THE BIOLOGY OF FOUR SPECIES OF SOIL-INHABITING COLLEMBOLA

Transcription

1 THE BIOLOGY OF FOUR SPECIES OF SOIL-INHABITING COLLEMBOLA by G.D.Sharma A thesis submitted to the Faculty of Graduate Studies and Research in partial fulfilment of the requirements for the degree of Master of Science. Department of Entomology, McGill University, Montreal. April 1962

2 TABLE 0 F C 0.N T E N T S Page I. INTRODUCTION... General remarks 1 2 Review of the Literature Acknowledgments.. II. MATERIALS AND METHODS 9 General 10 Isotoma notabilis 12 Folsomia similis 13 Pseudosinella petterseni and f.alba 14 Il III. ISOTOMA NOTABILIS SCHAFFER, 1896 (ISOTOI<rDAE) General Remarks 16 Description of Adult 19 Food 22 Ma ting 23 Parthenogenesis 24 Oviposition 25 Eggs Hatching 28 Postembryonic development 28 Moulting 30 Predators 31 Field Observations 33 Comparison of!ife-histories of Isotoma species

3 Page IV. FOLSOMIA SIMILIS BAGNALL, 1939 (ISOTŒ~DAE). 39 General remarks Description of Adult General 42 Genital Apertures Variation in the Setae of the Manubrium... Food.... Eggs..... Hatching Postembryonic development Comparison of!ife-histories of Folsomia species 57 " V. PSEUDOSINELLA PETTERSENI BORNER, 1901 AND PSEUDOSINELLA ALBA {PACKAftD, 1873) ( ENTOMOBltYIDAE) 62 General remarks 63 f.petterseni 63 P.alba Description of Adult The male genital plate 66 f.petterseni 67 P.alba Spermatophore 68 f.petterseni 68 f ~

4 Food...f.petterseni P.a1ba Eggs Page t.petterseni 72 f.alba Postembryonic development t.petterseni P.a1ba -- Effect of temperature on postembryonic stages P.petterseni 77 t.a1ba 79 Field Observations 80 Other Observations S1 Comparison of the life-histories of f.pettersen~ and P.alba 81 VI. CONCLUSION General discussion Summary General!_.notabi1is F.similis - t.petterseni and P.alba VII REFERENCES V:III. TABLES IX. FIGURES

5 I. INTRODUCTION

6 2 GENERAL REMARKS Soil animais, of which micro-arthropods constitute a very high proportion, do not merely play a subsidiary role in the formation of different humus forms, but a most decisive one (Kubiëna, 1955). In order to know more about this role, however, it is necessary to understand something of the biology of these animals. Now, the Collembola are among the most abundant soil arthropods both as regards numbers of species and numbers of individuals, but although some published information on the biology and!ife-histories of some species exists, it is surprising how little is in fact known. The present study is an attempt to augment to some extent our meagre knowledge of this group with the study of four species commonly found in soil and litter in eastern Canada; namely Isotoma notabilis Schaffer, 1$96 and Folsomia similis Bagnall, 1939 (Isotomidae), and Pseudosinella petterseni Borner, 1901 and Pseudosinella alba (Packard, 1$73) (Entomobryidae). Information has also been obtained regarding Pergamasus ~rassipes (Linné, 175$) a common predator mite which attacks at least one of the species, I.notabilis. REVIEW Q!!li LITERATURE Biological studies on collembola seem largely to have escaped the attention of biologists till the

7 3 beginning of the present century. However, even 120 years ago Nicolet (1842) gave some general information including the description of various types of eggs. Eighty years ago Lemoine (1883) described the method of copulation in Smynthurus [= Allacma] fuscus (Linné) and, at the turn of the century, Lie-Pettersen (1900) did the same for Smynthurus [= Bourletiella] novemlineatus (Tullberg). Hoffmann (1911) experimented with the rearing of Tomocerus flavescens (Tullberg}, and Folsom (1919) described differences between the young and adult forms of Tomoeerus vulgaris (Tullberg) and change of coloration during the postembryonie development of Anurida maritima (Guérin). Macnamara {1919; 1924) makes some observation on the biology of collembola in general and on their feeding habits, and Handschin (1926) gives some information on certain species, including Isotoma saltans (Nicolet). Holdaway (1927) and Davies (1928) throw some light on the bionomic and economie status of Smynthurus viridis (Linné). Strebe1 (1929; 1932; 1938) gives some account of the biology of Hypogastrura purpurescens (Lubbock), Sminthurinus niger (Lubbock) and Tomocerus vulgaris. Ripper (1930), besides showing the effect of moisture on Hypogastrura manubrialis (Tullberg), also describes the various life stages of this species. Davidson ( b) presents extensive information on the influence of temperature, rainfall and other

8 4 environmental factors on the eggs of Smynthurus viridis, and MacLagan (1932) has published ecological studies on the same species. James (1933) describes the Collembola of the Toronto region with notes on the biology of Isotoma [= Isotomurus] palustris (MÜller) and Davis and Harris (1936) give a detailed life history of Pseudosinella violenta (Folsom) [~ ~.petterseni]. More recently, Lindenmann (1950) has described the postembryonic development of several species of Orchesella, Britt (1951) has published information on Hypogastrura armata {Nicolet), Schaller ( ) describes the structure of spermatophore and its deposition on the substratum by Orchesella villosa {Geoffroy), Brown (1954) has observed Isotoma viridis Bourlet, feeding on nematodes, and Cassagnau (1955; 1956) has experimented with the affects of temperature on several species. Paclt {1956) reviews nearly all physiological and biological literature on collembola until that date. Within the last few years an increasing interest has been taken in the biology of collembola. For example, information regarding Tomocerus minutus Tullberg is given by Uchida and Abukawa (1956) and by Uchida and Chiba (1958; 1959). Murphy and Doncaster (1958) note collembola feeding on the cysts of injurious plantparasitic eelworms, and Poole {1959; 1961) has given a good account of the feeding habits and vertical distri-

