Ent. exp. & appl. 1 (1958): 66~72. North-Holland Publishing Co.

, Amsterdam

THE BEHAVIOUR OF APHIS FABAE I N SELECTING ITS

HOST P L A N T S , ESPECIALLY D I F F E R E N T VARIETIES OF
VICIA FABA
BY

H. J. MULLEK
Institut fiir Pflanzenztichtung, Quedlinburg, Deutschland

The different initial infestation of the two field bean varieties Rastatter and Schlanstedter
(in the relation 1 : 3 till I : 5) by Aphis (Doralis) fabae Scop. is caused by a preference
behaviour of the alates, which land on both varieties of bean in equal numbers, but whereas
99% leave Rastatter in a few minutes, only 90% leave Schlanstedter.

According to PAINTER (1951) the hereditary resistance of plants against in-

festation by insects may depend on three different ecological principles: Firstly
on the t o l e r a n c e of the host plant, that is not weakened or damaged by
the infestation. Secondly on a n t i b i o s i s, that is on the check of the growth
or reproduction of the pest by an insufficient nutritive quality or by toxic effects
caused by special properties of the plant or thirdly on p r e f e r e n c e, that
is on individual choice of the insects, which prefer distinct species or varieties
of plants more or less refusing others.
The fact, that the different varieties of beans are infested by the black aphids
in a very different measure, has been known for a long time, but nearly nothing
is known about the causes of these differences. N o tolerance could be observed
as yet. As the bean aphid must be looked upon as an extremely polyphageous
insect, it seems improbable that the more or less distinct resistance of field bean
varieties might be based on preference. DAVIDSON (1922, 1925) who, in the
twenties, was the first to investigate the causes of the resistance of beans against
aphids, therefore based his research entirely on an examination of the antibiotic
effects of the plant. He placed a certain number of young adults on plants of
the tested varieties and after a fixed time he counted the numbers of their off-
spring. As these numbers were strikingly different in the tested varieties, he
considered that they gave a measure of resistance.
With a bean assortment at Quedlinburg, too, it was not possible to get a
reproduceable series from resistant varieties to highly susceptible ones based on
the reproduction rate. It appeared possible that there existed some mechanism of
preference. In this we confined our investigations on the much susceptible variety
Schlanstedter (S) and on the variety Rastatter (R) considered to be relatively
resistant, thirty plants of either being cultivated repeatedly in an alternating double
row of fifty to fifty centimetres. By a daily sampling of all Mate aphids and the
colonies of young larvae left by them respectively, we determined on every
morning the mean infestation resulting during a period of 24 hours on each plant.
 BEHAVIOUR OF APHIS FABAE 67

All repetitions yielded the difference of 1 : 3 to 1 : 5 between Rastatter and

Schlanstedter. In the course of 147 test-days during the growing period in 1950,
695 initial infestations have been established on 60 Rastatter plants but 3427 on
60 Schlanstedter, that means 83% of the entire infestation being found on
Schlanstedter. How constantly this primary settlement kept appearing is also made
clear by the graph (fig. 1). Considering the frequency distribution on the days on
which observation were made during the test years 1949 and 1950 of the height
of the percentage infestation of both varieties, there is a maximum near 85% of
S (sample mean 69,5 • 29,1%). Obviously the winged aphids are able to
distinguish the two varieties from one another and at the initial infestation they
prefer the S 3 to 5 times more than the R. Considering the progressive repro-
duction of aphids it is evident that this different preference is the primary cause
of the resistance observed by growers and farmers, even if an antibiotic effect
could be proved as well.

ii
fil
I n,

.!
fii!,i
II -'="
II nl
II. I
fulfill
liB l
IIm.I
Illlll
_11 I I IIIII.
% III U lllllllll IIIIIII

an_N.lnmm,, ~
il ili'!
I I I I I .lllllllllmlMIII]~nni~ll lltill~MtmmnmL IN h . n l
I
n~n
nLnw
2 # ~ # miler ~ , J ~ m N # l ~
~f~j
Fig. 1. The daily numbers of the initial infestations by Aphis ]abae on 60 Rastatter (white)
and 60 Schlanstedter (black) beans during the summer 1950

