MEASURING OF THE NECTAR YIELD OF RED CLOVER HEADS

Juhani Paatela and Hannes Heinrichs

Department of Plant Husbandry, University of Helsinki

Received February 2, 1960

Red clover is almost entirely self sterile so that cross pollination is necessary
for the development of seeds. The pollinators of red clover gather from the flowers
apart from pollen in particular nectar. The obtaining of nectar can be difficult
because of the length of the corolla tube or the small quantity of nectar in the
flowers. Much more attention has been paid to the length of the corolla tube and
its variations in different varieties, climatic conditions etc. than has been paid to
the amount of nectar and its variations in corresponding cases (2, 4, 6—12). This
may arise from the lack of a suitable method of measuring the nectar yield.

The water extraction method of measuring the nectar yield gives only the
amount of sugar contained in the nectar. The micro-pipette method (14) is prefer-
able for many purposes. The centrifugal method is, however, more rapid than
this and gives more useful results. The main principle is as follows: the red clover
head is set in a glass tube and centrifuged, whereby the nectar under the influence of
the centrifugal force is liberated from the flowers and accumulates in a column at
the bottom of the narrow lower part of the tube, where its volume or weight can be
easily measured (3, 4, 12, 13).

At the University Farm at Viik an attempt was made in 1954to develop the cen-
trifuge method into a simple and quick routine-method. This was necessary to make
an investigation of a great number of samples possible, and particularly, owing to
the great variation in the amount of nectar in red clover heads, within each sample
which thus comprised some 20—40 heads. Besides the conditions prevalent on the
field, and the grown varieties, the size and stage of development of the inflorescence
can affect the variation of nectar content. In what follows attention is paid to some
of the most important factors affecting the nectar yield and the centrifuge method
itself.

The taking and handling of the sample

Probably the most important factor contributing to the fluctuation in the amount
of nectar and its sugar content is the loose rain and dew between the flowers. There-
fore the sample should be taken only at those times of the day when there is no more

https://www.c-info.fi/en/info/?token=Sh9uKjb8L-vzSCbc.jgP-mtV3dMfpQtHh6ED2Hw.QajYpkudN8Z6KHF_zdPFz1-Zn8KgePqf1EL-31tMObgFVtBrcMfU3a8T12ydofUuIGCaQCdK75K-rd4LT3wusFm05qMqbRFNFerM-0HO5QV-R1jcmNRyQypvRsGJDoRHgssNADGBLWDjaoV1M-09EtbIAqFyLdEM1QNVUNBElNWEWGFwylZKOLPgpCG7Kr7c3Q


89

loose water in the beads. Therefore the heads harvested for a sample were shaken
a few times against the back of the hand. If no drops of water appeared on the skin,
the head was sufficiently dry.

The heads harvested for a sample had also to be of the same size and at the
same known stage of development, as well as grown in the same microclimatic condi-
tions, in other words, at the same height and in equally dense stands. The size of
the heads and the blossom stage were estimated ocularly. The checking calculations
indicated that the size of the head could be estimated on an average with an accuracy
of about ten florets. An effort was made to take the samples of heads middle-sized
inflorescences cons: sting of about a hundred florets. To illustrate the significance
of the size of inflorescence (number of florets) for the nectar yield, 48 large and
small-sized heads it the same blossom stage were gathered. The average nectar
yield of the 48 small heads was 18.2 ± 1.80 /d and that of large 33.4 ± 2.24 /A.
The difference between the average nectar yields of the large and small heads was
highly significant (t = s.3o***). The average number of florets on the small heads
was 83 and the average length of the corolla tubes 9.35 mm, and correspondingly
112 and 9.73 for la rge heads. Calculated per hundred florets there was more nectar
in the large (29.8 ± 4.42 /d) than in the small heads (20.9 ± 4.79 /d). The average
nectar yields differed from each other highly significantly (t = 9.47***). The sugar
content of the nectar of the large inflorescences was also higher (34.9 %) than that
of the small heads (30.4 %). This result indicates that the secretion of nectar and
sugar is more rapid in the large heads than in the small sized ones grown in the same
stand. Thus for obtaining representative results with a moderate number of heads the
sample should he composed of inflorescences of as near as possible the same (medium ) size.

