Development of Selection Criteria for Pecan

Transcription

Development of Selection Criteria for Pecan
EHGTES
fit 145
HRS
no. 96
3tates
nent of
ure
Agricultural
Research
Service
ARS-96
W:s,
Pecan Husbandry:
Challenges and
Opportunities
i:,ll g:,i, i
LTor
ary
December 1991
First National Pecan
Workshop Proceedings
Unicor State Park, Georgia
July 23-24, 1990
s€P
17 33
PRODUCTION
siblings (having the same pollen and pistillate
parent) or half siblings (having different pollen
parents). Full sib seedlingp may be the result of
THE DEVELOPMENT OF SELECTION
CRITERIA FOR PECAN SEEDSTOCKS
L.J. Graukel
INTRODUCTION
At best, a review evaluates the past from the
perspective of the present in hopes of
anticipating the challenges of the future. For
pecan rootstock development, this process is
complicated by the necessity to integrate
biological and economic realities, even as both
undergo unforeseeable changes. A full
appreciation of the subject requires 1) an
apprecialion for the genetic base of the species;
2) an understanding of the historic interactions
between that base and a developing technologl
(which is driven by economics); and 3) a critical
assessment of the current status of the industry,
both in terms of practices and problems. With the
above as a foundation, it should be possible to
develop productive long-term strategies for
rootstock development while anticipating, and
hopefully avoiding, potential problems.
In order to discuss the history, current status,
and potential of rootstock development in pecan,
several terms must be defined. In the process of
definition, concepts may be delineated which are
crucial to resolution of both the problems and the
potential solutions.
A 'rootstock' is the root
upon which a nscion" or cultivar is grafted, or
more simply, the root used as a stock. The
compound genetic system of scion and stock is a
nstionn.
The composition of stions is given in
this paper as 'scion'/seedstock, where the is
represented by the cultivar name, followed by "sd"
(eg. 'Schley'/'Moore' sd refers to ,schley' scions
vegetatively propagated onto'Moore' seedling
rootstocks). A "seedstock" is the source of seed
planted to produce seedling rootstocla. Seedlings
arising from the same seedstock are considered a
"familyn. Seedlings of a family may be full
lResearch Horticulturist, USDA-ARS. Pecan
Breeding and Genetics Research, Route 2 Box
133, Somen'ille,
TX
77879
either cross- or self-pollination. Seed produced
without the intervention of man is termed
"open-pollinatedn seed, the parentage of which
will depend on the composition, spacing and
dichogamy of associated trees. A 'populationn is
a loosely defined spatially associated assemblage
of trees, assumed to be capable of cross
fertilization, and therefore (for regenerating
stands) more genetically related than distant
assemblages. "Provenance' is the area of origin
of seed.
GENETIC BASE
Specles Distributlon
Pecan is distributed along the Mississippi River
and its tributaries from northern Illinois and
southeastern Iowa to the Gulf Coast. Isolated
populations occur as far east as southeastern
Ohio, northern Kentucky and central Alabama, and
as far west as Jeff Davis C-ounty, Texas, and
Chihuahua, Mexim. The species is abundant on
rivers and streams of central and eastern Oklahoma
and in Texas west to the Edwards plateau. pecan
occurs in regenerating stands as far south as
Taachila, Oaxaca, Mexico (Thompson and Grauke
1eeo).
The geographic and climatic variation across this
area is great, as evidenced by data for collection
sites in the USDA Pecan Provenance Test (Table
1). In addition, the region includes areas where
topography accentuates isolation, and where the
environment produces a particular stress . This
increases the potential for genetic adaptation
within populations across rhe range (Callaham
reTo).
Dlspersal By Man
Accurately mapping the native distribution of the
pecan is complicated by man's introduction of the
species to non-native areas. Even in undisputed
native pecan populations, such as the Devil's
River area of Val Verde County, Texas, the first
recorded occurrence of the species is in
association with human artifacts. pecan seeds and
leaves have been archeologically recovered from
Baker's Cave in the Devil's River area from strata
dated from about 6100 B.C. up ro abour 3000 8.C..
Pecan has not been reported from excavations of
the Golondrina hearth, which is the oldest level
of human occupation in the cave and is dated about
7000 B.C.. Juglans microcarya, a riverine
125
species
like pecan, is recovered from the
Golondrina hearth (Dering 1977, Hester 1981;
Hester, penonal communication). Whether this
indicates the introduction of pecan by man between
7000 and 6000 8.C., the natural extension of the
range of the species at that time, or is merely a
gap in the archeological recovery process has not
been determined.
Bettis et al. (1990), reported the presence of a
fossil pecan from floodplain sediments of the
Mississippi River near Muscatine, Iowa which was
accelerator-dated at 72n+-l20 yr B.P. They noted
that pollen profiles in the Mississippi River
Trench indicate higher peroentages of Carya
pollen in the valley than in the uplands during
the early Holocene, and suggest that the riverine
pecan may be responsible for thos€ peaks. If so,
the species was present near the present northern
limits of its range as early as 10,3{X) yr B.P.
Bettis et al. (1990) suggest that the abundance of
pecan in Early Archaic archeological sites and the
co-occurrence of early Holocene Carya pollen
peaks with the arrival of the Dalton artifact
complex in the Upper Mississippi Valley implies
early dispersal of pecan by humans.
