ment . . . some of the other people seem to like
to have the shrubs in an orderly catalogue col-
lection as they are at present."

The next year, however, he
wrote, "I can assure you that
none of us like the rigid
arrangement of the shrub col-
lection, but if we are going to
maintain the shrub collec-
tion on the Arboretum
grounds there is not much
alternative." She argued the
need "to consider the even-
tual elimination of these
rigid beds, situated where
Professor Sargent felt they
were not wisely placed from
the point of view of design."

By 1950 Sax was asking if
Farrand had "figured out how
our 600-700 shrubs can be
properly arranged in the
available space?" She replied
that "indeed I am struggling
with the monumental prob-
lem of how to rearrange the
600-700 shrubs in a less grid-
iron manner" and asked, "Do
the beds really need to be 20'
wide or could 10' beds and 8'
walks be substituted?" She
had calculated that she could
fit in 665 shrubs, feeling that
the number "can probably be
squeezed up to the 700 by
crowding some of the smaller
species," and she sent him a
quick drawing. Receiving no
reply she wrote, "Was the
plan so bad that it has caused
you to faint and fall by the
wayside, or have you any
comment on which we can
start our drawing?"

Actually faced with a plan,
even though it was only a
sketch, Sax changed the
thrust of his argument, now
emphasizing maintenance costs as the reason
for squelching the project. "Certainly the
revised plan of the shrub collection is a great
improvement in design, but the initial cost
would be high and the subsequent maintenance

Shrub and Vine History 15

would be increased at least three-fold." His next
letter assured her that "we realize the need for
re-vamping the shrub collection . . . but it looks
as though we are due for a continued and per-
haps severe inflation. We do not want to develop
projects which would add materially to future
maintenance costs until we can see some way
of meeting these financial obligations." She
replied, "Indeed your dilemma and problem are
most sympathetically understood here by a very
aged and devoted friend of the Arboretum/' but
"Perhaps — and this is but a suggestion — it
might be worth thinking over doing over a small
section at a time ..."

Only one more mention of the shrub collec-
tion appears in their correspondence. Sax
wrote, in 1951, that another problem had arisen
that would further complicate her revision: the
wisteria collection had to be moved from the
grounds of the Bussey Institution, requiring
that a new arbor be quickly constructed adja-
cent to the shrubs. As Farrand's plans had
always included a design for the arbor — a very
delicate design — the wisterias' needs finally
brought an end to the project.

Wyman's recollections in "Horticulture at
the Arnold Arboretum, 1936-1970" ( Arnoldia
vol. 30, no. 3) support the maintenance issues
and raised additional objections that were per-
haps not voiced in the 1950s. "When [Farrand]
recommended that the shrub collection be
removed and the shrubs planted in small groups
in the same area, it became obvious to both Dr.
Merrill and Dr. Karl Sax that some of her ideas,
if carried out, would cost far too much in day-to-
day maintenance. It has always been my strong
conviction that landscape planning should be
left to those on the staff who have had landscape
training and who understand the diverse practi-
cal problems of maintaining the plantings."

Wyman used the shrub collection as a teach-
ing tool and as an important resource in his
famous series of Friday morning Arboretum
"walks." But with his retirement in 1969 the
collection lost its spokesman and defender, and
the next decade saw a decline in the number of
new plantings. The diversity of the collection —
always noted for its encyclopedic content — began
to diminish, with large gaps appearing when
aging plants died and were not replaced. The

sharply edged grass paths had long since disap-
peared. Mechanized cultivators (rather than
more careful human hands) had been used since
the mid 1940s to keep weeds between the rows
to a minimum; nonetheless, woody weeds
began to overtake some of the specimen plants
and the entire collection took on a forlorn and
unkempt appearance.

THE FINAL CHAPTER

Peter Ashton's arrival from Britain in 1979 to
replace the retiring director, Richard Howard,
coincided with a renewed interest in the design
legacy of Frederick Law Olmsted. Assessing
the grounds from this point of view, and with a
knowledgeable and practiced eye for naturalistic
landscapes, Ashton was astounded by the sight
of the shrub collection. It had been so much a
part of the Arboretum's landscape for so long
that no one on the staff questioned its existence,
location, or design, but Ashton felt strongly that
it not only jarred with the Olmsted design it
was, in fact, an eyesore.

Investigations at what was soon to become
the Frederick Law Olmsted National Historic
Site turned up a full set of plans for the Arbore-
tum that showed Olmsted's original intentions
for this part of the grounds. Ashton was excited
by the opportunity to finally put the rose family
in its rightful place, considering this a project
that could initiate a restoration of the entire
grounds to the Olmsted-Sargent plan. Through
the generous bequest of Eleanor Cabot Bradley,
a longstanding member of the Arboretum's vis-
iting committee and an avid plantswoman in
her own right, the garden of rosaceous plants
took shape on the site of the shrub collection. It
had taken over a hundred years to implement
the last piece of Olmsted's plan; it would take
two more decades, a new director, and another
generous woman, Frances Leventritt, to find a
new site and to perfect a design for the shrub
and vine collection that realized Sargent's goal
for "a collection for investigation, which need
not necessarily be permanent, and which should
be arranged in a manner to permit the admis-
sion of . . . new forms and the removal of others
which have served their purpose."

