[ 135 ] ON INTRASPECIFIC AND INTERSPECIFIC COMPETITION IN ... · BY A. C. CROMBIE From the Zoological Laboratory, Cambridge (Received 20 October 1943) (With Five Text-figures) The

[ 135 ]

ON INTRASPECIFIC AND INTERSPECIFIC COMPETITION INLARVAE OF GRAMINIVOROUS INSECTS

BY A. C. CROMBIE

From the Zoological Laboratory, Cambridge

(Received 20 October 1943)

(With Five Text-figures)

The behaviour of the beetle, Rkizopertha dommica,is adapted to the needs of its larvae in that the sensa-tions which induce the females to oviposit wouldusually lead to eggs being laid only in environmentswhere sites suitable for larval development are to befound (Crombie, 1942). But the rate of oviposition,even at the extremes of adverse conditions possiblein actual populations, is so great that a considerablemortality must occur in the immature stages beforethe offspring are reduced to numbers which theenvironment could support (Crombie, 1942). Thefemales of the moth, Sitotroga cerealella, appear like-wise to have little control over their rate of oviposi-tion (Crombie, 1942). Their ability to select, undernatural conditions in the field, an oviposition sitesuitable for larval development is imperfectly known(King, 1918; Simmons & Ellington, 1927, 1933), butin a confined space it seems adequate (Simmons &Ellington, 1924; Crombie, 1942). The females ofneither species avoids ovipositing on already infestedgrains. In the course of the experiments on ovi-position already published (Crombie, 1942) it wasfrequently noticed that the eggs of either specieswere deposited next to or even on top of previouslyoviposited eggs of the same or of the other species(cf. Ullyett, 1936; Salt, 1937; Lloyd, 1939, 1940).Wheat grains already containing larvae of eitherspecies also seem to be as acceptable to the femalesof both species as uninfested grains. Even if thegravid females did avoid ovipositing on previouslyinfested grains her efforts would be vitiated both bythe size of the egg batches often placed on one grain,and by the habit of the active first instar larvae ofboth species of wandering away from the place wherethey have hatched. The rudimentary nature of' parental care' in these two species of insect meansthat the larvae must be adapted, particularly duringthe early instars, to an independent existence. Thiswill expose them, in choosing and occupying situa-tions in which to complete their development, to thecompetition of other larvae similarly engaged. Inthis paper the term superinfestation will designatethe cases in which one wheat grain is inhabitedsimultaneously by two or more young of the samespecies of insect. The term multiple infestation willdesignate the cases in which one wheat grain is in-

habited by the young of two or more different speciesof phytophagous insect (cf. Smith, 1929). It is im-portant to know firstly, whether the larvae themselvesavoid multiple or superinfestation, and if so, bywhich senses they perceive the larva already in occu-pation of the grain; and secondly, if multiple andsuperinfestation are not avoided, how overcrowdingaffects their development and what is the outcomeof competition between the two species in mixedpopulations (cf. MacLagan, 1932; Timofeeff-Res-sovsky, 1933, 1935; L'Heritier, 1937; Bodenheimer,1938). Overcrowding may evidently affect livingorganisms through the limitation of food, oxygen orspace (leading to reflex stimulation caused by theperception of other individuals, e.g. by actual con-tact), or the accumulation of excretory products('conditioning' of the medium) (Allee, 1931, 1934;Hammond, 1938, 1939; Clements & Shelford, 1939;Park, 1939, 1941; MacLagan, 1941). When insectlarvae are exposed to such conditions this may leadto the death of individuals before reaching maturity(MacLagan & Dunn, 1935; Salt, 1936); the reduc-tion in size of the pupae and resulting adults(Weidling, 1928; Holdaway, 1930; Salt, 1932a), orof the vigour and longevity of the latter; the loweringof the fecundity of the adult females (Hofmann, 1933;Dunstan, 1935); retardation (Landowski, 1938) orstimulation (Michal, 1931), of the rate of development;changes in the sex ratio of the emerging adult popu-lation (Herms, 1928; Brandt, 1937; Flanders, 1939;Graham, 1939); or failure to make full use of thefood reserves of the medium because of competitionfor space. The latter may lead to survival being lowerat high than at low densities.

The experimental conditions in these experimentswere the same as those described in a previous paper,to which reference should be made (Crombie, 1942).All the experiments were performed in a dark incu-bator at 30° C. and 70% R.H.

The larvae of both species usually bore into asuitable object (e.g. a wheat grain) in the first instar,and generally remain there, feeding upon the interiorof the grain, until they have completed their de-velopment (Barnes & Grove, 1916; Fletcher, 1920;Back, 1922). There are usually four larval instars inRhtzopertha followed by a prepupal stage which doesnot involve moulting. This moults to form the pupa,

136 A. C. CROMBIE

which is followed by the adult. The larval instars arewith practice easily distinguished by eye by meansof the characteristics described by Potter (1935), butwere checked by measuring the head capsule. Underour experimental conditions the average duration ofthe instars was as follows: egg, 6 days; first larvalinstar, 6 days; second, 6 days; third, 5 days; fourth,4 days; prepupa, 1 day; pupa, 7 days; total, 35 days.In Sitotroga there are also usually four larval instars,which are with practice readily distinguishable byeye. As before, they were when necessary checkedby measuring the head capsule. The average durationof the instars under our experimental conditions wasas follows: egg, 3 days; first larval instar, 6 days;second, 6 days; third, 5 days; fourth, 7 days; pupa,5 days; total, 32 days (Crombie, 1943). The periodfrom hatching to emergence was thus 29 days foreach species. This period was sometimes abnormallyextended (vide infra). In both species the pupa isfound in the cavity made by the feeding larva.Escape from the grain by the adults of Sitotroga isaccomplished by the act of emerging from the pupa.The adults of Rhizopertha usually remain inside thegrains for a few days while their cuticle hardensbefore eating their way out. The larvae of Rhizo-pertha can enter intact wheat grains only in the firstinstar. This can be accomplished by the larvae ofSitotroga in all instars, the resulting mortality de-creasing with age (Back, 1922). The mortality amongfirst instar larvae of both species as a result of suchan attempt may be as high as 50 % .

The results of all statistical tests in this paper aregiven in terms of the value of p, which represents theprobability of obtaining as great a deviation byrandom sampling as that observed in the experiment.A probability (p) of 5 % is usually taken as thearbitrary division between significance and non-significance, so that when f><o-os the differencebetween the values compared may be regarded assignificant (Simpson & Roe, 1939).

IIChoice of food by larvae. Single larvae were con-

fined in cells made by gumming a cover-slip over aglass ring. A series of such cells was stood on aground-glass plate. Each cell contained besides thelarva two objects, each composed of different sub-stances, between which the larva had to make achoice of which it was to enter. After 6 hr. in theincubator the cells were examined and the positionsof the larvae noted. The latter always bored into oneor other of the objects. Twenty Rhizopertha larvaeof each instar were offered the choice of each of thefollowing pairs of substances: a piece of wheat grainand a piece of cork cut to the same shape and size;a false grain made of plaster of Paris mixed withwheat flour and a false grain made of plain plasterof Paris (Crombie, 1941). In both experiments alltwenty larvae of each instar chose the wheat. Thedistinction made between the false grains of plaster

containing wheat flour and those composed of PJalone suggests the existence of a chemorecepBfesense, since the only difference between them isprovided by the food. The first experiment wasrepeated with twenty Sitotroga larvae of each of thefour instars, and in each case the wheat was preferredto the cork twenty out of twenty times. Twenty firstinstar larvae of each species, respectively, were nowgiven the choice of a piece of bean seed and a pieceof cork of the same shape and size, and of a piece ofwheat grain and a piece of bean seed of the sameshape and size. All the Rhizopertha and eighteen ofthe Sitotroga larvae chose the bean to the cork;seventeen of the Rhizopertha and all of the Sitotrogachose the wheat to the bean. The larvae of all instarsof both species can thus recognize objects containingfood, and the first instar larvae of both species areable to distinguish between different kinds of food.

IllAs already stated, the proportion of first instar larvaeof either species which manage to enter intact wheatgrains may be as low as 50 %. With such a highmortality the experiments to be described in thispaper could not have been performed. It wasnecessary therefore to damage the test in somestandardized and repeatable way, so that the larvaecould enter more easily. This was done by gentlytapping the grains until the test was just broken bya crack. The larvae almost invariably entered thegrains in the damaged regions and mortality duringentry was reduced almost to zero. The grains weretreated in this way in all the experiments describedin this paper. This may appear to introduce anunnatural condition, but we are chiefly interested inthe competition inside the wheat grains, and not inmortality which is the result of the accident of theirpossessing a hard test. This study must be regarded,however, as a study in competition for 'crackedgrains', and not for intact wheat.