Pecan populations exist
in Ttachila,
Oaxaca,
Mexico, an ancient ceremonial and administrative
center of the Olmec Indians. Valuable information
on the antiquity of the pecan population there may
be provided by ethnobotanical research in
association with the excavation of the site.
Unfortunately, no reports are currently
available. Pecan has not b€en recovered from
archeological excavations in the Tehuacan Valley
(smith t967).
Manning (1949) reported that field notes on pecan
collections by Dewey in Jaumave, Tamaulipas,
suggested introduction of the trees from Texas.
Obviously, the older
a population is, the harder
it will be to retrieve information
concerning
possible introduction. However, ancient isolated
populations, even though introduced, may harbor
genetic adaptations to the region.
It is necessary to determine the extent of genetic
adaptation to climatic and edaphic variables
within native (or long naturalized) populations.
Where these populations can be observed to have
adaptations to their region, directed selection
within the limits of that region should be the
most productive approach for rootstock development
and may have implications for cultivar
development. To accomplish this objective, seed
has been collected and provenanoe tests have been
established in somerville, fi and Byron, GA
(Grauke et al. 1S9a).
126
Spocies
Varlation Over Range
Nut characterlstics In relation to
origin.
seedstock
Nuts collected as seedstocks in the
above mentioned provenanc€ tests were studied.
Five trees from each of 19 populations (table 1)
were targeted. Ten nuts from each tree were
individually measured for weight, buoyancy,
length, width parallel to the suture, and width
perpendicular to the suture. The nuts were then
cracked and kernel weight and shell thickness were
determined for each nut. Results indicate clinal
patterns of variation in several nut parameters,
including shell thicknass (which is negatively
correlated with latitude) and length to width
ratio (which is positively correlated with
latitude).
In 1989, nuts were collected from native
populations in the U.S. ranging from Illinois
south to Louisiana and from Alabama west to Val
Verde Co., TX. Ten nuts from each of 5 trees in
each population were measured for weight,
buoyancy, length and widths. The data on length
to width ratios are consistent, with the greatest
ratios occurring in the north and east, with the
lowest values occurring in the west. Given the
stratery of both collections, the data obviously
reflect both phenotypic and genotypic responses to
variation over the range.
Thompson et al. (1989), reported signiftcant
variation in nut parameters of named cultivars
from different locations in the U.S, which were
attributed to genotype-environmental interaction.
Nuts of named cultivars had the highast length to
width ratio in Baton Rouge, with reductions in
that ratio to the west. Nut length was more
greatly reduced in the western populations than
was nut width for a given cultivar, which is
consistent with these observations. Since the
variable of genotype was controlled in these data,
the obsewed variation is phenotypic response.
Nut shape is, therefore, phenotypically influenced
by area of origin, with nuts in the north and east
being longer in relation to their width than nuts
in the south and west. Nut width is evidently
more highly conserved under conditions of moisture
stress than is nut length. Emergence of pecan
seedlings collected in 1989 was significantly
related to length to width ratio, with nuts from
the north and east being slower to emerge than
nuts from the west (Table 2). Whether nut shape
has a direct or indirect role in seedling
performance (i.e., selective value which muld
lead to genotypic differences between provenances)
or is selectively neutral is unknown at present.
Another nut characteristic which has been
associated with seedling performance in this test
is percent kernel, which is positively correlated
with seedling height and diameter (Fig 1).
Variations in percent kernel were not associated
with variations in length to width ratios.
Assertions that 'small nuts gave as good results
as larger ones for nursery purposes" (C. A Reed,
cited in Gray, lV27) or that "nurserymen are
justified in using inferior, cheap nuts" (Burkett
t932) are inaccurate. Within a seedstock family,
tree vigor has been positively related to nut
weight (Hinrichs 1965), and specific gravity
(Grauke and Pratt 1%5). Between families,
seedling performance has been positively related
to nut weight (Harris and Tauer 1987), perc€nt
kernel (current research), and to less specific
criteria of size and
fill
Seedling Performance
Origin
(Campbell 1978).
In Relation To Seedstock
Variation in the performance of seedstocks is
attributable to several sources. The provenance
of the seedstock, parentage, site variability, and
culture all influence seedling performance. Clear
resolution of the influence of each variable
requires control of the other variables, which
complicates the research. In some cases,
different levels may be mntrolled (e.g.,
provenance, pollen parent, pistillate parent) but
interpretation apparently oversimplifies results.
Hanna Onz\ and Madden (1974) worked with the
same trees, which were the result of reciprocal
crosses between parents representing different
provenances. Northern pollen parents reduced the
growth of either western or eastern seedstocks.
Both researchers found no significant effects of
pollen parent on growth of seedlinp apart from
the effect of provenance or selfing. [Vigor is
generally reduced in seedlings arising from a
selfed seedstock as compared to the same
cross-pollinated seedstock (Romberg and Smith
1946), an example of heterosis.l Madden (1974)
concluded that 'staminate parent greatly
influenced seedling vigor, thus seedlings from
open-pollination may be variable depending on the
staminate parent'. It is more aocurate to
conclude that the influence of staminate parent is
evident in crosses involving northern pollen on
either southeastern or southwestern pistillate
parents, but that no significant effect of pollen
on growth was observed were independent of the
north-south provenane effect. Ou (1990) found
that northern pollen resulted in delayed
germination of southern cultivars of pecan,
corroborating the north-south provenance effect by
pollen parent.