Sheila Connor is horticultural research archivist at the
Arnold Arboretum.

Ordering and Terracing in the Leventritt Garden

Douglas P. Reed and Gary Hilderbrand

We think of gardens as one of our funda-
mental means of cultural expression:
they embody particular ideas of use,
or cultivation, commemoration, or other
aspects of human enterprise. The Arnold Arbo-
retum's new M. Victor and Frances Leventritt
Garden for sun-loving shrubs and vines, as an
enclosed garden — a hortus conclusis — within
the larger Arnold Arboretum landscape, gives
expression to an essential question of botanical
and horticultural purpose: How should plants
be arranged for study and display?

The garden's layout pursues the question
with directness in two ways. First, the land, a
fan-shaped parcel that slopes more than thirty
feet, has been reshaped into level terraces of
varying width to fit the site's irregular outline.
This reshaping is rooted in the ancient practice
of leveling ground for cultivation where the land
is not naturally flat. Second, plants are arranged
in rows or within geometric groups, a familiar
method stemming from the most basic of horti-
cultural traditions. Geometry aids in our work
with nature.

Anyone who has grown their own lettuce or
apples or corn knows the importance of this
practice: for planting, for harvesting, for pruning
and irrigating, some system of order is required;
planting in rows on level ground makes sense.
During the sixteenth and seventeenth centuries,
when emerging medical faculties in European
universities began to build knowledge in plant
culture and taxonomy, they organized "physic
gardens" — at St. Galls, Oxford, Leiden, Padua —
in rectilinear or curving rows, with an eye to the
aggregate pattern as an artful exercise in order.
The Italians, who possessed almost no flat land,

Bethany Grasso, shrub and vine gardener, planting
the Leventritt Garden. The lath house of the Larz
Anderson Bonsai Collection is at center, to the left
of an imposing sugar maple. Metal trellises, soon to
be covered in vines, range along the top two walls.

Design

17

Ordering and Terracing in the Leventritt Garden

Douglas P. Reed and Gary Hilderbrand

We think of gardens as one of our funda-
mental means of cultural expression:
they embody particular ideas of use,
or cultivation, commemoration, or other
aspects of human enterprise. The Arnold Arbo-
retum's new M. Victor and Frances Leventritt
Garden for sun-loving shrubs and vines, as an
enclosed garden — a hortus conclusis — within
the larger Arnold Arboretum landscape, gives
expression to an essential question of botanical
and horticultural purpose: How should plants
be arranged for study and display?

The garden's layout pursues the question
with directness in two ways. First, the land, a
fan-shaped parcel that slopes more than thirty
feet, has been reshaped into level terraces of
varying width to fit the site's irregular outline.
This reshaping is rooted in the ancient practice
of leveling ground for cultivation where the land
is not naturally flat. Second, plants are arranged
in rows or within geometric groups, a familiar
method stemming from the most basic of horti-
cultural traditions. Geometry aids in our work
with nature.

Anyone who has grown their own lettuce or
apples or corn knows the importance of this
practice: for planting, for harvesting, for pruning
and irrigating, some system of order is required;
planting in rows on level ground makes sense.
During the sixteenth and seventeenth centuries,
when emerging medical faculties in European
universities began to build knowledge in plant
culture and taxonomy, they organized "physic
gardens" — at St. Galls, Oxford, Leiden, Padua —
in rectilinear or curving rows, with an eye to the
aggregate pattern as an artful exercise in order.
The Italians, who possessed almost no flat land,

Bethany Giasso, shrub and vine gardener, planting
the Leventritt Garden. The lath house of the Larz
Anderson Bonsai Collection is at center, to the left
of an imposing sugar maple. Metal trellises, soon to
be covered in vines, range along the top two walls.

Design 1 7

1 8 Arnoldia 6212

Q

O

D

C

Q

The vine supports in Claude Monet’s garden at Giverny, France, is one of
many precedents for simple, straightforward, metal trellises.

At Monticello Thomas Jefferson designed a thousand-foot-long terrace
for his vegetable garden, which was both practical — a source of food —
and experimental.

mastered the art of terracing as their primary
spatial means in display gardens, meanwhile
developing a tendency toward aggrandized pro-
cessional space across and between terraces for
spectacle and ceremony. French seventeenth-
century garden designers elaborated these pat-
terns to the point of extreme formalism, but
always rooted their gardens in the measured
logic of ordered planting.

