Ann. Bot. 46, 313-321, 1980
Development of the Almond Nut {Prunus dulcis
(Mill.) D. A. Webb). Anatomy and Chemical
Composition of Fruit Parts from Anthesis
to Maturity
J. S. H AWKER and M. S. BUTTROSE
CS1RO Division of Horticultural Research, G.P.O. Box 350, Adelaide 5001, Australia
ABSTRACT
The growth of the fruit of two varieties of almond (Prunus dulcis (Mill.) D. A. Webb) was studied from
anthesis (week 0) to maturity (week 32). The dimensions, fresh weight, moisture content, anatomy and
chemical composition of the pericarp, testa, embryo, endosperm and nucellus are recorded diagrammatically, graphically and by micrographs for one variety. Of the two ovules present at flowering only
one normally developed further. By 12 weeks after flowering the whole fruit had reached full size. The
space enclosed by the pericarp was filled by nucellus until week 10, with subsequent enlargement of both
endosperm and embryo. From week 16 to week 20 the embryo increased to full size with a concurrent
decrease in the size of the endosperm. Sixteen weeks after flowering, the embryo began to accumulate
protein and lipid, little of which originated from either the nucellus or endosperm. The embryo contained
no starch or reducing sugar but up to 3 % sucrose in the early stages which decreased as lipid and protein
increased. Starch and sucrose levels were high in the testa at week 16 but subsequently dropped, starch
more rapidly than sucrose. The role of the testa in transport of metabolites to the embryo is discussed.
Key words: Prunus dulcis, almond, fruit development, anatomy, embryo, endosperm.
INTRODUCTION
The almond fruit {Prunus dulcis (Mill.) D. A. Webb) is a drupe with a leathery dry and
tough mesocarp. Unlike fleshy drupaceous fruits, it has only two phases of growth as
measured by total weight and size, stage III or the final rapid growth of the mesocarp
being absent in the almond (Brooks, 1939). The seed or kernel of the almond reaches
full size at the end of stage I, but within the seed the endosperm increases in size in
stage II, then decreases as the embryo increases. Various aspects of almond development
and composition have been studied by several workers over the years (Antoni, 1969;
Benken and Rikhter, 1971; Brooks, 1939; Chuvaev, Semenova and Shirshova, 1962;
Galoppini and Lotti, 1962; Pavlenko, 1940; Sequeira and Lew, 1970; Souty et ai, 1971;
Winton and Winton, 1932), but a comprehensive study covering the complete growing
period has not been made.
The present work attempts to remedy that situation by detailed observation of the
growth and development of two varieties of almond fruits at the macroscopic and
microscopic level and by analysis of the major chemical components of the parts of the
almond fruits. To avoid undue repetition of previous reports the results for the growth
of only one variety are presented in enough detail to allow an understanding of the
development of the kernel within the pericarp. Data for the other variety is only given
when it differs from that of the first variety.
O3O5-7364/8O/O9O313 +14 $02.00/0
II
© 1980 Annals of Botany Company
BOT 46
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
Accepted: 28 November 1979
314
Hawker and Buttrose—Development of the Almond Nut
MATERIALS AND METHODS
RESULTS
Growth of whole fruit
Trees flowered during the last week of July and the first week of August. In this paper
fruit age is specified in terms of weeks after flowering, and 30 July was taken as Day 0.
Growth of fruits is illustrated in Fig. 1. Overall dimensions reached a maximum by
10-12 weeks after flowering, as did the space finally occupied by the kernel or embryo.
However the fruits were not mature until some 28 weeks after flowering when the
mesocarp dehydrated and abscission began. The space enclosed by the pericarp (epicarp
+ mesocarp+endocarp) was filled by nucellus during the enlargement phase up to
10-12 weeks. The endosperm then proliferated for a period of 6 weeks at the expense
of the nucellus with a concurrent growth of the embryo. At 18 weeks there was no sign
of nucellus. Due to continuing embryo growth the endosperm decreased in size between
16 and 20 weeks, but in the material examined it never fully disappeared. The embryo
reached final size at some 20 weeks after flowering. The pericarp remained undifferentiated until about 16 weeks after flowering when secondary thickening of cell walls of
the endocarp was noted.