Early growers found that s€edlingp arising ftom
southern seedstocks quickly winterkilled in the
north, while trees from northern sources performed
well (Heigas, 1.896, p. 51). The use of Mexican
seedstocks in Texas was discontinued based on fear
of increased susceptibility to cold injury to
seedlings which began growth earlier and ended
active growth later than local seedlings (Gray,
1927). Burkett (1932) noted rhat rhe longer
period of active growth of Mexican seedlings
resulted in larger trees, as compared to trees
arising from l,ouisiana and Georgia seedstocl$,
while seedlings from West Texas were comparable in
size.
Phenologl of pecan seedlings is, at some level, a
function of provenance, with inception and
duration of foliation increasing in seedstocks
from southern sources. Hanna (1y72) observed
that the inception of bud growth in the spring
varied as a function of both pistillate and
staminate parents in reciprocal crosses between
cultivan selected from different provenances.
Seedlingls having either parent from northern
provenance began growth later than seedlings whose
parents were from either western or eastern
provenances. Seedlingp of northern parents also
significantly smaller than sepdlingp from eastern
or western provenanoes, with inception of growth
being correlated with tree height. If extended
phenologl is associated with increased size, then
southern seedstocls are advantag@N, within
limits.
The limits to the advantage offered by extended
period of growth are s€t, at least in part, by
frenzn damage to tender tissues. We have observed
phenological differences within open-pollinated
families of seedstock which were related to
su@uent frenze injury (Grauke and Pratt
19E7). Furthermore, scions grafted on families
of seedstock with earlier growth were
significantly more advanced in growth than when
worked on other stocks (Fig. 2), and onsequently
experienced increased frezn damage (Fig. 3).
Under the conditions of our study, no seedlingp
were killed by the trenze.
Hanna QnU associated patterns of ion uptake
with particular pistillate parents. Seedlings
arising ftom reciprocal crosses w€re distinguished
by ion uptake characteristics associated with the
female parent. This lead Hanna Qnq to suggest
the possibility of cyloplasmic inheritance.
Miyamoto et al., (1S5) studied gr@'th and ion
of pecan trees grown in lpimeters
irrigated with saline water. They found that
'Rivenide' scedling rmtstocks absorbed lesser
accumulation
r27
of Na and produced equal or greater root
than 'Apache' or 'Burkett' seedling
amounts
mass
rootstocks. They concluded that 'Riverside' might
be better suited :rs a rootstock in salt-affected
areas. We have observed patterns of nutrient
accumulation in a test orchard which varied
significantly as a function of scion, seedstock
and tlock (Grauke et al. 1989b). The pattern has
been consistent for two years. Differences in
nutrient accumulation as a function of pistillate
parent could be involved in the regional patterns
of seedstock performance.
Nuts from our 1989 U.S. mllection have been
planted in a randomized block test in a
greenhouse. Each of four greenhouse benches,
bounded by border rows, includes 6 seedlings of
each of 53 familie.s which represent 13
populations. Observations have been made on
inilial emergence and early growth (Table 2)'
Several points should be noted in these data: 1)
Populations of trees can be distinguished from
other populations on the basis of days to
emergence, height, and diameter, despite
variability between families within a population'
or variation between seedlings of families; 2)
when considered as a group, nursery seedlings made
the greatest early growth in height and had the
greatest diameters; 3) populations collected in
ittinois and Alabama were the latest to emerge and
had the least height growlh at 54 days; 4) the
only nut characteristic significantly related to
seedling performance was percent kernel, which was
directly correlated with both seedling height
(Fig. 1) and seedling diameter.
Our results are consistent, in general, with those
of Harris and Tauer (1987), who also found
significant variation in the performance of pecan
seedlings as a function of the origin of seed,
with significant differences occurring between
populations. Those researchers collected five
ieedstocks each of 23 native populations (termed
"standsn in their study) in Missouri, southeastern
Kansas, throughout Oklahoma, southwestern
Arkansas, north lruisiana and north east Texas.
Seedlings were grown in a randomized complete
block design. Seed characteristics, and seedling
growth and phenolo$/ were statistically
evaluated. They reported differences between
populations which were related to north-south
clinal trends, with increased seedling size being
associated with stands tocated south of the
growing region. Seedling size was positively
related to seed size in their test.
These observations reinforce the recognition of
nnorthernn and nsouthernn pecan regions
distinct
and imply the potential for selection of distinct,
r28
advantageous populations as well as seedstocks
within regions. In order to efficiently make
directed selections for a particular region,
increased resolution of the boundaries between
distinct growing regions is needed. In black
walnut, for instance, recommended regions for the
collection of seed are within precisely delimited
geographic units (Deneke et al., 1980). What are
the northern extremes to which a given pecan
seedstock should be transported from its native
range? The northern region accounts for only a
small proportion of commercial nut production,
with the vast majority of that production coming
from native trees. The incentive for the
development of appropriate northern rootstocks is
therefore less than in the south. However,
attention is being focused on the subject
(Campbell, L978; Bill Reid, Personal
communication).