Avoidance of super- or multiple infestation. 150wheat grains were placed into each of three dishesso that they were one layer deep. Into one of thesedishes (a) were introduced 100 first instar Rhizo-pertha larvae, into a second dish (b) 100 first instarSitotroga larvae, and into the third (c) 50 first instarlarvae of each species. After 24 hr. in the incubatorthe grains were dissected and the positions of thelarvae noted. The results are given in Table 1. Thenumber of grains containing o, 1, 2, 3, etc., larvae,assuming that the latter distribute themselves atrandom among the grains, were calculated from theStoy formula where .v is the number of larvae andN the number of grains (Appendix to Salt, 19326).Each of the observed values (a), (b) and ((c): total)were compared with these calculated values by meansof x1. and none of them was significantly differentfrom the latter (p > 0-05). In (c) the members of eachspecies entered the grains in practically equal num-bers. It may be concluded therefore tht>t the larvae

Intraspecific and interspecific competition in larvae of graminivorous insects 137

*ither species avoid either superinfestation orpie infestation (cf. King, 1918; Simmons &

Ellington, 1925 ; Bodenheimer, 1930). Each species,however, possesses a kind of behaviour by whichovercrowding may be prevented, viz. migration. Thehabit of migration of Sitotroga larvae from the grainsthey were inhabiting has been remarked upon byBack (1922), but no reference has been found to itin Rhizopertha.

Causes of migration of Rhizopertha larvae. Thenon-avoidance of superinfestation by Rhizoperthalarvae was confirmed by the following experiment:each of a number of first instar larvae was given thechoice between a fresh grain and a grain (markedwith a spot of ink) which contained three livinglarvae. The two grains were placed side by side ona ground-glass plate, and covered by one of the cellsdescribed above. The larvae cannot escape from suchcells, a series of which were arranged on each glassplate. After-6 hr. the grains were dissected by slicingwith a sharp scalpel under a binocular microscope,

Table 1. Distribution of larvae among wheat grains

Larvae

Calculated(a) Rhizopertha(6) Sitotroga(c) Rhizopertha

with SitotrogaTotal

No

0

77747i

—

74

grains containingx larvae

1

52596T293261

2

171412

66

12

3

43421

3

and the positions of the larvae noted. In approxi-mately half the cells (23 out of 50) the fresh grainscontained one larva while in the other half (27 outof 50) the infested grains contained four larvae,showing that half the larvae had entered each set ofgrains. There was therefore no avoidance of theinfested grains. The point to be decided now waswhether or not it is the presence of other larvae in thesame grain which causes a larva to migrate. Fifty firstinstar larvae were placed separately under cells withone wheat grain each, and the whole put into theincubator for 6 hr. The cells were then examinedand all the larvae were found to have entered thegrains. A fresh grain (marked) was now introducedinto each cell and the whole returned to the incubatoronce more. Every 2 days until adults emerged thefresh grains were removed, dissected and replacedby other fresh grains. No larvae were found in thefresh grains and eventually forty-two adults emergedfrom the infested grains. Those from which adultshad not emerged were then dissected: dead firstinstar larvae were found in four of them, dead larvaeof later instars in two more, while the other two con-tained dead, somewhat deformed adults. It may beconcluded then that migration occurs only when two

or more larvae enter the same grain. Mortality mayoccur, however, in larvae occupying grains singly.

If larvae migrate thus owing to the influence ofother larvae occupying the same grain, it may beasked by what sense a larva perceives that there isanother larva present. There are three possibilities:chemoreception, 'hearing' or stimulation from me-chanical vibrations (cf. Minruch, 1925) and stimula-tion resulting from actual contact. Now if larvaemigrate after detecting the presence of other larvaeby chemoreception, freshly killed larvae ought per-haps to cause migration in the same way as livinglarvae. A number of freshly killed larvae were intro-duced into a hole made in a wheat grain, a livingfirst instar larva placed beside each such grain, andboth covered with a cell. After 6 hr., during whichthe larvae entered the grains, a fresh grain (marked)was introduced into each cell. After 48 hr. thegrains (50 in all) were dissected and all the larvaewere found still in the grains containing the deadlarvae. Chemoreceptive or other stimuli from thelatter do not therefore cause migration.

The effect of mechanical vibrations was investi-gated as follows. Two grains, each containing threefirst instar larvae, were firmly cemented together.Fifty such pairs were set up. Each of these, togetherwith five fresh grains to each, was then placed on aglass plate and covered with a cell. After 2 days thegrains were dissected and the positions of the larvaenoted. Now as described below, the number oflarvae which migrate from or die in a grain increaseswith the initial number present (Table 2). Me-chanical vibrations are not likely to cause death, sowe may omit the dead larvae for the moment. Ifsuch vibrations were the cause of migration onemight expect a similar increase in migration withan increase in this factor. Now one would expectthat the amount of vibration or noise produced bysix larvae in two grains cemented together, wouldbe greater than that produced by three larvae in asingle grain. However, the average number of larvaemigrating per grain from the cemented grains was0-32, which is not significantly different from thenumber, 0-31 per grain, of larvae migrating fromsingle grains each containing initially three larvae.There was also no significant difference between thetotal number of larvae killed and migrating per grainin the two experiments, this value being 0-4 and 0-38for cemented pairs and single grains respectively(Table 2). Mechanical vibrations do not thereforeseem to be the cause of migration.

The probable importance of actual contacts be-tween larvae as a cause of migration becomes im-mediately apparent. The reaction of one larva to thepresence of another was observed as follows. Whentwo larvae in the first, second, third or fourth instarswere put together into a small hole drilled in a wheatgrain and watched under a binocular microscope,they were often seen to attack each other with theirmandibles, and eventually either one or both left the

i 3 8 A. C. CROMBIE

hole. When a larva entered such a hole it alwayswent to the bottom and turned round so as to faceoutwards. Other larvae trying to enter the hole werefiercely attacked. Sometimes such combats resultedin the body wall of one of the antagonists becomingpunctured and its bleeding to death. In their tunnelsin wheat grains larvae of all instars were alwaysfound curled up with the head facing towards theway they had entered. Furthermore, in all grainsdissected during the experiments to be described,whenever two larvae were found in the same tunnelat least one of them was always dead. It thus seemsprobable that whenever two larvae meet within agrain they will attack each other, with the result thateither or both will migrate or be killed.

period of time. This procedure was followedlarvae in the first, second, third and fourth iThe first instar larvae had recently hatched, theothers having recently moulted. The results are givenin Table 2. With initially only one larva per grain nomigration or death occurred during this 2-day period.During this period of their lives both these effectswere therefore apparently due to the presence ofmore than one larva in each grain, and the increasein migration and death with increased initial num-bers was due to crowding.

Now if larvae do not perceive other larvae (e.g. bychemoreception or 'hearing') before actual contactoccurs, or, more precisely, if they do not react toperception at a distance, if it occurs, by dying or

Table 2. Effect of density on the reduction in numbers of Rhizopertha larvae competingfor the same wheat grain

( 1 )

Initial no. larvaeper grain

(«)

Instar I 12

3458

Instar II 234

Instar III 23

Instar IV 23

( 2 )

JNO.

grains

1 0 01 0 6

4 238131 0

52H13

4 126

2413

(3)

X otal no.larvae

1 0 0

2 1 21 2 6

1 5 2

6580

1 0 4

4252

82784839

(4)

Totalno. larvae

killedand

migrating

0

1716353346

12

918

1828

1 418

(5)Observedaverage

no. larvaelulled andmigratingper grain

(Pi)

00 160 3 80 921'694 6

0 2 3

0 641 3 8

0-441 080 5 81 3 8

(6)Observedaverage% larvae

killed andmigratingper grain

(Pilt)

81 2 7

2 33457-5

i i - 51 6 334 52 2

36

2946

(7)Calculated

average% larvae


(PJi)

8162 4

3256

11 5233452 2

442958

(8)