These results raise the question, how these differences of preference come
about. Till now we have not succeeded in analysing the behaviour of the bean
aphid in choosing its host plant in the laboratory or in the greenhouse by experi-
ment, as it is hardly possible to keep winged aphids in the infestation stage. The
young winged aphids are easy to breed, but they show a strong flying impulse
connected with a distinct positive phototaxis and do not react either on host
plants or on any stimulation to land. But while outdoors after a shorter or Ionger
flight this phase turns over again to negative phototaxis and to a tendency to
settle on host plants, the alates in the greenhouse will stay in the brightest place
of the roof, starving rather than showing normal reactions to their host plants.
Therefore the further analysis of the preference resistance of Vicia faba to
Aphis fabae could be tried only in the field, where the experimental conditions
are not constant and the opportunities for observation are seldom favourable, but
where instead of that the normal natural conditions prevail.
First of a11 we proved whether the preference applied to certain parts of the
bean plant, or if all organs equally caused the difference in the initial infestation
of the two varieties. It became evident in a series of tests with 30 decapitated R-
68 H . J . MOLTER

and S-plants each, that plants deprived of all vegetation points indeed showed a
smaller total infestation, but the S - - as in the control - - kept showing 90% of
the initial infestation. The same result came about when, reversely, all leaves were
removed and only the sprouts were left intact. The suspicion that the extra-floral
nectaries might cause the difference, was removed by experiments on plants
without stipulae, for here also 80% of the initial infestation developed in S-
plants. Even the reduction to one single primary leaf did not change the dif-
ference. There was the same result with cropped leaves freshened in water and
arranged in the field in the usual mixed double row. Not matter whether primary
or secondary leaves were used as a material, the S got about 80 to 83% of the
daily found colonies.
Neither does the age of the leaves influence the difference in the preference
of the two varieties. It is true: on tests on plants with 3 young, middle aged
(ripe) and older (yellowish) leaves each, the total infestation decreased to 49,
33 and 18% respectively. But the S unvariably attracted 75 to 80% of initial
infestations. Neither particular organs nor the physiological stage of the plants
or leaves change the preference relation.
Before we know anything about the behaviour of the aphids infesting a new
host plant it appeared possible that the initial infestation observed in 24-hour
intervals was coming about by a primary choice, that is: at first arrival, at the
first landing on a plant. One could believe it might be caused by the different
smell of the varieties, so that the corresponding organs of the aphids allowed
them to make a choice at a distance.
In the course of the last decade however some observations of KENNEDY
(1950 a, b), and MOERICKE (1955) and the author (H. J. MOLI.ER 1951 u.
1953) showed us that there is a great restlessness among the aphids in the in-
festing stage, and that of the numerous landing aphids only very few settle there
definitely. The first hint in the direction of a secondary choice was brought about
by the comparison of the initial infestation of the single plants with catches by
yellow dishes that were observed simultaneously in a R- as well as in a S-plot at
24 hours' intervals. While the initial infestation was as usual with S, the number
of Aphis fabae in the yellow dishes of both plots were practically always the same
(fig. 2). Both plots had obviously been flown over by the same number of
aphids in the infestation stage; however the S had been settled 5 times more
frequently than the R. It remained still undecided whether the aphids had landed
in equal numbers on both varieties and then most of them had left the R or if
they had primarily avoided the latter.
Shortening the control intervals to 1 hour and finally to 20 minutes did not lead
to any final proof. The percentage of the initial infestation on the S went down
accordingly from 87 to 70 and 65% respectively, but this difference is not
statistically significant.
All we could do was to catch uninterruptedly all the bean aphids alighting on
two comparable plants of the two varieties over a period of many hours. During
a period of more than 40 hours a total of 874 Mate aphids were captured, 53,3%
on R and 46,5% on S. The same plants, when exposed to colonisation by aphids
both before and after these observations on alighting had been made showed the
typical differences, i.e. 85% of the total infestations were on the S plant. This
 BEHAVIOUR OF APHIS FABAE 69

was definite proof that both varieties were primarily infested by equal numbers
of infesting stage aphids and that the difference of preference exclusively arose
from secondary choice, namely from "trying" or testing. This conception was
confirmed by 27 hours' continued observation of, altogether, 600 Mates alighted
on two similar plants of both varieties during 9 different control periods in July

16
15
ld
12
11 ffne't;~g Colonies
9

Segdow dz~de~
11
fd_

N I t 1 I I I I I [
6 15 g5o/o35 45^55 G5 ~5 85~ ,95

Fig. 2. Comparison between the daily numbers of the initial colonies on beans and the
alate aphids (Aphis fabae) captured in yellow dishes in separated Rastatter (R) and
Schlanstedter (S) plots (of 32 plants each) by the distribution of the control days accor-
ding to their percentage of colonies on S-plants (black) respectively of specimens in the
dishes of the S-plot (hatched).