The east-west geographic limits for the
development of rootstocks are more ambiguous than
those for the north-south axis. The southern
pecan region has historically been divided into
qrstern and western sections on the basis of
climate and associated tree culture, with
rootstock being one of many areas of divergence.
Gray (7927) offers anecdotal justification for use
of western stocks in the West:
"If Nature, by natural selection, has
given
the East trees largely scab-resistant, is it
not reasonable to suppose that she has also
given the West trees that will produce stocks
better suited to western conditions? While
more evidence along these lines is needed the
western planter can scarcely go far wrong by
using western stock; and obviously other
stocks are not, without more evidence, to be
condemned.'
These observations are consistent with more recent
observations (J. B. Storey, personal
communication) of increased tree size and improved
performance of 'Western'fRiverside' sd stions as
compared to 'Western'fElliott' sd stions, when
grown at Delicias, Chihuahua, Mexico. If seedling
rootstocks arising from western seedstocks are
better adapted to western conditions, what is the
basis of that selective advantage? If real
differences exist, they should be demonstrable and
could form the basis of regional selection
criteria.
HISTORY OF ROOTSTOCK DEVELOPMENT
Early efforts to develop orchards by planting
selected seed advocated the use of'large nuts
with soft shells" (Heiges, 1996, p 50), stressing
the "importance of planting only the very best and
finest nuts obtainable" (Wm. Nelson, cited in
Parry, 1897). These efforts produced orchards
with substantial tree to tree variation in nut
size, shape and production. Although that method
of orchard establishment was soon abandoned,
numerous nimproved cultivan" were selected from
the orchards planted with selected seed in Texas,
Louisiana, Florida, and Mississippi (Crane et al.,
te37).
The development of vegetative propagation by
grafting and budding allowed for increased orchard
uniformity, and less emphasis was placed on the
quality of the seedstock Reed (cired in Gray,
lV27) tracrt the early history of nunery
seedstock selection:
"In the beginning of nunery work Texas and
Louisiana seedlinp, on account of their
cheapness and availability, were used. The
nuts were usually graded and the larger sizes
sold at a profit; the smaller ones being
planted. It was found that, apparently,
small nuts gave as good results as larger
ones for nursery purposes.
As competition increased, nunerymen began to
use nuts of improved varieties for seed and
make capital of the fact in their
advertising. The Stuart nuts, which were
chiefly used in the beginning, proved to be
slow growen. This practice finally resolved
itself mostly into the use of nus of the
Moore and Waukeenah varieties... The actual
superiority of these stocks remains yet to be
determined...'
In an effort to determine if those stocks were
superior, Sitton and Dodge (1933) provide one of
the only substantive comparisons of orchard tree
performance as
a function of rootstock. Using
paired trees to reduce variability in initial tree
size, they compared the growth and nut production
of 'Schley' and 'Stuart' on 'Moore' and
'Waukeenah' seedling rootstocls. Eight year old
'Schley'fMoore' sd stions made more grov/th and
produced greater yields than'Schley'fWaukeenah'
sd stions. Seven year old 'schley'fMoore' sd
stions made significantly greater production than
even aged 'Schley'/ 'Waukeenah' sd stions,
although differences in size were not
significant. 'Stuart' made significilntly more
growth on 'Moore' than on 'Waukeenah' seedling
roo[stocks, but yields were so small and erratic
that no significance was attached to the
concomitant increase in fleld. This effort to
establish linkage between nunery selections and
orchard performance has not been punued.
Currcnt Practices
The commercial pecan nursery industry currently
relies on open-pollinated s€ed for the production
of rootstocks. Selection criteria for
commercially used seedstocks are empirically
derived by both nurs€rymen and researchers on the
basis of performance in the nursery rather than in
the orchard. Variation between different
seedstocks is sufficiently great that selections
have been made by nurs€rymen. Selection has been
in favor of seedstocks producing vigorous, uniform
seedlinp (Gray 1V27, Traub 1931, Yarnell 1934,
Madden and Malstrom 1975, Madden ln6, ln8).
Vigorous seedlings are capable of being grafted
at an earlier age, reducing the time between
planting and sale. Furthermore, sinc€ grafted
trees are priced by size, uniform vigorous stands
require less labor and return more.
The effectiveness of selection for vigor by the
nursery industry can be demonstrated by comparison
of the growth of representatives of the nursery
seedstock to growth of native seedling;s collected
throughout the u.s. (Table 2). The increased
growth of seedlings from nursery seedstocks is
confirmation of the high heritability for seedling
height and diameter growth calculated by Harris
and Tauer (1987).
Seedstock selections made by nurserymen show a
stratified geographic disrriburion (Gast and
Overcash 1980, Thompson 1990). In the
southeastern U.S., the predominant seedstocks are
'Elliott', 'Curtis', and 'Moore'. In the
southwest, the predominant seedstocks are
'Riverside', 'Burkett', and 'Apache'. In the
northern pecan areas, native nuts and 'Giles' are
largely used (Thompson 1990). Each of these
seedstocks originated in the geographic region
where it is now used (Thompson and Young 1985),
although the original source of the Florida
seedlingB ('Elliott', 'Curtis', and 'Moore') is
obscure (Taylor 1894, Heiges 189i, Blackmon
1926). 'Apache', a controlled cross from the USDA
pecan breeding program with both western
('Burkett') and eastern ('schley') parentage, is
used primarily in the region of origin of is
pistillate parent.
r29
Root Problems
Little s)4stematic research has been performed to
link the criteria of rootstock selection for pecan
to production. The lack of attention is due in
part to the lack of dramatic problems associated
with pecan rootstocks. In the sister genus
fughns, root rot caused by Phytophthora
spp. is the most serious challenge facing walnut
growers (McGranahan 1987). There are no
comparable widespread, devastating root problems
on pe€n (Hepting tnD.