% larvaei n

column 4which

migrated

898166553 i

83785°7868

8672

The effect of density upon the migration and deathof Rhizopertha larvae. Different numbers of larvawere introduced into wheat grains as follows: therequisite number of larvae was placed with a wheatgrain on a ground-glass plate and covered with a cell.A number of such cells were set up. After 6 hr. inthe incubator the grains were examined. All thelarvae had usually entered them by this time, grainsin which they had not done so being rejected. Fivefresh grains (each marked with a spot of ink) werenow introduced into each cell and the whole returnedto the incubator for 48 hr. At the end of this periodthe grains were all dissected by slicing with a sharpscalpel (the original infested grains being examinedfirst in order to minimize errors due to the extensionof the 48 hr. period by the time taken in dissection)and the positions of the larvae noted. This provideddata about the number of larvae killed in and mi-grating from each original grain during a constant

migrating, then contacts between larvae in the samegrain would occur only as a result of random en-counters. The relationship between the initial num-ber of larvae present and the number killed andmigrating would then depend on the relative proba-bility of their encountering each other when therewere different initial numbers of larvae per grain.The determination of the relative probability de-pends simply on a consideration of the number ofencounters that could take place between any twoindividual larvae in a grain containing t larvae. Thisnumber of encounters is equal to t (t— 1), and if thenumber of larvae killed and migrating per grainis pi, then

p( = ki(i-i), p,/i=k(i-i), (1)

which means that the proportion of larvae killed andmigrating is linearly related to the initial number oflarvae per grain. The value of k may be calculated


value of t from the observed value of pt, andle of k may then be used to calculate the

values of pt at other values of i. The reasonablenessof the above hypothesis will then be measured bythe agreement between the observed and calculatedvalues of pi or ptji. The values of k calculated fromthe data for first, second, third and fourth instarlarvae in Table 2 when i=2 are, respectively,008, 0-115, ° ' 2 2 an<i 0*29. The values of pt/i(expressed as percentages) calculated, with differentvalues of *, from these values of k respectively,appear in column 7 of Table 2, while the ob-served percentages of larvae killed and migratingper grain are shown in column 6. An inspectionof Fig. 1, which sets out the relation between »' andpiji, shows that the agreement of observed andcalculated values is very close for first instar larvae,but less so for the older larvae. The goodness of fitbetween the observed and calculated value of p{

60

eop

50

30

20

10

2 3 4 5 6Initial no. larvae per grain (t)

8

Fig. 1. The relationship between density and the reduc-tion in numbers of Rhtzopertha larvae developing inwheat grains: instars I, II, III and IV alone (Table 2,column 6), and instar I competing with Sitotroga(Table 4, column 7). The lines are drawn through thetheoretical points (Table 2, column 7).

(which are not shown here) was tested by calculatingthe values of x*, and there proved to be no significantdifference between the two sets of values for eitherthe first or the second instar. This test could not beapplied to third and fourth instar larvae. These re-sults show that at any rate with first and secondinstar larvae the hypothesis that death and migrationoccur as a result of random encounters within the grainis a possible one, and bearing in mind the absence ofany other probable explanation of the experimentaldata it is reasonable to adopt it as correct.

Two further points of interest in Table 2 may benoted. First, as larval density increases the propor-tion of larvae killed as a result of encounters in-creases, while the proportion migrating decreases(column 8). Secondly, the number of larvae killedand migrating per grain increases with later instars(column 5). This may be because the larger size of

later instars increases the probability of encounter.Finally, it may be mentioned that the number oflarvae killed and migrating from grains which hadbeen cut either in half or in quarters was significantlygreater than from the whole grains. These observa-tions are of interest merely in that they confirm theforegoing results, so that no details will be given.

The effect of density upon the migration and deathof Sitotroga larvae. The existence of larval fightingin this species was observed nearly 200 years ago byDuhamel & Tillet (1762). These authors remark thatthey often found three or four dead larvae in grainsof which only one living larva had taken possession.On one occasion two larvae were seen apparentlyfighting. After several minutes one was dead andthe other was seen working its way into the grain.They were never able to find more than one livinglarva per grain. Similar observations have beenmade by Flanders (1933). Fighting, sometimes re-sulting in death, has been observed here, also, amonglarvae confined together in a small hole bored in awheat grain. In all the grains dissected during theexperiments described below two living larvae werenever found in the same tunnel, although a livingand a dead larva, or two dead ones, were sometimesfound together. The experiments described abovewith Rhtzopertha were now repeated with thisspecies, but only with first instar larvae. As theresults were similar to those obtained with theformer species, and negative, no details will be given.The larvae did not avoid entering grains alreadycontaining other larvae, and chemoreceptive and'auditory' stimuli were, as before, apparently notcauses of migration. An experiment to determinewhether migration or death occurred in grains con-taining only a single larva was carried out as withRhtzopertha. No migration occurred, and from thefifty grains forty-three adults eventually emerged(cf. Barnes & Grove, 1916; Fletcher, 1920). On thedissection of those from which adults had notemerged four dead first instar larvae, one dead thirdor fourth instar larva, one dead pupa, and one dead,deformed adult were found.

The average reduction in numbers of larvae pergrain by migration or death during a fixed periodof time under different degrees of crowding was nowdetermined as before. The results are shown inTable 3. Seven out of a hundred larvae died, duringthe 2-day experimental period, in the grains con-taining initially only one larva. This means that 7 %of the larvae died, without encounters with otherlarvae, in the first 2 days after hatching. There seemsto be no reason why such mortality should not occurat any density, so that we must subtract from thetotal number of larvae killed and migrating at anydensity (column 4), 7% of the initial number oflarvae introduced into the grains. The remainder isthe number of larvae killed and migrating as a resultof random encounters (column 5). The value of kwas calculated as before from the observed value

140 A. C. CROMBIE

of p{ when t = 2, and in this case k = 0-09. The valuesof pji (expressed as percentages) calculated fromthis value of k at different values of i are given incolumn 8, while the observed percentages of larvaekilled and migrating per grain appear in column 7.

result of random encounters within the grairbefore, the proportion of larvae killed as a resiSBRencounters increases with increasing density, whilethe proportion migrating decreases (column 9). Therate of migration and death of Sitotroga is signifi-

Table 3. Effect of density on the reduction in numbers of first instar Sitotroga larvae competingfor the same wheat grain

(1)

Initial! no.

larvaeper

grain(0

123468

(2)

No.grains

1006940201619

(3)

Total no.larvae

1001381208096

152

(4)

Observedtotal no.

larvaekilled andmigrating

722302554

116

(5)

Total no.larvae

lulled andmigrating

after chanceencounters

012-421619-44i-3

105-7

(6)

Observedaverage no.killed andmigratingper grain


(Pi)

0 1 8054097258554

(7)Observedaverage

% larvaekilled andmigratingper grain


(/>«/«")

91824-343693

(8)Calculated

average% larvae

killed andmigratingper gram

afterchance

encounters

(.Pth)

918274563

(9)

% larvaein

column 5which

migrated

8170574219

3 70

I50

S 40

so

20

10

> AloneOWith RhUopenha

0 I 2 3 4 5 6' 7 8 9Initial no. larvae per grain

Fig. 2. The relationship between density and the reduc-tion in numbers after chance encounters of first instarSitotroga larvae developing in wheat grains alone (Table 3,column 7) and in competition with Rkizopertha (Table 4,column 7). The line is drawn through the theoreticalpoints (Table 3, column 8).

An inspection of Fig. 2 shows, as before, the closeagreement between observed and calculated per-centages. The goodness of fit between observed andcalculated values of p{ (not shown here) was testedby calculating the value of x*> and there proved tobe no significant difference between them. It istherefore reasonable to believe that the eliminationof Sitotroga first instar larvae also takes place as a

cantly higher than that of Rhizopertha (x* corre-sponds to p<o 01).

The effect of competition between Rhizopertha andSitotroga larvae upon the migration and death of bothspecies. The above experiment was now repeatedexcept that first instar larvae of both Rhizoperthaand Sitotroga were introduced into the grains to-gether in equal numbers as shown in Table 4. Thevalues for Sitotroga are consistently higher thanthose for Rhizopertha, although the difference be-tween them is not statistically significant. Theexpected values of pji were calculated from thevalues of k obtained from Tables 2 and 3, respec-tively, and are shown in column 8 of Table 4. Aninspection of Figs. 1 and 2 shows that, as before,there is a close agreement between observed andcalculated percentages. The calculation of x* showedalso that there is no significant difference betweenobserved and calculated values of pi for eitherspecies (not shown here). The values of pji are (forboth species) closely similar whether they are com-peting intraspecifically or interspecifically, whichshows that the degree to which the larvae of eitherspecies are affected depends upon that species ratherthan upon the competing species. The Rhizoperthaappear to be slightly the more successful in compe-tition. It should be mentioned that the larvae ofeither species will attack those of the other. In allthe grains dissected in the experiments describedbelow, whenever two larvae were found in the sametunnel at least one of them was always dead.