1952. Out of 312 alates alighted on R, only 3 definitely settled, that is: they bore
some larvae (2 other cases remaining doubtful because of the nightfall), all but
1% left sooner or later. But on the S-plant 29 out of 289 alighting alates ( + 5
doubtful cases) remained for birth of larvae, that means 10%. It is therefore
secondary choice that decides the difference of preference, but it is not simply
that on the preferred variety all alighted aphids remain sitting and only from the
less preferred plants the aphids fly away, but arrival and departure on both
varieties are different in percentage. This significant difference of 1 : 1 0 of
plants under constant observation is greater than at the daily counting of the
initial colonies, where it is 1 : 5. This face is clear if we consider that with the
70 H . J . Mi)LLER

latter method a great number of aphids are caught that have not yet definitely
settled and that would soon have left. This amount however, being the same on
both varieties, decreases the difference relation.
The fact that on the highly infestable and most preferred S bean only 10%
of the alighted aphids remain sitting demonstrates that the choice of the host
plant must be masked in a highly effective and variable manner by other factors
or other ways of behaviour. The final settlement will always be smaller than the
arrival. We know however that the intensity of the infestation flight is strongly
dependent on the temperature, on the humidity and on the wind-speed. Ac-
cordingly the host preference is strongly dominated by two antagonistic reactions:
the flying impulse and the infesting impulse, both of them being influenced
contrarily by factors of environment. High temperature and humidity increase the
flying impulse, lower ones increase settling. Low wind-spee& make possible un-
restricted flying off and high wind-speeds may force an aphid to settle when it
would have left the plant under calm conditions. A tired aphid on the contrary
may settle under favourable flying conditions on a plant that it would have
refused in a former phase.
It is very difficult therefore to tell anything definite about the real causes of
the acceptance or refusal of one of the two varieties by the alighted aphids. The
searching time of the future settlers, that is the time between the landing and the
final settling, was with the 3 R-settlers 13, 14 and 46 minutes; with the 29
S-colonists it varies between 2 and 32 minutes with an accumulation between 2
and 17, especially between 2 and 10 minutes, the mean being 11 minutes. The mean
staying time of visitors, that is the time between landing and taking off again,
was on the contrary longer for those on S beans (61/2 minutes) than for those
on R beans (31/2 minutes); these differences, notwithstanding the variability of
the periods in different cases, are statistically significantly different. Therefore
most of the visitors are leaving R-plants sooner than S-plants (fig. 3).

50

tlO 9 o,, R.,z=g,,O&m,=~.,*mln

~A on ~ r;=gl~ m,=gG rain
30

to

o-* ]g a * 5|a ~ 8 s w|tl 73 15 w t5 ~G 1~ la 19 $#],~ Zgmia. a n e l # o n ~ e r

Fig. 3. Frequency distribution of the staying times of alatae (from alighting till take off)
on a Rastatter and a Schlanstedter bean plant in percent of the total numbers of visitors
during altogether 27 hours of flight.

During the searching and the staying time the Mates wander about a great
deal on the surface of the plant, however interrupting their wandering very often,
setting up their proboscis and laying back their antennae. We could see by the
 BEHAVIOUR OF APHIS FABAE 7I

remaining salivary sheaths that the so-called test feeding punctures (which mostly
do not take more than 20 tot 40 seconds) very seldom penetrate beyond the
epidermis and obviously penetrate into the intercellular tissue and constantly end
blindly in the intercellular spaces. It seems improbable that on doing so they might
suck up or taste anything.
The significance of the test feeding punctures is still unknown. As they end
blindly it is doubtful whether the aphids imbibe by them any material for tasting.
According to unfinished histological investigations the real feeding punctures
pass always intercellularly to the phloem through tissues with no or little inter-
cellular spaces. Therefore I should suppose the test feeding punctures to be a
search for tissues having no intercellulars. These are normally found in young
organs and in the surroundings of the vascular bundles, which are enveloped by
sclerenchymatic or collenchymatic sheaths reaching to the epidermis. These are the
very places, where aphids like to settle finally. Here the punctures automatically
lead to the phloem. Probably the stylets are able to indicate by lever effects minute
differences of pressure in the tissues. Therefore the aphid continues to insert its
sty/ets only where it feels a close contact with the neighbouring cells. If we could
prove this, we could suppose that the causes of the difference of preference were
to be seen in the size and structure of the intercellular systems of the varieties. W e
could assume that the aphids succeed more often on susceptible varieties, less often
on resistant ones to get to tissues with no or very few intercellular spaces, in
order to find the phloem and so to be induced to settle down. In this way the
statistical distribution of the infestation would be understandable.
Some of the aphids (alates and apterae) will change places once more or twice
after the birth of the first or of a few young larvae. There is a distinctly per-
ceivable tendency with the aphids to leave R-plants more frequently than S-
plants, and at the next choice even the R-infestors prefer S-plants. Here perhaps
an antibiotic effect is t o be observed.
As we can now see the differences in resistance of the tested varieties of field
beans arises first by a different host preference of the aphids. Moreover antibiotic
effects may play a minor role too. The causes of the different preference are not
clear as yet. It is sure only that the preference behaviour with varieties develops by
a secondary choice not before the landing and is influenced by manifold exogenous
factors (as plant condition, weather and so on) and by endogenous ones too (as
age, physiological stage and motivation of the aphids).