Cotton root rot (Phymatotrichum omniuorum) is
a factor in some areas of Mexico, but occun only
in isolated instances in the U.S. (Brinkerhoff and
Streets 1946, Hepting 1971). The lack of severe
problems ftom cotton root rot may be due to the
fact that the disease is endemic in Texas within
the area of native distribution of pecan, and
protection is afforded by the genetic diversity of
the regionally selected rootstocks used by the
commercial nursery industry. Such conditions
should promote the greatest diversity of plant
defenses (Harlan 1977).
of a potentially new Meloidogtne
on pecan (Johnson et
al. 1975, Grauke and Pratt 1985), in each instance
associated with observations of differential
susceptibiliry by families of open pollinated
rootstocks. The extent of the problem has not
Nematodes
species have been reported
occur toward this goal, linkage must be
established between seedling characteristics and
orchard performance. Screening methods will be
established along with rhat linkage.
Review of the literature reveals ambiguity
concerning the interaction benreen tree size and
yield. At the simplest level, increasing size
increases yield (Hinrichs 1965, Sitton and Dodge
1938) with the result that mean trunk cross
sectional area is the best predictor of tree yield
potential. Perhaps more significant are
observations such as those by Lutz (1938) who
found initial tree size more correlated to
subsequent yield than to subsequent growth. The
direction of seedstock selection by the nursery
industry is consistent with this information. On
the other hand is the realization that dwarfing
rootstocks could be advantageous (Hanna 1987, Wood
1987). The advantage would be due to reduced
tree size, which would allow increased stocking
densities, while maintaining adequate per tree
production for net increases in per acre
production. Whether the goal of reduced size in
pecan trees is best achieved by selecting dwarfing
rootstocks or by selecting scion cultivars with
patterns of reduced growth in relation to stable
yields is debatable. Until tests are designed
which both control variables of rootstock and
cultivar and adequately monitor parameters of
growth and components of yield, the debate will
generate more heat than light.
been sptematically ascertained.
Root problems are so limited in scope and so
regional in nature, while rootstock development is
such a long-term process, that the rnost effective
strateg/ may be reliance on the broad genetic base
of regionally adapted rootstocls. It follows that
one problem to be avoided is the creation of
epidemic root problems by undermining the probable
basis of current protection - genetic diversity.
Kormanik and Ruehle (1989) reported increased size
and performance of sweetgum (Liquidamber
styaciflua L) which was related to increased
numbers of first order lateral roots (FOLR).
Furthermore, seedstock families differed in the
production of certain classes of FOLR. These and
SELECTION CRITERH FOR PECAN SEEDSTOCKS
other reports (Kormanik 1989) suggest FOLR may
a potentially useful criteria to evaluate families
of pecan seedstocks. If this selection criteria
is found to be useful, it will be consistent with
suggestions made by the pecan nursery industry,
of the nursery industry in
selecting seedstocks appropriate to its ne€ds,
some questions remain. Are the needs of the nut
and incorporated by particular nurseries, over 50
years ago. More importantly, it could serve as a
screening tool for large numben of seedstock
families which could improve the power of the
Despite the success
production industry, as well as the nurs€ry
industry, served by these seedstocks? If so,
could the process of selection be enhanced by
rootstock research? If production needs are not
served by current selection, what is the
appropriate criteria of selection?
The ultimate criteria of improved rootstock
performance is increased profitability of the
orchard. In order for effective selection to
130
selection process.
Another potentially useful selection criteria may
be the carbohydrate @noentrations of seedlings,
lvithin families. Wood (1989) reported that yield
in test rees of 'Schle1l, and 'Stuart' (on unknown
seedling rootstocks) was closely oorrelated with
January root starch concentrations. This could be
a very useful tool for s€lection, if families
differ in root starch lerrels and patterns are
apparent at an early age.
be
SUMMARY
Recent reviews of pecan rootstock research (Hanna
1987, Wood 1987) have stressed the need for
The history of rootstock development has b€en
driven largely by the doreloping technolos/ of
"improved orchard culture'. The early emphasis on
the use of selected seed succumbed, with the
advent of vegetative propagation, to the use of
undifferentiated or erren 'cull' seed. Marketing
based on the use of known scion cultivars as
seedstock lead to the recognition of families of
seedstocks which, within each region, could be
relied upon for production of uniform vigorous
seedlinp. This stage of development was reached
in the 1930's and little significant change has
occurred since. Effors to establish linkage
between nursery selection and orchard performance
have not been punued, despite an increasing body
of er/idenc€ which iustifies the strates/.
reliable techniques of clonal propagation. Once
achieved, the powerful tool of clonal propagation
muld be directed toward the development of
"dwarfing, nematode resistance, greater yields,
precocity of production, selective ion absorption,
salinity tolerance, and a reduction in variability
of orchard trees' (Hanna 1987). The pursuit of
clonal rootstocks is a continuation of the trend
of increasing orchard uniformity through the
application of developing technologt. It is a
direction which carries considerable risk for
catastrophic disease epidemics (Marshall 1977),
even for clonal rootstocks selected for
advantageous trais. The risk is increased if
"clonal propagation potential" becomes a "major
evaluation criteria" (Wood 1987) in itself.
