THE BIOLOGY OF CONOMELUS ANCEPS (GERMAR)
(HOMOPIERA : DELPHACIDAE).
by
G.H.L.Rothschild, B.Sc.
A Thesis submitted in part fulfilment of
the requirements for the Degree of Doctor
of Philosophy in the University of London.
May, 1962. Imperial College Field Station,
Silwood Park,
Sunninghill,
Ascot, Berkshire. ABSTRACT.
This thesis is divided into three sections. The first deals
with the general biology of Conomelus, including the life history and
descriptions of the immature stages. Sexual maturation of the adults and
oviposition in relation to host plant characters, have been studied in
detail. The internal and external effects of pipunculid and Strepsipteran
parasites on adult hosts are described.
The second section deals with the biology of the predators and
parasites associated with Conomelus. Various species have been studied in
detail, including a nymphal pipunculid parasite, two predatory mirid species,
and a fungal egg parasite, the life histories of which were previously un— known. The seasonal occurrence of 91 species of predator in the rush plot
is outlined, with reference to their association with Conomelus.
The third section deals with studies on a Conomelus population between autumn 1959 and autumn 1961. Particular reference is made to the mortality factors affecting the egg, nymphal and adult stages, including parasitism and predation. The precipitin reaction has been used to evaluate predation in the field. Dispersal of nymphs, and brachypterous and macropterous adults, within and from the rush habitat, has been studied. TABLE OF CONTENTS.
Page
INTRODUCTION 1
SECTION 1. BIOLOGY OF CONOMELUS ANCEPS.
Description of the stages 3
Habitat and host plants 5
Methods of rearing 9
Life history.
(i) Egg stage 10
(ii) Nymphal stages 17
(iii)Adult stage 19
Sexual maturation 19
Effects of parasitism on the reproductive organs 30
SECTION 2, BIOLOGY OF THE PREDATORS AND PARASITES OF C.ANCEPS.
Methods 43
Parasites of the egg stage 44
Parasites of the nymphal and adult stages 52
Predators of the egg stage 61
Predators of the nymphal and adult stages 65
SECTION 3. POPULATION STUDIES.
Methods 76
Population estimates
(i) Numbers of eggs 86
(ii) Numbers of nymphs 87
(iii) Numbers of adults.., 90 SECTION 3 CONT. Page,
Estimates of mortality.
(i) Egg mortality 92
Discussion 101
(ii) Nymphal and adult mortality 106
Discussion 133
Dispersal
Methods 141
Nymphal movements 143
Adult movements 143
SUMMARY 154
PLATES 158
ACKNOWLEDGMENTS 177
REFERENCES 178
APPENDIX 184 INTRODUCTION
The object of this work has been to study the biology of Conomelus
anceps (Germar) and its predators and parasites. C.anceps feeds only on
Juncus spp., and populations within a patch of Juncus are, therefore, self-
contained and suitable for quantitative studies. Later discussion in
Section 3 will, however, indicate the difficulties of sampling the rush
habitat, and how these difficulties affect the estimation of delphacid and
predator numbers. The only published account giving both biological and
quantitative data for a delphacid species in the field is that of Kanervo
(1957). This work on Delphacodes pellucida (Fab.) does not include any
data on egg and nymphal populations or any estimations of predation. None
of the other work is quantitative but deals almost entirely with the biology
of the delphacids and their parasites, in particular the pipunculids and
Strepsiptera. This literature includes the studies of Hassan (1939),
Lindberg (1939), Morcos (1953) and Williams (1957). There are also numerous references to the sugar-cane delphacid Perkinsiella saccharicida Kirk. and its associated natural enemies - Perkins (1905 a, b, c, d, 1906 a, b, 1907).
In 1961 the present author came across the unpublished work of Whalley (1958) on the homoptera associated with Juncus in North Wales, including various leafhopper species, C.anceps, Delphacodes fairmairei (Perris) and a jassid
Tettigella viridis (L.). Whalley gives estimates of nymphal and adult populations and shows the effect of parasitism on the numbers of delphacids in the field. This author also deals with the biology of C.anceps, and in particular the egg stage. 2.
Most other references to leafhopper populations deal with jassids
of economic importance; these include the 'cork of Carter (1930), Hanna
(1950), Mulla (1957), and Joyce (1961). In Finland KontKannen (1950)
sampled leafhopper populations over a wide area to obtain data on the
habitat preferences of over 20 species of delphacids and jassids. The
present work on C.anceps is divided into three sections; the first deals
with the biology of the delphacid; the second section deals with the biology
of the parasites and predators; the third section includes the population
studies and an account of the importance of the natural enemies; this is
followed by a general discussion.
This work was conducted at Imperial College Field Station, Silwood
Park, near Ascot, from September 1959 until May 1962. 3.
SECTION I.
The biology of Conomelus anceps (Germar).
Introduction.
China (1950) has dealt with the nomenclature of this species
listing the single British member of the genus as:
Conomelus Fieber 1866
anceps (Germar 1821)
limbatus (Fabricius 1794 nec Oliver 1791).
Edwards (1896) describes the delphacid as Liburnia limbata (Fab.). Le
Quesne (1960) states that the species is widespread and common on Juncus spp.
in the British Isles, the remainder of Europe and North Africa. The only
previous work on the biology of C.anceps is, firstly, that of Hassan (1939)
who describes the egg and fourth inster nymph, and outlines the life cycle,
and, secondly, Whalley (1958) who deals particularly with the egg stage. •
Description of the stages.
The Egg: Figured by Hassan (1939). Mean length 1.09 mm. (10 eggs),
greatest width 0.22 mm.; elongate, curved, tapering towards the operculum.
Chorion smooth, translucent. Yellow yolk body present at either pole depending upon the age of the egg.
Nymphal stages: The present author has described the nymphal stages using measurements of head—width and femur—length to separate the instars. The characters cif the hind tibiae and tarsi used by Williams (1957) to separate nymphs of Perkinsiella saccharicida were also present in C.anceps. After 4. completion of these nymphal descriptions the author came across the unpublish- ed work of Whalley (1958) who also described the nymphs, using tibial lengths to separate the stages. The data in Table 1 show that measurements of head-width and femur-length are suitable for separation of the instars, no overlapping occurring at any stage; body-length, however, is rather variable.
First instar (Plate 1, Fig.l): length 1.05 mm.; grey, inter- segmental areas, and foveae of he.ad and thorax, pale; median facial 'keels darker grey; legs pale grey with terminal part of femur, base of tibia and tarsal points dark; abdomen may be white or yellow ventrally; eyes reddish; hind tibiae and tarsus (Plate 2, Fig.l), tibia with four spines distally; tarsus is two-segmentedf with the first segment bearing four spines distally.
Second instar: length 1.30 mm.; as first instar, but borders of thoracic and abdominal sclerites darker grey; hind tibia and tarsus (Plate
2, Fig.2), tibia bears two small spines, one basally and another in the distal one-third, largest tibial spur now articulated.
Third instar (Plate 1, Fig.2): length 1.53 mm.; greyish-brown, with pale areas on abdominal tergites; eyes reddish-brown; hind tibia and tarsus (Plate 2, Fig.3), articulated tibial spine itself bears two to three small spurs, five non-articulated tibial spines distally; first tarsal joint also with five distal spines.
Fourth instar: length 2.05 mm.; colouration as third instar, hind tibia and tarsus (Plate 2, Fig.4), articulated tibial spine bears four 5.
to five spurs; second tarsal segment now has two to three median spurs.
Fifth instar (Plate 1, Fig.3, 4): length 2.62 mm.; pale greyish,.
brown, pale markings on most tergite ventral side of abdomen of female
nymph white, and male nymphs yellow; nymphs can also be sexed by the
terminal abdominal segments (Plate 2, Fig.6, 7); wing pads of macropterous
nymphs (Plate 1, Fig.3) longer than brachypterous forms (Plate 1, Fig.4);
hind tibia and tarsus (Plate 2, Fig.5), articulated tibial spine serrated
for most of length; three tcrsal segments, the first bears five to six
distal spines, and the second segment three to four spines.
Adult: is described by Edwards (1896) and Le Quesne p960).
Table 1. Nymphal measurements.
Body-length Head-width Femur-length
Instar Mean .Min. Max. Mean Min. Max. Mean Min. Max.
First 1.05 0.77 1.20 0.31 0.27 0.35 0.15 0.15 0.17
Second 1.30 1.17 1.47 0.42 0.40 0.45 0.22 0.22 0.25
Third 1.53 1.27 1.70 0.52 0.50 0.55 0.31 0.30 0.32
Fourth 2,05 1.70 2.25 0.66 0.62 0.70 0.42 0.40 0.45
Fifth 2.62 2.45 2.90 0.85 0.80 0,90 0.60 0.60 0.65
Measurements in mm. (10 individuals of each instar).
Habitat and host plants.
The plOt studied in the present work is situated in an area of
Silwood Park known as Pond Field. The centre of the field is 6.
poorly— drained and supports a dense stand of Juncus, forming an area of
1267 square yards. Unlike the open tussock—growth of Juncus in better—
drained pasture, the Pond field plot was covered by a uniform cover of rushes.
The edges of the rush area merge with well—drained grassland and here the
open Juncus tussocks are found. In the wettest parts of the plot the
commonest plants are:
Juncus effusus L.
J.articulatus L.
Stellaria alsine Grimm.
The following species are found in other parts of the plot:
Holcus lanatus L.
Poa trivialis L.
Deschampsia caespitosa (L.)
Cirsium arvense (L.)
C.palustre (L.)
Lotus uliginosus SchKuhr.
Sonchus asper (L.)
Chengpodium album L.
Epilobium sp.
Juncus effusus is the dominant rush species with J.articulatus more frequent
in completely ititer-logged soil. During the spring and summer, from May
to September, little surface water is visible, although single heavy rain—
storms lead to vaterlogging. Throughout the winter months much of the plot
is under several inches of water. 7.
Host plants.
Eggs were only found in stems of the food plant, J.effusus. No
oviposition sites were discovered in J.articulatus or any of the other
plants listed on page 6. The term oviposition site is used in this work
to describe the actual opening made by the ovipositor of the female in the
rush stem, while eggs are being laid. Whalley (1958) has found Conomelus
eggs in Juncus inflexus, J.maritima, and J.conglomeratus. Oviposition is
thus restricted to Juncus species. Although Conomelus nymphs and adults
feed on J.articulatus in the laboratory, no eggs are laid in this rush
species. J.articulatus differs from J.effusus in that the stems of the
former are hollow and do not contain any pith. A few tests were carried
out to determine the influence of the pith on oviposition.
Table 2. Relationship between presence of pith and oviposition.
J.articulatus J.effusus.
pith (from pith absent pith present pith removed J.effusus) inserted.
No. of replicates 5 5 5 5
No. of replicates 3 0 5 0 plus eggs
The results in Table 2 indicate that the pretence of pith, even in
J.articulatus, may initiate egg-laying after the preliminary probe with the ovipositor has been made. Probe marks were found on all four groups of stems in the experiment. In one case the pith was removed from half of a S.
stem of J.effusus, and it was observed that eggs were only laid in the
portion containing pith. In the field, however, it appears that Conomelus
females do not attempt to oviposit into J.articulatus stems as no probe
marks were.ever found. Some other factors must, therefore, be involved in
deciding the host plant preference.
Food plants.
In the present study Conomelus was observed only to feed on
J.effusus in the field, although J.articulatus was accepted in the laboratory.
Hassan (1939) reared Conomelus nymph through to the adult stage on grass
stems (species not stated), but found that adults died after a few days
when restricted to this food. The actual process of feeding is similar
in both nymphs and adults, although adults and later instars generally feed on all parts of the Juncus stems, while the early instars are mainly found in the basal 6 inches. Initially the delphacid explores the surface of
the stem with the tip of the rostrum; when a suitable point is reached the rostrum is gradually inserted over a period of 60 - 180 seconds until the anteclypeus rests on the surface of the foodplant. Feeding delphacids always face up the stem gripping the surface with their tarsal claws.
Nymphs and adults feed at the same point for 38 - 180 minutes unless disturb- ed, when they rapidly run downwards, or move out of sight to the opposite side of the stem. Feeding delphacids are not easily dislodged from their foodplants as they must first remove their stylets; this may account for their capture by less-active predators.
Other delphacids occurring in the habitat: Several other species were occasionally taken in the area, i.e. Megamelodes venosus (Germar), 9.
Deiphacodes pellucida (Fab.), D.fairmairei (Perris) and Dicranotropis hamata (Bohemen).
Life history of Conomelus anceps.
Methods of rearing.
C.anceps was reared in large numbers from rush tussocks, containing eggs, kept in glass. tanks, measuring 14 x 12 x 13 inches. By keeping these mass cultures at a temperature of 25 — 30°C. delphacids were readily obtained throughout the winter months. In experiments requiring the rearing of individual delphacids, a number of other methods were used. Hoppers were kept in 3 x 1 inch glass tubes containing moist sand and several pieces of rush; the tubes were covered with muslin, and the food changed every two to three days. Although Hassan (1939) used this method successfully high mortalities occurred in the present study because of excessive condensation in the tubes. Plaster of Paris was, therefore, used instead of sand, considerably reducing the mortality. In other rearing experiments whole
Juncus plants were groin in 3 — 6 inch flower pots; the cages over these pots were either glass bell—jars or cellulose acetate cylinders, fitted with muslin covers. Attempts to rear Conomelus in cellulose acetate or muslin sleeves around rush stems in the laboratory, and the field, were unsuccessfuL
Nymphs were also reared from eggs kept on moist filter paper in petri-dishes.
Although moulds often developed on the cultures, these were readily suppress— ed by applying 1% nipogin solution to the filter paper.
Life history.
The life history may be outlined as follows: the eggs are laid in
Juncus stems in late summer, and overwinter in this stage. The egg—laying 10.
curve is shown in Plate 7. Nymphs emerge in the following May, and after
passing through five instars the first adults appear in late June. Eggs
are then laid again and the adults may survive until mid—November. There is
one generation per annum.
The egg stage.
The oviposition sites of Conomelus have been described by
Bakkendorf (1934), Hassan (1939) and Whalley (1958). Wagner (1913) incorr—
ectly figures the egg site of a Delphacodes sp. as that of C.anceps. The egg sites of Conomelus are visible as small circular or oval openings in the epidermis of the rush, measuring from 0.12 to 0.30 mm. in diameter; the holes produced merely by probing do not measure more than 0.05 mm. in diameter. The plant tissues around these openings die and darken, making the sites fairly conspicuous.
Type of stem chosen for oviposition: The diameters of 90 rush stems, containing Conomelus eggs were measured, and the distribution of the sites in relation to stem size is given in Table 3.
Table 3. Relationship between stem diameter and number of egg sites.
Stem diameter of total stems Mean No. of % of total sites. in mm. sites/stem
0.84 — 1.05 12.36 108 15.32
1.05 — 1.26 31.46 125 17.73
1.26 — 1.47 29.31 132 18.72
1.47 — 1.68 21.47 214 30.35
1.68 — 1.89 4.49 127 18.01 11.
If the eggs are laid at random, the two columns of percentages should be
comparable. However, the results show that more eggs are laid in the stems
having a diameter of 1.47 to 1.89 mm. than would be expected from random
oviposition. This may simply be due to physical reasons and not actual
selection by the delphacid, as there is less stem area available for ovi-
position at the base of a small stem than a large one (eggs are only laid
in the basal 6 inches of the rush as described below). This explanation is
partly borne out by the fact that In the early stages of the oviposition
period egg-laying appears to be at random; the two columns of percentages
in Table 4 are more comparable than those in Table 3, in spite of an
unexpectedly high number of eggs in the 0.84 - 1.05 mm. stems. Table 3
was compiled after the completion of oviposition in the field (20/9/61),
while the data in Table 4 was obtained earlier in the season (2/8/61)5.
Table 4. Numbers of egg sites in relation to stem diameter.
Stem diameter % of total stems. Mean no. of % of total stem. in mm. sites/stem.
0.84 - 1.05 12.36 4.5 25.00
1.05 - 1.26 31.46 3.6 20.00
1.26 - 1.47 29.21 4.7 26.11
1.47 - 1.68 21.47 4.2 23.33
1.68 - 1.89 4.49 1.0 5.55
Part of stem chosen for oviposition: Most eggs are laid within 6 inches of
the tip of the leaf sheath as shown in Plate 3. Occasionally eggs are laid under the leaf sheath, vbile at the other extreme egg sites have been found 12,
near the stem tip, up to 24 inches above the leaf sheath. Reasons for the
basal position of the oviposition sites are not altogether clear. The
females may prefer to lay their eggs within the clump to avoid full daylight
conditions, or because of the higher humidity in this region. Conditions
are certainly most favourable for the eggs in the basal 6 inches, as the upper parts, of the stem tend to dry out or rot, and break—up by the follow— ing spring; eggs collected between 14 and 16 inches above the leaf sheath had invariably collapsed because o desiccation. Whalley (1958) has recorded a similar pattern of egg distribution in Conomelus.
Numbers of sites per stem: The mean number of sites per stem varied in the three seasons, particularly between 1959 and 1960, as shown in Table 5.
Table 5. Mean number of oviposition sites per stem.
Season Mean no, of sites (plus eggs) Max.no. of sites (plus eggs) per stem. per stem.
1959 35.38 280
1960 93.71 393
1961 123.33 316
The figures for the total number of sites per stem are rather higher as a certain percentage of the sites do not contain eggs; these are not included in the estimates in Table 5. Whalley (1958) has also recorded unoccupied sites, and considers that these are caused by mutual disturbance among the ovipositing females. He has obtained data to confirm this by confining varying numbers of gravid females to single rush stems, and recording the number of unoccupied sites produced. His results are given in Table 6. 13.
Table 6. Relationship between number of females per stem and number of
oviposition sites produced (Whalley's data).
Replicates No, of females/ Total sites Total unoccupied rfA occupied stem produced in sites 3 days
1 5 27 3 90.00
1 10 59 7 89.39
1 20 37 15 71.15
The present author also carried out some tests of this type but did not confirm Whalley's results, as shown in Table 7.
Table 7. Relationship between number of females per stem and number of
oviposition sites produced.
Replicates No. of females/ Total sites Total unoccupied % occupied stem produced in sites 5 days
3 3 26 2 92.3
3 4 39 1 97.4
3 5 34 3 91.2
3 10 63 6 90.5
3 12 69 5 92.7
The egg-laying data from the field seems to show a slight trend for the number of unoccupied sites to increase throughout the oviposition period, as shown in Table 8. 14.
Table 8. Relationship between number of occupied and unoccupied sites.
Date No. of stems Total sites/stem No. of occupied % occupied examined sites/stem sites
21/7/61 50 0.2 0.2 100.0
29/7/61 50 2.5 2.5 100.0
2/8/61 22 9.4 7.5 79,8
9/8/61 15 11.0 9.0 81.8
16/8/61 15 28.1 25.9 92.0
23/8/61 15 44.5 39.6 97.7
1/9/61 15 86.2 71.9 83.0
7/9/61 15 140.3 120.3 87.7
13/9/61 15 196.7 141.7 72.0
20/9/61 15 169.6 123.3 72.7
There was, however, no correlation between the total number of sites per stem and the percentage of unoccupied sites (r = +0.0050). In stems with a high egg density, over 700 per stem, the number of unoccupied sites was high, i.e. up to 50% of the total. This could not be due to increasing numbers of females, as suggested by Whalley, as the delphacid population is rapidly dwindling at this time. A possible explanation is that with the high density of oviposition sites in the basal parts of these particular stems, female delollacids find it increasingly difficult to insert their ovipositors without encountering eggs already present within the stem.
This would lead to withdrawal of the ovipositor, and would be followed by repeated attempts at oviposition thereby leaving unoccupied sites. This 15.
explanation would also account for the data in Table 7 in that although the
number of females per stem may be high, the actual density of sites per stem
after 5 days is still very low and would not impede oViposition.
Numbers of eggs per stem: The eggs are laid in batches of 1 to 7 with a
mean of 2.86 eggs per site. In contrast to other deiphacid species no
protective substances are secreted over the opening of the egg site, The
relative numbers of different batches in the field are given in Table 9,
based on 350 oviposition sites.
Table 9. Numbers of eggs per oviposition site.
Numbers of eggs per site % of total sites
3 27.0 2 25.0 4 20.0 1 16.9 5 6.4 6 3.2 7 0.9
The number of eggs laid per stem are given in Table 10.
Table 10. Numbers of eggs per stem.
Season Mean number of Max.number of eggs No. of stems examined eggs_pei stem per stem
1959 101.2 632 59 1960 266.8 1124 108 1961 352.6 904 40 16.
Between 1959 and 1960 the number of eggs laid per stem has more than doubled,
while the increase for 1960 to 1961 is considerably less. The relationship
between the number of eggs, the numbers of female delphacids ultimately
produced from these eggs, and the number of eggs laid by the latter is
discussed in Section 3.
Overwintering of the egg: Willer (1951) has studied the development of the
Conomelus egg, and Whalley (1958) has made a study of the egg diapause and
the factorsnecossary for breaking the condition.
Whalley's findings may be summarised as follows:
The newly—laid egg has an opaque yolk body at the posterior pole; after several days this body moves to the opercular end of the egg and becomes yellow in colour. The egg remains in this diapause condition until the
following spring when the yolk body returns to its original position. Red eye spots then appear anteriorly, and full differentiation of the embryo occurs. A 100% hatch was recorded from eggs that were kept at temperatures of 3°C. to 10°C. for several days and were then restored to 25°C. Of several hundred eggs kept at a constant temperature of 25°C., without a cool spell, only 2 hatched. Embryonic development took place in eggs kept at
33°C., but no hatching occurred; at 37°C. and 45°C. the eggs failed to develop. Whalley's work therefore, shows that most eggs enter a complete diapause in the field, but that the condition can be broken without freezing temperatures. As a small percentage of the eggs hatch in the laboratory without entering diapause, it is probable that some nymphs may appear in the field before winter months; this may account for the few third instar 17.
Conomelus nymphs that have been found at the end of September in Pond field.
(ii) Nymphal stages.
Hatching: Nymphs begin to emerge in early May and continue to do so for
at least 4 to 5 weeks in the field. Hatching.was observed on many occasions;
the egg becomes elongated anteriorly and the operculum is gradually forced
off by the expansion of the nymph. Pumping movements are visible in the
pharyngeal region which may be due to the nymph swallowing the amniotic
fluids — Wigglesworth (1953). Vaves of muscular contraction pass along the
nymph which then emerges from the chorion still 'enclosed in the embryonic
membranes. After a. short resting period the nymph appears to take in air,
visible as a continuous stream of bubbles, resulting in distension, partic—
ularly of the head, and the final rupturing of the embryonic membranes..
At this stage pulsatile organs are visible in the tibiae. Tibial pulsatile
organs are mainly recorded from the Hemiptera, including aphids — Richardson
01918) and Notonectidae — Crozier and Stier (1927). The newly—emerged nymph is almost translucent, becoming white and finally grey after a minimum
of 30 minutes. In the laboratory a 5% hatching failure was recorded. In most of these cases the nymphs had emerged from the chorion, but were unable
to break through the embryonic membranes. Examination of oviposition sites in the field showed that almost all nymphs emerged successfully.
Most first instar nymphs appear to aggregate on the basal 6 inches of the stems, but it is not clear whether this is due to actual gregarious— ness or merely due to the high density of egg sites in this part of the stem.
Nymphs emerging from egg sites situated high up on the rush stems always 18. moved down into the basal regions immediately after attaining their final colouration.
Moulting: Nymphs nearing a moult are easily recognisable in that the thoracic tergitos are widely separated along the mid—line. In the laboratory, nymphs moulted on the rush stems or on the muslin covers of the cages; in the field exuviae were always found on the stems. Nymphs face upwards, when moulting, and grip the stem only with the tarsal claws, although Hassan (1939) records that the rostrum, also, is often inserted into the stem.
Length of instars: Hassan (1939) has noted the duration of the various stages of Conomelus, and his data is given, in brackets in Table 11. Due to the considerable mortalities involved in rearing individual hoppers the data in Table 11 is based, on the pooled results of all breeding work.
Table 11. Length of instar in days at 16 4- 2°C.
Instar Minimum Mean Maximum
First 5 (8) 10.1 (11) 19 (14)
Second 4 (6) 9.0 (9) 15 (11)
Third 6 (4) 8.0 (7) 12 (10)
Fourth 6 (2) 10.9 (3) 15 (4)
Fifth 10 (4) 10.8 (7) 16 (9)
Total 31 (24) 48.8 (37) 77 (48)
Hassan's (1939) data in brackets (no rearing temperatures given)
The duration of the individual instars in the field could not be estimated as the interval between taking successive samples was too long, i.e. 7 days. 19.
However, as the estimated duration of the total life cycle in the field
(1961 data) approximates to the insectary figures, it is probable that the
instar lengths are also similar. In 1961 Conomelus nymphs were first seen
in the field on the 1st May and the first adults on 22nd June, an interval
of 53 days, in 1960 the dates were the 9th May and the 7th of July respect—
ively, an interval of 59 days. The difference between the insectary and
the field figures is probably due to the lower field temperatures, partic—
ularly earlier in the season, during the first and second instars. The
differences in the length of the life cycles in 1960 and 1961 cannot be
related to temperature. The mean monthly temperatures for May and June in
1960 were 55°F. and 60°F. respectively; in 1961 the corresponding temper—
atures were 52°F. and 58°F. Whalley (1958) in North Wales records an
interval of 50 days between the appearance of the first nymph, and the first
adult.
(iii) The Adult.
The newly—moulted adult is white, becoming grey after about 30 minutes,
and attaining the full adult colouration after 60 to 120 minutes.
Sexual maturation: The reproductive organs of delphacids have only been
fully described by Lindberg (1939) in a Chloriona sp., and by Strubing
(1956), who describes the female organs of about a dozen species including
C.anceps. Hassan's (1939) work includes a generalised figure of the
delphacid reproductive system.
Female reproductive organs (Plate 8, Fig.6): consist of a pair of ovaries, each with an average of 11 ovarioles (minimum 10 and maximum 14, based on 20.
72 females). In a few instances ovariole asymmetry was recorded, with 11
ovarioles in one ovary and 12 in the other (also 10 + 11 ovarioles). Each
ovariole may contain up to 4 follicles. From the ovaries the lateral
oviducts run into the short vagina; also opening into the vagina are two
other ducts, one leading into the spherical accessory gland, and the other
to the sperm receptaculum, which itself terminates in the spermathecal
gland. The mature ovaries fill most of the body cavity and the ovariole filaments pass into the anterior region of the thorax.
In the view of the present author, Lindberg (1939) and Hassan
(1939) are incorrect in their interpretation of the accessory gland and spermatheca. Both authorities describe the accessory gland (present author) as the spermatheca, and the spermatheca (present author) as the accessory gland. This is obviously incorrect as the 'accessory gland' of
Lindberg and Hassan contains masses of spermatozoa after fertilisation, while their 'spermatheca' is generally only filled with a viscous fluid, which the present author regards as secretory material. Strubing (1956) correctly names the spermatheca, but describes the accessory gland (present author) as a bursa. As no spermatozoa were ever found in this structure, the view of the present author is that it cannot be described as a bursa.