ZUSAMMENFASSUNG

Als prim~ire Ursache des unterschiedlichen Befalls der beiden Ackerbohnensorten Rastatter
und Schlanstedter durch die Schwarze Bohnenlaus, Aphis (Doralis) fabae Scop., wird
Pr~iferenzreslstenz nachgewiesen. In mehrj~hrigen Beobachtungsreihen ergab sich bei 24-
stiindigen Kontrollen ein Unterschied in der Anzahl der Initialkolonien yon 1 : 3 bis 1 : 5
auf Rastatter und Schlanstedter. Direktes Abfangen bzw. eingehende Beobachtung der auf
den Pflanzen gelandeten Gefliigelten zeigte, dab diese Initialbefallsdifferenz erst sekund~ir
entsteht. Die Pflanzen beider Soften werden statistisch in gleicher H~iufigkeit beflogen,
jedoch siedeln sich auf Rastatter nut 1%, anf Schlanstedter 10% der Gelandeten fiir l~ingere
Zeit (bis zum Absetzen wenigstens einer Junglarve) an. Die Bedeutung der sogenannten
Probesaugstiche sowie der Gr/SBe und Verteilung der Interzellularr~iume in den Pflanzen-
geweben ftir das Auffinden des Phloems llnd sornit fiir die Wirtswahl wird diskutiert.
 72 H. 1. MffILLER

R E F E R E N C E S

DAVIDSON, J., (1922), Biological studies of Aphis rumicis L. Reproduction on varieties of

Vicia faba. Ann. appl. Biol. 9, 135--145.
- (1925), Biological studies of Aphis rumicis L. Factors affecting the infestation of
-

Vicia faba with Aphis rumicis. Ann. appl. Biol. 12, 472--507.
KENNFDY, J. S., (1950a), Aphid migration and the spread of plant viruses. Nature, Lond.
165, 1024.
-- (1950b), Host-finding and host-alternation in Aphides. Verb. VIII. Intern. Congr.
Entomol., Stockholm, 4 pp.
MOERICKE, V., (1955), f3ber die Lebensgewohnheiten der gefliigelten BIattli~use (Aphidina)
unter besonderer Berficksichtigung des Verhaltens beim Landen. Z. angew. Ent. 37,
29--91.
MOLLER, H. J., (1951)~ Elber die Ursachen der unterschiedlichen Resistenz yon Vicia faba L.
gegentiber der Bohnenblattlaus Doralis fabae ScoP. III. Ober das WirtswahlvermSgen
der Schwarzen Bohnenblattlaus Doralis fabae StoP. Ziichter 21, 161--179.
- (1953), IV. Das Zustandekommen des unterschiedlichen Initialbefalls. Ziichter 23,
-

176--189.
PAINTER, R. H., (1951), Insect resistance in crop plants. Macmillan Company New York,
52O pp.

DISCUSSION
H. J. DE FLUITER : W i e tief versenken die Blattl~use beim ,,Probieren" ihre Stechborsten
ins Pflanzengewebe, bevor sie sie wieder zuriickziehen ?
H. J. MOLLER: Die Probesaugstiche enden fast ausschlieBlich zwischen den Epidermis-
zellen bzw. in den Interzellularriiumen der unmittelbar darunter liegenden Zellschicht,
jedoch niemals i n den Zellen selbst. Sie erreichen in keinem Falle ein Leitbiindel.