Clonal propagation could contribute to the
development of durable, efficient rootstocks only
if the methods of screrning for critical yield
limiting parameters are incorporated. The
literature indicates that we can control the
variable of rootstock sufficiently well to design
test sj6tems which will help in establishing
rootstock screening methods, while yielding other
valuable information on genotype/environment
interactions. We should establish a cooPerative
inter-regional group to design and establish test
systems. We might discover that judicious
selection for seedstocks will accomplish the goals
of tree size control, nematode and disease
resistance, increased precocity and productivity,
selective ion absorption and salinity tolerance,
from a more genetically diverse, and therefore
more resilient, base, and within a technolory
which our industry can immediately incorporate.
The most significant rec€nt development has been
the addition of the 'Apache' seedstock to those
commonly used by western nurserymen. The
development of 'Apache' as a seedstock is an
appropriate and almost evolutionary culmination to
this period of rootstock development: 'Apache' is
a progeny of the prominent western nursery
seedstock family of 'Burkett'; it resulted from
the usDA cultivar improvement program, but was
recommended as a seedstock on the basis of
empirical observations of seedling vigor (the
selection criteria of ommercial nurserymen),
without being tested for orchard performance as a
rootstock.
The role of research in rmstock dwelopment has
not been one of insightful leadership. Excellent
beginningp have been made, but rather than
pursuing tantalizing leads, we have jumped to
conclusions. The most common conclusion is that
genetic variability of seedstocks is an adversary
to be overcome through the development of clonal
propagation, rather than an ally to enlist in the
development of imProved culture. We have often
ignored the variable of roostock hmily and have
almost completely failed to explore the structure
of families within poPulations across the species
distribution, within the aontertr of edaphic and
climatic environment. As a result, we have
overlooked a valuable opportunity to learn of
genotyp€/environmental interactions which have
implications, not only for rmtstock development'
but for cultivar development, for resour@
conservation, and potentially, for the dwelopment
of improved management strategies in the face of
changing climatig biological, and eonomic
environments.
LITERATURE CITED
Bettis, E.A, III, R.G. Baker, B.K Nations and
D.W. Benn. 1990. Early Holocene pe*an Carya
illinoensis, in the Mississippi River Valley
near Muscatine, Iowa. Quaternary Res.
33:1U2-107.
Blackmon, G.H. l9?5. Status of the Pecan
industry in Florida. Proc Ann. Fla. State
Hortic. Spc.39:7$77.
Brinkerhoff, L..{. and R.B. Streets. l%.
Pathogenicity and pathological histologt of
Phynatotrichum omnivqun (the fungus causing
cotton or Texas root rot) in a woody perennial,
the pecan. Univ. of Ariz Agri. Enp. Stn. Tech.
Bull. 111:101126.
Burkett, J.H. l93L The pccan in Texas. Ter
Dep. of Agric Bull. 111.
131
Callaham,
R.Z LnO.
Geographic variation in
forest tre€s, pp. 43-47. In: O.H. Frankel
and E. Bennet (eds.) Genetic Resources in Plants Their Exploration and Consemation. F.A. Davis
Co., Philadelphia, PACampbell, R.D. 1n8. Pecans for the true north.
Annu. Rep. North. Nut Grow. Assoc. 69:55-59.
Harlan, J.R. ln7. Sources of Genetic Defense.
pp 35-356. In: P.R. Day (ed.) The Genetic
Basis of Epidemics in Agriculture. Ann. NY Acad.
Sci., Volume 287.
Harris, ICD. and C.G. Tauer. 19{37. Variation
among open pollinated provenances of pecan grown
in an Oklahoma nursery. OK State Univ. Res. Rep.
P-892.
Reed and M.N. Wood. 1937. Nut
breeding. USDA Yearbook Separate No.
H.L, C.A
Crane,
1590:827-890.
Heiges, S.B. 1896. Nut Culture in the United
States; embracing native and introduced species.
U.S. Dept. Agric., Pomology Div., Govt. Printing
Office, Washington D.C.
Deneke, F.J., D.T. Funk and C. Bey. 1980.
Preliminary seed collection zones for black
walnut. U.S. For. Serv. Northeastern Atea.
Hepting,
NA-FBA,f-4.
Trees
For. Serv. Agri. Handbook No.
P.
1977. Baker Cave plant macrofossil
report. Unpublished manuscript on file with the
Dering,
G.H. 1971. Diseases of Forest and
of the United States. U.S. Dept. Agric.
Baker Cave Project, Texas Archeological Research
Laboratory, Univ. of Texas at Austin.
Gast, R.T. and J.P. Overcash. 1980. Pecan
rootstocks and scion cultivars produced by
commercial nurseries in the Southern United
States. Pecan South 7(2):14-18.