Observations on migration and death throughoutdevelopment. By the usual method four freshlyhatched Rhizopertha larvae were introduced into


a number of wheat grains. Five fresh grainsd) were then placed in each of the cells con-

taining one infested grain, and the whole returnedto the incubator. The fresh grains were dissected,being replaced by an equal number of new freshgrains, once every 4 days, until adults had ceased

migration took place in the first and second instars,and migration had practically ceased by the 20thday. By this time most of the larvae would beentering the prepupal stage. Of the original fourlarvae per grain the average number which survivedto become adults was only 1-14 or 28-5%, 26-1%

Table 4. Effect of density on the reduction in numbers of larvae of Rhizopertha and Sitotrogacompeting for the same wheat grains

(l)

Initial no.ist inatar

larvaeper grain

(0

Rhiz. 1Sit. 1Total 2Rhiz. 2Sit. 2Total 4Rhiz. 3Sit. 3Total 6Rhiz. 4Sit. 4Total 8

( 2 )

No.grains

——72

——

48——33

——

10

(3)

Total no.larvae

7272

144

9696

192

6969

138

4040

80

(4)

Observedtotal no.

larvaekilled andmigrating

612

18

1931

50

233356

2328

Si

(5)

Total no.larvae

killed andmigrating


67

13

192 4 3

43'3

2327-5S°-S

2325-24 8 2

(6)Observedaverage

no. larvaekilled andmigratingper grain


(Pt)

0 0 80 1 0

0 1 8

0 3 90-510 9 0

i - o1-2

2-2

2 32-524-82

(7)

Observedaverage% larvae



(P.H)

8-39 7—

1 9 825'4—

33'44O'O

—

57-56 3 0

—

(8)Calculated

average% larvae


afterchance

encounters(Pifi)

89

—

2427—

4045—

5663—

(9)

% larvaein

column 5which

migrated

8386

85

797878 s48555i

394442

Table 5. Observations on the migration of Rhizopertha from wheat grains throughout development(initially four 1st instar larvae per grain; 43 grains)

Ins tar

IIIIIIIV

Total

No. larvae migrating after x days

4

21

21

8

44

8

12

31

4

16

214

7

2 0

12

3

Total no. adults emergingTotal no. larvae migrating + adults emergingTotal no. larvae introduced into the 43 grains.'. no. immature insects killed

32

11

2

Total

2611

35

45 = 2 6 i %

49 = 28-5%94

17278 = 45-4%

Averageper grain

o-60 260 07O-I2

1 05

1142 194-0I -8I

emerging from the infested grains. The number oflarvae migrating to the fresh grains over the wholedevelopmental period was thus observed, instarsbeing checked by measuring the width of the headcapsule. Most of the adults had emerged by the40th day but a few emerged later. The infestedgrains were all dissected on the last day of observa-tion. The results are shown in Table 5. Most of the

of the larvae migrating and 45-4% being killed ordying. Some of the latter, in various instars in-cluding one dead, deformed adult, were found ondissecting the infested grains at the end of theexperiment.

This experiment was now repeated with Sitotroga.The results (Table 6) here are similar to those justdescribed. Instars were checked as before by mea-

142 A. C. CROMBIE

suring the width of the head capsule. Most of theadults had emerged by day 36, but some appearedafter longer periods. Migration took place chieflyin the first and second instars and ceased by day 20.The larvae of this species would be in the fourthinstar and perhaps beginning to pupate by this time.Of the original four larvae per grain an averagenumber of 1-27 per grain or 31-6% survived toemerge as adults, 22-5 % of the larvae migrating

average number which survived to become adwas 1-23 per grain or 30-8%. Of these sunf48-7% were Rhizopertha and 51-3% Sitotroga. Theproportion of the original larvae migrating was25-8% of which 45% were Rhizopertha and 55%Sitotroga; while the proportion dying or being killedwas 43-3 % of the original number, of which 54%were Rhizopertha and 46% Sitotroga. As beforesome of the latter, in various instars including one

Table 6. Observations on the migration of Sitotroga from wheat grains throughout development(initially four 1st instar larvae per grain; 30 grains)

Instar

IIIIIIIV

Total

No. larvae migrating after x days

4

14

14

8

1

1

1 2

1

3

4

16

1

32

6

2 0

1

1

2

Total no. adults emergingTotal no. larvae migrating + adults emergingTotal no. larvae introduced into the 30 grains

. no. immature insects killed

Total

i7631

27 = 2 2 5 %

38 = 31-6%65

120

55=45-9%

Averageper grain

0-57O-2

o-iO 03

0 9

1-272-1740183

Table 7. Observations on the migration from wheat grains of Rhizopertha and Sitotroga (in competition)throughout development (30 grains)

Instar

IIIIIIIV

Total

No. larvae migrating after * days

4

R.

7

7

S.

8

8

8

R

1

1

2

5.

21

3

1 2

R.

1

1

S.

3

3

16

R.

1

2

3

S.

1

1

2 0

R.

1

1

S

11

2

% migratingTotal no. adults emerging% adults emergingTotal no. larvae migrating + adults emergingTotal no. larvae introduced into the 30 grains.'. no. immature insects killed% immature insects killed

Total

R.

92

3

>4

2 3 318

3°326028467

S.

1 1

41

1

17

284!931636602440

Total

2064i

3 i

2S 83730868

1205243'3

Average per grain

R.

030070 1

0-47

06

1 07

2

o-93

5.

o-37013003003

0 56

0-63

ra2o-8

Total

067O-20130 03

I 03

1-23

22741 73

and 45-9 % dying or being killed. Some of the latter,in various instars including one dead deformed adultand one dead pupa, were found on dissecting theinfested grains at the end of the experiment.

A third experiment was now carried out whichwas a repetition of the preceding two except that ofthe initial four freshly hatched larvae per grain twowere Rhizopertha and two Sitotroga. The results aregiven in Table 7. As before migration occurredchiefly in the first and second instars and had ceasedby day 20 Of the original four larvae per grain the

dead pupa of each species, were found on dissectingthe infested grains at the end of the experiment.

What has been observed here is what would happenin a culture in which a large number of wheat grainswere present, such as in experimental populationsgrowing in wheat media (unpublished) or in storesof wheat in which these insects are pests. Theexperiments have provided data about the amountof reduction in numbers for which each of the twomethods, migration and death, is responsible, theinstars during which migration occurs, the number


*rvae capable of successfully completing theiropment in one grain, and the outcome of com-

petition between Rhizopertha and Sitotroga.

IVThe effect of larval density upon the survival of

Rhizopertha and Sitotroga. Increasing numbers offreshly hatched larvae were introduced into differentwheat grains and each wheat grain placed in a smallvial. The whole was then placed in the incubator.

separately. In both cases when there was initiallyone larva per grain only 80 % of the larvae, or of8per grain, survived to become adults. Now in theexperiments carried out to investigate whether mi-gration would occur with initially only one larva pergrain (vide supra), with both species four of thecorpses found in the grains from which no adultsemerged were first instar larvae. If these had sur-vived 46 out of 50 or 92 % of the Rhizopertha larvae,and 47 out of 50 or 94% of the Sitotroga larvae

Table 8. The effect of density upon the survival of the larvae of Rhizopertha andSitotroga developing in wheat grains

( 1 )

Initial no. larvaeper grain

(')

Instar I 12

3468

1 0

142 0

Total (2-20)

Instar II 12

34

Total (2-4)

Instar I 12

3468

1 0

142 0

Total (2-20)

Instar II 1

( 2 )

No. grains

1787274754 04040

3°36

407

1 0 0402918

87

1 3 0806663494440

3°3°

402

1 0 0

(3)

Totalno. larvae

(4)

Totalno. adultsemerging

(5)Average

no. adultsemergingper grain

(a) Rhizopertha alone1781442 2 230O2 4 03204OO42O72O

2766

IOO808772

239

1438392874941SO3240

474

94433926

108

(6) Sitotroga alone

1 3 01 6 01982 5 2

2943524 0 04 2 06 0 0

2676

I O O

1 0 4

8867696049522932

446

96

o-81-151-241 1 61 221-03

1-25I 07I I I

1-16

0'941 0 8i-34i-441-24

o-8i " i

I'OI

I I

1-22

I-II

I 30 97I 07I - I

0 9 6

(6)

% larvaedevelopinginto adults

( i - P ) x 100

8057-74i 52920-512 812 57 65-6

17-2

94544536

135

80523427'52 0 5

1 4136 95-6

16

96

(7).% survivors

withdevelopmentretarded byover 21 days

1 52 41 0 8i - i 54 14 90

3' i2-5

i-5

2'12O

O

3 852 8

3 91 4 87-5

n - 615 0

1 6 321 0

0

6 2 5

1 2 1

3-i

There were no fresh grains, each of the vials con-taining one infested grain only. The vials wereexamined at regular intervals and the numbers anddates of emergence of adults recorded. The adultswere sexed and weighed. We thus have records ofthe total number of adults emerging, the numberemerging per grain, the developmental period, sexratio and weight. In Table 8 are given the resultsof experiments on the effect of larval density uponthe survival of (a) Rhizopertha and (6) Sitotroga

1KB 20, 2

would have become adults. When the larvae wereintroduced into the grains in the second instar, withinitially one larva per grain 94 % of the Rhizoperthaand 96 % of the Sitotroga larvae survived to becomeadults (Table 8). Therefore there is a certain mor-tality in the first instar which is independent oflarval encounters.