Swanson and Shuel (13) have investigated the effect of the blossom stage of the
head on the quantity of the nectar yield. During the first day after full blossom
no change in the amount of sugar was noticeable, but from the second day and
during the third t ie sugar yield per head decreased by two thirds. At the University
Farm the samples were taken when three uppermost verticils were just opening,
in other words, the head was approaching the full bloom stage, which was easy to
verify, even though the finding of a large number of heads might produce difficulties.
An attempt was nade to take the samples from normally dense uncovered stands
of about the same height. Perhaps more even results can be obtained from covered
plots, because the bumble-bees and bees are prevented from sucking nectar from
the florets. Howe ver, in these too the hereditary differences in nectar yield cause
a noticeably large variation.

The harvested heads were gathered in plastic bags and taken quickly to the
laboratory in order to prevent noteworthy changes in the content or the quality of
the nectar yield. Swanson and Shuel investigated these changes and 132 clover
heads were split lengthwise and one lot of half-heads was centrifuged one hour after
harvesting, the other lot two hours after harvesting. The correlation coefficient
for the nectar yield between the respective lots of half heads, based on a sample
of 32 paired halves, was 0.73. This value is the same as that obtained for the simul-
taneous centrifugation of paired halves (0.76 for a sample of 132 paired halves). The
results for sugar concentration were similar. At the University Farm the harvesting.
2



90

handling and examining of 48 samples lasted on an average about two hours, whereby
during this period hardly any statistically notable changes occurred in the amount
and quality of the nectar. In the handling of the samples the bract of the head was
first removed. In accordance with the method employed 4—6 mms were cut off
from the tip of the florets to facilitate the excretion of the nectar. This procedure,
however, proved to be laborious and lengthy. Even with careful cutting of the
florets nectar stuck to the scissors. Moreover the floret snippets remained at the
head, from which they were difficult to remove without damaging them. The
necessity of this procedure was tested by centrifuging 40 cut and 40 uncut inflores-
cences. From the former an average of 35.0 /A of nectar was obtained and from
the latter 33.9 [A. The average sugar content of the nectar was correspondingly
33.4 ± 0.98 % and 35.0 ± 0.93 %. The cutting of the heads was thereby shown to be
unnecessary, at least when sufficiently large glass tubes were used.

The glass tubes and the setting of heads in them

The method employed by Swanson & Shuel (13) is in its main features the
following: the harvested inflorescences are set inside a frame made from two metal
wires about 2.5 cms high and 1.5 cms wide fixed in the cork. From the opposite
side a thin needle is inserted through the cork and into the floral axis. The cork
is then inserted into the end of the prepared glass tube with the head pointing
downwards. In the centrifugation the nectar exudes from the florets and collects
in a column at the bottom of the narrow lower part of the tube, in which the volume
is easily measured. The method proved in certain respects slow and laborious.
Firstly it was difficult to insert the needle into the floral axis without breaking it,
nor was the needle able to hold the inflorescence in position during centrifugation,
since it easily became pressed against the frame beneath. Several florets were thus
broken nor was the nectar able to flow freely from them. In addition part of the
florets came loose twisting in the frame. The broad upper part of the glass tube, on
which the heads were placed upside down, was not sufficiently large at least for large
and uncut inflorescences resulting that part of the florets easily became attached
to the glass wall of the tube thus preventing the nectar from flowing freely from
all the florets. Another hindrance to the flowing of the nectar into the narrow
lower part of the glass tube, into the capillary tube, lay in the fact that the upper
part narrowed too sharply, so that part of the nectar remained on the wall of the
upper part. The method employed by Stapel & Götzsche (12), which forms a
basis for the method employed by Gubin (4) and Evert (3), differs from the above
particularly in the way of fastening the inflorescence; here part of the stem of the
head is threaded through a hole made in the cork. The cork is placed in the glass
tube inflorescence first and a second cork is inserted into the hole, whereby the stem
of the inflorescence remains between both the corks. In inserting the upper cork
the lower cork easily becomes so tightly wedged in the tube that its extraction is
difficult, in addition to which the stems easily break.