Grauke. L.J. and J.W. Pratt. 1985. Pecan
rootstock research. Proc. SE Pecan Grow. Assoc.
Shade
385.
T.R. 1981. l:te Paleo-Indian occupations
at Baker Cave, southwestern Texas. Bull. Tex.
Arch. Soc. 53:101-119.
Hester,
Hinrichs, H.A 1965. Pecan investigations in
Oklahoma. Annu. Rep. North. Nut Grow. Assoc.
56:23-25.
Johnson, J.D., H.P. Smith, W.H. Thames, Jr., L.R.
Smith and others. 1y75. Nematodes can be a
problem. Pecan Q. lO(2):5-7.
78:123-136.
Grauke, L.J. and J.W. Pratt. 7987. The effect of
rootstock on bud growth in pecan. HortScience
22:718.
Grauke, L.J., J.A Payne and B.W. Wood. 1989a.
North American pecans: A provenance study. Annu.
Rep. North. Nut Grow. Assoc. 80:
Grauke, L.J., R.S. Sanderlin and R.D. O'Barr.
1989b. The effect of rootstock on nutrient
accumulation in pecan. HortScience 24
(Suppl.):127.
Gray, O.S. 197. Requisites of good pecan
nursery stock. Proc. Texas Pecan Grow. Assoc.
7:33-40.
Hanna, J.D. 1972. Absorpton and accumulation of
chloride ions by pecan (Carya illinoensis
Koch) seedling rootstocks. Ph.D. Dissertation,
Texas
A&M University, College Station,
Hanna,
J.D. 1987.
Iz; R.C. Rom (ed.)
Amsterdam.
Kormanik, P.P. and J.L. Ruehle. 1989.
First-order lateral root development: something to
consider in mother tree and progeny assessment.
Proc. South. For. Tree Improv. Conf. 2O:220-227.
Lutz, H. 1938. The effect of size of young pecan
trees on their subsequent growth and yield. Proc.
Am. Soc. HortSci. 36:335-346.
McGranahan, G. H. 1987. Research priorities in
genetic improvement of Penian walnuts and their
rootstocks. Annu. Rep. North. Nut Grow. Assoc.
78:73-77.
Texas.
G.D. 1974. Breeding for the development
pecan Carya illinoensrs (Wang.) IL Koch,
seedling rootstocks. Ph.D. Dissertation, Toras
Madden,
Pecan rootstocks, pp.
Rootstocks for fruit
crops. J. Wiley and Sons, New York.
r32
Kormanik, P.P. 1989. Importance of first-order
lateral roots in the early development of forest
tree seedlings, pp. 1.57-169. In V.
Vlastimil and F. Kunc, (eds.) Interrelationships
between Microorganisms and Plants in Soil.
Dwelopments in Soil Science 18. Elsevier,
nl-410.
of
A&M University, College Station,
Texas.
Madden,
Pecan
G.D. ln6.
Q.
Methods improve
lO(2\:5-7.
rootstock.
Madden, G.D. 1978. Progress report on pe€n
seedling rootstock investigations 1971-1978, at
Brownwood, Texas. Oklahoma Pecan Grow. Assoc.
48:6G65.
Madden. G.D. and H.L. Malstrom. 1975. Pecans
hickories, pp. 420-438. In: J. Janick and
and
J.N. Moore (eds.) Advances in Fruit Breeding.
Purdue Univ. Press. West l-afayette, IN.
Manning,
W.E. 1949. The genus Carya in
Mexico. J. Arn.
Arb.3O:425-432.
Thompson, T.E. 1990. 1990 update: Pecan
cultivars: current use and recommendations. Pecan
South 24(1):12-17,20.
T.E and LJ. Grauke. 1990. Genetic
in pecan and other hickories. In:
Thompson,
resources
J.N. Moore and J.R. Ballington (eds.) Genetic
Resources of Temperate Fruit and Nut Crop. Intl.
Soc. for Hort. Sci.
Thompson, T.E. and F. Young. 1985. Pecan
Cultivars - Past and Present. TX Pecan Grow.
Assoc. Inc. 255pp.
Thompson, T.8", F. young, W.O. Mcllrath, H.D.
Petersen and othen. 1989. Pecan nut and kernel
characteristics show genotype_environment
interactions. J. Am. Soc. Hort. Sci. 114:?06-711.
Marshall, D.R. 7977. The advantages and.hazards
of genetic homogeneity. In' P.R. Day (ed.)
The Genetic Basis of Epidemics in Agriculture.
Ann. NY Acad. Sci., Volume 287.
Traub, H.p. 1931. The variation of seedlings
Pecan Grow'
tre€s' Proc'
Miyamoto, s., c.R. Gobran and rc piela.
Salt effects on seedling growth and ion uptake of
three pecan rootstock cultivars. Agron. J.
yarnell, S.H. lgt4. pecan seedling rootstock
77:383-388.
studies. proc. fi pecan Grow. Assoc. l4:L5-17.
1ess.
Ou, S.IC 1990. Pre-harvest sprouting in pecans.
A&M Univ., College
Ph.D. dissertation. Texas
Station, fi
J.R. 18C7. Nus for Profit: A treatise.
on the propagation and cultivation of nut-bearing
trees adapted to successtul culture in the United
states with extracts from leading authorities.