Now let i be the initial number of first instarlarvae per grain, P be the proportion dying, and a bethe average number of insects surviving per grain

A. C. CROMBIE

when i > i. Then if the larvae are eliminated atrandom (i.e. the probability of survival of any insectis inversely proportional to the initial number pre-sent) but a remains constant, the proportion oflarvae surviving will be

When the value of a has been determined the ex-pected values of (1—P) for any value of * above 1may be calculated. The agreement of expected andobserved values of (1— P) will depend on the con-stancy of a. The value of a for Rhizopertha was 1 • 16and for Sitotroga 1 • 1. It is obvious from column 5of Table 8 that for both species there is no regularchange in a with density (cf. Jacobi, 1939). The ob-served values of 100 (1 — P) shown in column 6 were,however, compared with the expected values foreach species (not shown) by means of x*» and thereproved to be no significant difference between themfor either species. There was also no significant

60

C

I 8 405 x

I 20Ik00 2 4 6 8 10 12 14 16 18 20

Initial no. 1st instar larvae per grain (1)Fig. 3. The relationship between density and the sur-vival of first instar larvae of Rktzopertha and Sitotrogadeveloping in wheat grains alone (Table 8, column 6)and in competition with each other (Table 9, column 7).The line is drawn through the theoretical values of 100(1 — P) when a= i-z (equation (2)).

difference between the observed values of (1 — P) forthe two species, so that the two observed values of amay be considered as equal. The observed relation-ship between i and 100 (1 — P) for each species isshown in Fig. 3. The value of (1 — P) approachesbut never reaches zero.

The majority of the adults of both species emergedwithin a few days of each other. In Rhizoperthaover 95 % emerged between 28 and 38 days afterhatching, the mean developmental period of theinsects emerging within these limits being 32 days.Over 86 % of the Sitotroga emerged between 26 and36 days after hatching, the mean developmentalperiod of the insects emerging between these limitsbeing 29 days. The term 'developmental period'refers here to the time elapsing between the hatchingof the egg and the emergence of the adult from thepupa. This can be accurately determined for Sito-troga since in the act of emerging from the pupa the

adult also escapes from the wheat grain in whicpupa is lying. But owing to the habit ofinside the grain for a few days after emergence fromthe pupa while the adult cuticle hardens, the precisemoment of emergence from the pupa cannot beaccurately determined for Rhizopertha which havepupated inside wheat grains. What is observed is themoment when the adult eats its way out of the grain,and consequently the values for developmentalperiod are too high. From comparison with dataobtained with insects developing in flour, wherethe precise moment of emergence from the pupacan be observed, the period between emergence andescaping from the grains is approximately 3 days,which makes the actual mean developmental periodof Rhizopertha approximately 29 days, or equal tothat of Sitotroga. The insects not emerging withinthe limits of the majority of each species as dennedabove, had widely divergent developmental periods,always longer than those of the majority. The longestdevelopmental period observed for Rhizopertha was90 days, and for Sitotroga 120 days. As the develop-mental periods of these minorities were widelyscattered, they were disregarded in calculating themean developmental periods at each density. Thelatter did not vary significantly at any density fromthe mean values given. Now in column 7 b aregiven the percentages of the surviving Sitotrogawhose developmental periods were more than 21 dayslonger than the average. That is, these insectsemerged after the 50th day from hatching. Thecorrelation coefficient between the proportion ofinsects with such retarded development (column 7 b)and the average number of adults emerging per grain(column 5 b) was calculated, and proved to be signi-ficantly different from zero at the 1 % level, andpositive. The two sets of variables are thereforehighly positively correlated. On the other hand,with Rhizopertha the number of survivors emergingmore than 21 days after the average (i.e. after day 53)was never greater than two at any density, and nocalculations can be based on such small numbers.The reason for choosing the period of 21 days willappear later. With Sttotroga, in the majority of thegrains from which more than one adult emerged,one of the insects had an average rate of develop-ment while the rate of development of the other wasretarded. The result was that the two insects whichdeveloped successfully had widely different rates ofdevelopment. In this experiment there were actually56 grains from which more than one Sitotroga adultemerged. The dates of emergence of the adultswhich had developed in one grain were separated bymore than 21 days in 27 = 48-2% of these grains, by14-21 days in 16 = 28-6%, and by less than 14 daysin 13 = 23-2%. In the grains in which only oneinsect completed its development the adults emergedwithin 14 days of each other in 390 = 90% out of434 grains. In Sitotroga, therefore, if two larvaewhich have entered the same grain have widely


»ent developmental periods it will be moreible that both of them will successfully com-

plete their development, than it will be if their de-velopmental periods are nearly equal. This is possiblybecause larvae of approximately the same instar areless tolerant of each other than are those of widelydifferent instars (cf. Flanders, 1933). One larva witha retarded rate of development in a grain containingseveral normal larvae would therefore have a greaterchance of survival than any of the latter. In column7 ft of Table 8 there is a general tendency for theproportion of retarded larvae among the survivorsto increase with increasing density, at least up toten larvae per grain. When there was initially onlyone larva per grain 3-9% of the survivors had re-tarded rates of development, which shows that thephenomenon is at least to this extent independentof crowding. Now if it is assumed that a fixedpercentage (e.g. 39%) of the larvae hatching haveconstitutionally retarded rates of development then,as the initial number of larvae per grain increases,the absolute number of retarded larvae will increasealso. If the survival rate of the latter is greater thanthat of normal larvae then, since the number oflarvae which survive per grain is limited, the pro-portion of retarded larvae among the survivors willincrease. It is also possible, of course, that crowdingis a cause of the retarded rate of development (al-though, as already stated, not always) and that thisis the explanation of the results in column 7 b.Above ten larvae per grain perhaps the probabilityof elimination early in development is so great thatthe possession of an atypical rate of development isno longer an advantage.

There is no similar phenomenon in Rhizopertha.The dates of emergence of the adults emerging fromone grain were separated by less than 7 days in48 = 78% of the 62 grains from which more thanone adult emerged, by 7-14 days in 9=14%, by14-21 days in 2 = 3-2%, and by over 21 days in3=4-8%. In the grains in which only one insectcompleted its development the adults emerged within14 days of each other in 470 = 96-7% out of 486grains. Furthermore, the proportion of retardedlarvae was always very small.

Overcrowding in the larval instars seems to havehad no other effect upon these insects than to in-crease mortality and to affect the numbers of larvaewith retarded development in the way describedabove. The average developmental period of themajority of the larvae remained unaffected. The sexratio of the emerging adults did not differ signi-ficantly from unity at any density for either species.The average weights of these adults at all densities•did not differ significantly from the following values:Rhizopertha males, 1-17 mg.; females, 1-3 mg.; Sito-troga males, i-8 mg.; females, 3-7 mg. The rates ofoviposition in wheat of ten of the females which haddeveloped at each density were measured at a densityof two grains per female (Crombie, 1942, Tables 1

and n ) . The average rates of oviposition for allthese females were 9-8 eggs per female per day forRhizopertha and 112 eggs per female per 5 days forSitotroga. The rates of oviposition of each speciesat any density did not differ significantly from thesevalues, respectively. The 143 grains from which oneRhizopertha emerged weighed 7-15 g. (50 mg. pergrain) before and 5-50 g. after the insects had de-veloped in them. The faeces and frass were removedfrom the still uneaten part of the grain before thesecond weighing. The loss in weight incurredthrough the development of one Rhizopertha pergrain was therefore 1-65 or 23% of the originalweight. The average amount of food eaten per larvaduring development (i.e. the average loss per grain)was 11-5 mg. This figure will be too high since somefeeding would have been done by the adults beforethey ate their way out of the grains after emergence.The loss in weight of the 104 grains (5-2 g.) fromwhich one Sitotroga emerged was 1-92 g. or 37% ofthe original weight. The average amount of foodeaten per larva during development was therefore18-5 mg. The average numbers of insects survivingper grain when t> 1 (1-16 for Rhizopertha and I-Ifor Sitotroga) are therefore less than the food presentin a grain could support if unhindered by otherforms of competition between the larvae (vide infra).

The effect of competition {with initially equal num-bers of larvae) upon the survival of Rhizopertha andSitotroga. The first experiment was a repetition ofthose just described except that larvae of the twospecies were present at the same time in equalnumbers, i.e. an equal number of each species wascompeting for the same grain in each vial. Theresults are shown in Table 9 and Fig. 3.