91

At Viik it was decided to fasten the heads as follows: the stem of the inflores-
cence is threaded through a hole in the rubber stopper and bent over the top of
the stopper. Then a frosted glass wedge about 4 cms long and a shade thicker than
the hole in the stopper is carefully inserted into the hole (Fig. 1). Because of the
soft and flexible wall of the rubber stopper, the stems are not easily crushed, and
the frosted surface of the wedge prevents the stem or the wedge itself from slipping
out of position. The rubber stopper is inserted head first in the glass tube (Fig. 2).

Figure 1. Glass tube with its capillary part, rubber stopper with hole, frosted glass wedge and brass
gauze.

Figure 2: Glass tubes in the frame after centrifugation. The rotation speed of 3,500 r.p.m. has caused
part of the stems to break. In the capillary tubes the nectar can be seen as a dark line, at the bottom

of the broad upper part of the tube lies the metal gauze.



92

Before this a small, concave, fine-meshed gauze (size of holes 0.7 mm) is placed at
the head of the capillary tube at the bottom of the broad part of the glass tube.
This metal gauze prevents dirt which is loosened in centrifugation from being
thrown into the capillary tube, which greatly facilitates the accurate measuring
of the height of the nectar column. In experiments employing the above method
of fastening it was proved that the head remained in position at a rotation speed
of even 3,000 r.p.m., which corresponded to 1710 gms centrifugal force. At greater
rotation speeds the stems began to break (Fig. 2).

Figures 1 and 2 also show the form of the glass tubes used: the total length is
16 cms, of which the broad upper part is 8 cms and the capillary tube 8 cms. The
diameter of the upper part is 3 cms and the glass 2 mms thick. About five cms
from its mouth the tube begins to taper off towards the narrow capillary tube,
which has a diameter of 1.5 mms. The inflorescences, even uncut, fit well into these
glass tubes and the nectar is allowed to flow from them into the capillary tube,
a process which is also aided by the gradual tapering off of the tube.

Swanson & Shuel (13) in their experiments on nectar yield used 465 gms
centrifugal force and a centrifugation time of 6 minutes. An International Clinical
centrifuge, in which there was room for four tubes, was employed. In the experi-
ments carried out at the University Farm it was shown that the centrifugal force
(corresponding to 1,500 r.p.m.) was insufficient for the extraction of the nectar,
at least from uncut florets. An International size 2 Model V centrifuge was used
in which 8 glass tubes fit simultaneously (Fig. 3). Several experiments were
made to ascertain the necessary centrifugal force and centrifugation time for extract-
ing the nectar. The nectar yields and sugar content of inflorescences of equal size

Figure 3. An International size 2 model V centrifuge used in the centrifuging of heads.



93

and at the same blossom stage were measured 1y2, 3,6, and 9 minutes after
centrifugation. A speed of 2,000 r.p.m. corresponding to 760 gms centrifugal force
was already sufficiently great to extract most of the nectar, but the centrifugation
time showed no noteworthy significance with regard to the yield and quality of
the nectar. Employing speeds of 2,500 and 3,000 r.p.m. corresponding to 1316 and
1710 gms centrifugal force, a slightly larger amount of nectar was obtained than
at a speed of 2,000 r.p.m., but with an increase in the centrifugation time the sugar
content decreased. The cause of this was believed to lie in the fact that other
liquids, in addition to the nectar, had become mixed with it and were diluting it.
Consequently, a speed of 2,500 r.p.m. was considered appropriate with a centrifuga-
tion time of 6 minutes of full speed.