ANATOMY
General
At flowering, the almond ovary contains two ovules (Plate 1 A), of which normally only
one develops (Plate 1 B). The ovary wall develops to form the pericarp of the mature
fruit, the outer and inner integuments of the ovule (Plate 1 c) develop to form the seed
coat (testa), the nucellus (Plate 1A, c) progresses and disappears as shown in Fig. 1,
and the endosperm and embryo develop from the embryo sac (Plate 1 c). A developmental stage is shown in Plate 4 c.
Pericarp
The ovary wall at flowering time was about 40 cells and 0-45 mm thick at a midlateral level, with many areas staining heavily with the PAS-Toluidene Blue treatment
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
Fruits of almond (cv. Chellaston and Johnston's Prolific) were collected at 2-week
intervals from anthesis to maturity (30 July 1974 to 11 March 1975; week 32). Samples
of almonds were taken from ten marked trees in a commercial orchard at Willunga, 40 km
south of Adelaide, South Australia (see Moss, 1964). Linear measurements and fresh
weights were taken on each of 50 almonds and their dissected parts. Only almonds with
normal kernels were used; shrivelled kernels were discarded. Moisture content, lipid and
starch were determined on triplicate 1 g samples of randomized tissue and duplicate
assays were carried out on triplicate samples of randomized tissue for protein, reducing
sugar and sucrose. Reducing sugar, sucrose and starch were determined as described
before (Walker and Hawker, 1976) and protein and total lipid as described by Hawker
and Bungey (1976).
For microscopy representative flowers of Chellaston were sampled on 16 July, and
developing fruits were sampled at regular intervals (at least fortnightly) until maturity
in early February. Entire flower ovaries and very young fruit were fixed whereas for
developing fruit different tissue portions were fixed separately. Tissue was fixed in 4 %
glutaraldehyde in 0025 M phosphate buffer at pH 7 0 for 4 h at room temperature,
dehydrated in an ethanol series and embedded in glycol methacrylate. Sections 2-25 /ira
thick were mounted on glass slides and stained in PAS reagent followed by Toluidene
Blue O. Some sections of maturing embryos were stained for protein by Fast Green.
Hawker and Buttrose—Development of the Almond Nut
40
\-_^ Pericorp
V/y\ Endosperm
H|j Nucellus
[^] Embryo
315
20
0
0
22
26
28
(Plate 1 A, B). By 3 weeks it was 100 cells and 2-2 mm thick (Plate 2D) and by 6 weeks
it was 150 cells and 4-5 mm thick (Plate 3 A). Thereafter there was little increase in cell
number although cells enlarged during the expansion phase up to 10-12 weeks. The
mesocarp cells enlarged more than those of the endocarp and signs of intracellular
differentiation were well recognized by 16 weeks. These signs were accumulations of
heavily stained contents in many mesocarp cells (Plate 3 B) and pockets of cells beginning
secondary thickening in the endocarp (Plate 3 c). This differentiation in the pericarp
progressed over the following weeks. At 22 weeks the endocarp could still be sectioned
but the pockets of secondary thickened cells were numerous and the thickening more
advanced (Plate 3D). At later times sectioning of the endocarp became impracticable.
The abscission zone at which mesocarp separates from endocarp at ripeness was not
evident until towards 28 weeks.
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
20
FIG. 1. Diagrammatic representation of growth of the principal parts of the Chellaston almond
fruit. Details of pericarp are shown only at the final date. Dimensions are the mean of 50
observations. Numerals beneath fruits refer to weeks after flowering (30 July).