The total number of insects emerging per grain(column 5, instar I) and the percentage of thetotal number of larvae surviving to become adults(column 7, instar I), respectively, do not differ signi-ficantly from the values of the same variables wheneither species was developing alone in the grains(Table 8). This means that the average value of a(equation (2)) here is really equal to those found inTable 8. There is no regular change in the value of awith larval density. The observed values of 100(1 — P)shown in column 7, total, were compared with thevalues calculated with a=i-zi (not shown), and x1

revealed no significant difference between them.When each species is considered separately theaverage values of a are 0-69 for Rhizopertha and0-52 for Sitotroga. The expected values of 100 (1 — P)were calculated with these values of a, respectively,and proved to be not significantly different from theobserved values for each species, respectively, whentested by means of x3- This means that each speciesreduces the probability of survival of the other indirect proportion to its own numbers, and that thereis at all densities a constant relationship between thenumbers of the two species which survive. Rhizo-pertha was always more successful in competition

146 A. C. CROMBIE

than Sitotroga, the average ratio between them beingRhizopertha to- Sitotroga as 57: 43 =1-3:1 . (Thevalue of x1 calculated in order to compare the num-bers of adults of the two species in column 4 (instar I)corresponded to p < o-oi, showing that the differencebetween the two sets of values is significant.) Atlarval densities above ten per grain this relationshipmay seem to change to the disadvantage of Sitotroga,but the agreement of observed and expected values of100 (1 — P) argues against this, and furthermore withsuch small numbers the ratio between the survivorsof the two species is very sensitive to random fluctua-tions. Approximately the same relationship is foundwhen the larvae of both species were introduced inthe second instar. The weight and sex ratio of theadults were, as before, unaffected by density, andwere not significantly different from the values

(x* corresponds to p< o-oi). Thus a large ^of the Sitotroga which survive after competitio^Borthe same grain with Rhizopertha had retarded de-velopmental periods. There were fifty-two grains inall from which individuals of both species emerged.The dates of emergence of the individuals of thetwo species developing in the same grain were sepa-rated by over 21 days in 15 = 29% of these grains,by 14-21 days in 26 = 50 % , and by less than 14 daysin 11 = 2 1 % . The developmental periods of all theadults were within 10 days of the same length in 300(= 90 %) of the 333 grains from which one individualof either species emerged. There were 41 (79%)grains in which the developmental periods of thetwo species differed by more than 14 days. In 39(i.e. 75 % of the total 52 grains) of these the rate ofdevelopment of the Sitotroga was retarded while that

Table 9. The effect of density upon the survival of Rhizopertha and Sitotroga larvae competing forwheat. Each species initially present in equal numbers

( 1 )

Initialno. larvae

pergrain

»

R. 5. Tot.

Instar I1

2

3457

1 0

1

2

34S7

1 0

2

468

1 0

142 0

Total

Instar II2 2 4

( 2 )

No.grains

75707060

5°3030

385

25

(3)

Total no.larvae

R.

751 4 02 1 0

2402502 1 0

300

H 2 5

5°

5.

751 4 02 1 0

2402502 1 0

300

1425

50

Tot.

150280420480500420600

2850

1 0 0

(4)

Total no.adults

emerging

R.

42465441382 2

24

267

23

s.

3241403535

98

2 0 0

' 4

Tot.

74879476733i32

467

37

(5)

Averageno. adultsemergingper grain

R.

0 5 60 6 60 7 70 6 80 7 6o-730 8 0

0 69

0 92

5. J Tot.

o-430 59O-570 5 80 7 00 3 00 2 7

0 5 2

0 56

0 99253426

[•46[•03[•07[ 21

1 4 8

(6)

%oftotal

adultscom-

prised0 1

R.

57535854527 i7557

62

S.

434742464829

25

43

38

(7)


( i - P ) x 100

R.

563327171 5 61 1

8-31 9 2

46

S

43291 7 6I4'61 3 83 72 3

1 3 6

28

Tot.

493i22-515-81 4 67 45 3

1 6 4

37

(8)%of

survivorswith de-

velopmentretardedby over21 days

R

2 32 2

3 52 42 60

0

2 O

O

s.

6 2 527-03 2 428-54 4 0

0

1 4 3

2 9 5

43

already given with each species developing in thegrains separately.

The effect of overcrowding upon the develop-mental period of the survivors was similar to thatalready described for each species separately. Thepercentages of the surviving insects with retardeddevelopment are shown in column 8. As beforethe number of adults emerging per grain (column 5,total) and the proportion of retarded Sitotroga(column 8) are highly positively correlated, thecorrelation coefficient being significantly differentfrom zero at the 1 % level, and positive. The pro-portion of retarded Rhizopertha was always very low,comprising only 2 % of the total survivors, whilethat of retarded Sitotroga comprised 29-5 % of them.The latter proportion is significantly higher than theproportion of retarded insects (3-9 %) when therewas initially only one Sitotroga per grain (Table 8

of the Rhizopertha was of the average value (28-38days), and in only two ( = 3'8s%) did the oppositerelationship hold. This suggests that the competitionof Rhizopertha with Sitotroga larvae whose rates ofdevelopment differ widely from their own is lesssevere than with those Sitotroga whose rates of de-velopment approximate their own. The proportionof retarded insects among the survivors of Sitotrogaincreases with increasing density up to ten larvaeper grain, when the value of 44% is reached. Asbefore there is a fall in this value with initially morethan ten larvae per grain. It is clear that the pos-session of a proportion of larvae with atypical ratesof development is of considerable survival value forSitotroga, enabling more individuals of this speciesto survive in the limited space of one wheat grain,both when in competition with Rhizopertha andwhen only Sitotroga larvae are competing for the

Intraspeafic and interspecific competition in larvae of graminivorous insects 147

The cause of retarded development in Sito-larvae is unknown.

The effect of non-contemporaneous entry into grainsupon the competition of Rhizopertha and Sitotroga.The initial larval density was kept constant at fourfirst instar larvae per grain (two of each species), but inone series of vials Rhizopertha was introduced 7, 14,21 and 28 days before Sitotroga (Table 10 a), whilein another series Sitotroga was introduced 7, 14, 21and 28 days before Rhizopertha (Table 106). Theresults are shown in Table 10 and Fig. 4. Statisti-

cessful than the Sitotroga. For instance, while thesurvival of Rhizopertha was unaffected by Sitotrogawhich follow it by 14 or more days (column 5, andTable 8 a), Sitotroga always had a lower rate ofsurvival when Rhizopertha followed it at any intervalof time than when it was alone (column 5 b, andTable 8 6). (d) As the time interval between theintroduction of the two species became longer (above7 days), the number of adults emerging per grainof the first species eventually reached a fairly con-stant value, while that of the second species gradually

Table 10 Effect of introducing one species at different periods of time before the other upon the competitionof Rhizopertha and Sitotroga developing in wheat grains. Each species initially present in equal numbers

(1)1st instarlarvae of

one speciesintroduced

x days beforethose of

the other

( 2 )

No.grains

(3)

Total no. larvae(2R. + 2S.per grain)

(4)

Total no.adults

emerging

(5)

Averageno. adultsemergingper grain

(6)No.

grainsfromwhichboth

speciesemerged

(7)%oftotal

adultscom-prised

of

(8)


Control:both intro-duced atsame time(Table 9)

R.

7 0 1 4 0

S.

1 4 0

Tot. R.

2 8 0 46

S. Tot R.

87 0 6 6

S.

o-59

Tot. R

1-25 53

5 . R.

47 33

S.

29

Tot.