Male reproductive organs (Plate 8, Fig.1): The male organs consist of paired testes, each comprising 3 pear—shaped follicles enclosed within a testis sheath. The paired vase deferentia widen terminally to form the vesiculae seminales, which in turn join the ductus ejaculatorius. Also joining the latter are a pair of convoluted accessory glands. The testis 21. sheath and also the hypodermal layer of the abdomen contain a yellowish pigment; Conomelus females contain little or no yellow pigment, and this feature enables nymphs to be sexed from the third instar upwards.
Development of the reproductive organs.
Males: Conomelus males mature feirly rapidly and mating may occur within
14 days of emergence. Sperms are already present in the testes of fifth instar nymphs, but these are not yet motile. Table 12 shows the growth rate of reproductive organs in relation to weight throughout the season, in the field. Although males can already copulate and transfer motile sperm after a 14 day maturation period, the reproductive organs continue to grow until the second or third week in August. At this stage the testes reach their maximum size; the accessory glands, however, increase noticeably in size throughout the entire season, while the ductus ejaculatorius and vasa deferentia also show a slight increase in length over this period. From
September onwards the testes decrease in size, particularly in thickness; the testes of males dissected in November had degenerated to such an extent that they were only distinguished with difficulty from the fat body; the fat body of Conomelus males is generally small and rather diffuse.
Females: In the female maturation is less rapid than in the male. The mean length of the pre—oviposition period of 40 females was 25.2 days, with a minimum of 18 days and a maximum of 36 days, at 16 2°C. In the field the estimated length of the pre—oviposition period in 1961 was 29 days; this figure was calculated from the appearance of the first adults and the first eggs in the field (22nd June to 21st July). Copulation may occur while. 22.
the ovarioles are still poorly developed, but mature males will not pair
with newly—emerged females.
The ovarioles of newly—emerged females are small and thread—like,
measuring as little as 0.03 mm. in vidth; they gradually increase in size
as maturation occurs, due to the development of eggs, as indicated in Table
13. In immature females much of the abdominal cavity is filled with a
rather diffuse fat body, which decreases in size as the ovaries develop.
Table 13 shows that ripe eggs first appear in the lateral oviducts in mid—
July, and that the mean number of eggs per female increases throughout the
season, reaching 12.0 during the main oviposition period. The maximum
number of eggs found in a female was 23, 12 eggs in one ovary and 11 in the
other. Maturation of successive eggs in a single ovariole is slow, and
the second oocyte is still poorly developed when the first egg ripens.
However, egg—laying is still continuous as all the ovarioles do not produce mature eggs simultaneously, and may be found in all stages of development.
Other parts of the female reproductive system also increase in
size throughout the season, particularly the accessory gland, which becomes distended with a viscous material; the spermathcca and oviducts also show slight increases in size during this period. The ovaries of Conomelus females collected at the end of the season, in October, showed no signs of degeneration and contained an average of at least 13 ripe eggs. Very few fifth instar nymphs were taken later than the 30th August, so that most of these females must have been at least 60 days old (insectary estimate 38 days). Whalley (1958) states females of this species may survive for over Table 12. Measurements of male reproductive organs.
MALES. Testes Accessory gland Ductus ejaculatorius Vas deferens Date Length Width Thickness Length Width Length Width Length Width Mean weight in mg. (of 20 males) 5/7/61 1.403 0.494 0.126 1.704 0.110 0.575 0.138 1.035 0.074 1.51 17/7 1.453 0.524 0.225 1.782 0.115 0.690 0.153 1.010 0.071 1.57 29/7 1.481 0.503 0.191 1.888 0.141 0.685 0.167 1.014 0.107 1.66 t\-1 16/8 1.736 0.506 0.253 2.280 0.177 0.784 0.200 1.449 0.102 1.87 .oa 30/8 1.610 0.713 0.253 2.851 0.268 0.989 0.184 1.610 0.150 1.88 13/9 1.462 0.680 0.234 2.736 0.208 0.966 0.201 1.748 0.128 1.94 , 13/10 1.205 0.552 0.115 3.246 0.274 1.035 0.243 1.771 0.117 1.81 1/11 0.846 0.368 0.064 3.238 0.300 0.975 0.280 1.692 0.108 not weighed
Measurements in mm. (mean of 10 individuals). Table 13. Measurements of female reproductive organs.
FEMALES.
Ovarioles Number of eggs Oviduct Accessory Spermatheca Mean weight in mg. Date gland Length Width per female Length Width Diameter Length Width (20 individuals)
5/744.10.922 0.054 0 0.614 0.112 0.582 0.577 0.130 1.74 17/7 0.830 0.047 1.37 0.623 0.069 0.506 0.531 0.115 1.73 29/7 1.136 0.049 2.00 0.593 0.092 0.499 0.464 0.108 2.00 5/8 1.729 0.061 5.60 0.761 0.097 0.637 0.618 0.128 2.09 16/8 1.913 0.200 9.00 0.586 0.139 0.655 0.568 0.128 2.48 30/8 1.757 0.115 11.80 0.749 0.110 0.788 0.616 0.145 2.79 13/9 2.081 - 12.40 0.736 0.120 0.906 0.630 0.143 2.90 ripe eggs present
3/10 2.063 13.60 0.759 0.119 0.980 0.577 0.138 3.07 ripe eggs present
Measurements in mm. (10 individuals) 25.
3 months. Degeneration of the ovarioles in old age or before hibernation
has been recorded in various insect groups. Phipps (1949) gives an account
of these changes in grasshoppers, suggesting that the degeneration is due
to lack of food at the end of the season. Richards and li:aloff (1954), however, consider that the age of the female affects the rate of yolk prod- uction, as old females with an ample food supply die with ovarioles contain-
ing little yolk. The apparently normal condition of the ovarioles, the increase in the number of eggs and in the weight of older females, indicates that the normal degeneration pattern does not occur in Conomelus. These females never laid eggs in the insectary in spite of their fully gravid condition. This failure to oviposit probably accounts for the accumulation of eggs and distension of these senescent females. The cause of this oviposition failure is not clear. It is possible that while, for an unknown reason the oviposition urge is lost, the female continues to produce more eggs resulting in the utilisation of any remaining reserve materials, and ending in exhaustion and death. Females collected in the field throughout October died, after varying periods of time, with their bodies always greatly distended with eggs. Johansson (1958) has observed this phenomenon in the Milkweed bug Oncopeltus fasciatus, although only in two instances. In both cases the bugs died with their abdomens so fully distended with eggs that their hind guts were constricted, and Johansson concluded that starvation was the cause of death. He does not, however suggest any reasons for the initial oviposition failure. 26.
Macropters.
Tables 14 and 15 show that the growth rate of the reproductive organs in macropters is very much slower than in the brachypters; this may be due to the flight muscles developing at the expense of the reproductive organs. Chapman (1956) records a direct relationship between the degree of development of the sexual organs and of the flight muscles in Scolytid beetles. Although the testes of the macropter are smaller than those of the wingless male, normal spermatozoa are present; the accessory glands, ejaculatory duct are also generally small. Table 15 shows that the reproductive organs of the female are smaller than those of the brachypters, and that the rate of maturation is much slower; females containing ripe eggs were first found in early September. In insectary-reared specimens, however, the pre-oviposition period averaged only 25 days. Reciprocal mating occurred readily between macropters and brachypters, but none of the nymphs emerging from the resulting eggs reached the adult stage.
Weight changes and sexual maturation.
The individual weights of 20 males and 20 females were taken at regular intervals throughout the season, and the data are given in Table 15 and Plate 4.
Males: Plate 4 shows that the weight in males reaches an average figure of about 1.85 mgo, relatively eaxly in the season, and that over the remain- ing period most individuals either remain at, or fluctuate around, this weight. Most of this increase is due to the growth of the testes and accessory glands. In spite of the testicular degeneration of senescent 27. Table 14. Measurements of male reproductive organs. MACROPTEROUS MALES.
Testes Accessory Vas Ductus Weight in mg. gland deferens ejaculatorius (No.of individuals Date Length Width Length Length Length in brackets)
17/7/01.092 0.460 1.345 1.150 0.598 1.29 (3) (1) 29/7 1.150 0.490 1.152 1.007 0.655 1.37 (5) (1) 16/8 0.990 0.582 1.587 1.050 0.590 1.41 (3) (4) 30/8 1.035 0.535 1.748 0.966 0.651 1.32 (5) (4) 13/9 1.173 0.460 2.415 1.403 0.644 (2) Measurements in mm. (No. of individuals in brackets)
Table 15. Measurements of female reproductive organs. MACROPTEROUS FEMALES. Ovariole Spermatheca Oviduct Accessory Weight in mg. No.of eggs gland (No. of individuals Date per female Type Diameter Length Length in brackets)
17/7/1,1- Thread- 0.476 0.414 0.460 1.58 (4) like (2) 29/7 - Thread- 0.411 0.561 0.523 1.57 (9) like (4) 16/8- Thread- 0.421 0.529 0.529 1.62 (10) like (9) 30/8 - Thread- 0.452 0.625 0.568 1.88 (10) like (4)
13/9 10.33 Containing 0.651 0.788 0.529 - (6) ripe eggs Measurements in mm. (No. of individuals in brackets) 28.
males, there is only a slight decrease in weight at the end of the season;
this can be related to the continued growth of the accessory glands. The
maximum weight attained by a male was 2.50 mg.
Females: Plate 4 shows that the weight of the females increases steadily
throughout the season. This increase is due mainly to the growth of the
accessory gland and the production of mature eggs in the ovaries. Because
of the retention of eggs by old females, the mean weight increase is
continued until the end of the sear;on, when a peak of 3.07 mg. is reached.
The maximum recorded weight of a Conomelus female was 3.92 mg.
Macropters: the weights of macropters (Plate 4) are lower than those of
the brachypterous forms, and can be directly related to the small size of
the reproductive organs.
Fecundity.
The mean number of eggs laid per female, in the insectary at
16 4" 2°C. was 29.9 with a maximum of 56 and a minimum of 10 eggs (25 females);
3 macropterous females laid an average of 23.00 eggs in the insectary, with
a maximum of 29 and a minimum of 18 eggs. An estimate of fecundity in the
field was obtained by calculating the numbers of eggs laid and dividing
this figure by the number of females produced during the season (calculated
by the regression technique of Richards and Waloff). In 1960 an estimate
of 42.02 eggs per female was obtained, and in 1961 39.61 eggs per female.
The difference between the field and insectary data is probably due to, the unnatural conditions in the artificial cultures.
The mean number of eggs laid per day was 2.43, with a maximum of 29.
22 eggs per day. The number of eggs laid per female was related to the
maturation (pre-oviposition) period, the egg-laying period and the longevity,
as shown in Plate 5, Fig.2, and Plate 6, Fig.l. The number of eggs laid
are not related to the length of the maturation period (correlation coeffic-
ient r = +0.0484 not significant). Plate 6, Fig.2 shows that most of the
eggs are laid after a mean period of 24.9 days with a maximum of 36 days
and minimum of lei days (25 females). = The pre-oviposition period averaged
25.2 days, with a maximum of 27 and a minimum of 23 days (3 females).
Plate 5, Fig.l shows that there is a general relationship between
the numbers of eggs laid and the duration of the egg-laying period of the
female although the figures are rather variable. This relationship is
significant (r = +0.6146 P 1:0.001). The same trend is shown when relating
egg numbers to the longevityiPlate 6, Fig.l (r = +0.6561 significant
P < 0.001), as the length of the egg-laying period, itself, is positively
correlated with the longevity (r = +0.8031 significant P < 0.001) Plate 5,
Fig.2. The mean duration of the egg-laying period was 23 days, with a maximum of 24 and a minimum of 1 day (25 females); the mean egg-laying period of three macropterous females was only 7.3 days, with a maximum of
13, and a minimum of 2 days. An estimate of the egg-laying period in the
field was obtained from the length of the peak oviposition period, i.e., 24 days (15th August - 8th September 1961).
The mean longevity of females in the insectary was 38.3 days with a maximum of 52 and a minimum of 27 days. The corresponding figures for the macropters were 32.7, 36 and 29 days respectively. In the field,.a 3C).
rough estimate of longevity was obtained by noting the date of appearance
of the first adult and the date when decline of the adult population was
first recorded; the estimate of longevity in 1961 was 43 days (22nd June —
4th August), which approximates to the insectary figure. In 1960 the
longevity of females in the field was estimated to be 36 days (5th July —
10th August). Longevity of the males in the insectary averaged 27 days
with a maximum of 42, and a minimum of 16 days. In the field the males
are found over the same period as the females, although they tend to emerge
and die earlier (see page LA). Conomelus adults may be collected in late
October and November, which indicates that the longevity in the field may
be at least 60 days, In 1960 several females were taken on the 8th January;
these may, however, have arisen from a partial second generation on account
of the exceptionally warm autumn.
Effects of parasitism on the reproductive organs.
Pipunculid parasitism.
The only accounts in the literature that deal with the detailed
effects of pipunculids on delphacids are those of Lindberg (1946) in
Finland and Williams (1957) in Mauritius. In the present study some data
on the effects of Pipunculus semifumosus Kowarz on Conomelus nymphs and
adults were obtained.
External effects: Nymphs and adults containing primary larvae were
indistinguishable from normal individualsl and only showed obvious symptoms
when mature second instar larvae were present. The parasitised delphacids
become greatly distended, although the female hosts are not generally much 31.
larger than normal gravid individuals. However, the intersegmental
membranes of the latter are white in colour, whereas those of the parasitised
deiphacids are yellow due to the colouration of the underlying larvae. The
sclerites of the parasitised adults are often poorly scierotised and lighter
in colour. To estimate the effects of the parasite on host size, measure-
ments were made of the head-width and femur-length of infected individuals
Table 16.
Table 16. The effect of P.semif=osus on head-width and femur-length.
Females Males
Head-width Femur-length Ovipositor Head-width Femur-length (11) (10) (15) (10) (11)
Normal 0.97 0.84 1.33 0.93 79
Parasitised 0.95 0.79 1.27 0.92 78 t 1.72 3.91 3.24 0.67
° Freedom 20 18 28 18 20
Significance Not Significant Significant Not Not significant significant significant
Value of P P <0.01 P < 0.01
The results show that the femur lengths, but not the head widths of the parasitised females are significantly smaller than those of normal individuals.
Males show no significant size differences when parasitised; this may be due to the fact that the males emerge before the females, in the field, and are, therefore, already in the adult stage when attacked by the parasite. Female deiphacids are mainly parasitised in the fourth or fifth instar and do not 32. acquire the normal adult dimensions after moulting. Fifth instar nymphs, containing mature larvae, are also reduced in size; the head widths of parasitised specimens measured 0.79 mm., and normal nymphs averaged 0.85 mm.
(t = 3.052 for 80 freedom P (0.02).
Lindberg (1946) records that the wings of Chloriona spp. (Delphac— idae) are altered in various ways as a result of parasitism by pipunculids; the number of cells in the forewings are reduced andthe normally macropter— ous males become brachypterous. In the present study no alterations were found in the wings of 67 parasitised adults. Macropterous forms in samples of parasitised delphacids appeared in the same ratio as in normal samples.
The reduction of the external genitalia of a delphacid, Chloriona sp., by a pipunculid parasite has been described by Lindberg (1946). The only visible effect on the external organs of Conomelus is the reduction in length of the ovipositor (see Table 16); the male parameres and aedeagus are unaffected by parasitism. Lindberg (1946), however, noted considerable reduction in the size of the parameres and aedeagus and also abberations in the structure of the terminal abdominal scierites. The difference in the effect of the pipunculid on Chioriona and Conomelus is probably related to the period of time that the parasite spends within the nymphal stages. The pipunculid attacks the second or third instar Chioriona nymphs and over— winters in the nymphal host from September until May or June — Lindberg
(1946); Conomelus is parasitised in the fourth or fifth instar (or adult) stage in late June or early July. The pipunculid is, therefore, only present within the nymphal host for a short period, as most nymphs have 33.
moulted to become adults by mid—July.
Internal effects of the parasite: The most obvious internal effects are
those resulting in reduction of the reproductive organs. In general the
degree of reduction is directly related to the size of the larval parasite
and, therefore, to the stage at which the host was initially attacked.
Dissection, however, indicates that larvae of similar size may produce
varying degrees of damage, as shown in Tables 17 and 18.
In the male host, damage to the sexual organs may be variable, see
Plate 8, Fig.', 2, 4 and Table 17. In most cases the mature second instar
consumes the entire testes, ducts andpart of the accessory glands; the
reduced testes generally contain normal spermatozoa, and copulation and
fertilisation of females may still occur. The gut and nerve cord remain
intact, but the former may be pushed to one side. When the larva emerges
from the host all remaining tissues are destroyed, resulting in the death
of the delphacid.
Reduction of the female reproductive organs often results in the
complete destruction of the ovarioles with the spermatheca and accessory
glands remaining intact, Plate 8, Fig.3, 5, 6. In most cases the ovariole
follicles never develop, and in only one instance were mature eggs found in a parasitised female — Table 18. The figures in Table 18 indicate that although parasite damage is variable, larvae measuring over 2.3 mm. in length generally destroy the entire reproductive system. Again, as with the external effects, the reduction of the internal organs of the females is greater than the reduction occurring in the male; this difference is 34.
Table 17. Relationship between length of larva and damage to male
sexual organs.
Larval length Length in mm.
(mm.) Testes Vas deferens Accessory gland
0.42 1.25 1.40 1.73
0.93 1.02 1.42 1.91
1.15 destroyed 1.50 1.84
1.85 0.92 0.97 1.34
1.95 destroyed destroyed 0.50
2.00 II II destroyed
2.05 0.55 0.87 1.62
2.10 destroyed destroyed destroyed
2.10 0.80 1.12 1.15
2.18 destroyed destroyed destroyed
2.22 1.20 1.50 0.82
2.25 destroyed destroyed destroyed
2.25 It 1.12 1.74
2.25 1.05 0.54 1-19
2.25 destroyed 1.87 2.04
2.31 If destroyed destroyed
2.35 0.85 1.87 1.90
2.46 destroyed destroyed destroyed
2.52 II 1.30 1.97
2.57 it destroyed destroyed
(These data were selected from 32 dissections) 35.
Table 18. Relationship between larval length and damage to female
sexual organs.
Larval length Length in mm.
in mm. Ovarioles Oviduct Accessory gland Spermatheca
0.52 Thread-like 0.56 0.45 0.67
1.05 ft 0.62 0.80 0.62
1.50 tt 0.52 0.37 0.62
1.70 ft 0.87 0.50 0.62
2.05 Destroyed Destroyed Destroyed Destroyed
2.12 tt It 0.40 0.67
2.20 ft It Destroyed Destroyed
2.20 Thread-like 0.70 0.55 0.83
2.30 It 0.45 0.42 0.50
2.30 Destroyed Destroyed Destroyed Destroyed
2.35 Well developed 0.50 0.75 0.30 2 ripe eggs
2.37 Destroyed Destroyed 0.55 Destroyed
2.37 if ►r 0.42 0.36
2.45 It 11 Destroyed Destroyed
2.50 ft 11 It ft
2.59 Thread-like 0.52 0.45 0.30 2.60 Destroyed Destroyed 0.45 Destroyed
2.75 , ft Destroyed 36.
probably also related to the stage at which the host is attacked.
Effects on host movement.
A few tests were carried out on parasitised adults to determine
whether their rate of movement was impaired. In spite of repeated
stimulation only 1 out of 10 individuals managed to leap at all — a distance
of 1 inch only. Healthy adults will readily leap from 12 to 36 inches
when disturbed. Walking movement was also reduced; 10 parasitised adults moved an average of 0.67 cm. per second, with a maximum of 1.33 cm. per
second, and a minimum of 0.17 cm. per second. Unparasitised Conomelus
adults generally leap, and do not readily walk, when disturbed; their rate of walking movement has, however, been calculated to be at least 1.50 cm. per second. The inhibition of leaping and walking can be related to the damage done to the springing muscles, and to the nervous centres, by the parasite.
Strepsipteran parasitism.
The effects of strepsiptera on deiphacids have been described by
Hassan (1939), Lindberg (1939, 1949) and Williams (1957). In the present study notes were made of some of the aberrations produced in Conomelus by the parasite, Elenchus tenuicornis Kby.
External effe.cts: The external effects vary with the sex of the parasite.
The female parasite uniformly distends the host abdomen, extruding the cephalothorax through the lateral intersegmental membrane; the sclerotised male puparium generally lies at right angles to the main axis of the host and, therefore, distorts the abdomen, which bulges towards the point of 37.
extrusion of the puparial cephalothorax. The lighter colouration of
parasitised delphacids is due to the poorer pigmentation of the sclerites,
especially those of the abdomen. The host exhibits no visible symptoms
while the parasite is in the triungulin or primary larval stages. The
head—widths and femur—lengths of parasitised adults are given in Table 19.
Table 19. Head—width and femur—length of parasitised adults.
FEMALES.
Male parasite Female parasite
Head—width (11) Femur—length (10) Head—width (6) Femur—length (6)
Normal 0.97 0.84 0.98 0.86
Parasitised 0.94 0.76 0.95 0.81
t 4.02 3.33 1.75 2.50
° Freedom 20 18 10 10
Significance Significant Significant Not Significant significant
Values of P P <0.001 P 4:0.01 P
MALES.
Male parasite Female parasite
Head—width (5) Femur—length (5) Head—width (7) Femur—length (8) Normal 0.93 0.78 0.93 0.79
Parasitised 0.92 0.74 0.91 0.75 t 0.73 1.85 1.32 1.78
° Freedom 8 8 12 14
Significance Not Not Not Not significant significant significant significant
Measurements in mm. Number of individuals in brackets. 38.
Although based on rather few specimens the data in Table 19 do show that the size of the females tends to be reduced by both male and female parasites; the male hosts are not affected in this manner. The explanation of the more marked effects on the female host may again be the same as that given on page 31; Conomelus is attacked in the fourth or fifth instar, or early adult, so that the rapidly developing male is less affected by the parasite than the female.
No reduction of wing venation was observed in parasitised adults.
Lindberg (1960) states that stylopisation may reduce the number of macropters produced in Deiphacodes spp. The percentage of macropters in the parasitis- ed samples of Conomelus was only 3.3% in 1960, compared to the 8% in healthy samples. No definite conclusions could, however, be drawn from these figures as the samples were too small.
The effects on the external genitalia are marked; in the female the ovipositor is often reduced in length (Table 20), while the males either exhibit great reduction or none at all, with few intermediate forms (Plate 9,
Figs.1-3).
Table 20. Mean ovipositor length of parasitised Conomelus females.
Male parasite (12) Female parasite (6)
Normal 1.32 1.35
Parasitised 1.18 1.32
t 2.16 0.64
0 Freedom 22 10
Significance Significant Not significant at P = 0.05
Measurements in mm. Number of individuals in brackets. 39.
Although this result is only based on small samples there is an indication
that male parasites have a much greater effect on the external genitalia than
the female parasites. This is probably due to the difference in structure
and growth—rate of the two sexes; the male puparium develops more rapidly
and is extruded some time before the female parasite. The genitalia of
Conomelus males were generally normal and the reductions shown in Plate 9,
Figs.2, 3 only occurred in about 1/6th of the specimens. In the extreme
forms of reductions the peremeres are either very small or absent, and the
shape of the genital segment or pygofer may become altered. These extreme
reductions are probably again related to the length of time that the parasite
remains within the nymphal host; the effects being greatest in those males
that emerge late in the season and have, therefore, been attacked in the
fifth instar. Lindberg (1939) and Hassan (1939) both deal with delphacids
that contain the overwintering parasite, and in most cases these show
considerable reductions in their external genitalia and terminal abdominal
sclerites; the genitalia of parasitised overwintering delphacids,
Dioranotropishalllataandpelphacodespe1lucida.taken in the present study
were almost always aberrant. This would again confirm the relationship
between the time spent by the parasite within the nymphal host and the resulting effect on the adult.
Internal effects: These effects can again be related to the stage at which
the host is attacked. Mature Conomelus females, with fully—developed
ovaries, occasionally contained triungulins, which had obviously not
affected the hosts. The nerve cord and gut are generally never attacked,
40.
although the latter is often displaced. The effects on the reproductive
organs are rather variable even when fully-developed male puparia and mature
female parasites are present; this variability is shown in Table 21, which
gives some representative figures obtained from 33 parasitised adults.
Table 21. Relationship between parasite and damage to reproductive organs.
MALES.
Lengths in mm.
Parasite Testes Accessory gland Vas deferens Ductus ejaculatorius
Female 0.62 2.22 1.25 0.50
if Destroyed 1.75 1.00 0.60
If 11 Destroyed Destroyed Destroyed
Male puparium 1.25 1.60 1.15 0.45
u Destroyed Destroyed Destroyed Destroyed
FEMALES.
Ovaries Spermatheca Oviduct Accessory gland
Male puparium 14 ripe eggs 1.37 0.80 0.46 (+ sperm)
u Reduced Destroyed Destroyed Destroyed
H Destroyed H /1 H
Female 26 ripe eggs 1.50 0.90 0.49
It Thread-like 1.50 0.75 Destroyed
u Destroyed Destroyed Destroyed if
7 out of the 19 female hosts contained ripe eggs; 3 had well-developed
ovaries and no eggs; in the remaining 9 the reproductive organs were 41. completely destroyed. The spermatheca of one of the females contained sperm) showing that males will still copulate with parasitised individuals. Of the males, 3 out of 14 had fairly well—developed testes containing sperm. In all parasitised adults, however, the size of even undamaged reproductive organs is smaller than in the healthy specimens; this also applied to delphacids parasitised by pipunculids.
Sex ratio of Conomelus.
The mean sex ratio of Conomelus adults produced throughout 1960 and 1961 is given in Table 22.
Table 22. Mean sex—ratio of Conomelus adults in 1960 and 1961.
Ratio Year Wing form Females Males
1960 Brachypter 1 : 1.55
1961- I' 1 : 1.62
1960 Macropter 1 : 1.31
1961 ri 1 : 0.36 42.
Table 23 gives the sex-ratios of Conomelus during the course of 1961.
Ratio Date No. of females No. of males Females Males
5/7 2 38 1 : 19.00
8/7 9 50 1 : 5.55
17/7 21 96 1 : 4.57
21/7 35 105 1 : 3.00
25/7 50 73 1 : 1.46
2/8 43 70 1 : 1.63
5/8 38 49 1 : 1.29
16/8 43 48 1 :. 1.12
23/8 53 44 1 : 0.83
30/8 53 47 1 : 0.89
13/9 65 53 1 : 0.81
20/9 62 39 1 : 0.63
Table 23 shows that the mean sex-ratios are fairly constant for the two years, and that the macropterous condition is far more prevalent in females than in the males. It is apparent that the males tend to emerge before the females, but that the ratio of females to males gradually increases throughout the season until the sex-ratio is reversed in late
September, i.e. 1 male : 1.59 females. 43.