316
Hawker and Buttrose—Development of the Almond Nut
Testa (seed coat)
Nucellus
Cells of the nucellus increased rapidly in number and size during the expansion phase
of fruit growth up to 10-12 weeks. The cells had non-staining contents and folded walls
(Plates 2 A and 4 A). AS the nucellus was displaced by endosperm, the nucellar cells
collapsed and a layer of adpressed cell walls was seen at the perimeter of the endosperm
(Plate 2 c).
Endosperm
Initially the tissue was non-cellular (Plate 4A), but thin cell walls soon formed and the
cells appeared similar to those of the nucellus (Plate 2 c). The distinct endosperm haustorium was readily seen at early stages on dissection (Plate 4 B), and the tissue at later
stages could be separated cleanly from the surrounding nucellus (Plate 4E) and from the
embryo (Plate 4F). AS the endosperm was displaced by embryo, the endosperm cells
collapsed and layers of cell walls were seen as described for the nucellus. Apical cells
of the endosperm were never displaced (Fig. 1) and at maturity these desiccated and the
cell walls adhered to the crumpled seed coat.
Embryo
Early development of the embryo was slow and at 6 weeks it was still only a group of
cells on the end of the suspensor (Plate 4D). By 8 weeks it had grown to heart-shape
(Plate 4B), by 12 weeks the cotyledons were well developed and the shoot and radicle
were beginning to form (Plate 5 A). By 14 weeks the radicle was well developed (Plate 4E)
and by 18 weeks the embryo was approaching anatomical maturity (Plate 4F). At this
stage, however, and even at 20 weeks (Plate 5 c), the cotyledon cells had stored little
protein or lipid. Storage occurred over the next 8 weeks after which cells were packed
with protein and lipid (Plate 5D). The suspensor maintained a cellular connection
between the developing embryo and a small surviving pocket of nucellar cells, which
was located at the apex of a non-cellular, heavily-stained region (Plate 5A, B).
CHEMICAL COMPOSITION
The moisture content, f. wt, protein, total lipid, starch, sucrose and reducing sugar
content for Chellaston almonds are shown in Figs 2-8. Qualitatively, similar results
were obtained for the variety Johnston's Prolific, and the quantitative results at week 32
near maturity were such that the embryo f. wt reached 2-4 g, and contained 320 mg
protein and HOOmg lipid. By comparison, at week 32 Chellaston embryos weighed
21 g and contained 280 mg protein and 860 mg lipid.
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
In a transect across the outer and inner integuments at flowering there were about
20 cells (Plate 1A, C) and this number scarcely changed throughout development (Plate 2 A,
B, c). Cells comprising the outer epidermal layer of the testa enlarged greatly during the
first few weeks after flowering and became filled with darkly-staining contents (Plate
2 A, B, c), but no development was observed after 8 weeks when final dimensions of the
testa were attained. Cells comprising the inner epidermis remained very small and
provided a strong barrier from which the nucellus was readily torn away (Plate 2B).
Between 12 and 16 weeks the cells contained heavily stained starch granules. The testa
was well provided with vascular tissue (Plate 2 B, C), and this was supplied by a massive
vascular strand from the pericarp (Plate 4 c). Degeneration of the testa was noted
between 14 and 16 weeks, especially collapse of the innermost cells (Plate lc). By 28
weeks it was wholly collapsed and formed a dry furry skin over the mature embryo.
Hawker and Buttrose—Development of the Almond Nut
16
20
Time (weeks)
24
28
32
Time (weeks)
FIG. 2. Moisture content of parts of the fruit of the almond (Primus dulcis, cv. Chellaston).
Before the 10th week, the embryo, testa, nucellus and endosperm were measured together and
are shown as the kernel. • , Testa; O, embryo; • , pericarp; D, endosperm; A, nucellus;
A, kernel.
Fio. 3. Fresh wt of the kernel and of its constituent parts of almond (cv. Chellaston) during
development. Prior to week 10, the nucellus and testa were not weighed separately. • , Testa;
O, nucellus; D, endosperm; • , embryo; A, kernel.