7 daysH .,21 ,,28 „

7 days14 >,21 ,,28 „

RIIIIIII

3030

3030

60606060

60606060

120

I2O

120

I2O

a) RJnzoperthc

24343i36

7132421

3i475557

before Sitotroga

081 131-031-2

023

O-430807

I 03

156I-8319

0

12

1721

78835663

22

174437

40

56-751-760

n-721-740

35

(b) Sitotroga before Rhizopertha30

303030

60606060

60606060

120

120

I2O

12O

IO

162325

25252626

35414951

0 33O-530-77

083

083083087087

1161-361-64i 7

511

1616

29374749

7i63535i

16726738-341-7

41-741-7

43 3433

(c) The larvae of both species were introduced at the same time in the instars showns.IIIIIII

20

20

20

20

4040

4040

4040

4040

8080

8080

20

20

914

12

18

1515

3238

2429

I-O

I 0

0450-7

o-609

075075

i-6

1 9

1-2

1 45

916

2

12

63533848

37376252

50

50

22-5

35

30

4537-537 5

25-739345747'5

29333-340T425

4047-53036-3

cally significant differences were discovered by calcu-lating the values of x* or t, and on the basis of thesethe following statements are made (in each casep<o-o$). (a) The relationship between the twospecies as observed by their survival (columns5 and 7), relative to their relationship when bothspecies were introduced at the same time (control),remained always to a greater or lesser extent infavour of the species introduced first (cf. Smith,1912; Picard, 1922). (b) For the second species themost unfavourable interval at which to follow thefirst was 7-14 days, (c) As in previous experiments,however, the Rhizopertha were relatively more suc-

increased (column 5). The total number of insectsemerging per grain thus increased also. This wasachieved by an increase in the number of wheatgrains from which more than one adult emerged(column 6), which suggests that the mortality whichoccurred in the immature stages of the competinginsects was not primarily due to lack of food. Addi-tional evidence in support of this conclusion comesfrom two sources. The first is the observation thatup to three and four normal-sized adults of eitheror both species have been known to emerge from asingle grain. Secondly, as already mentioned, theaverage loss in weight of the wheat grains as a result

148 A. C. CROMBIE

of the successful development of one insect per grainis only 23 % of the original weight of the grain forRMzopertha and 37 % for Sitotroga. Now the speciesintroduced first eats food, uses up oxygen and condi-tions the inside of the grain with its faeces and otherwaste products. However, if these were the principalcauses of the results one would expect that the longerthe interval between the introduction of the twospecies the greater would be the depression of thespecies introduced second. This was not so, so thatthe principal cause of the results must have been theactive competition of the larvae inside the grains forspace (cf. Pemberton & Willard, 1918; Lloyd, 1940).

7 14 21Interval between two species in days

28

Fig 4. The effect of introducing one species into thegrains at different periods of time before the other uponthe survival and competition of Rfnzopertfia and Sitotrogalarvae (Table 10, column 5). (a) Rhizopertha introducedfirst, (6) Sitotroga introduced first.

On the other hand, 28 days after introduction thefirst species would usually be in the pupal instar.It is therefore not likely to have engaged in activecompetition. Now the food provided by one grainis as already mentioned adequate for the completedevelopment of at least two larvae. The reason whythe second species survived better in an otherwiseuninhabited grain (Table 8) than in a grain in whichthe first species was preceding it by 21 days or more,is therefore likely to be that the first species condi-tioned the inside of the grain into which the secondspecies enters. A considerable amount of faeces andfrass was always made by the larvae (vide infra),(e) Sex ratio and weight of neither species wasaffected by competition.

Now in all except two of the 98 grains (column 6)

from which both species emerged the interval o^between the emergence of the individuals ofspecies was greater than 21 days. As already men-tioned, as the interval of time between their intro-duction becomes longer (above 7 days), there is astatistically significant increase in the number ofgrains from which both species emerged (column 6).Since the rates of development of the majority ofindividuals of both species would be approximatelyequal, the members of each species would be in morewidely different instars as the interval of time sepa-rating their introduction increases. The resultstherefore agree with those obtained in previous ex-periments (Tables 8, 9), in which it was found thatthe mortality of Sitotroga and its competitors whichwere developing in the same wheat grain was greateramong insects with approximately contemporaneousdevelopment than in those whose developmentalperiods differed by a great enough margin (21 days).

Why should the species introduced first have arelative advantage over the other species, and whyshould it be more unfavourable for the second speciesto be introduced 7—14 days after the first than eitherat the same moment or after longer intervals of time?The following experiment was performed in orderto answer these questions. Two larvae of each specieswere introduced at the same time into each grain,but in each experiment the larvae of one specieswere in a more advanced instar than those of theother, as shown-in Table 10 c. The numbers involvedare rather small, but the results are remarkablysimilar to those shown in Table 10 a, b, the initiallymore mature larvae here bearing the same relation-ship to their competitors as those introduced first inthe previous experiment bear to theirs. The mor-tality of the insects introduced in the first instar wasin both species greater when the other competingspecies was introduced in the second instar thanwhen the latter was introduced in the first or thirdinstar. Now in both species the second instar occu-pies the period from about 6—12 days after hatching,so that in the previous experiments the larvae intro-duced first by 7-14 days would be in the second orthird instar when the second species was introduced.Those introduced first by 21 days would be in thethird or fourth instar by the time the second specieswas introduced. Furthermore, both the total numberof insects emerging per grain, and the number ofgrains from which both species emerged, weregreater when one species was introduced in the firstinstar and the other the third instar, than when onewas introduced in the first instar and the other inthe second instar, or both in the first instar. It seemsreasonable to suggest that these differences in theseverity of competition between larvae of differentages may be the explanation of the greater survival oflarvae whose development was not contemporaneouswith that of their competitors than of those in whichit was (Tables 8, 9). As the larvae enter the grainsafter increasingly different periods (Table 10) the


luring which they are in competition of courseaes shorter, but this does not explain why one

species should suffer more when introduced daysafter, than at the same time as, the other. If thelatter result is correct, it must be concluded thenthat the toleration of the larvae for each other at anymoment varies with their relative ages. There is,however, another possible explanation. As men-tioned above, Rkizopertha adults remain inside thegrain for a few days after emergence before eatingtheir way out. It is possible that these adults maydamage other insects in the same grain.. If the latterwere following them by only a short interval theywould be more likely to be in the pupal stage when

duced into another series of vials (Table 11 b). Theresults are shown in Table 11 and Fig. 5. The totalnumber of adults of both species emerging per grainand the percentage of the total larvae developingwith adults are approximately equal to the values ofthese variables at comparable densities in Tables 8and 9. More adults of the initially more numerousspecies always emerged per grain than of the initiallyless numerous species but, as in Table 9, Rkizo-pertha was more successful in competition thanSitotroga. The number of Rhizopertha which sur-vived when it was in a minority is significantlygreater than the number of Sitotroga which survivedwhen it was in a minority (x1 corresponded to

Table 11. Competition between Rhizopertha and Sitotroga developing in wheat grains withdifferent initial numerical relationships

( 1 )

Initial no.ist instar

larvaeper

grain

R. S. Tot.

Control(Table 9)1 1 2

(a) R.>S.2

48

16

1

1

1

1

359

17

(6) R.<S.1

1

1

1

2

48

16

359

17

Total(*) + (*)

( 2 )

No.

75

80808080

80808080

7 i 5

(3)

Total no.

R.

75

160320640

1280

80808080

1360

S.

75

80808080

1 6 03 2 0

6401280

1360

Tot.

1 5 0

2 4 04 0 07 2 0

1360

240400720

1360

2720

(4)

Total noadults

emerging

R.

4 2

58757069

3618192 0

407

S.

32

24171 1

9

48666765

339

Tot

74

82928178

84«48685

746

(5)Average

no. adults

per grain

R.

0 5 6

O730 9 40 8 7086

O'450 2 30 2 40-25

—

5 .

°-43

o-30 2 1

0 1 4O'I2

0 60-820 840 8 1

—

Tot

0-99

1 03i - i 51 01

0 9 8

1 051-051 081 06

—

(6)% sur-vivors

(observed)comprised

of

R

57

7i8286-4894

432 1

2 2

24

54-9

S.

43

2918

n-61 0 6

57797875

45-i

(7)


R.

56

362 3 51 1

5'4

45252425

—

S.

43

302 1

i V 811 2

3 02 0 51 0 4

5 1

—

Tot.

49

342311-2

5'7

352 1

1 2

6 3

—

(8)Expected

% sur-vivors

comprisedof

R.

57

738491-595'5

4025148

—

S.

43

27168-S4'5

60758692

—

these adults emerged than in the larval stage, andperhaps pupae are more susceptible to such damagethan larvae. On several occasions dead and partlyeaten pupae were found in grains from which otherinsects had previously emerged as adults, but thedata were too meagre to be able to draw any definiteconclusions. This explanation would not of courseapply to the results obtained when Sitotroga wasthe first species introduced into the grains.

The effect of inequality of initial numbers upon thecompetition of Rhizopertha and Sitotroga. One Sito-troga first instar larva with increasing numbers ofRhizopertha first instar larvae were introduced intoeach grain in one series of vials (Table 11 a), whileone Rhizopertha first instar larva with increasingnumbers of Sitotroga first instar larvae were mtro-

p<o-oi). The percentages in column 8 were calcu-lated on the assumptions that the insects will beeliminated at random (i.e. that the probability ofany insect surviving is inversely proportional to theinitial number present) but that the probability ofsurvival for Rhizopertha compares with the proba-bility of survival for Sitotroga as 57 :43 = i'3 : 1.The latter is the average ratio of the survivors be-longing to each species when there were initiallyequal numbers of larva of each species in each grain(Table 9). Let the initial number of Rkizoperthaequal r, and of Sitotroga equal s. The probabilitythat one Rhizopertha will survive is then equal to

r + 1-3/ i"3

A. C. CROMBIE

Similarly, the probability that one Sitotroga will

survive is . The relative probability of sur-i-3r + s

vival is therefore RMzopertha to Sitotroga as i-^r : s.The agreement between expected and observedvalues was tested by calculating the value of Xs,and there proved to be no significant differencebetween them.