SECTION 2.
Biology of the predators and parasites.
The biology of some of the predators and parasites associated with
Conomelus is almost unknown. During the present study the life histories
and immature stages of the egg parasite Aprostocetus, the nymphal and adult
parasite Pipunculus, and the mirid predators Tytthus and Cyrtorhinus, were
studied. The biology of most of the other natural enemies has been des-
cribed, at least briefly, in the literature. However, some notes on the
occurrence of the predators in the field, and their feeding habits, are
included in this account, as these are relevant to the mortality estimates
discussed in Section 3.
Methods.
Predators were obtained from the field by beating and hand porting; mirid and nabid nymphs were reared in the laboratory from rush clumps kept in glasstanks. Adult egg parasites were also obtained from these Juncus cultures, and survived on a diet of sucrose solution or pieces of raisins.
Pipunculids and Strepsiptera were bred from visibly-parasitised hosts. collected in the field, and the immature stages were studied by dissecting
Conomelus samples throughout the season. Spiders and Opiliones were identified with the keys of Locket and Millidge (1951-53) and Todd (194B), respectively. Immature stages of the parasites were mounted in polyvinyl- lactophenel plus acid fuchsin, or Hoyer's mountant. Preparations of fungal material were made in lactophenol, using Cotton Blue stain. 44.
Parasites of the atistage.
Anagrus incarnatus (Haliday) (Mymaridae).
Although there are numerous references in the literature to this
species, the systematic position of this and other Anagrus species is still
rather obscure. Bakkendorf (1926) has reviewed the group, and considered
that all the 34 species of Anagrus known at this time were one and the same
species, namely A.incarnatus, This concept of one cosmopolitan species is
no longer held, and many species have been described. Whalley (1958) gives
a detailed account of the biology of A.incarnatus, and his work includes a
key drawn up by Hincks to separate this species from A.etomus (L.), another
mymarid bred from homopteran eggs. In the present study, mymarids fitting
Hinck's description of A.incarnatus have been reared from the eggs of
Conomelus, and a jassid Tettig(Ila viridis (Linn.); the individuals
parasitising the delphacid eggs, however, possess long ovipositors. Whalley
(1956 and 1958) considers that the long-ovipositor type is merely a different
race of the same species. The view of the present author is that these
types represent separate species, as they have such distinct host preferences.
Description of the stages: The immature stages have been described by
Ganin (1869), Henriksen (1918-19), Bakkendorf (1926) and Whalley (1958).
Life history.
Copulation and oviposition were observed in the present study.
Newly-emerged females mate readily without feeding beforehand; the males generally locate the females within 60 - 120 seconds in a 3 x 1 inch tube, and copulation lasts from 15 to 20 seconds. Ovipositing females run over the surface of the Juncus stem, exploring the surface with the antennae; on 45.
finding an oviposition site, the insect moves about 0.5 mm. from the opening
before inserting the ovipositor into the eggs lying obliquely within the
stem; oviposition never occurs directly into the opening. Anagrus females have been observed ovipositing in the field; the females fly slowly among
the rush tussocks, settling on the upper parts of the stems, and then'run downwards until the egg sites are reached. Generally only one egg pet site is- parasitised, but up to 4 eggs per site has been recorded.
Only one parasite is found in each host egg; the conspicuous orange larvae develop rapidly, and prepupae have been found only 10 days after oviposition at 58 — 68°F. (14.4 — 20°C.); adults emerge after a pupal period of 4 — 7 days, leaving a small circular opening in the chorion and a similar exit—hole in the rush stem; these neat circular holes are very characteristic, and differ from the larger ragged openings left by
Aprostocetus adults. The duration of the entire life cycle at 58 — 68°F.
(14.4 — 20°C.) is 16 days. In the field the period is probably longer, 3 to 4 weeks, and the mymarid could, therefore, pass through 4 generations per annum. In the present study it was confirmed that a second generation occurred on the Conomelus eggs in spring, as adult mymarids were already taken on the 20th April in 1960, and the 27th April in 1961, and primary larvae were found as late as mid—June. It was also apparent, in a few instances, that an extra generation occurred on the Conomelus eggs laid in autumn, although most of the parasites overwinter as larvae and do not emerge until the following spring.
Bakkendorf (1926) considers that Conomelus is the primary host 46.
of A.incarnatus, and that the alternate hosts are found among other
Homoptera and the Odonata. The results obtained in the present study
did not support this view, as the percentage of parasitised Conomelus eggs
was generally very low (0.12 — 0.32%) in Pond field. In another part of
the field station where other leafhopper species were more abundant, a
higher proportion of the Conomelus eggs was attacked (2.55%), this indicates
that the primary host or hosts, of A.incarnatus were present in this area.
Aprostocetus mandanis (Walker) .CEtliovhi.42..a.)
Graham (1961) has published the following synonymy of this species:—
A.mandanis (Walker)
= Cirrospilus mandanis Wlk.1838
= Annellaria conomeli Bakk.1933
= Tetrastichus conomeli Bakk.1953.
Whalley (1958) has reared the parasite from Conomelus eggs, and identified
it as a species near Tetrastichus_conomeli; he incorrectly describes the
species as a member of the Entodontinae.
Description of the stages: Bakkendorf (1934) described the adults, under
the name Annellaria conomeli, and then redescribed the species in 1953, as
Tetrastichus conomeli. The immature stages have not been fully described previously, although Whalley (1958) provides some notes on larval develop—
ment.
Egg: No eggs were found in the present study, but Whalley (1958) states
that they are pedunculate in shape.
First instar larva: Only one specimen was obtained; this was damaged, and 47.
it was not possible to figure the mouthparts; the mandibles are minute
and poorly selerotised; the larva is translucent, and measures 0.25 mm.
in length.
Second instar: The length of this stage varies from 0.55 mm. (early phase)
to 1.25 mm. (late phase). Plate 10, Figs.l and 2; 13-segmented, trans- lucent in the early stages, becoming opaque and whitish later, gut contents yellow, becoming deep brown later; the head and three thoracic segments each bear a pair of minute dorsal and ventral spines, the caudal segment bears a pair of dorsal spines; mouthparts (Plate 10, Fig.4), strongly scierotised mandibles articulate with the anterior and posterior pleurostom- al processes, epistoma well-developed, pleurostoma indistinct and hypostoma indefinable, labial and maxillary palps visible as small tubercles; the larva is unusual among the Tetrastichinae in that no spiracles or tracheae are present. A great increase in size occurs during this stage (Plate
10, Fig.' and 2) which suggests that there may be a third larval instar; however, no differences in the size or structure of the mouthparts or other characters, were observed, and also no extra exuvia were ever found; a third instar could, therefore, not be confirmed.
Pupa (Plate 10, Fig.3): Mean length 0.95 mm.; whitish before attaining the adult colouration; caudal segment bears the remains of the larval integument.
Life history.
Mating and oviposition were not observed, in spite of many attempts to do so. Oviposition is probably similar to that of Anagrus, 48. the ovipositor being inserted through the rush stem into the deiphacid egg below.
Only one larva is found in each egg, and generally only one egg per site is parasitised, but in one instance four out of four eggs contained
Aprostocetus larvae. As oviposition was not observed in the laboratory, the duration of the egg stage could not be estimated. However a rough estimate was made from the field data; primary larvae were first noted on the 16th August 1961, 16 days after the first host eggs were available in any number; hatching may, therefore, occur within 16 days of oviposition.
The larva moults to the second instar, devours the entire contents of the host egg, and then emerges by chewing its way through the chorion at the posterior pole. The remains of the host egg are distinct from those left by Anagrus because of the extensive damage to the chorion, and the round excretal pellet or meconium left within the egg—shell. After leaving the host egg the larva becomes predatory and feeds on other Conomelus eggs as well as the eggs of Tettigella viridis, and other Aprostocetus larvae; eggs parasitised by mymarids are also destroyed, as larvae with orange gut contents have occasionally been found. The larva tunnels through the pith of the Juncus stem for distances of 1 cm. or so, seeking further egg sites.
The maximum number of eggs destroyed by a single larva was 9, including the original host egg; the mean number of eggs taken by Aprostocetus in the field was 7.97 per larvae, i.e. 199 larval tunnels contained the remains of
1587 eggs in 1961. Pupation occurs after an initial prepupal stage during which the larvae becomes rigid, and the segmentation more distinct; the 49. pupa lies in a cavity in the pith, generally parallel with the rush stem.
Before pupation, at least 50% of the larva become infected and killed by a
fungus, Paecilomyces sp. It is not clear whether the larvae are killed by
the fungus, or whether they are infected saprophytically after death. The adult emerges by chewing its way through the stem wall ) leaving an exit hole
that is larger and more ragged than that left by Anagrus.
The life cycle can be summarised as follows: Aprostocetus over- winters as the larva within the host egg, or as a free-living predator within the rush stem; most larvae have left the eggs by mid-March, and pupae are found from mid-April until early June. The first adults have been taken as early as the 15th June. The parasite may then pass through another generation on an alternate host, as no Conomelus eggs are available until the end of July, or may survive this period as the adult. In the laboratory at 62 - 72°F. (16.7 - 22.2°C.) the period of larval maturation, after emergence from the host egg, is 20 days; the duration of the prepupal stage is 1 - 2 days, and that of the pupa 16-24 days. Adults survived for
5 days when fed on sucrose solution.
Bakkendorf (1934) and Whalley (1958) consider that Aprostocetus only passes through one generation per annum; this is borne out by their data which show.; that the emergence of the adult parasites coincides with the beginning of the oviposition period of Conomelus. Results obtained in the present study show that a second generation may occur, as adult parasites emerge long before any Conomelus eggs ere available. 50.
Fungal parasite Paecilomyces sp. (Moniliales, Fungi imperfect).
Dr.M.F.Madelin kindly identified the fungus infecting Conomelus eggs as a
Paecilomyces sp., similar to a species obtained from the eggs of a Delphac—
odes sp. at Rothamsted. No records of this genus have been found in the
literature, although there are numerous references to Spicaria, a very
closely related genus. Whalley (1958) records that large numbers of
Conomelus and Tettigella viridis eggs were attacked by a fungus, but has not identified the species. The only other British record for a fungal parasite of homopteran eggs is that of Morcos (1953) who obtained an
Acrostalagmus sp. from the eggs of the jassid Tettigella viridis.
Nature of the infection: The mode of infection has been observed in the laboratory; conidia in contact with a Conomelus egg send out germinating hyphae which penetrate the chorion; the hyphae grow rapidly within the egg, which collapses and becomes opaque; after a while the hyphae burst out through the chorion, forming masses of conidiophores, which once again produce conidia. In the field the process is probably similar; the conidia may settle in, or near, the oviposition sites, and send out germinat— ing hyphae which pass into the egg batches. After development within the host eggs is completed, the conidiophores grow out through the opening of the oviposition site. These may be seen in the field, giving the stems a whitish speckled appearance. In heavy infestations thee conidial masses coalesce and form a dense 'sleeve' around the stems. Infection of adjacent sites on the same stem may occur by the mycelium growing out of one site, passing over the surface of the stem, and into the next.opening. All the 51.
eggs in any one site are destroyed by the fungus, and the incidence of the
disease is greatest near the base of the stem, where the humidity is high.
The life cycle may be summarised as follows: The first infected eggs are
found at the end of August; the fungal infection gradually builds up
during the winter and spring months (Section 3, pagemo). How the fungus
survives through the period from May to August when no Conomelus eggs are
available is unknown; an alternate host may become infected, or the conidia
may remain as resting stages over the period, or the fungus may develop
saprophytically on vegetable or dead animal matter.
Pathogenicity: Dr.Madelin has kindly provided me with the data obtained in
establishing the pathogenicity of Paecilomyces. Seven petri—dishes,
containing 2% plain agar, were each flooded with 0.5 ml. of spore suspension,
and another set were left untreated as controls. The surfaces of Conomelus
eggs were sterilised with dilute sodium hypochlorite solution for 20 — 30 minutes, and 7 — 15 eggs were placed in each dish; the cultures were
incubated at 18°C. The results are summarised in Table 24.
Table 24. Pathogenicity of Paecilomyces to Conomelus eggs.
Treatment Total no. of eggs No. plus Paecilomyces % plus Paecilomyces
Inoculated 40 39 97.5 with fungus
Control 39 0 0
The data clearly shows that Paecilomyces is pathogenic, as most of the healthy embryonated eggs used in the experiment were attacked; in all cases very vigorous growth of the fungus occurred, with hyphae initially breaking 52.
through the micropylar region of the egg, and later enveloping the entire
surface.
Nymphal and adult parasites.
Pipunculus semifumosus (Kowarz) (Diptera; Pipunculidae).
Although there are numerous breeding records for pipunculids, very little
work has been done on the biology of the parasites. The only recent work
is that of Williams (1957) who describes the immature stages of a pipunculid
from Mauritius. Lindberg (1946) deals almost entirely with the effects of parasitism on the host. The only British work is that of Keilin and
Thompson (1915) who describe the larva of Chalarus spurius (Fallen). There are several British breeding records for P.semifumosus; Whalley (1958) has
obtained the species from C.anceps and Delphacodes fairmairei, and Morcos
(1953) has bred the parasite from the bracken delphacid Criomorphus pteridis
(Spinola).
Description of the stages.
Eggs: No eggs were found in the present study in spite of dissecting nearly
2000 Conomelus nymphs and adults, and 1100 specimens of other delphacids.
No gravid females were obtained in the field so that it was not possible to examine the ovarian egg. The only record of a pipunculid egg (species unknown) is that of Loew (1841) who describes the ovarian egg as flask—like in shape.
First instar larva (Plate 11, Fig.3 and 4): Length 0.37 mm. (early stage)
— 2.35 mm. (late staga), 12 indistinct segments, including the head,terminal abdominal segment forms a vesicle, segments 2 — 10 bear minute spines 53. ventrally; translucent; tracheal system indistinct, but late first instar is clearly metapneustic; mouthparts (Plate 11, Fig. 1 and 2) well-developed; cephalopharyngeal skeleton consists of a basal pharyngeal sclerite, which bears a pair of sclerotised ventral and dorsal arms, these being joined by a dorsal arch; lateral or mandibular sclerites are articulated with the pharyngeal sclerite, each bearing a sharp downwardly-directed tooth or mandible terminally; also articulating with the pharyngeal sclerite is a median basal sclerite, i.e, the labial sclerite.
The mean length of 35 first instar larvae was 1.39 mm., ranging from 0.37 mm. to 2.35 mm. (final phase); the size of the cephalopharyngeal skeleton is the same in the smallest and largest larvae, and hosts parasitis- ed by mature larvae never contain more than one exuvium. Keilin and
Thompson (1915), and Williams (1957), consider that only twc larval instars occur in pipunculids, which appears to be confirmed by the present work. In view of the considerable increase in size during the first instar, the author is, however, of the opinion that an intermediate moult (as yet undetected) does occur.
'Second instar' or Mature larva (Plate 12, Fig.3): Mean length 2.31 mm., maximum 2.76 mm. and minimum 1.95 mm. (38 larvae); general cyclorrhaphan characters; stout, cylindrical, tapering anteriorly at the head and posteriorly at the caudal vesicle; segmentation very indistinct because of the considerable folding of the integument, integument smooth, with no spines or tubercles; general colouration yellowish —from colour of gut contents; mouthparts (Plate 12, Figs. 5 and 6), the cephalopharyngeal 54.
skeleton consists of a pair of strongly sclerotised triangular pharyngeal
scierites; these are produced forward into two less-sclerotised ventral
arms, which probably represent the once-separate hypostomal scierites;
articulated with the anterior arms is a pair of mandibular sclerites, each
bearing a downwardly-directed tooth or mandible; lying between the arms is
a basal labial plate which bears spines on its surface and along the anterior
edge; the pharyngeal sclerites are joined by an indistinct dorsal arch.
Head segment, when extruded, bears a pair of papillae on either side of the
mouth; these probably represent the antennae and maxillary palps. Tracheal
system amphipneustic; two main trunks run laterally from the anterior
prothoracic spiracles to the posterior abdominal spiracles, joined anteriorly
and posteriorly by large transverse trunks and numerous smaller tracheal
commissures; anterior prothoracic spiracle (Plate 12, Fig.2) consists of a
weakly-sclerotised spiracular chamber from which arise four protuberances,
each terminating in a pore, the whole spiracle is covered by a translucent
integumental 'cap'; posterior abdominal spiracles are borne on a strongly-
sclerotised spiracular plate (Plate 12, Fig.].); each spiracle consists of
three pores opening into a circular less-sclerotised area; the median posterior part of the plate bears a funnel-like depression while anteriorly
two short sclerotised processes pass into the body.
Puparium (Plate 12, Fig.4): Mean length 2.40 mm.; reddish-brown becoming
blackish-brown later in the stage; prothoracic spiracles project through
the puparium anteriorly, and posterior spiracular plate is also presen'.0 distinct fracture lines are present anteriorly. 55.
Adult: Is described by Verrall (1901) under the name of P.strigulipes
(Verrall).
Life history.
Mating and oviposition: neither of these activities were observed in the present study in spite of repeated attempts to do so. Jenkinson (1903) and Williams (1931) describe the oviposition behaviour of Hawaiian species; the female grasps a leafhopper nymph with her legs, sometimes lifting the host from the substrate, and inserts the ovipositor between the interseg— mental membranes of the abdomen. In the present study adult pipunculids were often seen hovering slowly among the Juncus tussocks, but never made any attempts to attack the numerous delphacids present.
Choice of host: only fourth and fifth instar nymphs, and adults, are attacked. Large numbers of first, second, and third instar larvae were dissected, but these never contained eggs or primary larvae. In a. few instances, fourth instar nymphs contained well—developed larvae) and these had probably been parasitised in the third instar. On the other hand, mature Conomelus adults containing larvae less than 0.5 mm. in length have been found; these have obviously been attacked in the adult stage. The second generation of pipunculids, however, attack the second or third instar nymphs of the overwintering host, Delphacodes pellucida, or the adults of
Megamelodes venosus (Germar).
Larval development.
The duration of the egg stage could not be determined as oviposit— ion was never observed. Estimates were, however, obtained from the time
56.
elapsing between the initial appearance of the adults in the field, and
the first appearance of larvae in the hosts; in 1960 these dates were
6th June and 30th June respectively; the estimated duration of the egg
stage is, therefore, 24 days. Only single larvae were found in the paras-
itised delphacids, but Keilin and Thompson (1915) have noted two larvae of
Chalarus spurius in a typhlocybid host, although only one matured. The
pipunculid larva increases in size, and gradually destroys the tissues of
the delphacid (details of the effects of parasitism on the host are given
in Section 1). Primary first instar larvae are found in the abdominal
cavity of host, and only rarely in the thorax. The orientation of the
larvae within the host is summarised in Table 25.
Table 25. Orientation of larvae within the host.
Instar Facing head of host Facing posterior part of host
Second 12 28
First > 1.5 mm. 11
First 1:1.5 mm. 8 7
The data show that the early first instar does not orientate itself in any
particular direction, the later first instar generally faces posteriorly,
and the second instar may face either way, but more often posteriorly.
Keilin and Thompson (1915) noted that the late first instar larva of
Chalarus spurius was orientated towards the head of the host, but that after
the moult the second instar faced the opposite way. Perkins (1905d) states
that the mature larvae of Pipunculus spp. always face the anterior part of 57.
the host. The mature larva fills the entire body cavity of the greatly—
distended host, and emerges after consuming any remaining tissues in the
delphacid. The larva forces its way out of the host by a series of
contractile movements, and may emerge within 10 minutes of the death of the host. Emergence may take place at any part of the host, but occurs most
frequently between the first and second abdominal sternites, and the third
and fourth abdominal tergites (Plate 8, Fig.7). The mature larva generally pupates within 24 hours after eme:gence, and only occasionally after 48 hours. Pupation in the field occurs at the base of the rush tussocks, and puparia were often found among the litter. Adults emerged after a mean period of 17 days in the insectary, and survived for only 4 days on a diet of sucrose solution. The sex ratio Was 1 female : 1.7 males.
The life cycle can be summarised as follows: the parasite overwinters as the larva in adults of Megamelodes venosus (3.57% of 371 specimens), and Delphacodes pellucida (0.21% of 479 specimens). The first adults emerge in early June of the following year, and attack the fourth and fifth instar nymphs of Conomelus, and peak parasitism is reached in mid—July. The second generation adults emerge in late July and early August, and attack the overwintering hosts.
A pupal parasite was bred out from a puparium in 1961; this was kindly identified by Mrs.J.A.J.Clark of the British Museum as a male of
Loxotropa atricrus Kieffer. 58.
Elenchus tenuicornis (Kirby) (Strepsiptera : Elenchidae).
The present systematic position of Elenchus species is still rather obscure. Hassan (1939) reviewliprevious work on Elenchus, and
concluded that Strepsipterous parasites of delphacids are all one and the
same species, namely E.tenuicornis. The most recent work on Elenchus is
that of Williams (1957) in Mauritius; this author describes his species as E.templetoni Westwood, but mentions, in his discussion, that he could
find no characters to distinguish the male from Hassan's description.
Lindberg (1939) described several different species of Elenchus for differ- ent delphacid hosts, but has since revised his opinions (1960), and considers that there is probably only one cosmopolitan species. Males taken in the present study all fitted Hassan's description.
Destription of the stages: These have been described fully by Hassan
(1939), Lindberg (1939), and Williams (1957).
Life history: The biology has been dealt with by all three workers above.
In the present study the parasite overwintered in second and third instar nymphs of Delphacodes pellucida (11.9%, 57 out of 479 nymphs parasitised), and Megamelodes venosus (0.81%, 3 out of 371 adults parasitised); over- wintering occurred as the triungulin or second larval instar stage. The parasite develops to the mature female or male puparium, which is generally extruded from the host in mid-June. By late June and early July, the
Elenchus females contain active triungulins, which are then liberated, and attack fourth and fifth instar Conomelus nymphs. The mode of attack was not observed, but in the laboratory triungulins leapt up to 50 mm., and 59. readily passed through the intersegmental membranes of Conomelus nymphs when placed artificially onto these hosts. Adult delphacids may also be attacked by the triungulins, as several mature Conomelus females containing these stages were found. Only one parasite per host was generally found. The strepsipteron develops within the Conomelus host, and the first adult females and male puparia are extruded in mid-July, and may be found until the end of September. Male puparia were extruded from nymphs and adults, but females only appeared when the host reached the adult stage. In late summer triungulins are liberated from the Conomelus hosts, and once again attack the overwintering delphacids.
The effects of strepsiptera on Conomelus nymphs and adults is discussed in Section 1.
Parasitic mites. Leptus Latreille species (Acarina : Erythraeidae).
These mites were kindly identified by J.G.Sheals of the British
Museum. Conomelus nymphs and adults occasionally carried these ecto- parasites, but were not adversely affected by them. The mites clung tenaciously to their hosts, and fed at various points including the wing- veins, between the femur and tibia, and the anal tube. Hassan (1939) has recorded Trembidium sp. from Delphacidae.
Fungal parasite Entomophthora sphaerosperma Fres.
This fungus, parasitising Conomelus adults, was kindly identified by Dr.M.F.Madelin. The species has a large host list including members of the Hemiptera, Coleoptera, Diptera, and Hymenoptera - Leatherdale (1958).
Steinhaus (1949) describes the biology of the fungus. In the 60.
present study only few infected individuals were obtained these being taken
between mid-July and October. The parasitised delphacids become greatly
distended, with creamy-white intersegmental membranes; before the host
dies, the body is filled with a thick mycelium, which is particularly dense
around the tracheal trunks; the reproductive organs and leg muscles are
greatly reduced or absent, and the haemolymph contains large oily droplets.
Shortly after the. death of the delphacid, the mycelium breaks up into numerous hyphal bodies; these prhduce conidiophores, which appear through the intersegmental membranes of first the abdomen, and then the thorax.
The dead insect gradually becomes covered by a dense /sleeve' of conidio- phores, which is initially white, and later becomes brown in colour. From the records in the literature it appears that the dead hosts are normally anchored by fungal rhizoids to the surfaces of stems or leaves; in the laboratory parasitised hosts always fell to the base of the containers after death, and in the field infected delphacids were never seen adhering to the rush stems. Development of the disease is rapid; healthy Conomelus adults confined with infected specimens in 3 x 1 inch tubes, died within 3 - 4 days.
Biology of the predators associated with Conomelus.
The biology of many of the predators taken in the present study is known, with the exception of the mirids; only brief notes are, therefore, given on the other groups, with reference mainly to their seasonal occurrence in the plot, and their association with Conomelus. 61.
Tytthus pygmaeus (Zetterstedt) (Heteroptera : Miridae).
This species and other members of the genus are predatory on various stages of leafhoppers. Little is known of the biology of these bugs, and only T.mundulus (Breddin), used in biological control work in
Hawaii, has been studied in any detail.
T.pygmaeus is widespread in marshy areas on rushes and grasses, and occurs throughout northern Europe - Carvalho and Southwood (1955).
Description of stages.
Egg (Plate 13, Fig.1): Mean length 0.86 mm., greatest width 0.23 mm.
(20 eggs); short, fairly curved, with a wide operculum; chorion tough, chorion rim and operculum (Plate 13, Fig,2); opaque, white when newly- laid, becoming reddish after several days.
Nymphs.
Table 26. Nymphal measurements.
Instar Head-width Antennal segments
I II III IV Rostrum
First 0.282 0.109 0.141 0.152 0.326 0.402
Second 0.374 0.140 0.206 0.209 0.381 0.553
Third 0.455 0.173 0.287 0.301 0.475 0.671
Fourth 0.568 0.215 0.431 0.396 0.537 0.781
Fifth 0.605 0.250 0.553 0.508 0.641 0.962
Measurements in mm. 10 individuals of each instar. 62.
First instar (Plate 13, Fig.3): spider-like, small abdomen and long legs; yellowish-orange to bright red; eyes deep red.
Second instar: as first, body more elongate in relation to legs; yellow- ish-orange.
Third instar: yellowish orange; extreme base of all tibiae, and base of first antennal joint, dark.
Fourth instar: as third.
Fifth instar (Plate 14, Fig.1): variable in colour, yellowish-orange or yellowish-green; first antennal joint dark except at apex; base of all tibiae black; apices of wing pads tinged with brown; rim of stink-gland opening also tinged with brown; eyes reddish-brown.
Adult: Described by Carvalho and Southmod (1955).
Life history.
Adult males and females mate readily in the laboratory, and may remain in copula for up to 3 hours. The eggs are laid in the upper parts of Juncus stems, generally within 5 inches of the stem tips; the eggs are inserted through the stem wall into the pith, with only the chorion rim visible, and are laid singly; several eggs may, however, be laid close together to form a row. Stems containing Conomelus eggs are generally chosen, and an average of 12.6 eggs per stem are laid, with a maximum of 60 and a minimum of-1 (based on 50 stems). Overwintering occurs in the egg stage; during the winter months the stem tips often decay and break up completely, and Tytthus eggs are often found among the basal litter.