Measuring the nectar yield

The nectar yield can be expressed per inflorescence in units of weight or volume,
or correspondingly per 100 florets or grammes of inflorescence. At the University
Farm it was decided to express the nectar yield in units of volume (microlitres)
per 100 florets. If, in connection with the taking of a sample, it is possible to take
approximately representative medium-sized inflorescences, it may be possible to
express the nectar yield per medium-sized inflorescence without a great deal of
inaccuracy. To obtain reliable results, however, the number of florets should be
counted or at leas'; the inflorescence should be weighed. The centrifuged nectar
yield was obtained by measuring the height of the nectar column in the capillary
tube with a millimetre scale (accuracy to 0.1 mm.) and by multiplying the height
with the coefficient indicating a volume of 1 mm. This was ascertained by intro-
ducing 0.1 ml of liquid into each tube with a pipette before use, and marking on
the broad upper part of the glass tube a number showing the length of the column
of liquid, e.g. 5.4 ans as in Fig 1. In this tube a 1 mm column corresponds to 1.79
ja\ of nectar.

Determination of the sugar content of the nectar

When the nectar in the capillary tube had been measured, it was sucked into
a fine pipette, front where it was transferred to a refractometre prism. A thermostat
attached to the refractometre gave the nectar a temperature of 20° C at the moment
of measuring (nl) 20). Having obtained the sucrose content in this way (5), the
actual sugar content of the nectar is also quite accurately shown, because by far
the greatest part of the nectar of red clover is sucrose (Park, 1932, ref. 1,
p. 263).



Summary

The paper gives an account of the method to measure the nectar yield of red
clover heads. The method was developed at the University Farm at Viik in 1954,
and based on two centrifugation methods employed earlier. The main attention
was paid to the taking of the sample and its fast handling, because the heterogeneous
nature of the material demands even in the case of one sample the examination of
a considerably large number of heads.

The secretion of nectar and sugar was more abundant in the large heads than
in the small-sized ones so that for the obtaining of realiable data with a moderate
number of heads the sample should be composed of heads of as near the same size
(medium size) as possible. The shortening of the corollas by cutting before centri-
fugation proved unnecessary at least where sufficiently large glass tubes were
employed. By changing the method of fastening the inflorescences and the form
of the glass tube, and by inserting a metal gauze at the bottom of the capillary tube
the clean nectar was easily directed into the capillary tube (Figs. 1 and 2).

REFERENCES

(1) Boetius, J. 1948. Über den Verlauf der Nektarabsonderung einiger Blütenpflanzen. Beihefte
zur Schweizerischer Bienenztg. 17: 258—315.

(2) Braun, E., Mac Vikar, R. M., Gibson, D. R., Pankiw, P. & Guppy, J. 1953. Studies in red clover
seed production. Part 11. Can J. of Agric. Sei 33: 437—447.

(3) Evert, R. 1939. Der Blutennektar. Leipziger Bienenztg. 54: 173—176.
(4) Gubin, A. F. 1936. Bestäubung und Erhöherung der Samenernte bei Rotklee Trifolium pratense L.

mit Hilfe der Bienen. Arch für. Bienenkunde 17: 209—264.
(5) Jakobs, M. 1938. Chemical analysis of foods and food products. 537 p. New York.
(6) Johnsson, L. H. 1946. Nectar secretion in clover. Effect of soil and climate on honey production.

New Zealand J. Agr. 73: 111 112.
(7) Paatela, J. & Heinrichs, H. 1959. Puna-apilan kukkien mesipitoisuuden merkityksestä sen

siementuotannossa. Summary: The effect of nectar content of red clover heads on the
yield of seed. Maatal. ja koetoim. XIII: 167—178.