Garden traditions were carried to the New
World by collectors and farmers alike. Planters
such as Thomas Jefferson and George Washing-
ton pursued an American model of husbandry

on large landholdings, carving mag-
nificent estates out of old growth
forest and patchwork farms. They
organized portions of their land to
capture views and exploit scenic
possiblities, and with the luxury of
vast holdings, they anticipated a
more romanticized view of the land-
scape for Americans. Yet when it
came to planting for production or
display, both relied on practical
linear arrangements and simple
terracing. The rational beauty of
these devices can be seen today in
the restored gardens of Monticello
and Mount Vernon.

For us, the task of designing a new
shrub and vine garden at the begin-
ning of the twenty-first century pre-
sented a clear challenge: to provide
an arrangement that would satisfy
the curatorial needs of the Arbore-
tum staff and to develop a memo-
rable spatial experience of the site.
Here we found a convergence of
horticultural science and landscape
design that evolved into a unique
expression of program and site.

The advantages of terracing the
site were obvious. Level terraces
accommodate the planting beds,-
slopes, ramps, and stairs provide
accessible routes throughout. A main
terrace, separated from the adjacent
drives and greenhouse area by a stone
wall nearly 500 feet in length, pro-
vides a setting for the garden's out-
door pavilion and principal space for
public events. This terrace also organizes the
majority of trellises for vine display. From here,
successively wider terraces descend toward the
lower portion of the site. The outlines of these
terraces reflect the topographic form of the site,
and the shrub beds also conform in shape and
proportion. A gently sloping arc of lawn, also
envisioned as a gathering space, cuts across
each terrace and provides a visual center to
the garden.

With the shape and orientation of display ter-
races established, the collections were orga-
nized according to horticultural criteria — soil

Design 1 9

The Leventritt Garden in plan. A is the central lawn-, B is left of the pavilion; C marks the vine structures-,
D the Larz Anderson Bonsai House; E the Dana Greenhouse complex.

and moisture requirements, exposure and shade
tolerance, size and growth habit — to reinforce
the spatial form of the site. A distribution of
small deciduous trees and groups of broadleaf
evergreens are distributed across the scheme to
amplify the garden's structure and add winter
interest. The garden's edges are used to fulfill
certain curatorial objectives, including the dis-
play of plants from the woodland edge commu-
nity, and also accommodate relocated plants
from the dwarf conifer collection that inhabited
the site prior to construction.

Seen from the vantage point above the garden's
main terrace, the scheme is expansive and direc-
tional, with the site's roughly triangular shape
emphasizing an arcing lawn and path that ges-
ture back toward the Arboretum's open lawns
and pervasive canopy trees. Yet when viewed
from below, as the true breadth of terraces and
masonry walls becomes more apparent, the

scheme's dual nature is realized: it is both a tra-
ditional terraced display and an active, modern
sculptural form whose essential expression is
derived from the site's specific conditions.

These design intentions bring coherence to
the garden and establish an identifiable spatial
image for the site. The earthwork and masonry
efforts required to achieve this were monumen-
tal, but they were well within the enduring
traditions of garden making. Planting the collec-
tions and bringing them to maturity will take
time, but the order and structure are visible
at the outset and the real work of the
garden — research and display, cultivating and
maintaining — is well under way.

Douglas Reed and Gary Hilderbrand are principals
of Reed Hilderbrand, the Leventritt Garden's land-
scape architects. They collaborated with Maryann
Thompson, Architect.

Sun-Loving Shrubs and Vines for the Leventritt Garden

Peter Del Tredici, Michael Dosmann, Tom Ward , and Julie Coop

The position of the shrub is distinctly secondary, and the burden of the inferior
race is upon it. A tree may be valued for what it is, but a shrub is rated for just
what it can do. It must render a service to compensate for its cultivation. This
service may be one of beauty, through its flowers; or of use, by its fruit; or its
foliage or habit of growth may be especially attractive, or of such a nature as will
give it value as a shield or a cover for waste and barren places.

— Harriet Keeler, Our Native Trees and How to Identify Them, 1 900

The Leventritt Garden adds a new compo-
nent to the Arboretum's collections — a
display of sun-loving shrubs and vines
suitable for southern New England. Unlike the
main Arboretum collections, which are princi-
pally intended for scientific research, the new
garden has been designed to demonstrate the
horticultural qualities of both species and
cultivars. Nonetheless, it will contribute to
the Arboretum's scientific mission by adding
genera — particularly vines and herbaceous
perennials — in botanical families that are
important to our research.