The nymphs hatch in May, after the eggs have been subjected to a 63.
cold period to break the diapause, and then restored to warmer conditions.
In the laboratory batches of eggs were kept at 1.5 ±0.5°C. for 36 days, and
then transferred to 21 4' 1.5°C. for 27 days; other samples of eggs were kept at a constant temperature of 21 ± 1.5°C. for the entire period of 63 days. Nymphs hatched from nearly all the eggs that had been subjected to
the cold treatment, but none hatched from the other group. The hatching period is fairly prolonged, and adults and first instars may be taken together. The duration of the various instars at 20°C. was: first instar
11 days; second instar 7 days; third instar 16 days; fourth instar 8 days; fifth instar 8 days, i.e. total life cycle 50 days. This data is based on only 5 individuals, as it proved very difficult to rear these mirids individually in the laboratory. These figures are, however, related to the field estimates where the total life cycle was completed in about 59 days (in 1961), i.e. first instar nymphs appeared on the 1st of May, and adults emerged on the 28th of June. The males emerge first, with females appearing about 7 days later; very few males are, however, found later than mid—August, whereas the females survive until October. The sex ratio is 1 male : 2.90 females.
The life cycle can be summarised as follows: The eggs are laid from mid—
August until October in the tips of rush stems; overwintering occurs in this stage. Nymphs emerge in May, and the first adults appear at the end of June.
Feeding.
Apart from the work on T.mundulus in Hawaii, little is known of the feeding habits of these bugs. T.mundulus is known to feed on delphacid 64.
eggs, but it has not been recorded whether the mirid also takes leafhopper nymphs, or plant food. The only record in the literature of feeding in
T.py9maeus is that of Massee (1954) who observed the mirid feeding on a young leafhopper nymph (species not identified). In the present study predation on Conomelus eggs and nymphs by this species was confirmed. It is clear that T.pygmaeus is a specialised predator, as all stages of the mirid active— ly search for the oviposition sites of the deiphacid. The bugs use the rostrum in their search for Conomeius eggs, exploring the surface of the rush stems by continual probing movements of this organ; on finding an egg site the tip of the rostrum is passed several times around the rim of the opening before fzeding occurs. The stylets are fully inserted into the site, the rostrum becoming looped during the process. 10 — 15 minutes is the mean time required to ingest the entire contents of an egg. After feeding on one egg the bug withdraws its mouthparts and cleans them with the fore legs; the rostrum is then re—inserted into the same site, and the process is repeated until all the eggs in the batch have been destroyed.
In the laboratory Tytthus adults destroyed an average of 3.71 eggs per day, with a maximum of 13 per day; very few feeding trials were carried out with mirid nymphs, but two first instar nymphs were each observed to suck two eggs in a day. Adults were presented with rush stems containing variable numbers of delphacid eggs to discover whether the predation rate was related to the number of eggs available — Table 27. 65.
Table 27. Predation rate in relation to number of available eggs.
Nos. of available eggs: 11 12 22 25 28 34 38 40 45 46 62
Nos. destroyed in 4.5 days: 6 3 10 10 16 9 15 16 13 6 13
It is evident that there is no correlation between the number of available
eggs and the number sucked.
Tytthus nymphs and adults were presented with the following range
of food: Conomelus nymphs and adults, collembola, aphids, mites, and
immature spiders; of this prey, only first and second instar Conomelus
nymphs were accepted, and these only in the absence of delphacid eggs.
However, the mirids readily attacked most cf the prey species listed above, when these were dead or moribund. Tytthus nymphs and adults survived for up to 10 days on Juncus stems, in the absence of animal food, but the nymphal stages rarely moulted to the next instar. It is, therefore, probable that at least some animal food is necessary for normal development.
As no Conomelus eggs are available from the end of May until the end of
July, the mirids may feed on dead animal material during this period.
There was no evidence of cannibalism in the present study, although again dead or moribund individuals were attacked.
Cyrtorhinus caricis_(Fallen) (Heteroptera Miridae).
This species was observed feeding on most stages of Conomelus, with the exception of the egg, in the field and the laboratory. C.caricis is widely distributed, occurring near the bases of sedge and rush clumps —
Southwood and Leston (1959). 66.
Description of stages.
Egg (Plate 13, Fig.4).: Has been described by Kullenberg (1943); similar to Tytthus; strongly curved; slightly larger, mean length 1.10 mm., greatest width 0.33 mm. (5 eggs); chorion rim and operculum (Plate 13, Fig.5).
Nymphs.
Table 28. Nymphal measurements. Instar Heedwidth Antenna! segments
I II III IV Rostrum First 0.275 0.112 0.162 0.150 0.175 0.500 Second 0.420 0.162 0.227 0.273 0.326 0.509 Third 0.530 0.193 0.344 0.381 0.364 0.655
Fourth 0.650 0.241 0.541 0.537 0.462 0.883
Fifth 0.763 0.359 0.804 0.746 0.496 1.040
Mean measurements in mm. 10 specimens of each instar.
First instar: similar to Tytthus, difference in ratio of third to fourth antennal segment is best diagnostic character; yellow to yellowish-orange; area around stink gland opening red; eyes red. Second instar (Plate 13, Fig.6): as first. Third instar: as second, but base of first tarsal joints dark.
Fourth instar: yellowish-green; red area around stink gland very conspic- uous; both tarsal joints dark. Fifth instar (Plate 14, Fig.2): green; borders of wing pads, tarsal joints, 67. and antennae, dark.
Adult: is described fully by Carvalho and Southwood (1955).
Life history.
Eggs were not found in the field, but in the laboratory females oviposited into stems of Holcus tanatus. Kullenberg (1944) records the eggs from the tips of stems of Scirpus spp. First instar nymphs hatch in the third week of May, and the first adults emerge in early July; the estimated duration of the life cycle in 1961 was 42 days, i.e. first instar nymphs hatched on
23rd May, and the first adults emerged on the 4th July.
Kullenberg (1944) considers that C.caricis is mainly phytophagous feeding on Scirpus spp. and Carex spp., but that dead animal matter may also be taken. In the present study Cyrtorhinus nymphs and adults frequent- ly fed on Juncus sap, but also attacked most stages of the delphacid; eggs were, however, not taken, and the bugs did not exhibit the characteristic searching behaviour of Tytthus. In the field Cyrtorhinus adults were also taken with immature spiders and harvestmen as prey, as well as nymphs of their own species; in the laboratory the mirids also fed on dead and moribund deiphacids and mirids.
Fieberocapsus flaveolus (Reuter).
This mirid was occasionally taken in the plot; fourth and fifth instar nymphs and adults were observed from mid-June until the end of July; the fifth instar is figured on Plate 14, Fig.3, to show how it differs from the other two mirids. Although this species fed on aphids in the laboratory, no stages of Conomelus were accepted. 68.
Other members of the Heteroptera included Anthocoris nemorum (Linn,)
Adults of this widespread polyphagous predator first appeared in the plot
in mid-July; these must have laid eggs in the rush area, as nymphs were
found in August. In the field and the laboratory, fourth and fifth instar
nymphs, and adult, anthocorids fed readily on Conomelus adults; the delphacid
egg sites were, however, not attacked. There are no previous records of
delphacids as prey of anthocorids, although Hill (1957) notes that jassids
may be attacked.
Adults of the predatory pentatomid Zicrona caerulea (Linn.) were
found in the plot in September, but did not appear to prey on Conomelus.
Three species of nabid were found in Pond field, Nabis flavomarg-
inetus Scholtz, Stalia major (Costa), and Dolichonabis limbatus (Dahlbom),
of which the first two were the most abundant. Nymphs of N.flavomarginatus
and S.major appeared in early May, with D.limbatus hatching near the end of
the month, and the adults of the first two species emerged from July to
August; Dolichonabis adults were not seen until September. All three nabids were observed to feed on most stages of Conomelus, except the egg, both in the field and in the laboratory. The early nabid nymphs were generally found on the litter among the baSes of the rush clumps, whereas the adults appeared to hunt for their prey among the upper strata of the vegetation; Fewkes (1958) considers that all stages move into the upper strata at night when seeking their prey. Positive results were obtained with the precipitin test for all three species.
The only known coleopteran predator of Conomelus in the field was 69.
Coccinella 7—punctata L. Various other members of the Coccinellidae also
occurred in the plot, these are listed in Section 3 page110. C.7—punctata
adults occasionally appeared in late May and June, but only reached their peak numbers in mid—July. The beetles fed voraciously on delphacid nymphs
and adults in the laboratory, and positive precipitin reactions were obtain—
ed from field samples. It was, however, observed that the Coccinellids usually remained on, or near, the flowering heads of the Juncus plants, and
that they did not wander down into the tussocks in search of prey; the beetles were apparently feeding on the nectar or pollen of these flowers.
Clausen (1940) considers that Coccinellids only feed on such plant material when animal food is scarce. Other beetles occurring in the Pond field plot included various species of Staphylinidae listed in Section 3, page 111. Most of these did not accept Conomelus prey in the laboratory, although one adult of Stenus impressus Gm. ate several first and second instar nymphs. These beetles generally inhabit the basal litter layer, and are not often found on the rush stems.
A total of 19 species of Carabidae were collected in the rush area
(listed on pagelio, Section 3). The most common of these were Feronia strenua (Panz.), F.madidus (Fabr.), Bradycellus harpalinus (Serv.) both
Dromius spp., and Leistus ferrugineus (Linn.). Most of these species fed on first. to third instar Conomelus nymphs in the laboratory, but no positive precipitin results were obtained with material from the field. It is, however, probable that predation did occur in the field, but too few beetles were tested to confirm this. Davies (1953) has shown that many carabids 70.
feed on a wide range of material; Feronia spp. may feed on Collembola and and other small arthropods, as well as plant matter; Leistus spp. feed on
Collembola and mites, while Nebria brevicollis, another species occurring in the plot,. may take Collembola, spiders, harvestmenl, and earthworms.
Cantharid larvae (species unknown) were often found on the litter layer in May. These are generally considered to be voracious predators, but did not readily feed on Conomelus in the laboratory; only one out of
40 larvae accepted a second instar nymph.
Various other insect predators were found in the plot: Collembola (Sminthur— idae) were observed feeding on first and second instar Conomelus nymphs in the field, and in the laboratory. No positive precipitin reactions were, however, obtained from field material.
Immature and adult specimens of Ectobius lividus (Fabr.) (Dictyop— tera Blattidae) appeared in Pond field from early July until mid—August; laboratory trials and the precipitin test showed that this species did not feed on Conomelus.
Three greenish—coloured syrphid larvae (Diptera : Syrphidae) were found in the rush area on the 4th July; of these, two reacted positively in the precipitin test. The species of syrphid was not identified.
Arachnids : Araneae.
A total of 38 species of spider was collected in Pond field, of which half were members of the Linyphiidae; the species are listed in
Section 3, page113.
Gnaphosidae and Clubionidae : Drassodes (Gnaphosidae) species were found 00 71. adults throughout the season, with immature stages appearing in late summer.
Clubionids were far more abundant, particularly the immature stages which were found from May until mid-June. Both groups of spiders inhabit the litter layer, and are only rarely found in the upper parts of the rush clumps. Positive precipitin results were obtained with immature and adult
Clubionids from the field, and the spiders fed readily on Conomelus nymphs and adults in the laboratory.
Thomisidae: Adults of Xysticus crIstatus (Clerck), X.lanio C.L.K., and
Tibellus oblongus (Valck) were common in Pond field from May to July, with the immature forms appearing from mid-July onwards. These spiders frequent the upper strata of the vegetation, and, therefore, prey mainly on the later instars and adults of Conomelus, rather than the early stages which only occur lower down. All three spiders fed on deiphacid nymphs and adults in the laboratory, and positive precipitin results were obtained from field samples.
Salticidae: One adult of Evarcha arcuate (Clerck) was collected, this gave a negative precipitin result.
Lycosidae: 6 members of this family occur in the rush area (listed in
Section 3, pageill); of these the most abundant are the Lycosa species;
L.pullata (Clerck) forms 50% of the Lycosa population, L.prativaga L.Koch
27.6%, L.nigriceps Thor. 13.8%, and L.amentata (Clerck) 8.6%. Adult Lycosids were found from late April until August, with the immature stages appearing in August; Lycosids were generally found in the open areas of the rush plot, and only occasionally within the tussocks; however, it was clear that 72.
Conomelus nymphs and adults were often taken in the field, as about 38% of the Lycosid samples reacted positively in the precipitin test. Immature and adult spiders fed voraciously on most stages of the delphacid in the laboratory.
Pisauridae: Pisaura mirabilis (Clerck) was the commonest member of this family, and was only found in the upper parts of the rush tussocks. The spider occurs as the adult from early May until mid-September, with the immature stages hatching in August. Both juveniles and adults feed on most stages of Conomelus in the laboratory, but in the field the spiders probably only take late instars and adults as they inhabit the upper zone of the vegetation, Field samples reacted positively in the precipitin test.
Mimetidae: one adult of Ere cambridgei Kulcz was taken in August 1960.
Theridiidae: Three members of this family were found in Pond field (see
Section 3, page113), of which Theridion bimaculatum (Linn.) was by far the most abundant. This species builds small 'scaffold' webs among the bases of the rush stems; none of the many webs examined contained Conomelus nymphs or adults, and no positives were obtained with the precipitin test. In the laboratory the spiders feed readily on second and third instar nymphs.
T.bimaculatum adults were most abundant from May until July.
Tetragnathidae: Two of the nine British species were found in the area,
Tetragnatha montane Simon and Pachygnatha clercki Sund. The former is a web-builder and inhabits the upper regions of the rush tussocks, whereas
P.clercki is a hunting spider and is found on the litter and among the stem bases. P.clercki adults were found from May until September, and T.montana 73.
from May until July. Although both species fed on most stages of Conomelus
in the laboratory, positive precipitin results were only obtained for P.
clercki (over 40% positives in 1961).
Argiopidae: The only member of this family taken frequently in Pond field
was Araneus cornutus Clerck. The webs of this species were found among the upper parts of the rush tussocks, and although these never appeared to trap any Conomelus nymphs or adults, one specimen reacted positively to the
precipitin test.
Linyphildae: 17 species of Linyphiid were collected in Pond field (see list,
Section 3, page113); this list is probably by no means exhaustive as many juvenile and sub—adult spiders could not be identified. Linyphiids were abundant and formed an average of 75.4% of any spider sample collected in the plot. The commonest species, in order of importance, were: Lepthyphantes tenuis (Bl.), Oedothorax gibbosus (B1.), Pocadicnemis pumila (B1.),
Hypomma bituberculatum (Wid.), Dismodicus bifrons (B1.), and Linyphia clathrata Sund. Of these species, L.tenuis and L.clathrata are web— builders, and are generally only found in the upper zone of the vegetation; the other species inhabit the litter layer, and may be found among the stem bases. All these Linyphiids occur as adults throughout the spring and smog". months. The precipitin test showed that L.tenuis and L.clathrata frequently fed on Conomelus in the field; only one positive was obtained from the ground—zone Linyphiids (H.bituberculatum) in spite of the fact that these species fed readily on first to fourth instar nymphs in the 74.
laboratory; this was probably related to the small number of specimens that
were successfully tested in 1961, because of the decomposition of the
antiserum (see Section 3, page 84?.
Opiliones.
Eight species of harvestmen were taken in the present study (listed in Section 3, page114); the commonest of these, in order of abundance, were Oligolophus agrestis (Meade), Platybunus triangularis (Hbst),
Leiobunum blackwalli Meade, and Mitopus morio (Fabr.). Immature stages of
0.a9restis were abundant from May until the end of July, when the first adults appeared; this species generally occurred on the basal litter and among the stem bases; Bristowe (1949) and Todd (1949) consider that these harvestmen, and other species, migrate into the upper strata of the vegetation at night, when the activity and predation peak is reached. In the present study, however, immature and adult specimens of Oligolophus were often seen feeding on Conomelus nymphs and adults during the day; other prey included Tytthus and Cyrtorhinus nymphs. Todd (1950) records
Kelisia fasciata (Kirschbm.) (Delphacidae) as the prey of 0.tridens
(C,L.Koch). Up to 40% of the Oligolophus samples from the field were shown to contain Conomelus antigen by the precipitin test. Platybunus adults were common in the plot from early May until July, and were again inhabitants of the basal litter layer. This species was often observed feeding on Conomelus nymphs in the field, which was confirmed by the precipitin test. Immature specimens of Leiobunum were found in the rush area from early May until August, the first adults appearing from late
June onwards; these very active harvestmen occurred both in the ground 75.
zone and in the upper regions of the rush tussocks, In the laboratory
Leiobunum spp. did not readily feed on any stage of Conomelus, but the
high percentage of positives obtained with the precipitin test (up to 45%)
confirmed that delphacid prey was often taken in the field. The records
in the literature show that Leiobunum, as well as Oligolophus, will take a
wide range of food material including jassids, aphids, inchneumonids,
lepidopterous larvae, numerous Diptera, woodlice, earthworms, and bird
droppings — Bristowe (1949) and Todd (1950).
Immature stages of Mitopus morio were first seen in early May,
and adults were found from mid—June until the end of September. This
harvestman is the most voracious of all the species present in the area,
and is extremely active, occurring in all strata of the vegetation; only
two specimens were tested for the precipitin reaction, and both gave
positive results. M.rnorio again accepts almost any food material, and its
feeding behaviour has been studied in detail by Phillipson (1960).
Acarina.
Two species of predatory mite were common on the basal litter
layer and among the rush stem bases, Eugamasus cornutus (Canestr.) and
Pergamasus crassipes (L.) (Parasitidae); these mites were occasionally
seen feeding on first and second instar Conomelus nymphs in the field, but no positive results were obtained with the precipitin test. 76.
SECTION 3.
Population Studies.
Methods of sampling the population.
Eggs.
To estimate the mean number of eggs per stem, a sample of 40 to
60 stems was collected at the end of the egg—laying period, in October.
To ensure that no oviposition sites were overlooked, the stems were cut
below leaf—sheath level, and kept in polythene bags until they could be
examined in the laboratory. To estimate the number of eggs per rush, the
stems were cut into sections, pinned onto a wax surface, and the epidermis
of each rush was stripped off, exposing the eggs embedded in the pith. The
mean number of stems containing eggs, per square foot, was estimated by
taking 50 x 1.8 inch diameter cores, counting the number of stems in each
sample, and converting these figures to numbers per square foot. This
estimate was multiplied by the mean number of eggs per stem to give the
.total egg population per square foot.
Nymphs and adults.
Most of the difficulties of sampling Conomelus nymphs and adults
arise from the growth habit of the Juncus plant, and the distribution of
the delphacids within the tussock. The rush tussock generally consists of
a mass of stems (up to 20 stems per square inch), woody rhizomes, and a
considerable amount of basal litter. The stems often grow to a height of
36 inches or more, and tend to collapse towards the end of the year, forming
a dense mat of dead material, up to 12 inches thick. This mainly occurs 77. in long—established Juncus habitats found in waterlogged areas such as Pond field. These features of the rush plant, and the presence of surface water or water—logged soil, constitute the main physical difficulties of sampling this habitat.
Other sampling difficulties are introduced by the variable distribution of Conomelus nymphs and adults within the rush mat or tussocks.
The first to third instar nymphs only feed in the basal regions, 6 inches above the leaf—sheaths, while the fourth and fifth instars, and adults, generally feed at any point on the stems. Sampling methods that involve sweeping and beating are, therefore, not practicable, as only those insects that inhabit the upper strata of the vegetation are obtained. The frequency distribution of delphacids in the samples was far from random, and could not be fitted to a poisson distribution (see Appendix Table 1). This aggregat— ion of the population accounts for the wide variation in the sampling data, shown by the 95% fiducial limits in Tables 32 and 33 (pages 10 and 91 ).
The only practicable method for sampling both nymphs and adults is to use a metal corer that is forced through the mat of vegetation into the soil below, removing a standard—sized core of soil and rush. Initially, in
1960, a corer, 5 inches in diameter, with a saw—toothed edge to assist cutting, was used. 30 to 40 minutes were required to examine each of these cores, which restricted the number of weekly samples that could be taken to about 20. It was decided, therefore, to reduce the size of the sampling unit and to take more samples each week. In 1961 a steel water pipe,
1.8 inches in diameter and 5 feet in length, was used as a corer. One edge 78.
of the pipe was sharpened to assist cutting through the rush mat. The
samples were removed by passing a metal rod through the corer, and were
placed in polythene bags for further examination in the laboratory. 50
cores were taken weekly from random sites throughout the plot, from early
May until the end of August. The sampling method could be considered non—
random in that Conomelus is never found in areas of bare ground or on plants
other than J.effusus, and only the latter was sampled. However, the rush
samples, themselves, were taken at random from the Juncus stands, i.e. at
fixed points along straight lines through the plot. In 1960 an attempt
was made to use the vacuum sampler for extracting nymphs and adults in the
field. This method was unsuccessful for three main reasons; firstly, the
large quantity of litter and other debris in the habitat generally filled
the sampling bag and the hose of the machine after only 30 seconds;
secondly the nozzle of the sampler could not be forced down into the base
of the rush clumps to extract the early nymphal stages, and, finally, the
vacuum sampler could not be used in the water—logged parts of the plot.
The delphacids were extracted from the cores by flotation.
Initially, magnesium sulphate solution was used, but this was later discarded
as excessive quantities of plant debris were extracted with the insects.
Attempts to use the Salt and Hollick apparatus to remove this plant matter proved unsuccessful as the soft—bodied delphacids were completely destroyed
during the process. Various other methods of separating animal and plant material were used, including mixtures of xylene or benzene and water.
These methods were also impracticable as the delphacids, particularly the 79. nymphs, tend to adhere to the litter below the interface of the two liquids.
The hoppers were extracted, finally, by simply using water, in which relat— ively little plant debris rose to the surface. Before removing the cores from the polythene bags a few drops of ethyl acetate were added, to inactiv— ate any delphacids present. The cores were placed, individually, into a glass tank (12 x 12 x 12 inches), and were broken up and agitated under water until all hoppers were extracted; a period of about 20 minutes was generally required for complete e;:traction. The insects were removed from the water surface with a fine brush, and placed into 70% alcohol for further examination and separation into instars.
Some tests were carried out to estimate the number of Conomelus nymphs and adults obtained by the coring and flotation methods. 20 cores, each 1.8 inches in diameter, were cut from a rush clump and re—set into an area of bare soil. 50 first instar nymphs were released on the rush stems arising from each core, and were then sampled in the usual manner, forcing the pipe sampler down over the core; the hoppers were extracted by flotation in water. The tests were repeated using 10 adults per core. The mean percentages extracted were 38.0% of the first instar nymphs, and 80% of the adults. The work on nymphal measurements show that the increase in size throughout nymphal life follows a near—geometrical progression. It has, therefore, been assumed that the extraction rates for the various nymphal instars follow a similar pattern — Table 29.
80.
Table 29. Numbers of each instar extracted by core and flotation method.
Instar % extracted
First 38.0
Second 46.4
Third 54.8
Fourth 62.2
Fifth 71.6
Adult 80.0
These percentages are used as correction factors when estimating the
population of nymphs and adults in the field. Whalley (1958) also used a
water—flotation technique for the extraction of delphacids, but obtained a
much higher extraction rate of 77.6% for first instar nymphs (84.4% for
adults). This difference is probably due to the fact that Whalley used a
6 inch corer from which the sample could be removed by shaking, or slight
pressure from above. The use of the pipe corer and extraction rod in the
present study led to considerable compression of the sample and the loss of
a larger number of nymphs. However, as discussed on page77, large corers
are of little use in sampling as relatively few samples can be dealt with at
a time. Whalley's population estimates are subject to much uncertainty as
only 3 to 6 cores were taken on each sampling day, giving estimates whose
limits were ± 85 — 97% of the mean. The 95% fiducial limits still show a
considerable range when, as in the present study, 50 cores are taken on each
occasion. 81.
Estimations of the total hatch of first instar nymphs were made by
the regression technique of Richards and Valoff (1954). This method is based on the assumption that there is a constant mortality of all stages
following hatching. The decline of the population will fit the formula
Y = nKt, where Y is the population on day t, n is the total number of nymphs
that emerge, and K is the fraction of the population that survives each day.
After the population has passed its peak, logarithmic values of Y should
follow a straight line as Log Y = log n t log K. A linear regression equation can be fitted to these logarithmic values of successive population estimates, combined with the corresponding day number t. By substituting a value of t that corresponds to the first hatching date an estimate can be made of the initial number of first instar nymphs. Adult emergence was calculated in a similar manner. Only rough estimates were obtainel ,by this technique as there was some evidence to indicate that mortality was not constant for all stages; the precipitin reaction appeared to show that more predation occurred in the fifth instar and adult stages than in the third and fourth instars (no estimates were obtained for the very early stages.
Sampling of the predators.
Few predators were obtained by the core and flotation methods as not many species floated in water. The difficulties of sampling the Juncus habitat have been described on page 77 and also apply to sampling of the predator populations. Attempts were made to estimate predator numbers by extracting all the insects from within 12 and 24 inch metal quadrats. In 82. practice it was found impossible to force the quadrats through the mat of rushes, even with the aid of a sledge-hammer, as excessive compression of
the vegetation occurred. The only practicable method was to cut blocks, measuring 6 x 6 x 6 inches, from the rush with a sharpened spade, and then
to sort through each sample on a white tray in the field. Most predators were extracted from a single sample in about 20 - 30 minutes, and 12 - 15 samples were collected at intervals of approximately 14 days. A successful attempt was made to sample an active low-density predator, i.e. Lycosid spp., by a marking and recapture method. The spiders had to be captured individ- ually by hand as pit-fall trapping was unsuccessful, and were marked on the cephalothorax with Britfix colour dope (Appendix Table 2).
Estimation of parasitism.
Parasitism of Conomelus nymphs and adults was estimated by dissect- ing samples of delphacids at regular intervals throughout the season.
Estimation of predation,
In order to estimate predation in the field, the following facts must be obtained:
(1) The number of predators present throughout the season; estimations
of predator populations are dealt with on page %I.
The number of predator species that regularly feed on Conomelus:
this requires the use of the precipitin test.
The percentage of the total population of each predator species
that regularly feed on Conomelus; this figure can be obtained
from the precipitin test. 83.
(4) The number of prey taken by each predator species in a given period
of time; these estimates are obtained by laboratory feeding tests.
The Precipitin test.
This test provides an estimate of the percentage of predators in
any sample that have recently fed on Conomelus in the field, and still
contain prey material in their bodies. The precipitin reaction is based
on the interaction between Conomelus antigen in the gut of the predator
and Conomelus antibodies produced in rabbit serum.
Preparation of the Conomelus antibodies (antiserum).