At about the 16th week after flowering, the f. wt of the embryo (the part of the almond
which develops into the edible portion) began to increase and the embryo started to
accumulate lipid and protein (Figs 3-5). The moisture content of the embryo fell as
lipid and protein increased (Fig. 2). The embryo contained less than 0-01 per cent starch
and no detectable reducing sugar but up to 3 per cent sucrose in its early stages of
development (Figs 6-8). The concentration of sucrose decreased as lipid and protein
accumulated (Fig. 7). In the testa at about the 16th week there were peaks of concentrations of starch (6 mg per testa) and sucrose (32 mg per testa) which subsequently
dropped, starch more rapidly than sucrose (Figs 6, 7). Reducing sugar concentration in
the testa (9 mg per testa) was relatively low at about the 16th week and the increase after
about the 20th week (Fig. 8) was due to the loss of water from this tissue. In contrast to
the testa, the pericarp at week 16 contained about 1 per cent sucrose, 3 per cent reducing
sugar and 0-2 per cent starch (Figs 6-8).
The nucellus and endosperm contained only low levels of protein and lipid and at the
very most < 70 mg of sucrose and reducing sugar (Figs 4, 5, 7, 8). Thus the 280 mg
protein and 860 mg lipid in the mature embryo did not originate, except in a small way,
from these two nutritive tissues. Similar changes occurred in the variety Johnston's
Prolific but generally some 2 weeks later than for Chellaston.
An experiment was carried out to determine whether the green fruit fixed sufficient
carbon by photosynthesis to provide a large part of the precursors for the developing
embryo. Girdling (removal of a 5 mm ring of bark with a knife) of small fruit-bearing
branches along with removal of leaves from these small branches at week 14 resulted in
poor development of the almond fruit. At week 26, the pericarps were shrivelled and the
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
12
317
Hawker and Buttrose—Development of the Almond Nut
318
400 _
800-
T
5T
/
{
\
300
5
/
/
1
-'!
}
/
1
1
4
400 -
\i
/
/
100
/
i
200 -
/
--•
s
•
0 -
\
•o
I
I
|
t
1
1
12
16
20
Time (weeks)
1
24
i
28
32
12
16
20
Time (weeks)
Fio. 4. Protein content of the fruit of the almond (cv. Chellaston) during development. • ,
Testa; O, embryo; • , pericarp; D, endosperm; A, nucellus; A, kernel.
Fio. 5. Total lipid content of the fruit of the almond (cv. Chellaston) during development. • ,
Testa; O, embryo; • , pericarp; D, endosperm; A, nucellus; A, kernel.
kernels weighed only 0-5 g compared to 2-7 g for the normal kernels, indicating that
either the adjacent leaves or transport of precursors from the rest of the tree were
essential for the filling of the kernel, i.e. the accumulation of lipid and protein by the
embryo.
DISCUSSION
The growth pattern of the almond fruit has been discussed previously by Brooks (1939).
Of more interest in the current paper is the pattern of accumulation of protein and lipid
in the embryo and the nature and pathway of movement of the precursors for these
compounds. Girdling experiments in the present work and comparisons with other
fruits suggest that little of the required precursors are obtained by fixation of the carbon
by the green pericarp of the fruit. Metabolites stored in the testa, nucellus and endosperm
during the first growth stage (up to about week 16) could only contribute about 10 per
cent of the storage material finally accumulated by the embryo. That fact, however, does
not detract from the importance of these tissues in providing precursors for the cellular
development of the embryo.
Obviously the metabolites for storage in the seed come from some part of the plant
other than the fruit tissues and the question remains by what route. From the dimensions
of the suspensor cells and the amount of lipid and protein accumulating in the embryo
between weeks 20 and 24, it can be calculated that if all of the substrate was moving via
the suspensor cells the specific mass transfer (SMT) would be about 15 mg cm~* s"1.