Selecting the Plants

Working from fall 2000 through spring 2001, our
plant selection committee — the authors, land-
scape projects manager Laura Tenny Brogna,
former assistant education director Ellen
Bennett, gardener Bethany Grasso, and intern
Stacey Berghammer — generated the list of
plants to be included, first establishing our cri-
teria. We agreed to give preference to plants
that meet at least two and preferably three of
the criteria in any of the three categories, horti-
cultural, botanical, and educational. Many of
our selections already exist in the Arboretum's
collections, and many more will be acquired
from other botanical collections and commer-
cial sources.

Criteria for Horticultural Display

Multiseasonality. In most instances, plants
of single-season interest will not be included.
To ensure interest year-round, preference will
be given to plants with a minimum of two, if
not three, seasons of ornamental interest,
whether bloom, fruit display, autumn leaf color,
or texture.

Proven performance. Most shrubs will be
selected for outstanding horticultural merit,
that is, nursery selections that are superior in
size, color, or persistence of bloom; in texture,
form, or color of foliage,- or form and stature of
habit. Choices will be determined by perfor-
mance in the Arboretum as well as in other
botanical gardens and landscapes. These will
include the best of the dwarf conifers previously
located on this site. Some of the vines will also
be chosen for horticultural merit; the genus
Clematis will be well represented, as will
Lonicera and Wisteria.

Botanical Criteria

Genera and species that are part of impor-
tant BOTANICAL FAMILIES BUT ARE NOT WELL
REPRESENTED IN THE MAIN COLLECTIONS. These

include shrubs that do not prosper in the main
collections because they lack sun or because
small stature or slow growth rate exposes them
to damage by dogs, humans, and mowers.

Left to right, top row: Enkianthus perulatus form and flower; middle row, Aconitum sinomontanum (Chinese
monkshood), Elliotia racemosa (Georgia plume), Hamamelis x intermedia ‘Diane’ (hybrid witch hazel); bottom
row, Caragana arborescens ‘Nana’ (low-growing pea tree) form and flower. Photographs by Peter Del Tredici,
Gary Roller, fohn Alexander III, Robert G. Mayer.

22 Arnoldia 62/2

From left to right. Viburnum opulus (European cranberry bush), V. hupehense (Hubei viburnum), V. phleobotrichum,
V. dilatatum (linden viburnum).

Examples of those that will find a home in the
raised beds of the Leventritt Garden are Buxus
and shrubby members of Fabaceae such as the
brooms, Cytisus and Genista.

Many of the vines will be selected for their
botanical value rather than ornamental interest.
Structures for growing vines in the Arboretum's
main grounds were heretofore limited to the
few remaining perimeter fences. The Leventritt
Garden will accommodate many vines new to
the Arboretum.

Several important botanical families in North
Temperate floras are represented primarily by
herbaceous plants, including some that were
cultivated at the Arboretum in earlier years.
The garden provides the opportunity to aug-
ment the botanical collections of these families
and genera.

Educational Opportunities

These selections will determine the opportuni-
ties for learning that the garden will offer. The
range of subjects will expand over time; below
are some of the areas of interest that our plant
selections will support. Orientation and
storytelling signs, informational pamphlets,
identification labels, and docent tours will be
used as interpretive aids.

Comparisons of wild taxa with their culti-
vated taxa. For instance, the flowering of
redbuds, Cercis canadensis and its cultivars
'Kovey' and 'Alba', can be contrasted in spring.

In summer, the differences between species
hydrangeas, clethras, and stewartias and their
respective cultivars can be studied.

Plant solutions for landscape problems.
Ilex glabra and Rhus aromatica can stop bank
erosion. Ericaceous plants grow well in acid
soils. Comptonia peregrina tolerates poor
soils. Vines can create wonderful vertical
screens, and for horizontal screens, shrubby
hedges are superb.

Special ornamental qualities. For ornamen-
tal winter interest, witch hazels lead the list.
Buddleias are just one example of a widely
adaptable plant that sustains flowering over
much of the summer. For fragrance, the
daphnes, jasmines, and honeysuckles are prime
performers. Plants with colored foliage and
bold textures will also be highlighted.

Biodiversity and biogeography. Biogeogra-
phy— specifically the study of disjunct popula-
tions in Asia and North America — has been
central to the Arboretum's research for more
than a century. It will be demonstrated on the
ground in the Leventritt Garden, where the
native Chionanthus virginicus will be com-
pared to its Asian counterpart, C. retusus, as
will Asarum canadensis with A. splendens and
Pachysandra procumbens with P. terminalis.

Plant conservation. Another area of long-
term interest has been the conservation of
endangered plants by growing them ex situ,

Plants 23

namely, here at the Arboretum. Elliottia
racemosa, Prunus depiessa, and Andromeda
glaucophylla head a long list of species in this
category,- among others to be included are
Amelanchier nantucketensis, Ilex collina, and
Abeliophylum distichum. At the other end of
the spectrum, plants that thrive too well — often
on what were the sites of presently endangered
species — will also be available for study, kudzu
and greenbriar among them.