The antiserum was very kindly prepared by Dr.Dempster who has given me details of the method employed. A total of 7000 Conomelus adults were starved for 24 hours to remove all Juncus material from their bodies; this was necessary to prevent any reaction with certain predators, such as mirids, which may occasionally take plant food. The delphacids were then killed, and the procedure described by Dempster (1960) was followed. The insects were pulverised in a pestle and mortar with 0.9% saline solution; this was then centrifuged, sterilised, and freeze—dried. The antigen was then dissolved in distilled water, and 0.4% potassium alum was added to bring the soluble proteins into suspension. After adjusting the pH of the suspension to 6.8, 2.5 ml. were injected intramuscularly into the hind leg of a rabbit. After 14 days blood was removed from the ear of the rabbit, and the serum was tested against a standard Conomelus extract. Repeated injections of the antigen suspension were given until the rabbit serum contained a sufficient number of antibodies to react with 1 : 3000 dilution 84.
of Conomelus extract. In 1960 antiserum sensitive to 1 : 3000 extract was
prepared, while the 1961 antiserum was initially sensitive to 1 : 4000;
the latter, however, deteriorated for an unknown reason to 1 : 10 sensitivity,
and led to the failure of almost all the 1961 serological work. A few
positive results were obtained later in 1961 by using 1 : 900 serum, which
was prepared by pooling less sensitive material.
Conomelus antiserum was tested for specificity against other leaf—
hopper species that occurred frequently in Pond field, particularly jassids.
The reaction to jassids was only about 1/300 as strong as that to Conomelus;
there was also no reaction with Juncus extract.
Collection of predators for testing.
Samples of predators were collected at regular intervals, and
identified, as far as possible; to species and instar. Whole smears were
made on filter paper of the smaller predators, such as mirids and small
spiders, and gut smears were made of carabids$ coccinellids and large
spiders. The smears were numbered end indexed, and then dried over
phosphorus pentoxide. These smears can be kept for up to two years without
deteriorating, and most tests were carried out in the winter months.
The Precipitin test.
The smears were cut from the filter paper and placed into 1 ml. centrifuge tubes together with normal saline solution; 0.1 ml. of saline was added to small predators, such as mirid nymphs and mites, and smears of larger predators were extracted with 0.2 ml. The tubes were then left for
24 hours at —4°C. to complete the extraction. 85.
Tests were carried out with a Multiple Dispenser apparatus; 0.02
ml. of smear extract was drawn up into a fine capillary tube, followed by
an equal volume of antiserum. The liquids were drawn up into the middle
of the tube, which was then sealed by forcing one end into a plasticine tray.
After leaving the tubes for two hours, they were examined in indirect light
against a dark background. Conomelus antigen in the gut of a predator
combined with the antibodies of the antiserum to form a precipitate; this
was visible as a white ring at the interface of the two liquids.
Feeding rates of the predators.
Trials were carried out in the insectary to determine the number
of prey taken per day by known predators of Conomelus, and also to discover
whether other unconfirmed predatory species would accept delphacid prey.
Various types of container for predator and prey were used, depending on
the size of the predator. Coccinellids, mirids, immature nabids and small spiders were kept in 3 inch diameter plastic dishes, while 5 inch petri— dishes were used for larger predators, such as harvestmen and Lycosid spiders.
Both types of dish contained a thin basal layer of plaster of paris which was occasionally moistened to maintain the humidity. Various stages of
Conomelus together with a few pieces of rush stem were placed into the dishes, and daily observations were made on the numbers taken by the predators.
More delphacids were added regularly to maintain prey levels. The results obtained under these artificial conditions are only rough estimates of what occurs in the field, but give an indication of the readiness with which different predators take the delphacid. 86.
Population estimates.
The population estimates for eggs, nymphs and adults, are given
in terms of numbers per square foot, and not the entire plot; to convert
the numbers per square foot to total population estimates, all figures have
to be multiplied by 11,403.
Numbers of eggs.
The numbers of eggs per square foot are calculated from the prodUct
of the mean number of eggs per stem and the mean number of stems plus eggs per square foot in Table 30.
Table 30. Numbers of eggs per square foot 1959 — 61.
Year Total eggs No.of stems Mean no.of eggs Mean no.of stems Total eggs per stem plus eggs per per sq.ft. sq. ft.
1959 3,643 36 101.19 13.63 1,379.22 (estimated)
1960 28,280 106 266.79 14.80 3,948.49
1961 15,867 45 352.60 12.45 4,389.87
95% Fiducial limits
1959 1,026.97 — 1,731.47
1960 3,049.10 — 4,898.50
1961 3,602.77 — 5,176.97
No estimates were made in 1959 of the number of stems plus eggs per unit area; the figure of 13.63 in Table 30 is purely hypothetical, and is the mean of the estimates for the two other years. The figures show that the 87.
number of eggs laid per square foot has increased each year, although it
must again be borne in mind that the 1959 egg population is based on a
doubtful stem—count estimate.
Numbers of nymphs.
Hatching occurs in May, and a hatching curve for an insectary
culture is shown on Plate 15; no curves were obtained from the field,
although the relative numbers of first instars give some indication of the
emergence rates. In 1961 hatching first occurred on the 1st of May, 9
days earlier than in 1960; this difference may have been due to the mean
temperatures for the previous months, i.e. in 1960, 43.8°F. (6.5°C.) in
March and 48°F. (8.9°C.) in April; corresponding temperatures for 1961
were 46°F. (7.8°C.) in March and 50°F. (10°C) in April. In the field
very few first instars were taken after the 10th June (1960) or 15th June
(1961); the hatching period was, therefore, 32 days in 1960 and 46 days in 1961. The hatching peaks in both the insectary and the field, reached
a similar period of time i.e. 14 days (55°F., 13°C.) in the insectary, and in the field, 14 days (54.87°F., 12.7°C.) in 1960, and 16 days (52.870F,
11.6°C.) in 1961; the data shows that the length of time required to reach
the peak is closely related to the temperature over this period. The duration of the entire hatching period is probably also related to the spread of the oviposition of the females in the previous year; peak oviposition occurred over a period of about 19 days in 1961. A few females, however, lay eggs for up to four weeks after most oviposition has finished, and these eggs may account for the first instar nymphs that have 88.
occasionally been found as late as mid—July.
Estimates of the total number of hatching nymphs can be obtained
from the egg population figures, after allowing for egg mortality, and also
by using the regression technique of Richards and Waloff (1954). The
estimates are given in Table 31.
Table 31. Estimated hatch of nymphal stages.
Year No.of eggs % egg mortality Expected hatch Estimated hatch per sq.ft. (regression method)
1959-60 1379.2 27.47 1005.2 836.2
1960-61 3948.5 37.33 2474.6 1124.4
Although the expected and estimated hatches for 1959 appear to be similar, it is necessary to add that the number of eggs per square foot is based on some arbitrary data (page $(), and also that the egg mortality is slightly underestimated as autumn mirid predation was not recorded. In 1961 3185 nymphs hatched from 22 stems in the insectary, an average of 144.7 per stem compared with the expected hatch of 167.2 eggs per stem; the difference in the two values falls within the normal variability of the stem samples, i.e. 95% fiducial limits of 23.27% of the mean; this confirms that the estimations of egg mortality are fairly accurate.
Numbers of each instar.
The numbers of each instar present throughout the season were estimated from the core samples. All figures were corrected by the extraction factor (Page $0) and converted to numbers per square foot, The nymphal data is given in Tables 32 and 33 and Plates 16 and 17. Discussion 89. of these data will deal mainly with the relative aspects, such as the fall— off rates of the various instars; the large fiducial limits of the estimates make detailed consideration of the absolute population numbers virtually impossible.
The numbers of first instars reached their peak of over 700 per square foot in 14 — 16 days, in both 1960 and 1961. The steep fall—off between the first and second instar stages is also similar in both seasons; in 1961 another considerable reduction of the population took place between the second and third instar. The mortality of stages following the third instar is relatively low — Plates 16 and 17. Although in 1960 the peak numbers of third, fourth, and fifth instars were generally three times as high as the corresponding estimates for 1961, the differences in the total numbers of adults produced were not as great; this was due to the fact that each instar emerged over a longer period in 1961. In 1960 it will be noted that the maximum number of third instars (130.22) is less than the corresponding figure for the fourth instar (151.74); similarly in 1961 the fourth instar peak (43.07) is lower than that of the fifth instar
(48.29). These differences are probably due to sampling errors, although it is also possible that, in 1960, the nymphs passed so rapidly through the third instar that the peak was missed by taking only weekly samples.
The data in Tables 32 and 33 also show that third and fifth instar nymphs must have been present in the plot before their first appearance in the samples, as third and fourth instars first appear together in 1960, and fifth instars and adults also appear simultaneously in 1961. This was 90.
probably due to the infrequency of sampling (weekly) and the relatively small number of samples taken (50 cores per week). Numbers of adults. The data in Tables 32 and 33 indicate that the peak number of adults in 1960 was considerably higher than in 1961, and that the adult mortality was greater in 1961. In comparing the figures for the two seasons it is, however, necessary to bear in mind that the fiducial limits
Table 32. Numbers of each instar per square foot, in 1960.
Date Instar Adult Total 95% fiducial I II III IV V limits 13/5 118.16 118.16 8.05- 228.27 23/5 667.7.4 76.00 743.74 134.40-1353.08 30/5 145.97 161.33 33.84 2.79 343.93 170.14- 517.72 10/6 34.67 125.33 130.22 82.33 19.69 392.14 311.66- 472.62 29/6 1.83 24.31 83.63 151.74 75.81 337.32 63.86- 610.78 5/7 1.83 6.08 36.03 43,51 107.32 194.77 78.76- 310.78 13/7 5.00 24.31 31.02 150.94 16.65 227.92 132.63- 323.21 20/7 3.68 27.37 26.44 57.49 35.36- 79.62 27/7 1.20 5.91 0.88 27.95 90.85 126.79 75.71- 177.87 3/8 0.88 19.69 58.77 79.34 52.68- 106.00 10/8 1.21 0.71 100.81 103.09 20.93- 185.25 17/8 1.07 45.02 46.09 25.16- 67.02 26/8 78.34 78.34 58.84- 97.84 91. Table 33. Numbers of each instar per square foot in 1961. Date Instar Adult Total 95% fiducial I II III IV V limits 1/5 18.56 18.56 0 - 37.12 4/5 79.02 79.02 38.70- 119.34 8/5 488.74 488.74 130.73- 846.75 12/5 409.38 409.38 309.42- 509.34 16/5 739.35 50.57 789.92 504.27-1075.57 24/5 231.86 91.15 323.01 171.07- 474.95 7/6 210.34 200.33 51.72 4.48 466.87 263.57- 670.17 15/6 74.56 44.88 57.95 1.79 183.36 100.80- 265.92 22/6 11.95 43.98 51.72 30.50 9.51 1.41 149.07 89.41- 208.73 28/6 2.97 19.54 37.25 43.07 15.83 118.66 40.53- 196.79 6/7 2.43 12.43 20.64 26.11 7.09 68.70 33.36- 104.04 13/7 2.43 8.28 12.57 48.29 12.75 84.32 53.84- 114.80 20/7 2.07 14.35 39.57 19.84 75.83 38.67- 112.99 28/7 1.79 30.08 42.51 74.38 42.31- 106.45 4/8 1.79 3.17 32.60 37.56 21.96- 53,16 11/8 3.17 23.39 26.56 6.84- 46.28 20/8 1.79 19.84 21.63 10.32- 32.94 28/8 9.91 9.91 0 19.82 92. of the 1960 samples are much Treater than those in the following year.
Estimates of the total number of adults produced in the field were obtained by the regression technique of Richards and Waloff (1954). The steeper
fall—off and longer emergence period of the adults produced an estimate of
2903 per square foot in 1961; the 1960 figure was 241.1 adults per square
foot; the latter was only a very rough estimate as the adult mortality
rate did not appear to be constant (the regression method is based on the
assumption that mortality is steady).
Estimation of mortality.
Egg mortality.
The factors causing overwintering mortality of Conomelus eggs fall
into four main groups: (1) Insect parasites.
(2) Insect predators.
(3) Fungal parasites.
(4) Unknown factors.
All four factors may act simultaneously on the egg population, although the relative importance of each individual factor may vary considerably from
sample to sample.
Insect parasites: Conomelus eggs are parasitised by two species of
Hymenoptera, Anagrus incarnates (Haliday) (Mymaridae) and Aprostocetus mandanis (Walter) (Eulophidae); the biology of these parasites is discussed in Section 2.
The percentage of eggs parasitised by Anagrus was very low in both seasons, 0.32% in 1959-60, and only 0.12% in 1960-61 (Table 34). Whalley
(1958) also records a figure of only 1% parasitism in Conomelus eggs. This 93. clearly indicates that Conomelus is not the primary host of the mymarid, as has previously been thought by Bakkendorf (1934). In another part of
Silwood Park, known as Marsh Bottom, 2.55% of Conomelus eggs were parasitised by Anagrus. In this area Juncus occurs as isolated tussocks among a variety of grass species and other plants; this habitat probably contains a greater number of the primary hosts of the mymarid, leading to a higher parasitism of Conomelus itself; these alternate hosts may be other species of delphacids or jassids.
Table 34. Numbers of eggs parasitised by Anagrus in Pond field.
1959-60 1960-61
Total eggs examined 3643 16,651
Number containing Anagrus 12 12
parasitism 0.33 0.12
In stems containing relatively few eggs (less than 100) up to 70 of the eggs may contain Anagrus larvae.
Aprostocetus is of far greater importance as an egg mortality factor than the mymarid. Although only one to two larvae are generally present per stem, the parasite becomes predatory after completing its endo- parasitic existence within the Conomelus eggs, and consumes, on average, a further 7 eggs; the biology is discussed fully in Section 2. The data for two seasons is given in Table 35. 94.
Table 35. Numbers of eggs destroyed by Aprostocetus.
1959-60 1960-61
Total number of larvae recorded in samples 60 199
Mean number of larvae per stem 1.67 3.21
Total number of eggs examined 3643 16651
Mean number of eggs per stem 101.2 266.8
Total eggs destroyed 479 1588
% destroyed 13.15 9.54
The results show that a higher percentage of eggs was destroyed in 1959-60 than in 1960-61; this was partially due to the fact that the ratio of total available eggs to Aprostocetus larvae was lower in 1959-60 than in the following season. In some rush stems Aprostocetus larvae destroyed up to 32% of the eggs, i.e. 36 larvae in one stem consumed 278 eggs out of a total of 872.
The increase in predation by Aprostocetus, throughout the winter months, was estimated by examining stem samples at the beginning and the end of the winter period. The data obtained from these samples are given in Table 36.
Table 36. Winter and spring predation by Aprostocetus.
Period Mean no. of % larvae killed Total eggs Total % destroyed Aprostocetus by fungus examined destroyed per stem
November 2.75 3.51 11,629 484 4.16 1960
April 3.21 56.78 16,651 1,158 9.54 1961 95.
The figures in Table 36 show that there was an increase in the mean number of larvae per stem betv:een November and April; this was probably due to
the continuing emergence of larvae from the eggs as newly—hatched larvae were recorded from September until March, The increase in larval numbers leads to greater predation of the eggs. The data in Table 36 also show that over 50% of the larvae are killed by a fungus in the spring months; as most of these larvae are, however, only killed in the mature stage, there is no great effect on the relative numbers of Conomelus eggs destroyed.
Insect predators: The only predator seen to destroy Conomelus eggs in the field was the mirid Tytthus pygmaeus Zett. Another mirid Cyrtorhinus caricis (Fallen) also occurred in Pond field, and is said to prey on eggs
Southwood and Leston (1959), Whalley (1958); this was not confirmed in the present study (the biology of both mirids is discussed fully in
Section 2). Eggs sucked by Tytthus are readily distinguished from those destroyed by Aprostocetus, as the chorions of eggs sucked by the former remain intact, whereas Aprostocetus tends to mutilate the egg shells leaving only the opercular caps. There are two periods of predation by
Tytthus; the first occurs in May when the mirid nymphs emerge and feed on the overwintered Conomelus eggs. This source of food, however, rapidly decreases as the delphacid nymphs are also hatching throughout this period.
No estimates of spring predation were obtained from the stem dissections as it was not always possible to separate egg shells from which nymphs had hatched from those which had been sucked. However, the mirid population reached a maximum of 71.9 nymphs per square foot in 1960, and 111.9 per 96.
square foot in 1961, which indicates that a large number of eggs may have
been taken; this predation may account for part of the discrepancy between
the expected and estimated hatching figures (page S8 ). In 1961 an attempt
was made to estimate the spring predation by the precipitin test. Unfortun—
ately the antiserum used in testing 573 mirids had decomposed, and gave no
positive results. One positive reaction was obtained in 34 tests using a
small quantity of 1 : 800 strength antiserum, and represented 2.94% of the
total mirid population. The predation estimate was obtained from the
product of the following: the mean number of mirids present each day during
the predation period (40.12 per square foot), the percentage of the
population containing Conomelus antigen (2.94%), the number of eggs sucked
per day (minimum estimate, 2), the duration of the predation period (31
days); the number of eggs taken by Tytthus was, therefore, only 73.13 per
square foot, i.e.( 1.85% of the total number present. In view of the weak
antiserum used in the test, this figure is almost certainly an underestimate.
No spring predation tests were carried out in 1960, as the use of this
technique was not suggested until mid—June of that year.
The second period of Tytthus predation takes place in late summer,
from August onwards, when the Conomelus females begin to oviposit.
Estimates of this summer predation are given in Table 37.
Table 37. Summer predation by Tytthus (from stem dissections).
1960-61 1961-62
Total number of eggs examined 16,651 15,867
Total number sucked 1,192 2,454
sucked 7.16 15.46 97.
The difference in predation between the two seasons is probably partly
due to sampling errors, but also to the greater numbers of Tytthus adults
present in 1961; the mean daily population of mirids during the 1961 predation period was 18.9 per square foot, and in 1960 the figure was 12.0 per square foot.
In the 1961 summer, samples of rushes were dissected at regular intervals throughout the delphacid oviposition period to estimate the rate of predation — Table 38.1T444-7.
Table 38. Rate of Tytthus predation during the Conomelus oviposition
period in 1961.
Date Mean no. of eggs per stem % of eggs destroyed
21/7 0.10 100.0
29/7 1.39 69.0
2/8 4.92 80.8
9/8 21.52 86.3
16/8 53.33 71.3
23/8 94.60 49.4
1/9 196.73 42.8
7/9 245.00 38.2
13/9 249.00 17.6
20/9 292.00 15.5
Although the data in Table 38 is again affected by sampling errors, the general trend in predation can be seen. Initially the mirids destroy the 98. eggs almost as soon as they are laid, but as more Conomelus females reach maturity and begin to oviposit, the mortality falls off. In 1960 an attempt was made to confirm the stem dissection estimates by means of the precipitin test. The predation estimate was obtained by the calculation described on page %, i.e. the product of the mean number of mirids per square foot present each day (12.00), the percentage of the population, containing antigen (0.71%, i.e. 1 positive out of 141 tests), mean number of meals per day, when a single egg remains detectable for 12 hours (2 eggs per day), number of days during which predator and prey occur together (40 days).
The estimate obtained by this calculation is only 6.81 eggs per square foot, i.e. 0.49% of the total egg population; this figure bears no relationship to the figure of 7.16% obtained by stem dissections. There was no obvious reason for this difference as the antiserum used in the 1960 work was sensitive to 1 : 3000 Conomelus extract. It is possible that assimilation of the antigen occurs much more rapidly in mirids brought in from the field, as they are continually disturbed in transit. This could not be confirmed in 1961 due to the failure of the serological work. Autumn predation tests on 300 Tytthus adults were also unsuccessful for the same reason.
Fungal parasitism: The fungus infecting Conomelus eggs in the field was identified by Dr.M.F.Madelin as a Paecilomyces species, and was shown by him to be pathogenic rather than saprophytic (details in Section 2). It is not always easy to establish whether an egg has been killed by the fungus or invaded secondarily; in the laboratory it was observed that Paecilo— myces developed rapidly on sterile eggs and eggs that had been punctured, 99. but not completely sucked, by Tytthus. However, the number of sterile eggs never exceeded 1% of the total egg population, and few partially— sucked eggs were found in the field. It, therefore, appears that the fungus generally acts as a pathogen. The percentage of eggs killed by
Paecilomyces in two seasons is given in Table 39.
Table 39. Numbers of eggs killed by Paecilomyces.
1959-60 1960-61 1961-62
Total number of eggs examined 3,643 16,651 2,930
Total number of infected eggs 498 2,537 453
% of total eggs infected 13.66 15.24 16.58
Fungal attack was generally first noted at the end of August, and gradually spread through the egg population, as shown in Table 40.
Table 40. Development of Paecilomyces infection in 1961.
Date % fungal attack Total number Total number of of eggs examined eggs plus fungus
23/8 0.017 1161 2
1/9 9.40 2007 189
13/9 13.10 2544 333
20/9 16.58 2930 453
In 1960-61 the increase of the fungal infection throughout the winter and spring months was shown by dissecting a sample of stems late in November, and another just before the hatching period in spring — Table 41. 100.
Table 41. Increase of fungal infection during the winter and spring months.
November 1960 April 1961
Total number of eggs examined 11,629 16,651
Total number of eggs with fungus 859 2,537
% of eggs with fungus 7.39 15.24
The figure for 1961-62 was not estimated as the author had to complete the
project before this period; as the percentage infection in late September
1961 was, however, already 16.58% it is probable that a very high mortality
will have occurred by April 1962, as in 1960-61 the figure for fungal
mortality had doubled by the spring. It was mentioned earlier (page95)
that over 50% of the Aprostocetus larvae were killed, or infected second-
arily, by Paecilomyces, although most larvae had completed their feeding
at the time of death.
Unknown factors of egg mortality: In 1959-60 it was possible to class
egg mortality into the first three groups given above, as the causes of
mortality were readily apparent, (although no record was made of autumn
mirid predation). In 1960-61, however, a number of eggs were infected by
an unknown 'disease'. The contents of the eggs became milky, and later
assumed a variety of colours ranging from black to violet. Smears of the
egg contents showed that large numbers of rod-forming bacteria were present,
but it was not clear whether these had entered the eggs as primary pathogens,
or merely saprophytically. It is possible that some of the infected eggs
were sterile; these are generally smaller than fertilised eggs, and do
not contain a yolk body; in the laboratory these eggs were readily attacked 101. by Paecilomyces, but did not develop the bacterial 'disease' symptoms described above. The estimates of mortality by unknown factors are given in Table 42.
Table 42. Mortality by unknown factors in 1960-61.
Total number of eggs examined: 16,651
Number of eggs killed by unknown factors: 877
% killed by unknown factors: 5.26
Discussion of egg mortality.
The total egg mortality occurring in three seasons is given in
Table 43.
Table 43. Total egg mortality in 1959-60, 1960-61 and 1961-62.
Season Total mortality Paecilomyces Aprostocetus Anarus Mirid Unknown (%) (%) (%) (%) (%) factors (%)
1959-60 27.47 13.67 13.15 0.33 - -
1960-61 37.33 15.24 9.54 0.12 7.16 5.27
1961-62 32.05 16.24 - 15.47 -
The figures for all three seasons are underestimates as, firstly, autumn mirid predation was not recorded in 1960, secondly, spring predation was not estimated in any of the years, and finally, in 1961-62, egg mortality was only studied in the autumn, and did not include any data for Aprostocetus,
Anagrus, or the extent of Paecilomyces infection in the spring.
The data in Table 43 show that the egg mortality has increased over the three seasons (the 1961-62 figure is only based on autumn results
102.
and must certainly reach at least 40% by the following spring). This rise
in mortality is attributed mainly to the increasing incidence of fungal
attack and egg predation.
An attempt was, therefore, made to relate the outbreak and
development of Paecilomyces to climatic data. It would seem likely that
the rate of fungal development is most rapid in the warmer months, i.e.
September and October in the autumn, and March and April in the spring.
The climatic data and fungal incidence for these months is given in Table 44.
Table 44. Relationship between climate and incidence of Paecilomyces.
Season Mean monthly Mean monthly % mortality temperature °F. rainfall (mm.) of eggs
First egg Sept.-Oct.1959 55 37.3 (not recorded)
generation Mar.-Apr. 1960 45.9 26.9 13.67
Second egg Sept.-Oct.1960 52.5 127.8 7.00
generation Mar.-Apr. 1961 48.0 34.7 15.24
Third egg Sept.-Oct.1961 54.5 72.8 16.58 generation
The figures for the outbreak of the fungus in September to October (none
are available for this period in 1959) do not suggest any relationship
between climatic conditions and magnitude of the outbreak. The very
considerable rainfall and average temperatures of September and October 1960
should have led to a high incidence of the fungus, as these approached the
optimal growth conditions for most entomophagous fungi. In fact the data
show that the greatest incidence occurred in the drier autumn of 1961. 103.
Although the spring mortality figures for 1960 and 1961 seem to bear some relationship to the temperature and rainfall conditions, it is not possible
to draw any definite conclusions from the data as no estimate of initial mortality in September—October 1959 was obtained.
In view of the successful use of Tytthus mundulus for the biolog— ical control of the sugarcane leafhopper Perkinsiella saccharicida in the
Hawaiian Islands it was of some interest to examine the predatory performance of a British species of Tytthus. T.mundulus was introduced into the
Hawaiian Islands from Queensland in 1923, and within a few seasons had reduced the deiphacid population to well below the economic level; as in most reports of biological work done at this period, no actual quantitative data on the percentage of eggs destroyed is given. The relationship between the number of generations per annum of the mirid and the deiphacid seems to provide the answer to the success of the predators. T.mundulus may pass through 10 generations per annum while Perkinsiella only has 3 to
4 generations each year; large numbers of mirid nymphs and adults are, therefore, present throughout the season. These readily find and destroy the eggs laid by the delphacids. As the mirids reproduce more rapidly than the leafhoppers, the latter are unable to lay enough eggs to outweigh the predation effects, and large reductions of the deiphacid population occur. Isolated 'pockets' of leafhoppers remain, and these small residual populations enable the mirids, which would otherwise die out, to survive in low numbers. In contrast T.pygmaeus, studied in the present work/ was univoltine, as was Conomelus. In late spring when the maximum number of 104.
mirids were present, the delphacid eggs were already hatching and in the
autumn, when eggs were again available, the mirid population hal already
fallen—off considerably. This explains the relatively small effect that
T.pygmaeus has on Conomelus egg populations, compared to the extensive
control of Perkinsiella in Hawaii.
The higher autumn predation in 1961 was probably due to the larger mirid population present in that year, i.e. in 1960 the mean number of mirids present each day during ..he. predation period was 12.0 per square foot; in 1961 the figure was 18.9 per square foot. No estimates were made of spring predation (page 95) , but in view of the large numbers of mirids present in May (both seasons) it was assumed that a considerable number of eggs were destroyed; these losses may have accounted, at least partially, for the difference between the expected and estimated hatching data, i.e. in 1960 1005.2 and 836.2 respectively, and in 1961 2474.6 and
1124.4 (expressed as numbers per square foot).