(8) Shuel, R. W. 1952. Some factors affecting nectar secretion in red clover. PI. Physiol. 27: 95—110.
(9) —* 1955 a. Nectar secretion in relation to nitrogen supply, nutritional status and growth

of the plant. Can. J. of Agricult. Sei. 35: 124—138.
(10) 1955 b. Nectar secretion. American Bee Journal 95: 229—233.
(11) Shuel, R. W. & Pedersen, M. W. 1952. The effect of environmental factors on nectar secretion

as related to seed production. Proceedings of the sixth international grassland congress
I: 867-870.

(12) Stapel, Chr. & Götzsche, O. 1941, Om Nektarsekretion og Kronrorslaengde mm. hos Rodkloveren
under forskellige Godningsforhold. Tidskr. for planteavl 46: 267 295.

(13) Swanson, C. A. & Shuel, R. W. 1950. The centrifuge method for measuring nectar yield. Plant
Physiol. 25: 513—624.

(14) Vansell, G. H, 1943. Alfalfa nectar and the honey bee. Amer. Bee J. 83: 106 107.

94



95

SELOSTUS:

PUNA-APILAN KUKINTOJEN MEDEN MÄÄRITTÄMISESTÄ

Juhani Paatela & Hannes Heinrichs

Yliopiston kasvinviljelytieteen laitos, Helsinki

Tutkimuksessa on tehty selkoa Yliopiston opetus- ja koetilalla Viikissä v. 1954kehitetystä meden-
määrittämismenetelmäätä, jokapohjautuu kahteen aikaisemmin käytettyyn sentrifugointimenetelmään
(3, 4, 12, 13). Pääasiallinen huomio on kiinnitetty näytteen ottoon ja sen käsittelyyn päämääränä nopea
mutta silti luotettava rutiinimenetelmä, koska erityisesti aineiston heterogeenisyys jo yhden näytteen
osalta vaatii jopa 20—40 kukinnon tutkimista.

Meden ja sokerin eritys oli suurissa kukinnoissa runsaampaa kuin pienikokoisissa, joten varmojen
tulosten saavuttamiseksi pienillä kukintomäärillä näytteen tulee ollakoostunut mahdollisimman saman-
kokoisista (keskikokoisista) kukinnoista. Teriöiden lyhentäminen leikkaamalla ennen sentrifugointia
osoittautui tarpeettomaksi ainakin kyllin suuria lasiputkiakäytettäessä. Myös kukinnon kiinnittämis-
tapaa ja lasiputken muotoa muuttamalla sekä käyttämällä kapillaariputken suulla metallilankaverk-
koa saatiin mesi helposti ja puhtaana valumaan kapillaariputkeen (kuvat 1 2). Käytettyä kiinnitys-
tapaa kokeiltaessa todettiin kukintojen pysyvän hyvin paikallaan vielä kierrosnopeuden ollessa 3000
kierr./min., mikä vastasi 1710 g:n keskipakoisvoimaa. Käytettävissä oli International-merkkinen
(koko 2, malli V) sentiifuugi (kuva 3). Nopeus 1500kierr./min. (vastaten 427 g:n keskipakoisvoimaa) ei
vielä ollut riittävä meden irroittamiseksi ainakaan leikkaamattomista kukista. Kokeilujen tuloksena
päädyttiin käyttämään nopeutta 2500 kierr./min. (= 1316 g) sekä 6 minuutin sentrifugoimisaikaa,
mikä laskettiin nopeuden saavuttamisesta lukien. Mesimäärä ilmaistiin mikrolitroina (yul) ja se saatiin
helposti mitatuksi kun tunnettiin kapillaariputken läpimitta ja mesipatsaan korkeus. Siirtämällä
hienopipetillä mettä kapillaariputkesta Abben refraktometrin prismalle ja mittaamalla meden taite-
kerroin (nD 20° C) saatiin meden sukroosipitoisuus (5), joka lähimain ilmaisee myös meden todellisen
sokeripitoisuuden, koska valtaosa puna-apilan medestä on sukroosia,