Placing the Plants

In autumn 2001, Reed Hilderbrand Landscape
Architects submitted a set of concepts for the
planting plan that included locations and distri-
butions of small evergreen shrubs as well as
small- to medium-sized trees (arrayed for visual

effects in spatial and rhythmic patterns) and
drawings of microclimate variations, optimum
shrub sizes (varying by bed and terrace position),
and traffic flow.

The planting design committee — the authors
along with Laura Tenny Brogna and Bethany
Grasso — began its task by considering the distri-
bution and size constraints of evergreens and
small trees. We placed larger plants at the sides
and toward the middle of each terrace, and slot-
ted a high proportion of low-growing plants in
the lower terraces.

Microclimate variations (sun/shade, dry/mesic
soils) in the garden are minor and we assigned
them a low priority with a single exception. We
designated the lower, northernmost end as a
frost pocket. Tender material, if included at all
in this garden, will be placed in the uppermost
terrace and above the great wall.

With these considerations in mind, the com-
mittee began to partition the entire plant list
into manageable groups and to locate individual
taxa and groups in plan.

Vines were sited in the upper terraces
according to their growing requirements and
their collections and ornamental value. Because
wisterias require a high degree of structural sup-
port, they are primarily clustered on and around
the pavilion; the remaining plants climb on trel-
lises, columns, and/or wall bases.

Small- to medium-sized trees were included
on our list specifically for design effect. We
followed the landscape architects' proposals in
locating most of them
within the terraces or
at the garden edges
with individuals ex-
tending from the ra-
vine area across the
terraces. Eastern red-
bud (Cercis canaden-
sis) and a few of its
variants were selected
as the primary tree
candidates. We sited
deciduous and ever-
green shade trees
along Centre Street
to serve as sound and
visual barriers.

Clockwise from top, Chionanthus virginicus (white fringetree) flowers, C. retusus
(Chinese fringetree) flowers and fruit.

24 Arnoldia 6212

Campsis radicans (trumpet creeper) at the Biltmore
Estate, North Carolina.

Akebia quinata on Cryptomeria fortunei (Japanese
cypress) in China.

Actinidia (kiwi vine) — four species and several
cultivars

Akebia — three species

Aristolochia (dutchman's pipe) — two species

Berchemia scandens (supplejack)

Celastrus (bittersweet vine) — three species
Lonicera (honeysuckle) — five to ten species
and cultivars
Pueraria lobata (kudzu)

Schisandra — two or three species
Trachlospermum asiaticum (star jasmine)
Wisteria — four species and many cultivars

Twining Herbaceous Vines

These vines need nylon mesh to climb.

Cocculus carolinus (Carolina moonseed)
Humulus japonicus (Japanese hops)

Growth Patterns of Vines for the Leventritt Garden

The plants that we call vines have in common a climbing habit, but how they climb varies widely.
The major mechanisms will be represented in the Leventritt Garden.

Twining Woody Vines

Some of these vines, including Celastrus,
Pueraria, Wisteria, require heavy posts or trel-
lises to climb on; others do better on a mesh
support such as chainlink.

Adventitious Root Climbers

These plants climb masonry walls or the trunks
of trees,- they can also grow as groundcovers.

Campsis radicans (trumpet creeper)

Euonymus fortunei (winter creeper) — many
cultivars

Hedera (ivy) — two species and many cultivars
Hydrangea anomala subsp. petiolaris (climb-
ing hydrangea)

Schizophragma hydrangeoides

Plants 25

Menispermum canadense and dauricum
(moonseeds)

Polygonium aubertii (silver fleece vine)

Vines with Coiling Leaf Petioles

A wall or other flat surface covered with nylon
mesh suits these vines.

Clematis — fifteen to twenty species and
cultivars

Vines with Coiling Tendrils

Most of these do well on trellises.

Ampelopsis (porcelain vine) — four species
Smilax (greenbriar) — species will be grown as
examples

Vitis (grape) — many species and cultivars

Vines with Adhesive Tendrils

These should be grown on stone or masonry.

Bignonia caprioleta (cross vine)

Parthenocissus (Virginia creeper and Boston ivy)

Wisteria intertwined with Parthenocissus at
Rockefeller University, New York City.

Scandant Shrubs (also known as Weavers)

The growth habit is sprawling but with support
these can "climb" to some extent.

Clematis — nonclimbing, "shrubby" species
Rosa — many species and cultivars
Tripterygium regelii

Many of these vines can cover large areas
and several are very vigorous, even rampant;
others — kudzu, Schisandia, Trachlospermum
asiaticum, Carolina moonseed, silver fleece
vine — lack vigor or die back in Boston.

This classification system is taken from "Physiological
ecology of mesic, temperate woody vines" by A. H.
Teramura, W. G. Gold, and I. N. Forseth, in The Biology
Berchmia scandens (supplejack). of Vines, ed. F. E. Putz and H. A. Moody (1991).