There were only 18 grains from which bothspecies emerged in this experiment; in n of thesethe Sitotroga emerged more than 21 days after theRMzopertha. As before the weight and sex ratio ofthe insects remained unaffected by crowding.

rated environment migration may lead tobut when the environment is limited inpoint will be reached when migration from one grainto another merely leads to death in another place.Then the only effect of overcrowding will be toincrease mortality. Migration from the grains tendsto decrease with later instars. Because of competi-tion for space the number of larvae of the same agewhich survive in one grain is less than that whichthe food present in the latter could support. Thisnumber (approx. 1-2 per grain) does not vary withdensity, so that the relationship between survivaland larval density is given by equation (2).

0 2 4 6 8Initial no. RMzopertha

16 0 2 4 6 8 16Initial no. Sitotroga

Fig. 5. The effect of the initial numerical relationship between them upon the survival and competition ofRhtzopertha and Sitotroga larvae developing in wheat (Table n , column 6). (a) Riiizopertha in a majority,(6) Sitotroga in a majority. The full line is drawn through the theoretical points (column 8).

V. SUMMARYThe females of both RMzopertha and Sitotroga ovi-posit in environments containing places suitable forlarval development, but the larvae themselves,usually during a period of migration in the firstins tar, choose the actual developmental site. Therate of ovipositdon of neither species of female bearsmuch relation to the amount of food present for thelarvae, and the latter do not refrain from multipleor superinfestation of wheat grains. The competitionwhich ensues is apparently wholly a struggle forspace, the limitation of food or oxygen, and the' conditioning' of the medium, being unimportant.Larvae (of any instar) will attack each other directlyafter encounters at random within wheat grains, andthe supernumerary individuals are either killed orforced to migrate. The probability that any par-ticular larva will survive is thus inversely propor-tional to the initial number present (equation (1)).Except that it tended to favour the survival of Sito-troga with atypical rates of development, over-crowding had no other effect upon the larvae ofeither of these species. In an unlimited or unsatu-

When the two species are competing the averageratio of the survivors is RMzopertha to Sitotroga asi-3r : s, where r and s, respectively, are the initialnumbers of larvae of each species. These representthe proportions of the 1 -2 larvae surviving per grainwhich belong to each species. This ratio remainsconstant at all densities when the larvae enter thegrains at the same time in the same instar. Eachspecies thus decreases the probability of survival ofthe other in direct proportion to its own numbers.

When first instar larvae of the two species enterthe grains at different times the above relationshipchanges in favour of the first comer. The most un-favourable time for the second species to enter thegrain is apparently when the first is in the secondor early third instar. With greater differences be-tween times of entry (i.e. of age) the severity ofcompetition for space decreases, so that more larvaeare able to survive and take advantage of the foodreserves of the grain. (The survival of the larvae ofthe second species is apparently then reduced tosome extent by the accumulation of the excretoryproducts of the first.) Sitotroga, but not RMzo-pertha, was able to take advantage of this decreased


ipetition because of the occurrence of larvae withpical rates of development in this species. The

latter were able to survive the competition of normallarvae of either species where other normal larvaewould have succumbed. Crowding to a certaindegree tended to increase the proportion of atypicallarvae among the survivors of this species.

Overcrowding in the immature stages had no effectupon the average developmental period of the larvae,

or upon the sex-ratio, weight or fecundity of theadults or either species.

I should like to thank Dr A. D. Imms, F.R.S.,and Dr W. H. Thorpe for their interest in theseexperiments and Dr A. R. Miller for his mathe-matical advice. This work was done while holdinga research scholarship from the University ofMelbourne.

REFERENCESALI.KK, W. C. (1931). Animal Aggregations. Chicago

Univ. Press.Arr.KR, W. C. (1934). Biol. Rev.9, 1.BACK, E. A. (1922). Fmrs' Bull. U.S. Dep. Agnc.

no. 1156.BARNES, J. H. & GROVE, A. J. (1916). Mem. Dep. Agric.

India, Chem. 4, 166.BODKNHEIMER, F. S. (1930). Monogr. angew. Ent. 10.BODENHEIMER, F. S. (1938). Problems of Animal Ecology.

Oxford.BRANDT, H. (1937). Z. angew. Ent. 34, 87.CLEMENTS, F. E. & SHELFORD, V. E. (1939). Bio-

Ecology. New York: Wiley.CROMBIE, A. C. (1941). J. exp. Biol. 18, 62.CROMBIE, A. C. (1943)- J- exp- Biol. 19, 311.CROMBIE, A. C. (1943). Nature, Lond., 15a, 246.DUHAMEL DU MONCEAU, H. L. & TlLLET, M. (1762).

Histoire d'un iniecte qui devore Us grains de I'Angoumoisavec Us moyens que Von peut employer pour U ditrutre.Pans.

DuNSTAN, G. G. (1935). Canad. Ent. 67, 89.FLANDERS, S. E. (1933). Ann. Ent. Soc. Amer. 36, 529.FLANDERS, S. E. (1939). Ann. Ent. Soc. Amer. 33, 11.FLETCHER, T. B. (1920). Mem. Dep. Agric. India, Ent.

6,69.GRAHAM, S. A. (1939). Ecol. Monogr. 9, 301.HAMMOND, E. C. (1938). Quart. Rev. Biol. 13, 421.HAMMOND, E. C. (1939). Quart. Rev. Biol. 14, 35.L'HERITIER, P. (1937). Arch. Zool. exp. gin. 78, 25s.HERMS, W. B. (1928). J. Econ. Ent. 31, 720.HOFMANN, C. (1933). Z. angew. Ent. 30, 51.HOLDAWAY, F. G. (1930). Nature, Lond., 136, 648.JACOBI, E. F. (1939). Arch, nterl. Zool. 3, 197.KING, J. L. (1918). J. Econ. Ent. n , 87.LANDOWSKI, J. (1938). Biol. Zbl. 58, 512LLOYD, D. C. (1939). PhtUis. Trans. B, 339, 275.

LLOYD, D. C. (1940). Proc. Roy. Soc. B, 128, 451.MACLAGAN, D. S. (1932). Proc. Roy. Soc. B, i n , 437.MACLAGAN, D. S (1941). Proc. Umv. Durham Phil.

Soc. 10, 310.MACLAGAN, D. S. & DUNN, E. (1935). Proc. Roy. Soc.

Edinb. 55, 126.MINNICH, D. E. (1925). J. Exp. Zool. 43, 443.MICHAL, K. (1931). Biol. generalis, 7, 631.PARK, T. (1939). Amer. Midi. Nat. 31, 235.PARK, T. (1941)- Quart. Rev. Biol. 16, 274, 440.PEMBERTON, C. E. & WILLARD, H. F. (1918). J. Agric.

Res. 13, 285.PICARD, F. (1922). Bull. biol. 56, 54.POTTER, C. (1935). Trans. R. Ent. Soc. Lond. 83, 449.SALT, G. (1932a). Bull. Ent. Res. 33, 211.SALT, G. (19326). Bull. Ent. Res. 33, 235.SALT, G. (1936). J. Exp. Biol. 13, 363.SALT, G. (1937). Proc. Roy. Soc. B, 123, 57.SIMMONS, P. & ELLINGTON, G. W. (1924). J. Econ. Ent.

17. 4i-SIMMONS, P. & ELLINGTON, G. W. (1925). J. Econ. Ent.

18, 309.SIMMONS, P. & ELLINGTON, G. W. (1927). J. Agric. Res.

34. 459-SIMMONS, P. & ELLINGTON, G. W. (1933)- Tech. Bull.

U.S. Dep. Agric. no. 351.SIMPSON, G. G. & ROE, A. (1939). Quantitative Zoology.

New York: McGraw-Hill.SMITH, H. E. (19:2). Tech. Ser. U.S. Bur. Ent. 19, 4, 33.SMITH, H. E. (1929). Bull. Ent. Res. 20, 141.TIMOFEEFF-RESSOVSKY, N. W. (1933)- Arch. Naturgesch.

N.F. 3, 285.TIMOFEEFF-RESSOVSKY, N. W. (1935). Arch. Naturgesch.

N.F. 4, 245-ULLYBTT, G. C. (1936). Proc. Roy. Soc. B, 130, 253.WEIDLING, K. (1928). Z. angew. Ent. 14, 68.

Documents

[ 135 ] ON INTRASPECIFIC AND INTERSPECIFIC COMPETITION IN ... · BY A. C. CROMBIE From the Zoological Laboratory, Cambridge (Received 20 October 1943) (With Five Text-figures) The