The third most important mortality factor was the Eulophid,
Aprostocetus. This parasite destroyed 13.15% of the total egg population in 1960, and 9.54% in 1961; as there were fewer eggs per stem in 1960 than in 1961 (101.2 and 266.8 respectively) the presence of a constant population of Aprostocetus in both years would account for the higher parasitism figure in the first year. Mymarids had a negligible effect on the egg population in the present study; in Hawaii a mymarid, Paranurus optabilis, may pass through two generations in one generation of host eggs, and may cause a 90% mortality (Muir, 1921). 105.
Whalley (1958) has estimated the total mortality of C.anceps in
North Wales, and his figures for two seasons are given in Table 45.
Table 45. Mortality of C.anceps (Whalley, 1958).
Year % Total mortality % Aprostocetus % Anagrus % fungus % unknown (not ident.) causes
1953-54 48.3 6.8 23.8 17.6
1954-55 42.0 6.3 0.1 11.6 24.0
Whalley's estimates are higher than those recorded in the present study
(Table 43, page 101), but are based on the examination of rather small
samples (1150 eggs per season). The figures for mortality by unknown
causes are relatively high, but include autumn mirid predation which was not estimated separately; it is not known whether the unknown mortality
was also due to the 'bacterial' disease observed in the present work, as
Whalley does not describe the appearance of eggs included in this category.
In summarising the egg mortality that occurred in the Pond field plot, the main points are:
(i) Generally at least one third of the total egg population was
destroyed in all three seasons.
(ii) The most important mortality factors, were, in order of importance,
the fungus Paecilomyces, the mirid Tytthus pygmaeus, and the
Eulophid Aprostocetus mandanis.
(iii) The incidence of the fungus could not be related to climatic factors.
(iv) The extent of the predation by the mirids and eulophids was thought
to be governed by the size of their populations in relation to
the number of available eggs. 106.
Nymphal and adult mortality.
Attempts were made in 1960 and 1961 to discover the role of natural enemies as possible factors involved in the fall-off of the delphacid population.
Parasitism.
It was possible to estimate parasitism fairly accurately as large samples of delphacids for dissection were readily obtained from the field throughout the season. There are two insect parasites of Conomelus nymphs and adults, Pipunculus semifumosus (Kowarz) (Pipunculidae) and Elenchus tenuicornis (Kirby) (Strepsiptera : Elenchidae), of which the former is more important. The biology of both parasites is discussed fully in
Section 2.
Pipunculus generally kills the adult stage only, and only occasion- ally does the mature larva emerge from the fifth instar nymph; no mortality occurs in the earlier nymphal stages although these may contain primary larvae. The highest percentage of parasitised adults are found in July; in 1960 6.84% of the adults were parasitised on 23rd July, and in 1961 an estimate of 13.04% was obtained from samples collected on the 12th of the month. Adult delphacids containing mature pipunculid larvae are generally incapable of reproduction because of the damage to the reproductive organs
(Section 1).
Parasitism by the strepsipteron, Elenchus, was very low in both
1960 and 1961, i.e. 1.92% and 0.1% respectively. One third of the adults containing mature parasites are still capable of reproduction, and it is, 107.
therefore, clear that Strepsiptera have a negligible effect on the delphacid
population.
The incidence of fungal parasitism was also very low in both
seasons, the number of delphacids showing external symptoms of disease
represented less than 0.001% of the population. Dr.Madelin identified the
fungus as Entomophthora sphaerosperma Fres (details of the biology are given
in Section 2). Although percentage infection was generally very low, 4.$35% INA of samples of Conomelus adults taken between 1st and 8th of August/were found,
on dissection, to be diseased. These delphacids contained masses of hyphal bodies which had not yet broken through the body wall to form the character— istic conidiophores; Dr.Madelin considered that these bodies were the early stages of an Entomophthora sp. Although samples of Conomelus were dissected throughout the season, diseased individuals were only found over this short period in August; this seems to indicate that climatic conditions were optimal for the rapid development of the fungus at this time. However the climatic records for late July and early August do not differ significant— ly from the remaining data for these months. No similar outbreak was recorded in 1960.
The only other parasitic organisms associated with Conomelus were ectoparasitic mites, Leptus sp. (identified by Mr.Sheals). These had no adverse effects on the hosts, and only about 0.001% of the delphacid population carry them.
Mortality due to predation.
Obtaining estimates of predation in the present study was far more 108. difficult than evaluating parasitism on account of the number of facts that had first to be known. These included:
(1) The number of predatory species, and their life histories. (ii) The numbers of each predatory species that occurred during the
predation period.
(iii) The percentage of predators in any sample that had recently fed
on Conomelus, as shown by the precipitin test.
The length of time that a.Conomelus meal (various instars) remains
detectable by this test.
(v) The number of prey taken per day (various instars).
(vi) The availability of alternative food.
Numbers of predatory species.
A list of the predatory arthropods occurring in the Pond field plot is given in Table 46. The precipitin test provided the necessary information to show which species actually fed on Conomelus in the field; laboratory feeding trials were carried out with predators that did not appear to feed in the field, to discover whether they would take the delphacids under less favourable conditions, i.e. lack of alternative food.
A total of 91 species of predatory arthropods were identified, of which spiders formed the largest number, i.e. 38 species; not all spiders were identified, particularly the Linyphiidae. The remainder of the list consists of the carabids 19 species, Heteroptera 8 species, harvestmen
(Opilionids) 7 species, Staphyl'inids 6 species, Coccinellids 5 species,
Acarina 2 species, and the remaining groups 1 species each. The list of 109.
species is by no means exhaustive, but represents those predators which
might feed on various stages of the delphacid in the field. 29 of the 91
species were shown by the precipitin test or observation to feed on Conomelus
in the field; another 25 species fed on the delphacid in the laboratory
feeding trials. As the predation estimates, discussed later, will show,
only about 10 species are of any importance, these being members of the
Araneae, Opiliones, and Heteroptera. The life histories of the predators
are discussed briefly in Section 2.
Numbers of each predatory species.
Predator population estimates were obtained from the 'block'
samples (page 8?,) see Table 47. The results in this table clearly show
the extent of the sampling errors; these are partly due to aggregation of
the predators, and also to their activity, many individuals escaping while
the block samples are removed from the rush clumps. The 951) fiducial
limits of many of the predator estimates were as high as 70% of the mean.
In calculating the predation estimates, the figure for the mean number of predators present each day of the predation period only, was used.
Table 46. Predatory arthropods occurring in the Pond field plot.
Collembola Sminthuridae Sminthurus sp. ?g t1
Dictyoptera Blattidae Ectobius lividus (Fab.)
Dermaptera Forficulidae Forficula auricularia L.
Cont.
110.
Table 46 Cont.
Hemiptera—Heteroptera
Pentatomidae Zicrona caerulea (L.)
Nabidae Nabis flavomarginatus Schltz. 44 t
Dolichonabis limbatus Dahlb. * I-
Stalia major Costa W f Anthocoridae Anthocoris nemorum (L.) * t
Miridae Tytthus pygmaeus (Zett.) * 41"
Cyrtorhinus caricis (Fall.) * t-
Fieberocapsus flaveolus (Reut.) A(
Hymenoptera Vospidae Vespula sp.
Coleoptera
Coccinellidae Coccinella 7—punctata L. w +
Adalia bipunctata (L.)
Propylea 14—punctata (L.)
Cantharidae Cantharis sp. (larva)
Carabidae Leistus ferrugineus (L.)
Carabus violaceous L.
Nebria brevicollis (Fab.)
Clivina fossor (L.)
Bembidion unicolor Chaud.
B.lampros (Herbst.)
Cont. Table 46 Cont.
Trechus quadristriatus (Schr.)
Harpalus rufipes (Deg.)
Acupalpus dubius Schilsk t A.meridianus (L.) t
Bradycellus harpalinus (Serv.)
Feronia caerulescens (L.)
F.nigrita (Fab.) t F.diligens Sturm.
F.strenua (Panz.) t
F.madidus (Fabr.) t Agonum fuliginosum (Panz.)
Dromius guadrinotatus (Panz.)
D.melanocephalus Dej.
Staphylinidae Stenus bimaculatus Gyll.
S.impressus Germ. t S.fulvicornis Steph.
Philonthus sp.
Quedius fuliginosus (Gray.)
Tachyporus sp.
Diptera Syrphidae Species unknown (larva) *
Arachnida
Opiliones
Nemastomatidae Nemastoma lugubre (Willer) Cont.
112.
Table 46 Cont.
Phalangidae Leiobunum rotundum (Latr.)
L.blackwalli Meade *
Mitopus morio (Fab.)
Oligolophus agrestis (Meade) * t
0.tridens (C.L.Koch) Platybunus triangularis (Herbst.) * t Acarina
Parasitidae Pergamasus crassipes (L.) t Eugamasus cornutus (Canestr.) *
Araneae
Thomisidae Xysticus cristatus (Clerck) *
X.lanio C.L.K.
Tibellus oblongus (Walck.)
Lycosidae Lycosa prativaga L.Koch t-
L.pullata (Clerck) *
L.amentata (Clerck) 1..nigriceps Thor.
Trochosa ruricola (Deg.) * 1- Tarentula pulverulenta (Clerck)
Pisauridae Pisaura mirabilis (Clerck) W
Dolomedes fimbriatus (Clerck)
Gnaphosidae Drassodes sp.
Clubionidae Clubiona subtilis L.Koch Cont. 113.
Table 46 Cont.
Clubiona sp.
Zora spinimana (Sund.)
Salticidae Evarcha arcuata (Clerck)
Mimetidae Ero cambridgei Kulcz.
Theridiidae Theridion ovatum (Clerck)
T.bimaculatum (L.)
Robertus lividus (Blackw.)
Tetragnathidae Tetragnatha montana Sim.
Pachygnatha clercki Sund. fi
Argiopidae Araneus sp. prob. cornutus Clerck
Linyphiidae (not all species identified)
Dismodicus bifrons (Blackw.)
Hypomma bituberculatum (Wid.) * 1
Oedothorax gibbosus (Blackw.)
Lophomma punctatum (Blackw.)
Micrargus herbigradus (Blackw.)
Savignia frontata (Blackw.)
Erigone atra (Blackw.)
Bathyphantes gracilis (Blackw.)
Floriona bucculenta (Clerck)
Lepthyphantes tenuis (Blackw.) t Pocadicnemis pumila (Blackw.) Cont. 114.
Table 46 Cont. Leptorhoptrum robustum (Westr.) Linyphia triangularis (Clerck) L.montana (Clerck) L.clathrata Sund. w
CIE obs.qTved on Conomelus in field, or by precipitin test.
1 observed feeding on Conomelus in laboratory.
Table 47. Predator estimates (numbers per square foot). 1960.
Date Tytthus Cyrtorhinus Carabidae Coccinellidae Nabidae Araneae Opiliones 23/5 71.91 30/5 42.30 29/6 15.51 13.39 3.52 28.90 2.82 5/7 11.28 2.11 4.23 2.11 17.62 7.05 13/7 4.65 0.77 10.93 2.33 13.25 3.10 27/7 11.91 1.62 2.68 3/8 6.42 1.76 1.13 1.13 0.56 10/8 9.66 3.88 0.77 6.07 17/8 11.76 24/8 17.92 3.04 0.16 0.16 0.16 3.84 3.84
Cont. 115.
Table 47 Cont. 1961.
Date Tytthus Cyrtorhinus Carabidae Coccinellidae Nabidae Araneae Opiliones
1/5 8.97 8.93 12.76 4/5 18.42 12/5 7.86 8.35 16/5 36.92 24/5 111.95 4/6 28.74 34.11 4.68 2.67 16.71 0.66 7/6 58.52 58.52 4.35 19.26 15/6 49.55 27.00 4.35 19.26 1.13 20/6 34.85 17.42 5.79 2.30 10.44 1.53 22/6 27.00 8.97 13.10 38.68 28/6 18.02 13.10 4/7 24.20 2.23 9.34 0.97 3.19 14.81 1.94 6/7 8.97 8.97 4.35 13/7 8.97 18 02 21.83 24/7 19.66 0.36 1.82 2.09 10.21 0.91 11/8 18.02 8.97 1.04 0.46 31/8 8.74 116.
The percentage of predators that had recently fed on Conomelus was estimated by
the precipitin reaction. As the number of predators tested in individual
samples was often rather small, the season was divided into three arbitrary
periods, and the results for each period were pooled. Three estimates of
predation were, therefore, obtained; one in the early nymphal period, May
until mid—June, a second for mid—June to the third week in July, and a third
for the adult stages, late July until the end of August. Although over 2000
predators were smeared and tested in 1961, few positive results were obtained
because of the decomposition of the antiserum (page 84- ). As a result of
the failure of the 1961 work the overall picture of predation in the field is
still very incomplete, and only limited conclusions can be drawn from the
data. The serological work in 1960 was also limited, as samples were not
taken before the end of June. Results for both years are given in Table 48.
The length of time that a Conomelus meal remains detectable in the gut
of the predator was estimated by smearing predators et varying intervals of
time after a known feed, taking into account the stage of development of both predator and prey. Over 300 feeding tests were carried out, and the data in
Table 49 are based on the maximum time for which a meal remains detectable, plus two hours, to give the nearest approximation to the actual time when the meal would no longer react. In some cases records were only obtained for late nymphal and adult prey, as in Dolichonabis limbatus; it was assumed, however, that the figures for early nymphal prey of other nabids would also apply to the former species; this same principle was applied to the other groups of predators. Only two specimens of carabidae,both Dromius melano— 117.
cephalus, gave positive results, but as these were smeared after only 12
hours no maximum detectability figures were obtained. The data in Table 49
shows that many predators, particularly spiders, retain a meal over a long
period, e.g. 4 days in Clubionidae. In some instances it was found that a
predator that had fed on an adult delphacid 10 hours or so previously gave
a very weak reaction, whereas a meal of an early nymphal stage was readily
detectable after 15 hours; this was probably due to the fact that the
predators did not always consume the entire prey, and was particularly
frequent among species that only suck the fluid contents of their prey.
Table 48. Pooled results of serological work for 1960 and 1961.
1960.
Second period (27th June — 21st July).
Predator Total number tested Number of positives % positives
Tytthus 111 0 0 Miridae Cyrtorhinus 29 0 0
Nabidae 74 13 17.57
Coccinellidae 6 0 0
Carabidae 5 0 0
Blattidae 5 0 0
Phalangidae 63 17 26.98
Lycosidae 7 1 14.28
Clubionidae 1 0 0
Thomisidae 5 0 0
Linyphiidae 20 0 0
Cont. 118.
Table 48 Cont.
Third period (21st July — 31st August).
Predator Total number tested Number of positives % positives
Tytthus 141 1 0.71 Miridae Cyrtorhinus 26 1 3.85
Nabidae 10 5 50.00
Coccinellidae 60 4 6.67
Phalangidae 93 40 43.01
Lycosidae 12 6 50.00
Clubionidae 12 3 25.00
Pisauridae 8 2 25.00
Thomisidae 8 2 25.00
Linyphidae ( 11 ) ( 5 ) 16.13 (Lepthyphantes and (estimated fig— Lin hia species ure from % of mainly . Lepthyphantes and Linyphia)
1961.
First period (1st Mey — 15th June).
Tytthus 34 1 2.94 Miridae Cyrtorhinus 29 1 3.45
Phalangidae 31 1 3.23
Linyphiidae 17 1 5.88
Cont. 119.
Table 48 Cont.
Third period (24th July — 20th August).
Predator Total number tested Number of positives % positives
Phalangidae 19 5 26.32
Linyphiidae 5 2 40.00
Lycosidae 7 1 14.29
Tetragnathidae 5 2 40.00
No samples tested in first period 1960, and no results obtained in second
period 1961.
Table 49. Maximum period (in hours) over which Conomelus antigen remains detectable.
Predator Prey
II—III instar V instar—adult No. of replicates
Heteroptera
Tytthus pygmaeus (adult) 12 (egg) 7
Cyrtorhinus caricis 24 9 (adult)
Nabis flavomarginatus 37 (nymph) 20
(adult) 48
Stalia major (nymph) 27 29 (adult) 55
Dolichonabis limbatus 30 (nymph) 8 (adult) 36
Cont. 120.
Table 49 Cont.
Predator Prey
II-III instar V instar-adult No.of replicates
Anthocoris nemorum 22 4 (adult)
Coleoptera
Coccinella 7-punctata 17 10 (adult)
Propylea 14-punctata 24 7 adult)
R It (larva) 24 1
Araneae
Xysticus cristatus 50 68 12
Tibellus oblongus - 46 6
Lycosa spp. 44 752 16
Pisaura mirabilis - 74 11
Clubiona spp. - 96 11
Theridion bimaculatum 45 50 24
Tetragnatha montana - 82 9
Pachygnatha clercki - 60 1
Oedothorax gibbosus - 30 3
Hypomma bituberculatum 58 - 2
Linyphia triangularis 38 - 2
L.clathrata 62 - 2
Cont. 121.
Table 49 Cont.
Predator Prey
II—III instar V instar—adult No.of replicates
Linyphiid species 54 10 (unidentified)
Opiliones
Oligolophus spp. 46 60 9
Mitopus morio 60 3
The results of the serological work give an estimate of the number
of predators that have fed on Conomelus within a given period of time; the
data do not, however, indicate whether the positive reaction for any predator represents one or more meals. The laboratory feeding trials provided rough estimates of the number of prey that might be taken daily
in the field, but these figures must then be related to the activity of the predators, the availability of Conomelus, and the amount of alternative food. Laboratory feeding estimates are given in Table 51. In order to estimate the degree of predation on alternative prey, the fauna of 28 x
5.0 inch diameter cores was analysed — Table 50. Arthropods from below the litter layer were not included in the counts, as it was unlikely that these would be encountered by predators of Conomelus which live on or above the litter.
122.
Table 50. Percentages,-of various arthropods in Juncus samples (main
families or groups listed in order of importance).
% total fauna % total fauna
Collembola 7.27 Isopoda 0.68
Coleoptera Staphylinids 11.55 Acarina 4.53
Carabids Araneae 7.19 (see list)
Chrysomelids Opiliones 1.45 (see list)
Homoptera - Jassids 1.73 Chilopoda 2.65 (excluding Conomelus)- Cercopids
Delphacids
Conomelus 56.80 Heteroptera - Miridae 4.95
Nabidae
Lygaeidae
Diptera - small Nematocera 0.68
Hymenoptera - parasitica 0.51 123.
Table 51. Mean predation rate in laboratory in numbers of prey per day.
Predator II-III instar No. of V instar- No. of prey replicates adult replicates
Tytthus adult (eggs) 3.71 11 Oar
Miridae nymph 0.71 10 (early)
Cyrtorhinus - - 0.25
Nabidae adult 3.11 10 1.66 4
" nymph (early) 2.53 9 - -
Coccinellidae adult 3.00 1 1.86 9'
it larva - - 2.50 2
Staphylinidae (Stenus) 0.62 2 - -
Carabidae 1.42 16 Araneae
Thomisidae - - 2.00 7
Lycosidae 2.54 12 3.27 3
Pisauridae - - 0.50 2
Clubionidae - - 0.40 4
Theridiidae 0.43 5 1.67 3
Tetragnathidae - - 2.00 2
Argiopidae 2.00 1 -
Linyphiidae 0.94 18 1.20 8 Opiliones
Phalangidae 1.94 16 1.81 4 124.
The figures in Table 50 were obtained in early June when Conomelus
numbers had already fallen appreciably; it could, therefore, be assumed
that the percentage of Conomelus was greater than 56% earlier in the season,
although mirid numbers were also rather higher at this time. Conomelus
may thus form at least half of the food available to predators, and although
the delphacid numbers fall off later in the season, the larger size of the
fifth instars and adults makes them more vulnerable to many predators,
particularly spiders; this was confirmed by laboratory feeding tests in
which predators were given both second—third instars and adults; phalangids,
lycosids, Pisaurids, and the larger Linyphiids generally fed on the larger
stages in preference to the others, and nabids (adults) showed a similar
preference.
Very few observations were made on the activity of predators in
the field on account of the density of the vegetation; any attempt to open
up the Juncus mat led to considerable disturbance of the predators.
Predation in 1960.
In 1960 estimates of predation were only obtained in the second
and third predation periods; the results of the serological work are
given in Table 48 (page 117).
Mirid predation: No positive reactions were obtained for Cyrtorhinus caricis in the second period, but in the third period 3.85% of the adults tested contained Conomelus antigen. An average of 2.33 Cyrtorhinus per square foot were present each day throughout the period of 40 days (21st
July — 31st August). A single delphacid prey (fifth instar — adult) 125. remained detectable for 1.0 days in the gut of the mirid. The total numbers of Conomelus taken during the third period is calculated from the product of the mean number of mirids present each day during the predation period (numbers/sq.foot), the percentage of the population containing
Conomelus antigen, the numbers of prey Laken daily, and the length of the predation period in days. The low feeding rate of Cyrtorhinus, i.e. 0.25 prey per day, means that each meal only represents the taking of one delphacid. The number of prey taken by Cyrtorhinus in the third period is, therefore, 0.90 per square foot.
Nabid predation: Positive reactions were obtained from the three species of nabid, of which N.flavomarginatus was the most abundant. Predation is estimated in the manner described above for Cyrtorhinus. In the second period:
Mean number of nabids per square foot 2.65
% containing Conomelus antigen 17.57
Detectability of antigen in days 2.00
At the rate of one delphacid per meal, the number taken in the second period (24 days) is 5.59 per square foot. In the laboratory nabids feed on 3.11 prey per day, and the total number taken at this rate would be
34.77 per square foot.
In the third period the number of delphacids taken was higher, although a meal of a fifth instar or adult remained detectable for 2.5 days.
The total number takeh in the third period was estimated from: 126.
Mean number of nabids per square foot 1.24
% containing Conomelus antigen 50.00
Detectability of antigen in days 2.50
The number of prey taken in the third period (40 days) is, therefore,
9.92 per square foot (single feed), or 41.17 (multiple feeds, 1.66 prey per
day).
The only other insect predators known to feed on Conomelus in 1960
were coccinellids, of which Cocciriella 7—punctata was the most abundant.
These predators occurred together with their prey for about 28 days (29th
July — 25th August). The number of prey taken was calculated from:
Mean number of coccinellids per square foot 0.69
containing Conomelus antigen ,.. 6.67
Detectability of antigen in days 1.00
The total number of Conomelus taken in the third period was, therefore,
1.29 per square foot (single feed), or 2.40 (multiple feeds, 1.86 per day).
Predation by Opiliones: In 1961, during the second period, the most
abundant species were Oligolophus agrestis (immature stages) and Leiobunum blackwalli (immature stages). The total predation was estimated from:
Mean number of Opiliones per square foot 4.32
% containing Conomelus antigen 26.98
Detectability of antigen in days 2.50
The number of delphacids taken by Opiliones is, therefore, 11.25 per square foot (single feeds), or 54.27 (multiple feeds 1.94 prey per day). In the third period 0.agrestis and L.blackwalli were again abundant, but L.rotundum 127. and Mitopus morio were also frequently found. The total predation was estimated from:
Mean number of Opiliones per square foot 2.36
% containing Conomelus antigen 43.01
Detectability of antigen in days 3.25
The number taken in the third period is, therefore, 12.59 per square foot
(single feed), or 74.29 (multiple feeds 1.81 prey per day).
Predation by Araneae: In 1960 1.,embers of the families Lycosidae, Clubio— nidae, Thomisidae, Pisauridae, and Linyphadae were shown by the precipitin test to have fed on Conomelus in the field. Although in the second period
Linyphads formed 84% of the total spider population, no positives were recorded. In the third period, however, 16.13% of the Linyphads were estimated to contain delphacid antigen; the reacting species were
Lepthyphantes tenuis, Linyphia triangularis and L.clathrata. The total number of prey taken was estimated from:
Mean number of Linyphlids per square foot 3.75
% containing Conomelus antigen 16.13
Detectability of antigen in days 3.00
Thus, in the third period a total of 7.98 delphacids per square foot (single feeds) were taken by Linyphads, or 29.04 per square foot (multiple feeds,
1.20 prey per day).
Clubiona species formed 22.6% of the spider population in the third period (no positives were obtained in the second period). Total predation was estimated from: 128.
Mean number of clubionids per square foot 1.37
containing Conomelus antigen 25.00
Detectability of antigen in days 4.00
The number of prey taken in the third period was, therefore, 3.42 per square
foot (single feed), or 5.47 (multiple feeds, 0.4 prey per day).
Predation estimates for the remaining spider species were low.
Predation by Lycosids, mainly L.pullata and L.prativaga in the second
period was estimated from:
Mean number of Lycosids per square foot 0.32
% containing Conomelus antigen 14.28
Detectability of antigen in days 2.50
Thus the number of Conomelus taken by Lycosids in the second period is
0.44 per square foot (single feed), or 2.78 per day (multiple feeds, 2.54 prey per day). In the third period, rather more predation occurred; due mainly to an increase in the spider population.
Mean number of Lycosids per square foot 1.37
% containing Conomelus antigen 25.00
Detectability of antigen in days 3.25
The number of delphacids destroyed by Lycosids in the third period is 1.49 per square foot (single feed), or 15.72 (multiple feeds, 3.27 prey per day).
The number of delphacids taken by Thomisid spiders, mainly )_cysticus species and Tibellus oblongus, in the third period was estimated from: 129.
Mean number of thomisids per square foot 0.46
% containing Conomelus antigen 22.22
Detectability of antigen in days 3.00
Thomisids, therefore, account for the loss of 1.35 delphacids per square foot (single feed), or 8.18 (multiple feeds, 2.00 prey per day). No positives were obtained in the second period.
The last family shown to have fed on Conomelus in the field was the
Pisauridae, of which Pisaura mirabilis was the most abundant; Dolomedes fimbriatus was occasionally found, but never contained Conomelus antigen.
Predation was only recorded in the third period.
Mean number of pisaurids per square foot 0.24
% containing Conomelus antigen 25.00
Detectability of antigen in days 3.00
Thus the number of delphacids taken by P.mirabilis is 0.79 per square foot
(single feeds), or 1.20 (multiple feed, 0.5 prey per day).
By totalling all the data it was estimated that between the 27th of June and the 21st of July 1960, 17.28 delphacids per square foot were taken by predators; this figure is based on single feeds only; if the predators fed at the laboratory rates in the field, the total loss by predation would be 91.82 per square foot. The numbers of Conomelus present at the beginning and end of the second period were 337.32 and 177.35 per square foot respectively, i.e. a fall—off of 159.97. The percentage of this mortality that could be attributed to predation was, therefore, 10.8%
(single feeds) or 57.4% (multiple feeds). In the third period from 21st 130.