26 Arnoldia 62/2

Clockwise from upper left. Daphne x burkwoodii ‘Carol Mackii’, D. cneorum ‘Eximia
(garland flower variety), D. genkwa, D. altaica.

We selected several broadleaf evergreen
shrubs (e.g., Buxus spp., Ilex crenata, Ilex
glabra) to act as green foundations year-round
throughout the garden, communicating much
the same way as the stone walls do. The chosen
plants share many features and characteristics
(fine texture, ease of manipulation, similar
form and habit). These evergreens were sited
linearly in plan.

Next, the remaining shrubs, both deciduous
and evergreen, were added to the plan. Because
their eventual sizes at maturity limit their
placement, we gave the largest shrubs first pri-
ority. Some genera on the plant list are heavily
represented and prominently featured (for
example, Daphne, Viburnum, and Hydrangea ).
Members of these genera will form patterns
that diffuse across and down the terraces,
interspersed with other shrubs. For example,

Viburnum primarily
extends down the
ravine and across
the eastern terraces,
while Hydrangea has
been massed at the
northernmost edge
of the garden and
then across the west-
ern terraces. Species
and genera with
fewer representatives
are integrated accord-
ing to their growing
requirements and
aesthetic compatibil-
ity with surrounding
plants and features.
Plants that grow best
and look best when
sited as individuals
(such as Japanese
maple) are so treated,
and those best used
in masses (for ex-
ample, brooms) are
massed.

A few areas have
been reserved for
specialized plantings.
Most of the sixty
dwarf conifers have been sited above the garden,
adjacent to the bonsai house and the large sugar
maple. As a group, these plants tend not to inte-
grate well with plants of different habits and
architectural forms, and therefore new plantings
of dwarf conifers will also be sited there. In the
lower garden terraces, several beds have been
prepared with acidic soils for ericaceous plants.

The last group — low shrubs, groundcovers,
and herbaceous perennials — will be placed at
the time of planting. Large and prominent
plants have been sited for specific reasons, but
flexibility is valuable when placing smaller
plants around these anchors.

Peter Del Tredici is director of living collections,-
Michael Dosmann is a former Putnam Fellow,- Tom
Ward is greenhouse manager and propagator; Julie Coop
is superintendent of grounds.

Genetic Piracy: A Newly Discovered
Marvel of the Plant World

Richard B. Primack

Plants have evolved a variety of reproduc-
tive systems. The trees of some species,
such as ash and ginkgo, are differentiated
by gender, with both a female and a male indi-
vidual required for seed production. Other spe-
cies, such as cherries and almonds, ensure
cross-pollination without separating individuals
by gender: each plant is self-incompatible, so
that stigmas cannot be fertilized by pollen
produced on the same plant. And then there
are species that don't require cross-pollination,
including the many weedy species and annuals
that can produce seeds even when stigmas and
pollen are from the same plant.

Yet another reproduction system, called apo-
mixis, involves seed production with no fertili-
zation at all. Apomixis occurs inside the ovary
of the flower when a cell divides to become an
embryonic seed. The resulting seed looks quite
normal but is in fact a clone that is genetically
identical to the parent plant, with slight differ-
ences arising from the peculiarities of cell divi-
sion. Apomixis occurs in many plant families
and is especially common in the raspberries of
the genus Rubus. It is thought to confer certain
advantages: no other plant is required for fertili-
zation, and the seeds produced belong to a geno-
type of proven success.

The cypress Cupressus dupreziana is found only in the desert of Algeria, fust 231 individuals of this endangered
species remain alive in the world.

28 Arnoldia 62/2

Cupressus dupreziana.

Genetic Piracy 29

Until recently, all known examples of apo-
mixis involved the production of seeds from
cells of the female plant, that is, the plant that
actually makes the seeds, although male
(pollen-producing) plants existed, they made
no genetic contribution to the embryo. Now,
however, a group of French scientists headed
by C. Pichot and M. El Maataoui have found an
example of apomixis in which the seeds are
identical to the pollen-producing plant, rather
than the female plant. The species involved is a
rare cypress, Cupressus dupreziana, that occurs
only in the Tassili N'Ajjer desert of Algeria, in
extremely hot, arid conditions. Because of its
small population size and poor regeneration
rate, it has been classified as endangered and
studied intensively. Field research has shown
that only around ten percent of wild seeds have
a viable embryo,- and since the same low rate has
been observed in trees being cultivated in
France, the problem was assumed to result from
the intrinsic nature of the species, rather than
simply from inadequate pollination.