Rates of SMT in phloem have been measured as between 016 and 1-75 mgcm~ 2 s - 1
(Canny, 1975), although rates as high as 50 mg cm" 1 s - 1 have been found in the phloem
of wheat roots (Passioura and Ashford, 1974). On the other hand, the rate of active
32
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
200 -
600-
Hawker and Buttrose—Development of the Almond Nut
319
0-6
16
12
16
20
Time (weeks)
20
Time (weeks)
2 4 2 8 3 2
FIG. 6. Starch content of the testa and pericarp of two varieties (cv. Chellaston (solid symbols)
and Johnston's (open symbols) Prolific) of almond during development. Less than 001 per cent
starch was present in the other parts of the almonds. • , O, Testa; • , Q, pericarp.
FIG. 7. The concentration of sucrose in almond (cv. Chellaston) during development. • ,
Testa; O, embryo; • , pericarp; • , endosperm; A, nucellus; A, kernel.
5-0
4-0
*
3O
I 2-0
1
1-0
8
12
16
20
24
28
32
Time (w«eks)
FIG. 8. The concentration of reducing sugar in the almond (cv. Chellaston) during development.
• , Testa; • , pericarp; Q, endosperm; A, nucellus; A, kernel.
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
12
320
Hawker and Buttrose—Development of the Almond Nut
ACKNOWLEDGEMENTS
Cheryl Mares, Kathy Scott and Jill Milln gave technical assistance. The use of the almond
trees in the orchard of H. C. T. Stace is gratefully acknowledged.
L I T E R A T U R E CITED
ANTONI, Z., 1969. Morphological investigations on fruit development in almonds. Szdld-Gyumolcsterm
5, 35-44.
BENKEN, A. A. and RIKHTER, A. A., 1971. Biochemical studies on almond fruits during maturation.
Yalta. Gosudarstvennyi Nikilskii bolanicheskii Sad. Trudy, SI, 125-43.
BROOKS, R. M., 1939. A growth study of the almond fruit. Proc. Am. Soc. hort. Sci. 37, 193-7.
CHUVAEV, P. P., SEMENOVA, N. A. and SHIRSHOVA, A. M., 1962. The dynamics of the accumulation and
translocation of carbohydrates in relation to the course of the accumulation of fats in the almond
and pistachio in Tadzhikistan. 7>. Ord. Fiziol. i Biofiz. Rast. Akad. Nauk Tadzhiksk. SSR 1, 156-85.
CANNY, M. J., 1975. Mass transfer. In Encyclopedia of Plant Physiology, New series, Vol. 1: Transport in
Plants 1. Phloem Transport, eds M. H. Zimmermann and J. A. Milburn, pp. 139-53. Springer-Verlag,
Berlin, Heidelberg, New York.
GEIGER, D. R., 1975. Phloem loading. Ibid. pp. 395-431.
GALOPPINT, C. and Lorn, G., 1962. The ripening of almonds with special consideration of the lipid
composition. Olearia 16, 164-7.
HAWKER, J. S. and BUNGEY, D. M., 1976. Isocitrate lyase in germinating seeds of Primus dulcis. Phytochemistry 15, 79-81.
HUTCHINGS, V. M., 1978. Sucrose and proton co-transport in Ricinus cotyledons. Planta 138, 238-41.
Moss, D. E., 1964. Growing almonds in Australia. Wld Crops 16, 76-83.
PASSIOURA, J. B. and ASHFORD, A. E., 1974. Rapid translocation in the phloem of wheat roots. Aust. J.
PI Physiol. 1, 521-7.
PAVLENKO, O. N., 1940. The chemical composition of the almond kernel during the ripening process.
Biokhim. KuV-tur. Rastenii 1, 461-6.
PREISS, J., 1978. Regulation of adenosine diphosphate glucose pyrophosphorylase. Adv. Enzymol. Relat.
Areas Mol. Biol., ed. A. Meister, Vol. 16, pp. 317-81. John Wiley, New York.