July until the end of August Conomelus numbers fell from 177.35 to 62.21 per square foot, a loss of 115.22. Total predation over this period accounted for 39.72 delphacids per square foot (single feeds), or 34.47% of
the total population. The estimates obtained by using laboratory feeding data gave a total predation figure of 177.57 delphacids per square foot, which was greater than the total mortality, and obviously incorrect.
Predation in 1961.
Because of the failure of the serological work in 1961, very few estimates of predation were made. By using 1 : 800 strength antiserum prepared by concentrating up less—sensitive sera, some data were, however, obtained. Estimates of predation were made in an arbitrary first period,
1st May until 15th June, and in a third period, 24th July until 20th
August; no results were obtained for the intermediate period, and the figures for the other two are based on small numbers of smears as little sensitive antiserum was available.
In the first period the only insect predator shown to contain
Conomelus antigen was the mirid Cyrtorhinus caricis. Estimates of total mirid predation were obtained from:
Mean number of Cyrtorhinus per square foot 14.95
% containing Conomelus antigen 3.93
Detectability of antigen in days 1.00
The total number of Conomelus nymphs taken by mirids in the first period is, therefore, 27.16 per square foot (single feeds). The laboratory data showed that not more than one delphacid would be taken daily. Other 131.
predators giving positive results included harvestmen and spiders.
Opiliones: a single positive was obtained from a Platybunus triangularis
adult, representing 3.23% of the harvestmen population in the first period.
The number of nymphs taken by Opiliones was estimated from:
Mean number of Opiliones per square foot 0.89
% containing Conomelus antigen 3.23
Detectability of antigen in days 2.50
Thus predation by Opiliones in the first period may account for the loss
of 0.53 nymphs per square foot (single feed) or 2.56 (multiple feeds,
1,94 prey per day).
Araneae: a single positive was recorded from a Linyphiid species Hypomma
bituberculatum, representing 5.88% of the total Linyphiid population.
Predation by Linyphiids was estimated from:
Mean number of Linyphiids, per square foot 6.70
% containing Conomelus antigen 5.88
Detectability of antigen in days 2.50
The number of prey taken by Linyphiids in the first period is, therefore,
7.25 per square foot (single feed) or 17.03 (multiple feeds, 0.94 prey per day).
In the third predation period only harvestmen and spiders were shown to contain Conomelus antigen.
Opiliones: Oligolophus agrestis and Leiobunum blackwalli adults reacted positively to the precipitin test; the number of prey taken by these predators was estimated from: 132.
Mean number of Opiliones per square foot 0.68
containing Conomelus antigen 26.32
Detectability of antigen in days 2.75
The total number of prey taken by Opiliones in the third period is, therefore,
1.74 per square foot (single feed) or 8.74 (multiple feeds, 1.81 prey per
day).
Araneae: In the third period Linyphid spiders, particularly Lepthyphantes
tenuis, were shown to have fed on Conomelus in the field.
Mean number of Linyphiids per square foot 3.15
% containing Conomelus antigen 40.00
Detectability of antigen in days 3.00
Thus Linyphiids may take a total of 11.23 delphacids per square foot in the
third period (single feeds), or 31.97 per square foot (multiple feeds, 0.94
prey per day).
Positive reactions vere also obtained from Lycosa species, and
Pachygnatha clercki (Tetragnathidae). The number of prey taken by the
former was estimated from:
Mean number of Lycosids per square foot 0.21
% containing Conomelus antigen 14.29
Detectability of antigen in days 3.25
The total number of Conomelus taken in the third period by Lycosids is,
therefore, 0.25 per square foot (single feeds) or 2.06 (multiple feeds,
2.54 prey per day).
For Pachygnatha the corresponding estimate was obtained from: 133.
Mean number of Pachygnatha per square foot 0.25
% containing Conomelus antigen 40.00
Detectability of antigen in days 3.00
Thus the number of prey taken by Pachygnatha in the third period is 0.89 per square foot (single feed) or 5.40 (multiple feeds, 2.0 prey per day).
By totalling all the above data, the loss of Conomelus nymphs by predation in the first period was 34.94 per square foot (single feeds) or
46.75 (multiple feeds). The fall—off of the Conomelus population during this period was considerable, 789.92 to 183.36 nymphs per square foot, i.e. a loss of 606.56 nymphs; the two predation estimates, therefore, only accounted for 5.76% (single feeds) or 7.71% (multiple feeds) of the total mortality.
During the third period, the Conomelus population fell from
75.10 to 21.63 delphacids per square foot, a decrease of 53.47. Total predation accounts for 26.39% of this mortality (single feeds), while the second estimate (multiple feeds) is too high and exceeds the total mortality figure.
Discussion of nymphal and adult mortality.
In both seasons the greatest reduction in the population occurred between the first and second instar, and also, in 1961, between the second and third instar. Mortality in later instars was lower, and occurred at a fairly steady rate — Plates 16 and 17. By using the regression technique of Richards and Waloff (1954) it was estimated that of 836.2 first instar 134. nymphs hatching in 1960, 241.1 reached the adult stage, i.e. 71.170 mortality; in 1961 the corresponding figures were 1124.4 and 290.3, i.e. a mortality of 74.18%. The losses occurring between hatching and the adult stage are, therefore, very similar for the two seasons.
Possible reasons for these losses could be: (1) factors of the habitat, particularly the Juncus plant (ii) intraspecific competition
(iii) dispersal (iv) climatic factors (v) other organisms in the habitat, particularly parasites and predators.
The Juncus plant does not alter to any apparent extent during the season except in length of stem, and is accepted by Conomelus at any stage of growth provided that the tissues are living. The delphacid was success- fully reared on 10 - 12 month old rush stems, the resulting adults having a normal fecundity. The amount of available Juncus in the habitat was constant throughout both seasons.
Competition for food is rare among phytophagous insects, and only a few examples have been found in the literature; these are mainly defoliat- ing species, which consume their entire food supply and then die of starvation, e.g. Chrysomela gemellata on Hypericum sp. in Australia -
Clark (1953). Among sucking insects, such as Conomelus, this is unlikely to occur unless the number of insects is so great that all host plants die because of the excessive removal of sap. No such effects on Juncus plants were observed in the field, and it was obvious that the food supply was far in excess of the requirements of the delphacid.
There is no evidence that any of the fall-off of the Conomelus 135. population can be attributed to dispersal. No large scale nymphal move— ments were observed, and only limited adult dispersal occurred; this is discussed more fully under dispersal (page 140).
There was also no evidence to suggest that the population reductions were related to climatic factors. Whalley (1958) considers that heavy rainfall is the main factor responsible for the considerable mortality between the first and second instal of C.anceps, but does not suggest how this could occur. The author believes that Whalley is incorrect in his assumption for a number of reasons. In the present study this first instar mortality occurred regularly each season, and could not be correlated with rainfall; in 1960 the greatest drop in numbers took place between the
23rd and 30th of May; the mean precipitation during this period was 0.2 mm. per day, with a maximum of 0.7 mm. per day. In 1961 no rainfall was recorded during this mortality period (16th — 24th May). Hanna (1950) and
Goodman and Toms (1956) have observed that heavy rainstorms of 10.6 mm., or more, will kill large numbers of the Cotton Jassid Empoasta lybica in the
Sudan Gezira; the mortality is probably caused by mud being splashed on to the plants, and also by the direct effect of the large raindrops. This is, however, very unlikely to occur in a Juncus habitat where the vegetation forms a dense protective mat below which the nymphs are found. The technique of extracting delphacids by water flotation indicates that these insects would not even be affected by flooding of the habitat, as nymphs and adults can readily walk over, or leap from, a water surface. It, therefore, seems unlikely that rainfall plays any part in the mortality of 136. first instar Conomelus nymphs.
An attempt was made in both seasons to relate parasitism and predation to the mortality occurring in the delphacid population. Parasit— ism was generally rather low in both 1960 and 1961, and never led to any mortality of the nymphal stages. Pipunculids accounted for 6.84% of the adult mortality in 1960, and 13.04% in 1961; very few fifth instars were killed before moulting to the adult stage. Percentage parasitism by
Elenchus was only 1.92 and 0.1% an 1960 and 1961 respectively, and was, therefore, of no importance as a mortality factor. The records of other workers show that the mortality caused by these two parasites is very variable, but may be of considerable importance. Hassan (1939) in his studies on Elenchus found that 55% of the adults of C.anceps were parasitis— ed, while other species Delphacodes pellucida and D.fairmairei were less affected, i.e. 17% and 6% respectively. Kanervo (1957), in Finland, found that 70% of a population of D.pellucida contained Strepsiptera, and Uilliams
(1957) has noted 76% parasitism of Perkinsiella saccharicida in Mauritius.
Records of Pipunculids in the European literature show that the incidence in delphacids is usually rather low. Lindberg (1946) found that a maximum of 2.42% of populations of Chloriona species contained pipunculid larvae, and Whalley (1958) records only 1% parasitism of Conomelus. In
Mauritius, however, over 30% of the nymphs of P.saccharicida may contain pipunculids — Williams (1957).
Mortality of Conomelus due to fungal attack was sporadic and generally very low, but in 1961 4.85% of the adults were found to be 137.
diseased.
In conclusion it is obvious that parasitism had no effect on the
nymphal numbers, but accounted for 8.8% and 17.6% of the adult mortality
in 1960 and 1961 respectively.
Estimation of the effect of predation on the numbers of Conomelus
was very difficult, for the reasons given on page 108; the only means of
determining the extent of predation in the field is the precipitin test,
and despite the set—backs which occurred during the use of this technique,
a certain amount of relevant data were obtained. From the precipitin results it was possible to determine which groups of predators were of
greatest importance in the field — Table 52.
Table 52. Predators of Conomelus in order of importance (from precipitin
work).
1960.
Predator June — July July — August
% of total predation % of total predation
Opiliones 65.2 37.1
Nabidae 32.3 23.0
Araneae 2.5 34.8
Cc)ccinellidae 3.0
Miridae 2.1
Cont.
138.
Table 52 Cont.
1961.
Predator May — June July — August
% of total predation % of total predation
Miridae 77.8
Araneae 20.7 87.7
Opilior.es 1.5 12.3
It is evident that spiders, harvestmen and nabids are generally the most
important predators over the predation period as a whole, while mirids
are important early in the season. The failure of the 1961 serological
work has meant that this early season predation could not be examined in
further detail. It is probable that certain other groups such as the
Carabidae also fed on Conomelus in the field, although this was not confirm—
ed by the precipitin reaction as only few beetles were tested.
The total mortality that could be attributed to predators is shown
in table 53.
Table 53. Estimated total predation in 1960 and 1961.
1960.
June — July July — August
% of total mortality % of total mortality
Single feeds 10.8 34.47
Multiple feeds 57.4 >100.00
Cont. 139.
Table 53 cont.
1961.
May — June July — August
% of total mortality % of total mortality
Single feeds 5.76 26.39
Multiple feeds 7.71 >100.00
The figures for May 1961 can virtually be disregarded as they are based
on tests carried out with rather poor antiserum (1 : 800), and are
certainly underestimates. The mid—seasonal figures for June — July in
1960 demonstrate the main difficulty that arises in estimating predation
by the precipitin test, namely the impossibility of confirming multiple
feeding in the field. Dempster (1960) has studied the predators of the
Broom beetle Phytodecta olivacea with the aid of the precipitin test, and
considered that multiple feeding was unlikely because of the scattered distribution of the beetle. Although the Conomelus population did show aggregation, the density of hoppers was high throughout the habitat, particularly in May and June, when 200 — 700 nymphs per square foot were present. The data in Table 50 (page i22) shows that in early June
Conomelus (although already decreasing in numbers) forms nearly 57% of the available food in the habitat; earlier in the season this percentage is certainly higher. These facts would suggest that multiple feeding does occur in the field, and that the higher predation estimates are more accurate than the single feed figures. The single feed predation estimates 140.
for July — August are, however, more acceptable as the number of delphacids has already decreased by at least 60% at this time, and it is less likely
that multiple feeding occurs. The 1960 figure of 34.47% is a more reliable
estimate than that for 1961 as more samples were tested, and the antiserum
was of normal sensitivity.
It is clear from the above discussion that only very rough
estimates of pr'dation have been obtained. There are three main reasons
for the errors involved; firstly the difficulties of accurately sampling
both predator and prey populations; secondly the fallability of the
serological technique, and, finally, the lack of knowledge as to the like—
lihood of multiple feeding in the field. There is, therefore, at present no satisfactory explanation for the considerable nymphal mortality of
C.anceps. The present author is however, of the opinion that predation
is the most important mortality factor, and that this would be confirmed by
carrying out extensive serological work, particularly in May and June.
Dispersal.
In an insect with specialised food requirements such as Conomelus
the ability to disperse is of great importance; as a Juncus habitat dries out the rushes become dominated by grasses and finally die off altogether.
Feeding experiments have shown that Conomelus can only survive and reproduce on Juncus spp. and must, therefore, disperse from the habitat when these unfavourable conditions arise. A certain amount of dispersal occurs in
Most animals while conditions in the habitat are still favourable; this may be regarded as an 'insurance' against the sudden onset of adverse 141.
factors — Elton (1950). An attempt was made in the present study to determine the degree of movement of Conomelus nymphs and adults throughout
the season.
Methods.
Marking and recapture.
Individual marking was not feasible for two main reasons; firstly the small size cf the deiphacids made the application of paint, without smothering the insects, extremely difficult; secondly the number of hoppers that had to be marked to obtain any recaptures was very high.
The mass—powdering method devised by Macleod and Donnelly (1957) for marking blowflies was, therefore, used. Delphacids were placed in a flask together with a tube containing 0.5 gm. of Rotor Blue B dye; air was then blown through a piece of tubing into the dye—stuff, creating a 'dust—storm' within the container. Tests showed that the dye particles adhered to virtually all the hoppers in any one batch. The deiphacids were then released in the plot, and samples were collected at varying intervals of time after the release date. The hoppers in these samples were killed, placed on sheets of filter paper and treated with acetone; a blue ring appeared around those individuals that had previously been marked, due to the acetone extracting the dye—stuff from the surface of the cuticle. This method could only be used with adults, as nymphs would lose the dye after a moult.
Sticky barriers.
To determine whether nymphs and adults dispersed by walking, the 142.
experiment shown in Plate DR was set up. A rush clump, 12 inches in
diameter, was set into an area of bare ground; a band of cellulose
acetate coated with 'Stictite' on both sides, 4 inches high and with a radius of 20 inches, completely encircled the tussock; glass plates also
coated with 'Stictite' were set at distances of 2 to 8 inches from the rushes. Most of the area around the clump was left as bare soil, but one
sector was planted with turf to discover whether dispersing nymphs or
adults preferred to move through cover or over bare soil. The 'Stictite' band and glass plates were examined at regular intervals.
Water traps.
Metal trays measuring 11.6 x 7.6 x 2.6 inches, painted white inside and black outside, were placed into holders attached to iron stakes
(3 — 4 feet in height). The trays were two—thirds filled with water containing a little detergent and formalin. 20 of these traps were placed around the plot at distances varying from 2 to 50 feet to study the dispersal of the macropters.
Suction traps.
No suction traps were used by the present author, but in 1959 two traps were set up for work on Thysanoptera and frit—flies in areas adjoining Pond field. The specimens taken in these traps supplied the only direct evidence for the aerial dispersal of Conomelus. Two types of trap were used, a 9—inch Vent—Axia with a through—put of 293 cubic feet per minute, and an 18—inch propellor type with a through—put of 2510 cubic feet per minute; the first trap was situated about 75 feet south—west of 143. the plot, and the second one 150 feet south of the area. The intakes of both traps were about 4.5 feet above ground level.
Estimation of nymphal movement.
The experiment, described under sticky barriers above, was set up, placing 400 first instar nymphs on the Juncus tussock. The 'Stictite' band and glass plates were examined at weekly intervals for two months.
No nymphs were found on the band or plates throughout the entire period, this seems to indicate that the nymphal stages remain on their food plant, and will not disperse over bare soil or through a grass zone; this was confirmed by laboratory observations where nymphs always remained on the rush plants, and were only rarely seen on the sides or covers of the cages.
There was no practicable method of determining nymphal dispersion within the rush area itself, but it was assumed that a certain amount of local movement occurred as the stems from any one plant were always intermeshed with adjoining tussocks. When disturbed the nymphs are capable of leaping 8.4 cm. (first instar) to 90 cm. (fifth instar); that this form of accidental dispersal does not often occur in the field was confirmed by the absence of any nymphs on the sticky bariers, or in water traps set within 6 inches of the ground.
Estimation of adult movement.
Most of the Conomelus adults are brachypterous and only possess tegmen—like fore wings. Dispersal in these forms can only take place by walking or leaping. The mass—powdering technique was used to estimate the movements of adults, and the results for 1960 and 1961 are given in 144.
Table 54.
In 1960 samples were only taken 6 days after the release date.
The data show that a large number of the delphacids, 42.47% of the sample,
had not dispersed further than 1.5 feet from the point of release; a few
individuals were also taken at 3 feet. In 1961.a similar pattern of
distribution was noted after 3 days, with a high proportion of marked
adults still within 1 foot of the release point; samples taken after 12
days, however, showed that the deiphacids had dispersed for distances up to
6 feet. It was impossible to determine whether the adults had covered this
distance by walking or leaping; however, the mean adult walking rate of
1.5 cm. per second, and the dense vegetational cover of the habitat, make
the likelihood of a dispersion of 6 feet in 12 days by walking virtually
impossible. As the adult delphacids are capable of leaping 3 feet when disturbed, it is probable that distances of 6 feet could readily be covered by this accidental means of dispersal.
Table 54. Results of mass-powdering experiment, 1960 and 1961.
1960.
250 females and 736 males released on 22nd July. Sample taken after 6 days. Females Males Distance from release Total No. Total No. point in feet Sample marked Marked Sample Marked Marked 0 14 4 28.57 59 27 45.76 1.5 12 2 16.67 20 6 30.00 3.0 26 0 0 52 2 3.85
3.5 33 0 0 80 0 0
Cont. 145.
Table 54 Cont.
1961.
1664 females and 4144 males released on 27th July. Sample taken after 3 days.
Females Males
Distance from release Total No. % Total No. point in feet Sample marked marked Sample marked marked
0-1 54 15 27.78 124 51 41.13
3.0 102 6 5.88 169 10 5.92
6.0 113 0 0 187 0 0
Sample taken after 12 days
0-1 38 4 10.53 59 9 15.25
6.0 145 7 4.83 151 7 4.64
Analysis of the percentage recapture for each sex (Table 54) shows that a higher proportion of males are taken near the release point in both seasons, and that the ratio of males to females decreases as samples are taken further out; it thus appears that the females have a greater tendency to disperse than the males.
Aerial dispersal of adults.
It was assumed that the only stage of Conomelus capable of active dispersal was the macropterous form. There are several records of aerial movements of deiphacids; Kanervo (1957) has annually observed large swarms of winged adults of Deiphacodes pellucida, which he considers may disperse over large areas with the aid of favourable winds. Williams (1957) has recorded sustained flights of Perkinsiella saccharicida and Dicranotropis 146. muiri in Mauritius, and has taken the adults at light traps. In all three species a high percentage of the adult population is macropterous and capable of flight, at least at certain periods in the season. The percentage of winged forms in Conomelus, however, is never greater than 7 - 10% of the total population - Table 57 (page 141).
Conomelus macropters are generally slightly larger than the short- winged forms, particularly the females; the thorax is darker in colour because of the underlying phragmata, and the pronotum and scutellum are structurally different (Plate 1%, Fig.l); the metanotum of the macropter bears several well-developed sclerites which are absent in the brachypter
(Plate 1%, Fig.2); the fore and hind wings are fully developed, and when at rest extend beyond the tip of the abdomen; well-developed flight muscles are present, particularly the median dorsal longitudinal muscles which fill most of the thoracic cavity. As these delphacids appear to have all the necessary adaptations for flight, it is all the more surprising to note that not a single record of a flying adult was obtained in the present study.
Macropters, suspended from the thorax by a small spot of paint and a blunt pin, were held in an air stream, but did not attempt to fly in- spite of varying the temperature conditions. Repeated disturbance of macropters in the field and the laboratory did also not induce flight.
The mass-powdering experiment also provided a few figures which again confirmed that these winged forms did not actively disperse - Table 55. 147.
Table 55. Local movement of macropters — mass powdering experiment, 1961.
214 macropters released 27th June.
No. in sample No. marked % marked Distance from release point in feet
First sample 17 14 82.35 0
(after 3 days) 11 2 18.18 3
Second sample 8 0 0 0
(after 12 days) 12 2 16.67 6
The percentage of marked macropters taken at 6 feet after 12 days, 16.67%,
is higher than the figure for brachypters at the same distance, i.e. 4.73%;
although the small size of the samples limit the conclusions that can be
drawn from the data, the differences in the two recapture rates may be due
to the greater local movement of the winged form.
The laboratory flight trials and field observations seem to
confirm that the macropter is incapable of flight, in spite of its well— developed wing muscles. In 1961 it was, however, discovered that two
suction traps had been set up in the vicinity of the plot in 1959, before
the author's arrival.Examination of the samples from these traps showed
that Conomelus macropters were present — Table 56. 148.
Table 56. Captures of macropters in suction traps — 1959.
Date Males Females
Trap I 30/7 — 1
(9—inch Vent—Axia)
Trap II 21/7 1
(18—inch propellor) 23/7 1
!I 24-26/7 1
tr 2-3/8 1
4/8 1
!I 6/8 1 2
7/8 1
Although the daily trap samples for June until October were examined,
macropters were only found over the 17 day period from 21st July — 7th
August. Attempts were made to correlate this short flight period with
climatic factors. It is probable that the delphacids were wind—assisted
in their flights, as the single capture in the south—western trap (Trap I)
was taken at a time when south—westerly winds prevailed; the second trap
was sited 150 feet south of the plot, and the direction of the wind over
the capture period was either variable or south—westerly. Kanervo (1957)
has observed the flight of Delphacodes pellucida, and records that the
threshold conditions were air temperatures greater than 25 — 27°C., and a
wind speed of less than 10 m.p.h.; the flight of the delphacids was slow, resembling that of aphids, and individuals were taken in nets up to 19 feet 149. above the ground. Conditions of temperature and rainfall did not vary greatly throughout July and August, and there was no evidence to suggest that these factors had induced the short dispersal flights of Conomelus.
No suction traps were used in the vicinity of the plot in 1960 or 1961, the nearest being about 200 yards distant; no macropters were found in the
July — August samples from all the Silwood Park traps for either of these years; the small numbers of macropters produced probably account for the lack of material in these more distant traps. However, analysis of the percentages of winged forms in samples taken at various dates in July and
August 1961, suggests that dispersal may again have taken place over a short period in August as a reduction in'numbers occurs at this time —
Table 57.
Table 57. Numbers of macropters in July and August 1961.
Date Total adults Number of macropters % macropters
8/7 64 5 7.81
21/7 151 11 7.28
2/8 119 6 5.04
5/8 94 7 7.45
23/8 100 3 3.00
30/8 101 1 0.99
6/9 120 1 0.83
20/9 102 1 0.98
No detailed estimates of the numbers of macropters were made in 1960, but the maximum percentage recorded exceeded that for 1961; of 1107 adults collected on 22nd July 121 were winged, i.e. 10.93%. 150.
As there is no obvious correlation between climatic factors and
flight, it is probable that aerial dispersal commences when the macropters reach a certain stage of development (the rate at which this is attained
will itself be indirectly related to climate). Samples of macropters were
dissected throughout the season in 1961, to discover whether there was any
correlation between flight activity and the condition of the flight muscles
and reproductive organs - Table 58.
Table 58. Condition of reproductive organs and flight muscles of macropters - 1961.
Females.
Date Ovarioles Accessory Median longitudinal Condition of gland flight muscle muscles
Length Thickness
17/7 (4) Thread-like 0.46 0.34 0.12 Normal
16/8 (10) il 0.53 0.38 0.14
30/8 (10) It 0.57 0.31 0.11
13/9 (6) Plus ripe 0.53 0.28 0.13 eggs
Trap sample (8) Thread-like 0.45 0.37 0.12 (1959)
Males.
Date Testes Accessory Median longitudinal Condition of length gland flight muscle muscles Length Thickness 17/7 (5) 1.09 1.34 0.35 0.11 Normal
16/8 (3) 0.99 1.59 0.38 0.14 ft 30/8 (5) 1.03 1.75 0.34 0.13 It 13/9 (2) 1.17 2.41 0.27 0.10 Trap sample (2) 1.06 1.18 0.31 0.11 (1939) Measurements in mm. Numbers of individuals in brackets. 151.
The condition of the adults taken in the suction traps shows that dispersal
takes place before the reproductive organs have matured; the flight muscles of these individuals are well—developed, particularly those of
the female. The 1961 data show that the flight muscles of both males and
females are most strongly developed in mid—August, which coincides with the
suggested dispersal period (see Table 57). The flight muscles of mature
females (13/9) are smaller than those of the immature specimens; as mentioned earlier (Section 1, page a(, )) this difference may be due to
the reproductive organs developing at the expense of the flight muscles.
There is no breakdown of the muscles after the flight period as noted in aphids by Johnson (1957, 1959), and the cross—striations and nuclei are clearly visible throughout the entire season.
Production of the winged form.
As only a small, and fairly constant, percentage of macropters was produced in both seasons, it was thought that the winged condition might be genetically controlled. Reciprocal crosses between macropters and brachypters took place regularly in the laboratory, and fertile eggs were laid; none of the resulting nymphs were, however, seared through to the adult stage because of the high mortality normally occurring in laboratory cultures. The ratio of macropterous females to males is generally higher than in the brachypterous form; i.e. 1961, 1 macropterous female : 0.36 macropterous males; 1 brachypterous female : 1.62 brachypterous males. It is, therefore, possible that the character is in some way sex—linked.
Vvigglesworth (1954) suggests that alary polymorphism may be due 152. to juvenile hormone acting directly or indirectly on the gene system. On the basis of this theory the longer an insect spends in the nymphal stage, the longer it is exposed to the juvenile hormone; this will result in the adult becoming short—winged, as this is essentially a juvenile character.
The length of time that the insect remains in the nymphal stage may in turn be related to other factors, such as temperature; Brinkhurst (1959) has shown that raising the rearing temperature for Gerrids leads to an increase in the numbers of macropters produced, as the insects spend less time in the nymphal stage; the same author, however, considers that some genetic factor is also involved. Willer (1959, 1960) found that the photoperiod affected the wing—length of jassids. Southwood (1961) has reviewed the whole problem, and considers that the effects of most of these environmental factors may directly or indirectly react with the juvenile hormone.
The macropterous condition in Conomelus was not so readily explained by this theory; only a small proportion of the population ever became macropterous in spite of the fact that the entire population was subjected to similar conditions of temperature and light. Also, although the female nymphs took longer to develop than the males, and were, therefore, exposed for a longer period to the juvenile hormone, macropterous females were far more common than winged males. Kisimoto (1956) found that over— crowding in the blown plant hopper, Nilaparvata lugens Stal (Delphacidae) prolonged nymphal development, and led to an increase in the percentage of macropters produced; this was particularly noticeable in the female, as the male is normally macropterous. Wilting of the food plant also resulted 153. in a lengthening of the nymphal period, and an increase in the number of winged adults produced. The greater incidence of female macropters in
Conomelus can, therefore, be explained by these findings, because of the longer nymphal period of the female. The mechanism of this process is not clear, and is not explained by the juvenile hormone theory.