But further study showed that the low rate
of seed viability was related to Cupressus
dupreziana’s atypical meiosis process. In most
diploid plant species (those with two sets of
chromosomes), meiosis of the male reproduc-
tive cell results in the cell splitting to form hap-
loid pollen grains— that is, grains with only one
set of chromosomes, half the number of the par-
ent plant. Seeds result from the merger of the
haploid pollen grains with haploid egg cells pro-
duced in the ovule. In C. dupreziana, however,
chromosome behavior and cell division are
erratic: sometimes a nucleus divides, but the
two resulting nuclei fuse together again; and
instead of producing a uniform mass of haploid
pollen, the male cones produce pollen of widely
varying sizes, many with no chromosomes at all
and some with one, two, or four sets of chromo-
somes. Laboratory studies have shown that only
the diploid pollen is capable of germinating.
This pollen is almost identical genetically to
that of the pollen-producing tree itself, with
slight variations arising during meiosis. And it
plays a role in seed formation that is unique
among known plants — male apomixis.

Analysis of seeds from a plantation of
Cupressus dupreziana trees showed that the

The pollen of Cupressus dupreziana is variable in size
and shape, as shown in this section taken through a
microsporangium before the pollen is shed. Only the
larger, round pollen would be viable; the other pollen
grains will not germinate. The epidermal wall of the
pollen sac is shown on the right side.

identifying DNA markers of the seeds did not
match those of the mother tree — that is, the
female tree on which the embryonic seeds
developed. In typical diploid species, each seed
contains half of the identifying DNA markers
of the mother and half of those of the pollen-
producing parent. The fact that the DNA mark-
ers of C. dupreziana’s seeds match those of only
one parent — the father tree — shows clearly that
this species uses apomixis to produce seeds. Pre-
sumably, a diploid nucleus in the pollen tube
enters the ovule within the young cypress cone
and, instead of combining with a female
nucleus, begins to divide on its own, taking on
the appearance of an embryo. If there is a female
nucleus present, it simply deteriorates.

The ability of the pollen of Cupressus
dupreziana to produce new offspring via apo-
mixis was further examined in a special planta-
tion of hybrid trees that had been created by
dusting pollen of C. dupreziana onto the recep-
tive female cones of C. sempervirens, the com-

30 Arnoldia 62/2

mon, or Mediterranean, cypress. The seeds pro-
duced by these crosses were germinated and
then grown for fifteen years. The resulting
"hybrids" were identical to C. dupreziana in
twig orientation, female cone size, and pollen
diameter; they bore no visible resemblance to
C. sempervirens, suggesting that no genetic
material had come from C. sempervirens. And
in fact, the DNA markers of the hybrids were
identical to those of C. dupreziana, but com-
pletely unlike C. sempervirens. These surpris-
ing results demonstrated conclusively that C.
dupreziana pollen is able to produce seeds with
no genetic contribution from the female plant,
not only within its own species but also when
"crossed" with other species.

This is the only known example of apomixis
involving the pollen parent. In human terms,
this is equivalent to a human mother giving
birth to a baby that is genetically unrelated to
herself, but genetically identical to the father.
The evolutionary advantages of this method of
reproduction are not evident. One possibility is
that isolated trees of this rare species can use
the female cones of other cypress species grow-
ing in the vicinity as vessels for producing cop-
ies of its own genotype. This "genetic piracy"
might allow the species to survive at low densi-
ties; however, at present no other cypress spe-
cies grows nearby in C. dupreziana’ s wild
habitat. But even if genetic piracy were feasible,
this unusual reproductive system has its short-
comings: it cannot generate the genetic varia-
tions that allow sexually reproducing organisms
to adapt as environmental conditions change.

References

El Maataoui, M., and C. Pichot. 2001. Microsporogenesis
in the endangered species Cupressus
dupreziana A. Camus: evidence for meiotic
defects yielding unreduced and abortive pollen.
Planta 213: 543-549.

Pichot, C., M. El Maataoui, S. Raddi, and P. Raddi. 2001.

Surrogate mother for endangered Cupressus.
Nature 412: 39.

, B. Fady, and I. Hochu. 2000. Lack of mother tree

alleles in zymograms of Cupressus dupreziana
A. Camus embryos. Annals of Forestry Science
57: 17-22.

Acknowledgments

Helpful comments on the manuscript were provided by
Elizabeth Platt, Spencer Barrett, and Christian Pichot.

Cd

Cs#1

CsflxCd

3 progenies

Cs#14

Cs#14 x Cd
3 progenies

Cs#22

Cs#22 x Cd
3 progenies

Cs#36

Cs#36 x Cd
3 progenies

This is a gel showing DNA markers of tissue taken
from trees of Cupressus dupreziana (Cd) and
C. sempervirens (Cs), as well as from trees produced
by cross-pollinating them. Pollen from one tree of
C. dupreziana was used to fertilize four trees of
C. sempervirens. Seeds from each of these
hybridizations were used to grow three progeny trees
from each mother tree. For each of the four crosses,
tissue taken from the offspring is genetically
identical to C. dupreziana, as shown by their DNA
bands being exactly like C. dupreziana but unlike
the DNA bands of C. sempervirens. These results
clearly demonstrate that the inheritance of genetic
material is strictly from the pollen-producing parent.