Sinnickson Chew, Printer, Camden, New Jersey.
Parry,
x*:"Ifi:1rr*n
fi
Wood, B.W. lgg7. Status and directions of pecan
rootstock research. Annu. Rep. North. Nut Grow.
Assoc. 7g:6_71.
Wood, B.W. 19g9. pecan production responds to
root carbohvdrates and rootatock. J. Am. Soc.
Hort. Sci. il,+,ZZl-ZZS.
Romberg, L.D. and C.L Smith. l9M. Effects of
cross-pollination, self-pollination, and
sib-pollination on the dropping, the volume, and
the kernel development of pecan nuts and on the
vigor of the seedlings. J. Am. Soc. Hort. Sci.
47:13O-138.
Sitton, B.G. and F.N. Dodge. 1938. Growth and
fruiting of thre€ varieties of pecans on different
seedling rootstocks. J. Am. Soc. Hort. Sci.
36:721-125.
Jr. 7967. Plant remains, pp.
220-255. In: D.S. Byers (ed.) The Prehistory
of the Tehuacan Valley. Vol. 1. Environment
Smith, C.E.,
Subsistence. Univ. of
and
Texas Press, Austin.
Taylor, W.A 1894. Report of the Assistant
Pomologist, pp.277-296. In: 1893 USDA
Yearbook of the Dep. of Agric. Ed. G. W. Hill.
r33
Table
1.
Climate indicators for pecan collection sites in the
U.S.1 and Mexico2
I.at.
(m.)
Long.
(cmlyr)
Elev.
I-ak., TN
Bow., TX
Wsh., MS
40
39
38
38
37
37
36
34
33
T.G., TX
31
v.v., TX
Gon., TX
30
30
29
29
23
22
20
20
94
97
94
88
97
95
90
94
91,
1m
101
97
100
100
99
100
99
rM
97
213
792
226
725
347
274
94
tt9
39
580
3r3
95
227
341
884
1219
1829
1372
1737
Mean
Site
Location
1.
Liv., MO
2.
3.
Jer.,
4.
5.
6.
7.
8.
9.
Web., KY
Cow., KS
10.
11.
12.
IL
Ver., MO
Chr., KS
13.
Kin.,fi
14.
Zav.,TX
15.
Jau.,
16.
t7.
s.c., Mx
bm., MX
18.
Say.,
19.
MX
MX
Oax., MX
t7
(c)
Prec.
(c)
Temp.
89
90
122
72.4
11.4
13.9
13.8
14.6
74.3
14.8
113
77.r
100
111
80
lm
132
46
44
98
54
26.4
24.9
25.6
25.1.
25.7
25.6
26.1,
17.4
22.7
23.0
1,8.7
21,.1,
21,.0
79.r
19.8
54.
2t.o
75
24.4
19.2
19.0
19.8
11.8
%
17.9
7.8
39
90
1,4.2
4.3
t9.L
8.0
20.6
5.1
&
27.9
lAnonymous. 1986. Climarological data. Annual Summary
(bystate) NoAA
2Garcia de Miranda, Enriqueta andZaidaFalcon de Gyves.
1980. Nuevo Atlas Porrua de la Republica Mexicana. 6th ed.
Editorial Porrua, S. A, Mexico.
134
Range
Table2. Population effects pecan seedling emergence, height and diameter.
Location
(days)
Emerg.ze
(cm)
Ala
Demopolis, AL
Greene Co., IL
Gulf coast, LA
18.2aw
17.9a
22.8
ru
lou
okl
Nur sdstk
Bra
Scu
TGr
Nue
Dev
LMo
SFe
Cha
OK
Various
Brazos R., TX
Scurry Co., TX
San Angelo, TX
Nueces R., fi
Devil's R., TX
[:s Moras, fi
Del Rio, TX
Champion, TX
S. Cent.
Dia.x
(mm)
Ht.Y
Population
g
g
23.8
16.6 b
1,6.2 b
1.4.9 c
27.7
26.4
14.8
L4.4
14.2
29.3
26.7
26.4
29.5
29.0
28.4
ef
cd
de
de
de
de
e
3.26
3.42
cd
ef
ef
cd
3.18
3.30
3.58
c
cd
cde
e
de
c
3.24
3.29
de
de
3.67 b
32.2 b
24.7
3.41
de
de
cd
de
4.26a
35.2^
c
cd
13.7
13.7
13.6
13.6
12.8
de
3.24
3.25
fg
3.18
e
t Duyr from planting to first emergence from soil.
Y
Heigtrt measured 54 dap after planting.
x
Diameter measured 55 days after planting.
w
Mean separation in columns by Duncan's multiple range test, 5Volevel.
135
40.0
c
-o
1nn
tr
E^-^
o'ZJ.U
'o,
I
c
! ZU,U
o
0)
a
c
n
o 15.0
Meon Percent Kernel by Fomily
Figure 1. The correlation between mean nul. percent kernel
by family and mean seedling pcrcent height by
family for 53 families of open-pollinated pecans.
3
o
'2
)
Figure 2. The effect ofrootstock on inception ofbud
growth of grafted pecan trees. Means separated by
Duncan's multiple range test, 0.05 level. Bars having
the same letter cannot be seoaratcd.
136