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
uptake into phloem is some 10000-fold lower than the rate of SMT within phloem
(Geiger, 1975). There seems no reason to believe that the suspensor cells act as sieve
tubes and we conclude that the metabolites entering the embryo do not pass exclusively
(if at all) via the suspensor cells. The testa as a pathway for the movement of compounds
from the stems to the embryo is suggested by the changes in metabolite present as the
embryo increases in size and in protein and lipid content. Starch reaches a peak in
concentration at week 16, subsequently drops while sucrose increases and then sucrose
decreases. It is possible that as the embryo becomes a larger sink, sugars are moved to
the embryo from the testa and starch synthesis is replaced by starch hydrolysis as the
supply of starch precursors decreases. Concentrations of substrates such as 3-phosphoglyceric acid and phosphorylated sugars which are known to activate ADPglucose
pyrophosphorylase might decrease and the concentration of inorganic phosphate could
increase, both events which would result in lower rates of starch synthesis by ADPglucose starch synthase and higher rates of starch degradation by phosphorylase (Preiss,
1978; Preiss and Levi, 1979).
There are no well formed vascular connections to the embryo from either the pericarp
or the stem but the testa which surrounds the embryo contains a network of vascular
tissue towards its outside. Metabolites must pass through the inner layers of the testa
to reach the nucellus, endosperm and embryo and the testa is obviously an important
organ in the movement of metabolites into the inner parts of the seed. The high concentration of sucrose present in the testa and the role of sucrose as a translocant in
plants suggests that sucrose is the form in which carbon moves into the embryo. Possibly
a proton-sugar-cotransport system operates as has recently been suggested for phloem
loading of higher plants (Hutchings, 1978).
HAWKER
AND
lllilROSt—Developirc-'
- > ' ''•:'
-!'—-••••/
V
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
PLATE 1
/l/i/i. But.
46. 313
321.
1980
(Facing p.
320)
HAWKER AND BUTTROSE—Development of the Almond Nut
mmm
mmm
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
:t
n. Bol. 46. 313 -?21 1980
PLATE :
HAWKER AND BUTTROSE—Development of the Almond Sut
•
•
•
•
"
#
;
«. So/. 46, 313-321, 1980
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
EN
PLATE 3
HAWKER AND BUTTROSE—Development of the Almond Nut
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
Aim. Bin. 46, 313-321, 1980
PLATE 4
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
,-i™. /?<;?. 46, 313-321. 1980
BUTTROSE—Development of the Almond Nut
HAWKER AND
PLATE 5
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
Hawker and Buttrose—Development of the Almond Nut
321
PREISS, J. and LEVI, C , 1979. Metabolism of starch in leaves. Chapter 23 in Encyclopedia of Plant Physiology, New series, Vol. 6: Photosynthesis II. Photosynthetic Carbon Metabolism and Related Processes,
eds M. Gibbs and E. Latzko. Springer-Verlag Berlin, Heidelberg, New York 282-312.
SEQUEIRA, R. M. and LEW, R. B., 1970. The carbohydrate composition of almond hulls. / . agric. Fd.
Chem. 18, 950-1.
SOUTY, M., ANDRE, P., BREUILS, LILIANE and JACQUEMIN, G., 1971. Etude sur la qualiti des Amandes
Amygdalus communis L: Variabilite de quelques caracteres biochemiques. Ann. Techno/, agric. 20,
121-30.
WALKER, R. R. and HAWKER, J. S., 1976. Effect of pollination on carbohydrate metabolism in young
fruits of Citrullus lanatus and Capsicum annuum. Phytochemistry 15, 1881-4.
WINTON, A. L. and WINTON, K. B., 1932. The Structure and Composition of Foods, Vol. 1, pp. 474-85.
John Wiley, New York.
EXPLANATION OF PLATES
PLATE 2
A. Section showing testa (T) and nucellus (N) at 8 weeks after flowering. Note the layer of small, thickwalled cells forming the inner epidermis of the testa. PAS-Tol.Blue O. Bar = 100 /tm.