Richard Primack is professor of biology at Boston
University. He is currently conducting research at the
Arnold Arboretum and in Concord, Massachusetts, on
how climate change affects the flowering time of plants.

Book Note

Karen Madsen

Wild Flowers of Yunnan and Central China by Beryl M.
Walden and Shui-ying Hu. Hong Kong: B. M. Walden, 2000.

The authors of this attractive volume tell us that of the
30,000 recognized plant species in China, 16,000 occur
in Yunnan, a province that comprises only four percent
of the country's land area. In an effort to promote conserva-
tion by increasing public awareness, the authors have given us
watercolor images and botanical descriptions of 516 of these
plants, most of them from cold-temperate areas and nearly all
from Yunnan Province. (The "Central China" of the title
refers to Jiangxi Province, where a small number of the book's
plants originate.)

Both images and descriptions testify to the lavish diversity
of the region. Included are seven each of asters, camellias,
irises, and thalictrums; eight viburnums; ten pedicularises;
thirteen each of gentians and roses; fourteen orchids; seven-
teen primulas; and an astonishing twenty-seven species of
rhododendron. Many of the 516 taxa — 150 of them — are
endemic to Yunnan, some occurring only in one portion of a
mountain range.

Yunnan's botanic diversity mirrors its spectacular range of
climates and topography, which includes "lateral and vertical
zones and mini-zones." The authors quote the English plant
explorer Frank Kingdon Ward as he described northwestern
Yunnan a century ago: "You will see flowers growing in reck-
less profusion . . . For days on end you may tramp over carpets
of flowers. Rhododendron in incredible variety, yet no taller
than heather in Scotland, though of every conceivable
colour." Already at that time 5,000 plants had been intro-
duced in the West from Yunnan.

Shiu-ying Hu, author of the botanical descriptions, has
had a long and distinguished career and is well known at
the Arnold Arboretum. An accomplished botanist and
fieldworker when she arrived in the U.S. from China in 1946,
she carried out her doctoral dissertation on the Chinese hol-
lies under the direction of E. D. Merrill, director of the Arbo-
retum, 1935-1946. Her many publications on the flora of
China, especially of Hong Kong, include works on Malvaceae,
Compositae, Orchidaceae, the Chinese material medica, and
food plants. Beryl M. Walden provided the watercolors of the
plants in Wild Flowers of Yunnan and Central China, as she
did for earlier collaborations with Hu, Wild Flowers of Hong
Kong (1977) and a two-volume Wild Flowers of South China
and Hong Kong (1983, 1987).

Beryl Walden’s watercolor o/Lilium
bakerianum delavayi, endemic to Yunnan.

32 Arnoldia 62/2

Arnold Arboretum Weather Station Data — 2002

Avg.

Max.

Temp.

(°F)

Avg.

Min.

Temp.

(°F|

Avg.

Temp.

(°F)

Max.

Temp.

(°F)

Min.

Temp.

(°F)

Precipi-

tation

(in.)

Snow-

fall

(in.)

JAN

43

27

35

68

10

2.86

9

FEB

46

24

35

65

9

2.62

1

MAR

50

29

40

71

16

4.05

4

APRIL

64

41

53

97

26

3.74

0

MAY

70

47

59

89

37

5.69

0

JUNE

80

57

69

99

42

4.68

0

JULY

80

64

72

105

50

2.45

0

AUG

90

65

78

104

53

1.75

0

SEPT

81

57

69

99

46

4.17

0

OCT

64

40

52

88

27

4.26

0

NOV

50

32

41

73

14

6.22

3.5

DEC

40

21

31

61

9

6.59

14.1

Average Maximum Temperature
Average Minimum Temperature
Average Temperature
Total Precipitation
Total Snowfall
Warmest Temperature
Coldest Temperature
Date of Last Spring Frost
Date of First Fall Frost

63°

42°

53°

49.08 inches
31.6 inches
105° on July 4

9° on February 12 and December 10
30° on April 27
32° on October 18

Growing Season 1 73 days

Note: According to state climatologist R. Lautzenheiser, 2002 was warm and sunny with above-normal levels of
precipitation; growing days numbered 173 compared to 132 in 2001. It was the warmest year since 1998 and
ranks tenth in 132 years of weather records. With a total precipitation of 49.08 inches we were 5.5 inches above
normal and well over the total of 45 inches in 2001. May and June were cool and moist, followed by record-
breaking temperatures in July, August, and September. Year's end — October, November and December — was very
cold; the first ten days of December averaged 10 degrees below normal.

3 9088 01912 7186