It was, however, thought unlikely that overcrowding occurred in the field, as only 7 — 10% of the adults became macropterous, and in the author's view, the condition is probably controlled by genetic factors. 154.
SUMMARY.
1. The first section deals with the general biology of C.anceps, including the description of immature stages and life history. Oviposition of
Conomelus has been studied for three seasons, relating numbers of egg sites to stem diameter, stem length and presence or absence of pith.
2. The mean number of oviposition sites per stem increased from 35.4 in
1959 to 93.7 in 1960, to 123.3 in 1961; 1 to 7 eggs were found per site, with a mean number of 2.86. Up to 30% of the oviposition sites did not contain eggs.
3. Hatching and moulting of the nymphs was studied. The duration of the five nymphal instars was estimated; the total period at 16 ± 2°C. was
48.8 days, field estimates were 59 days in 1960 and 53 days in 1961.
4. Sexual maturation of Conomelus adults is discussed. Males matured within 14 days of emergence, while the pre-oviposition period of females was 25.2 days at 16 ± 2°C. The seasonal development of the reproductive organs of field specimens has been studied.
5. The growth of the reproductive organs has been related to weight changes of field specimens. Males quickly attain a mean weight of 1.85 m9.9 mainly due to the growth of the testes and accessory glands. Females continued to gain weight until the end of the season, when their mean weight was 3.07 mg., this was due to the retention of eggs by senescent females,and the growth of the accessory glands. Females laid an average of 29.9 eggs in the insectary, and an estimated 29.6 to 42 eggs in the field.
The mean longevity of females and males in the insectary was 38.3 and 27.0 155. days respectively.
6. The external and internal effects of pipunculids and strepsipterep parasites on Conomelus adults have been studied.
7. Section 2 deals with the biology of the predators and parasites associated with Conomelus, particularly of the nymphal parasite, Pipunculus, and the mirid predators, Tytthus and Cyrtorhinus, which were previously almost unknown. The life histories of other natural enemies including three egg parasites, Anagrus, Aprostocetus (Hymenoptera) and a fungus
Paecilomyces, two nymphal parasites, Elenchus (Strepsiptera) and a fungus
Entomophthora, have also been studied.
8. The seasonal occurrence of various predators is discussed, including nabids (Hemiptera), Coccinellidae, Carabidae (Coleoptera), spiders including
Lycosidae, Thomisidae and Linyphiidae, and species of harvestmen.
9. Section 3 deals with studies of the Conomelus population including mortality factors of the eggs, nymphs and adults, and dispersal.
10. The number of eggs per unit area has been studied in three seasons; the numbers of nymphs and adults were estimated in 1960 and 1961 by taking weekly samples throughout the season. The data showed that considerable losses occurred during the egg stage, and the first and second nymphal instars.
11. Egg mortality increased from 27% in 1959 to about 40% in 1961, mainly because of the greater incidence of fungal attack and predation. The mean mortality figures were: 13.7 to 16.6% loss due to the fungus; up to
15.5% mortality because of mirid predation; Aprostocetus destroyed up to 156.
13% of the eggs.
12. Nymphal and adult mortality was estimated in 1960 and 1961. Of the
parasites only Pipunculus was of importance, causing 7 to 13% mortality in
late fifth instar nymphs and adults. Elenchus and the fungus Entomophthora
were unimportant as mortality factors.
13. Attempts were made to estimate predation by the precipitin reaction.
14. 91 species of predatory arthropods were identified, including 38
species of spider, 19 Carabids, 7 Opilionids, 6 Staphylinids, 5 Coccinellids,
and single species of various other groups. Only 10 of these predators
were important, these being Araneae, Opiliones and Heteroptera.
15. The percentage of predators containing Conomelus antigen was often
high, i.e. up to 50% of th e Lycosid spiders and 40; of the Opilionids.
Meals were retained for varying lengths of time ranging from 12 hours for
an egg meal in Tytthus, to 96 hours for an adult meal in a Clubionid spider.
The likelihood of multiple feeding occurring in the field was lour early
in the season when up to 57% of the entire arthropod fauna in the plot
consisted of Conomelus nymphs.
16. Mortality between the first instar and the adult stage was 71% in
1960 and 74% in 1961. There was no evidence to suppose that this mortality
was due to climate, dispersal, orintra—specific competition.
17. Because of the lack of knowledge as to the likelihood of multiple
feeding, the fallability of the precipitin test, and the difficulties of sampling both predators and prey, only very rough estimates of predation have been obtained, i.e. 57.4% of the total mortality of first to third 157. instar nymphs in 1960, and 34% of the total mortality of late nymphal stages and adults in the same year. In 1961, because of the decomposition of the antiserum, estimates of predation were only obtained late in the season.
18. Field experiments suggested that little or no nymphal dispersal occurred; there was some positive evidence for the local movements of brachypters within the rush plot. Suction trap material indicated that macropters emigrated from the plot for a short period in August, when they were still sexually immature. :t:J.aH I, C£on.Ra'D' iiDQ'pg lV'Dlpha.l stages.
Figure 1. ~~st inetar DYmph Figure. 2. Third. 'instar zvmph Figure 3 •. It'1fth 1nstar nymph (macrpptorous)
}l"tigure 4~ Fifth inStar nymph (brachypteroue) 0.2111111 I
3
0.5 mom 2 0.511111 fl~\i at !2aaowtJ.aa.1vympllal stagGs.
~1.cura I.11ind t1bitl.·.!U'ld. taruu.s of first instar ~p}1...
F1~e2. 11 ~ 'u ff .eon! hat4U" l.\YGlph.
Flgu:re a. ~1 11 fl It third in.atar nymp~.•
1?1gare 4 •.ff 11 U ft four'th !nBtar nymph.
.F1sure 5. 2, ft 11 n fifth inStar2'.\Ylnph •
.Flgul~ G.l(bdomoD off1fth 1notM-(male)wntral view.
Firo,tre 1. tt " . tf . . ff (tomale)ventral view.
0.5se
.1.
6 fh1i, 3, Relationship betveen· distance above leaf-sheath and numbers of CQ1l984v.1 ovipositiOn sites in llmgllg stems. IGO
60
5
12 15 r= 21
Distance of ovip.dsition site above leaf sheath in inches Plate 4. Seasonal weight changes n field eathPlos
3.1..
2.9_
21 .
to 2S _ E 2 _ cp 23 _ 1. c - , 21
.
-----A t% ie
1.3 I 1 1 1 11 1- F I 11 1 1 1 1 1 1 1 8 July.18 28 6Aug. 16 26 SSent. IS 25 5 Oct. between numbers of e 60- • • • 50 - • • N 40- • • • • in 30 - • . • on cr • • • • • • • 20 - • • s- • • ▪E 10 — • z
0 II I II I II I 2 6 10 14 18 22 26
26- Egg-laying period in days • • In 22 - a -a c 18- • • • • •
•L.•- 14 - • • a. • • • • • • CPc 10 • 17. O $1 71, 6- • •
2- ••
32 36 40 44 48 52
Longevity in days and duration of maturation period. 60- • • •
• • 1540— E 1— • • • • • • 30— • a. • • •• CP • • to 20-- • • •
• • • 1:1E 10-
0- 1 1 i 1 1 1 32 36 40 44 48 52 56
Longevity in days 60 • • • 50 • •
• • • • • • •
• • • • • • • • • is • • •
I ' I ' 1 I 1 1 18 20 22 24 26 28 30
Maturation period in days Plate 7. OvipositiOn.aim* of , Gonomeluct showing percentage prOation, b Tnthas. % of eggs destroyed by Tytthus e--e 0 0 0 0 0 co '13 v " r I I I I
4,---• wais Jad sb6a to JagwoN
165
' - . . : • • , • •
•••• ••••
V5
1-0 mm
7 6 i-o mm p.lau 2. The etfeats of etrepsipteron para.sites on aouoMluD adults. Figure 1.Abdomen or malerAAQlt4UI parasitised b,.a temale ilG;ImSJ only slight reduction of genitalia. at .. anal tube .
. . fs .,.' ext.rUded cepbalothoru or female IJrUD9hus.
gs ~ genital segment
pa. - paramere , xl - eleventh abdominal. segment Figures 2-3. Further. reduction of ·m.ale genitalia. me '. cephalothorax, ot, male ilIwUmv .Figure 4. Female host showing lateral extrusion ot jQe;gw§ male puparium.· " , . go- gonoplao(ovipositor sheath) Figure 5.MeJ.e host shoving lateral extrusion otoepbalothorax ot' female ~§ASllmt1. 166
PO mm.
I m m. , ElF. 10, . Immature stages of ImmGAQltUg ~Atd.Q l~lk. Figure 1. Primary larva.. , Figure 2. t>1ature lana. Figure 3. Pupa:.
It''1gure 4;. 1~1outhparte or mature larva. (Folloldng nomenolature of Short,1952 ) , ,a.p .p. - anterior pleurostomal process ep - ep1stoma lb .plp.- labial palp
m.d -ma.tld1ble mx.plp.- 'maxLll8l7 palp p.p.p. - posterior pleurostomal"process 167
0.25mm
•-....."..- •
I bplp 4 0.05 mm
wW 0'1 I t
wujz(>0 c I 1
1 Z W W E 0.0
SW ql PJ.lti J.&. Immature sta~s pt fJ.I?imalJla _=t:tIIOfllQ k~. Figure 1. Posterior spiracular pla.te of mature larva•. Figure 2. Prothoraaic spiracle or mature larva.
Figure 3. !~ature lirva. figUre 4. Puparium shoving. fracture ·lines anteriorly. Figure 5. Houthparts.ot ma.ture larva,dorsal viev. lb - lab1o.l ooler1te m - mandible me - mandibular solerlte ps - pltar3ngeal scler1te . Figure 6. J.fouthparts otmature la:rva,la.teral viev. ww 0.1
Z
E wS00
I LULU S0'0 ww 0.1 PlaW ·J.I. Immature stages ot IYttli\8 WiIi§:Gs Zett. and
~1Qm1§ car1gts (Fall.) Figure 1. Egg or t'zUb\l1 Figure 2. Operculum ofIrt1;lm.g egg,dorsalViev. , Figure 3. First instar tntWI' fllmph.
Figure 4. Egg ot· ~htmia
Figure 5. Operculum .0£ Qn:tQhb1nua egg,dorsal viev.
Figure 6. Second instar G.VJ:tarll;l.wa ~ph.
170
5
4
171
,f r • l, ..• ..1%., • I ... Ir. • I ... .i. A v • / • .... % • 1 •• 11.• It NT • # 1 c t , • .1/4 , # I 1 - , • I • I • , \ • t t A .r A I fel..441: .. % , t I 1 •• 1, I _ • r q, .• A . % • .4:A 4..A.3.1)6.0.4A..• % f't 71 t‘1,(N• `4. • s'll :t .... ,.t.../..4, , , , . ;:dar.0.14.4441,. AA.,.:2: & ear ,„ • ti tli it ii ". illaptopisAlorm•44.4AuLMA # ft A AM_ 10 i Ptli.l figsAis tag 4.1;14$4,1:4:1,:;‘ ‘‘ 1. t f ik.;01:41.14A114 .1" % ill I% i ' % r' IAA. •44.6,A 1, % #I • •# 1.,-* nlte J,§~ Hatching curve for an inseotS17 culture ot 92DAmllrll,-·
.... " Number of hoppers hotching per 22 stems •iti 5 0 8 0 0 0 0 i II
I I I 1 1 I 10' , g g 8 4.3 Uf aJnicuadwai PJ.sJj. Population·estimateainstarbyinatar and. total,in Pond Field 1960. 700
600
500
400
300 Total
20 ti 170- o ,50
s6 130 a in 6- 110 I Z 90-
70-
50
30
10-
10 20 31 10 20 30 10 20 31 10 20 31
May June July Aug
174-
600
500
40
30
200
170 0 .2 1 a 0.130
no
90
E
i 1 I I 10 20 30 10 20 31 10 20 31
May June July Aug
175-
•••111•11
1.0 mm.
2 Z10019.Field rmphat 176
glass plates
grass zone
Juncus clump
bare ground
Stictite band
1...... i 12 ins. 177.
ACKNOWLEDGMENTS.
I wish to express my sincere thanks to the following persons for
assistance which I received during the course of this work:
To my supervisor, Professor 0.W.Richards, for granting facilities at Silwood Park, for his advice and criticism of both the experimental and
the written work, and for the loan of literature.
To Dr.J.P.Dempster for advice and assistance with the precipitin reaction work.
To Miss T.Dedman for technical assistance with the precipitin work.
To Dr.T.R.E.Southwood for advice and the loan of literature and specimens.
To Mr.M.J.Way for advice and loan of literature.
To Mrs.J.A.J.Clark and Mr.J.G.Sheal of the British Museum for identifying insect and mite material.
To Mr.A.M.Wild for identifying most of the Linyphiid spiders.
To Dr.M.F.Madelin of the Department of Botany, University of
Bristol, for identifying the fungal parasites.
To Miss J.Turner for typing the thesis.
To Mr.J.W.Siddorn for producing negatives for the plates.
To the Nature Conservancy for the loan of a thesis (Whalley,1958).
Throughout this work I have been the recipient of an Agricultural
Research Council Studentship, whose generosity is gratefully acknowledged.
178.
REFERENCES.
BAKKENDORF, O. 1926. Recherches sur la biologie de l'Anagrus incarnatus Hal. Ann.Biol.lacustre., 14: 249-270.
- 1934. Biological investigations into some Danish Hymenop- terous egg parasites, especially in Homopterous eggs, with taxonomic remarks and descriptions of new species. Ent.Medd., 19: 1-134.
- - 1953. Description of three species of Tetrastichus (Micro- Hym.) with a host list. Ibid., 26: 549-576.
BRINKHURST, R.O. 1959. Alary polymorphism in the Gerroidea (Heteroptera). Proc.R.ent.Soc.Lond. (C)., 24: 17 (quoted in Southwood, 1961).
BRISTOWE, W.S. 1949. The distribution of harvestmen (Phalangida) in Great Britain and Ireland, with notes on their names, enemies and food. J.anim.Ecol., 18: 100-114.
CARTER, W. 1930. Ecological studies of the beet leaf-hopper. Tech.Bull.U.S.Dept.Agric., 206: 1-114.
CARVALHO, J.C. and SOUTHWOOD, T.R.E. 1955. Revisao do complexo Cyrtorhin- us Fieber Mecomma Fieber (Hemiptera - Heteroptera, Miridae). Bol.Mus.Goeldi, Para., 11: 7-72.
CHAPMAN, J.A. 1956. Flight muscle changes during adult life in a Scolytid beetle. Nature, 177: 1183.
CHINA, W.E. 1950. A check list of the British Hemiptera Homoptera Auchenorhyncha. Ent.Mon.Mag., 86: 243-251.
CLARK, L.R. 1953. The ecology of Chrysomela gemellata Rossi and C.hyperici Forst., and their effect on St.Johnts Wort in the Bright district, Victoria. Aust.J.Zool. 1: 1-69.
CLAUSEN, C.P. 1940. Entomophagous insects. New York, 688 pp.
CROZIER, W.J. and STIER, T.B. 1927. Temperature and activity, accessory hearts, Notonecta (Hemiptera). J.9en.Physiol. 10: 479-500. (quoted in Roeder, 1953).
DAVIES, M.J. 1953. The contents of the crops of some British Carabid beetles. Ent.Mon.Mag., 89: 18-23. 179.
DEMPSTER, J.P. 1960. A quantitative study of the predators on the eggs and larvae of the Broom beetle, Phytodecta olivacea Forster, using the precipitin test. J.Anim.Ecol., 29: 149-167.
EDWARDS, J .E. 1896. The Hemiptera — Homoptera of the British Islands. London, 271 pp.
ELTON, C. 1950. The ecology of animals. 3rd.Ed. London, 97 pp.
FEWKES, D.W. 1958. The biology and feeding habits of some British Nabidae (Hemiptera — Heteroptera). Ph.D. Thesis, London University.
GANIN, M. 1869. Beitrgge zur Erkentniss des Entwickelungsgeschichte bei den Insekten. Z.Wiss.Zool., 19: 381.
GOODMAN, A. and TOMS, A.M. 1956. Effect of rainstorms upon the cotton jassid in the Sudan Gezira. Nature, 178: 436.
GRAHAM, M.h.R. de V. 1961. The genus Aprostocetus Westwood Sensu lato (Hym.Eulophidae). Notes on the synonymy of European species. Ent.Mon.Mag., 97: 34-64.
HANNA, A.D. 1950. The effect of rainfall on the cotton jassid Empoasca lybica (de Berg) in the Gezira, Anglo—Egyptian Sudan. Bull.ent.Res., 41: 359-369.
HASSAN, A.I. 1939. The biology of some British Delphacidae (Homoptera) and their parasites, with special reference to the Strepsiptera. Trans.R.ent.Soc.Lond., 89: 345-384.
HENRIKSEN, K.L. 1919. De europaeiske Vandsnyltehliespe og deres Biologie. Ent.Medd., 12: 137-251.
HILL, A.R. 1957. The biology of Anthocoris nemorum (L.) in Scotland (Hemiptera Anthocoridae). Trans.R.ent.Soc.Lond., 109: 379-394.
JENKINSON, F. 1903. Verrallia aucta and its host. Ent.Mon.Mag., 23: 222-223. (quoted in Lindberg, 1946).
JOHANSSON, A.S. 1958. Relation of nutrition to endocrine—reproductive functions in the milkweed bug Oncopeltus fasciatus (Dallas) (Heteroptera : Lygaeidae). Nytt Mag.Zool., 7: 1-132.
JOHNSON, B. 1957. Studies on the degeneration of the flight muscles of alate Aphids. I. A comparative study of the occurrence of muscle breakdown in relation to reproduction in several species. J.ins.Physiol., 1: 248-256. 180.
JOHNSON, B. 1959. Studies on the degeneration of the flight muscles in elate Aphids. II. Histology and control of muscle breakdown. J.ins.Physiol., 3: 367-377.
JOYCE, R.J.V. 1961. Some factors affecting numbers of Empoasca lybica (de Berg)(Homoptera : Cicadellidae) infesting cotton in the Sudan Gezira. Bull.ent.Res., 52: 191-232.
KANERVO, V., HEIKINHEIMO, O., RAATIKAINEN, M. and TINNILA, A. 1957. The leafhopper Delphacodes pellucida (F.) (Hom.Auchenorhyncha) as the cause and distributor of damage to oats in Finland. Valt.Maatalour.k.Julk. No.160, 56 pp.
KEILIN, D. and THOMPSON, W.R. 1915. Sur le cycle evolutif des Pipunculides (Dipt'eres) parasites intzacoelomiques des Typhlocybes (Homopthres). C.R.Biol.. Paris, 78: 9-12.
KISIMOTO, R. 1956. The effect of crowding during the larval period on the determination of wing-form of an adult plant-hopper. Nature, 178: 641-642.
KONTKANEN, P. 1950. Quantitative and seasonal studies on the leaf-hopper fauna of the field stratum and on open areas in North Karelia. Ann.Soc.(zool.-bot.) fenn.Vanamo, 13: 91 pp.
KULLENBERG, B. 1943. Die Eier der schwedischen Capsiden (Rhynchota) II. Ark.Zool., 34A: 8 pp.
- - 1944. Studien fiber die Biologie der Capsiden. Zool.Bidr., 23: 1-522.
LEATHERDALE, D. 1958. A host catalogue of British Entomogenous fungi. Ent.Mon.Mag., 94: 103-105.
LE QUESNE, W.J. 1960. Hemiptera Fulgoromorpha. Hndbk.Ident.Brit.Ins., II (3). R.ent.Soc.Lond.
LINDBERG, H. 1939. Der parasitismus der auf Chloriona - arten(Homoptera - Cicadina) lebenden Strepsiptere Elenchinus chlorionae n.sp. sowie die Einwirkung derselben auf ihren Wirt. Acta.Zool.fenn., 22: 179 pp.
1946. Die Biologie von Pipunculus chlorionae Frey and die Einwirkung von dessen Parasitismus auf Chloriona - Arten. Acta Zool.fenn., 45: 1-50.
181.
LINDBERG, H. 1949. On stylopisation of Araeopids. Acta Zool.fenn., 57: 40 pp.
1960. Die Strepsiptere Elenchus tenuicornis Kby. and ihre Wirte Calligypona propinqua (Fieb.) and C.anthracina (Homoptera: Araeopidae). Comm.Biol., 23: 1-7.
LOCKET, G.H. and MILLIDGE, A.F. 1951, 1953. British Spiders I-II. London, 310 pp., 449 pp.
LOEW, H. 1841. Horae anatomicae Beitrgge zur genaweren anatomischen Kenntnis der Evertebraten. Abteilung 1. Entomotomien.
MACLEOD, J. and DONNELLY, J. 1957. Individual and group marking methods for fly-population studies. Bull.ent.Res., 48: 585-592.
MASSEE, A.M. 1954. Personal communication (quoted in Carvalho and Southwood, 1955).
MORCOS, G. 1953. The biology of some Hemiptera - Homoptera (Auchenorhyncha) Bull.Soc.Fouad ler Ent., 37: 405-439.
MUIR, F. 1921. The Sugar cane leafhopper and its parasites in Hawaii. Hawaiian PlantersRecord, No.3f 108-123.
MULLA, M.S. 1957. The biology of Typhlocyba prunicola Edwds. and T.quercus (Fabr.)(Cicadellidae Homoptera). Ann.ent.Soc.Amer., 50: 76-87.
MULLER, H.J. 1951. Ueber das schlupfen der Zikaden (Homoptera-Auchenor- hyncha) aus dem Ei (2.. Beitrag zur Biologie mitteleuropgischer Zikaden). Zoologica: 37: 1-41.
1959. Photoperiod affecting insects. Verh.Dtsch.Zool.Ges.Frankfurt, 1958: 76-84.
- 1960. Die bedeutung der Photoperiode ins lebens-lauf der Insekten. Z.angew.Ent., 47: 7-24.
PERKINS, R.C.L. 1905a. Leafhoppers and their natural enemies. I.Dryinidae. Bull.Hawaii.Sug,.Ass.ent.Ser. 1: 1-69.
1905b. Leafhoppers and their natural enemies. II. Epipyropidae. Ibid. 1: 75-85.
1905c. Leafhoppers and their natural enemies. III. Stylopidae. Ibid. 1: 90-111.
182.
PERKINS, R.C.L. 1905d. Leafhoppers and their natural enemies. IV. Pipunculidae. Ibid. 1: 123-157.
- - 1906a. Leafhoppers and their natural enemies. VIII. Encyrtidae, Eulophidae, Trichogrammidae. Ibid. 1: 241-267. 1906b. Leafhoppers and their natural enemies. Introduction. Ibid. 1: i - xxxii.
1907. Parasites of leafhoppers. Ibid. 4: 5-59. PHILLIPSON, J. 1960. A contribution to the feeding biology of Mitopus morio. J.Anim.Ecol., 29: 35-43.
PHIPPS, J. 1949. The structure and maturation of the ovaries in British Acrididae (Orthoptera). Trans.R.ent.Soc.Lond., 100: 233-247. RICHARDS, 0.W. and VJALOFF, N. 1954. Studies on the biology and population dynamics of British grasshoppers. Anti-Locust Bull.No.17: 1-182.
RICHARDSON, C.H. 1918. Pulsatile vessels in Aphidae. Psyche, 25: 15-17. - rd in Roeder, 1953). Erratum: Page 156 line 15, For "lower" read "higher" 100 pp. Addendum: SHORT, J.R.T. 1952.. The morphology of the head of hanism of wing larval Hymenoptera with special reference to Lond. (A) 36: 63-66. to the head of the Ichneumonoideatincluding a classification of the final instar larvae i of the British Isles. of the Braconidae. Trans.R.ent.Soc.Lond., 221: 27-84
.4,41...1,11" Li L.11. 7. pduiviJyY. New York, 757 pp.
STRUBING, H. 1956. Uber Beziehungen zwischen Ovidukt, Eiablage and naturlicher Verwandtschaft einheimischer Deiphaciden. Zool.Beitr. 2: 331-357.
TODD, V. 1948. Key to the determination of the British harvestmen (Arachnida, Opiliones). Ent.Mon.Mag., 84: 109-113.
1949. The habits and ecology of the British harvestmen (Arachnida, Opiliones), with special reference to those of the Oxford district. J.Anim.Ecol" 18: 209-229. 1950. The prey of harvestmen (Arachnida, Opiliones). Ent.Mon.Mag., 86: 252-254. 183.
VERRALL, G.H. 1901. British Flies. VIII. London, 691 pp. WAGNER, W. 1913. Ober die Biologie von Conomelus limbatus Fab. Z.wiss.InsektBiol., 9: 120-122.
WHALLEY, P.E.S. 1956. On the identity of species of Anagrus (Hym. Mymaridae) bred from leafhopper eggs. Ent.Mon.Mag., 92: 147-149.
1958. Studies on some Homoptera associated with Juncus. Ph.D. Thesis. University College of North Wales.
WIGGLESVORTH, V.B. 1954. The physiology of insect metamorphosis. Cambridge, 152 pp.
1953. The principles of insect physiology. 5th Ed. London, 546 pp.
WILLIAMS, F.X. 1931. The insects and other invertebrates of Hawaiian sugarcane fields. Honolulu, 400 pp.
J.R. 1957. The sugarcane Delphacidae and their natural enemies in Mauritius. Trans.R.ent.Soc.Lond. 109: 65-110. 184. APPENDIX.
Table 1. The distribution of delphacids in core samples compared with the Poisson series.
No. of delphacids Frequency (0) Expected frequency (0 E)2 per core Poisson series (E)
0 22 11.38 9.91
1 13 16.84 0.88 2 8 12.46 1.60 3 1 6.15 4.31
4 1 2.28 0.72
5 0 0.67 0
6 3 0.17 47.06 7 0 0.04 0
8 1 0.01 98.00
9 0 10 0
11 0
12 1
Total 50 50 162.48
x2 for 7° freedom = 162.48 Therefore, the observed distribution is significantly different from a
Poisson distribution (P -4:0.001), and the delphacids are distributed non— randomly.
185.
Table 2. Rough estimation of Lycosid spider population by marking and
recapture, using simple Lincoln index calculation.
First release 7/6/1961 Number marked and released 141
Captures 9/6/1961 Number captured 90
Number marked 10
Number in plot = 141 X 90 10
1269
Number per square foot = 0.11
Second release 9/6/1961 Number marked and released 77
Captures 12/6/1961 Number captured 97
Number marked 5
Number in plot 77 x 97 5 1494
Number per square foot 0.13