B. Section of the testa (T) at 8 weeks showing a clean separation from the nucellus, and the enlarging
outer epidermal cells filled with densely-staining contents. Note vascular bundles (arrows). PASTol.Blue O. Bar = 100/tm.
c. Section at 18 weeks showing testa (T), nucellus (N) and endosperm (E). The inner cells of both testa
and nucellus had collapsed. Note vascular bundle (arrow). PAS-Tol.Blue O. Bar = 100 ;im.
D. Median transverse section of pericarp at 3 weeks, with epicarp (EP), mesocarp (ME) and endocarp
(EN). PAS-Tol.Blue O. Bar = 100/tm.
PLATE 3
A. Median transverse section of pericarp at 6 weeks. ME = mesocarp, EN = endocarp. PAS-Tol. BlueO.
Bar = 100/tm.
B. Transverse section of epicarp and mesocarp at 16 weeks. PAS-Tol.Blue O. Bar = 100/tm.
c. Section of endocarp at 16 weeks. Note the smaller cell size compared with that of the mesocarp in B,
and the onset of secondary thickening (arrow). PAS-Tol.Blue O. Bar = 100 /tm.
D. Section of endocarp at 22 weeks. Secondary thickening was more advanced (arrows). PAS-Tol.Blue O.
Bar = 100 /tm.
PLATE 4
A. Section at 3 weeks showing endosperm (E), apparently still with free nuclei (arrows), nucellus (N),
and testa (T). PAS-Tol.Blue O. Bar = 100/tm.
B. Endosperm dissected from fruit at 8 weeks showing the haustorium extension. Inset shows heartshaped embryo from a corresponding endosperm (location indicated by arrow). Bar = 1 mm.
c. Median longitudinal section of a developing fruit at 16 weeks with scale included. P = pericarp,
EM = embryo, E = endosperm, N = nucellus, T = testa. A massive vascular bundle enters the
testa from the pericarp at the double arrow.
D. Longitudinal section at 6 weeks showing the embryo (EM) on the end of the suspensor (S), surrounded
by endosperm (E). PAS-Tol.Blue O. Bar = 10/tm.
E. Embryo (EM) with attached endosperm (E) after dissection from a developing fruit at 14 weeks.
Bar = 1 mm.
F. Embryo (EM) with detached endosperm (E) at 18 weeks. Bar = 1 cm.
PLATE 5
A. Median longitudinal section at 12 weeks showing embryo (EM) with its cotyledons surrounded by
endosperm (E). Portion of the persisting suspensor appears in this section (double arrow). Note the
small pocket of nucellus cells (N) followed (in a basal direction) by a non-cellular, dark-staining area.
The numerous dark dots in the testa (T) are starch grains. PAS-Tol.Blue O. Bar = 100/tm.
B. Detail of the persisting suspensor (double arrow) connecting embryo (EM) and the small pocket of
nucellus cells (N) from a fruit at 10 weeks. E = endosperm. PAS-Tol.Blue O. Bar = 100/tm.
c. Section from the centre of cotyledons at 20 weeks. Note the general lack of staining. Bar = 100 /im.
D. Section from centre of cotyledons at 28 weeks. Protein bodies stained with Fast Green. Bar = 10 /tm.
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011
PLATE 1
A. Transverse section of almond ovary one week before flowering. The ovary wall (OW) encloses two
ovules (O). I = integuments, N = nucellus. PAS-Tol.Blue O. Bar = 100/tm.
B. Transverse section of ovary within one week after flowering showing ovary wall (OW), two ovules
(O), integuments (I) and the enlarged nucellus (N) of a fertilized ovule. PAS-Tol. Blue O. Bar = 100 /tm.
c. Median longitudinal section of ovule just before flowering showing micropyle (M), outer integument
(OI), inner integument (II), embryo sac (ES) and nucellus (N). PAS-Tol.Blue O. Bar = 100 /tm.
Downloaded from aob.oxfordjournals.org at University of California, Davis on March 29, 2011