9 5 bution of several species of collembola. Milne (1959; 1960), besides dealing with the ecology of collembola, a1so gives a brief account of life histories of seven species of collembola, inc1uding the observation that ~nychiurus latus Gisin changes its color at 12 C. He further tested the chemical nature of the pigment of the same species. Bellinger (1960) suggests that the contrasting patterns of sorne species of collembola are examples of warning coloration. Goto (1960a) has shown that Folsomia candida Willem can be parthenogenetic; Goto and Ogel (1961} discuss variation in the mucro; Torne (1961) did some experimenta on the food requirements of the same species, and Marshall and Kevan (1962) confirm the occurrence of parthenogenesis in!.candida and discuss this phenomenon for the Co11embola generally. They also i11ustrate the embryonic deve1opment in F.candida and indicate the effect of temperature on various life stages of the same species. A we11 known habit of certain co1lembola is that of aggregation in vast numbers which has been recorded numerous times- see, for examp1e, Davies (1932), Wray (1945}, or Park (1949), who give accounts of aggregation in Podura aquatica Linné, Achorutes armatus [: Hypogastrura armat~ and Hypogastrura bengtssoni (Agren), respectively. Pac1t (1956) reviews the subject with other topics up to that date.

10 6 References to natural enemies of collembola are not very numerous, but Steinbock (1931) and Lipovsky {1951) give sorne information; Paclt (1956) also reviews the subject. Karg (1961) has very recently given a detailed account of the feeding habits of thirteen species of gamasid mites which attack collembola. His information is based on comparing the depths of occurrence, structural characteristics and population fluctuations of predator and prey in natural environmental conditions as well as under the influence of insecticides. These predatory mites were seen feeding on collembola by the same author. Detailed information on the morphology of collembola is rather scanty and most of what is known can be found in general entomological texts, such as Imms (1957). Denis (1949) probably gives the fullest account, but Paclt (1956) includes some more recent references. One of the earliest authors to discuss morphology was Lubbock (1873), and among notable contributions to our knowledge of the morphology of collembola are those of Hilton (1914; 1936) who describes the nervous system of Neanura gigantea (Tullberg) and for other collembola also, of Willem (1900} and Davies (1927) who give details of the tracheal system in Smynthurus fuscus and S.viridis respectively, and of Janin (1947) and Carpentier (1947; 1949) who have studied the

11 7 structure of the thorax. Philiptschenko (1912) gives a good account of embryology in Isotoma cinerea (Nicolet) after which no such description is available for any species, although superficial accounts of the gross embryology have been published, such as that for Folsomia candida by Marshall and Kevan (1962). Publications regarding ecological and taxonomie studies on collembola are too numerous to cite here, but a most valuable bibliography of the Collembola from all aspects is that of Salmon (195la; 1956).

12 ACKNOWLEDGN~NTS I am grateful to Professer D.K.McE.Kevan who supervised this work, for his valuable advice, help and encouragement and for the facilities which he made available tome. I wish to record my thanks to Dr.H.Gisin of Geneva and to Dr.W.R.Richards of Ottawa for taxonomie assistance. I am indebted to the National Research Council of Canada for financial assistance in the present work.

13 9 II. MATERIALS AND METHODS

14 10 GENERAL During the months from June to October, 1960 and 1961, soi1 samp1es, 2.4 inches (6.1 cm.) in diameter and 5 inches (12.7 cm.) in depth, were taken from mu11- type soil under sugar-map1e trees (Acer saccharum L.) in the Morgan Arboretum, Ste. Anne de Bellevue, Quebec, and extracted by means of Tullgren funnels (Haarl~v, 1947; 1955). Some samples of pot-soil from the greenhouses at Macdonald College, Quebec, were also extracted. Numerous species of small arthropods were collected in the living state on wet filter papers in glass jars (8.2 cm. in diameter and 9.5 cm. in height), and from these filter papers the various species of collembola and mites required for study were transferred to glass rearing jars (6.2 cm. in diameter and 7.2 cm. in height), using a fine camel hair brush under binocular microscope. These rearing jars had tight-fitting screw caps and were basically similar to those described by Edwards {1955) and Goto (1960b). The rearing jars were filled to a depth of 2.0 cm. with sand in order to occupy the excess space while at the same time helping to supply moisture to a plaster-charcoal layer 2.0 cm. deep, which was poured over the sand. Each species was reared separately. The plaster layer was moistened at monthly intervals and remained fairly constant at 95 to 100 per cent. relative humidity (Hubber, 1958).

15 11 At first on1y yeast was added to the rearing jars as food for all species of collembola, but later it was found necessary in the case of f.similis and P.alba to add fragments of partly decayed leaves of sugar-maple (~ saccharum L.) and American elm (Ulmus americana L.) in order to establish strong colonies. As a source of food for the predatory mite, Pergamasus crassipes (Linné) (Acarina: Parasitidae), which was also reared in the laboratory, a supply of living collembola (Isotoma notabilis) and mites (Tyrophagus putrescentiae(schrank~ was provided. For individual rearing, cells similar to these described by ~~rshall and Kevan (1962) were used, except that those employed in the present work were slightly 1arger - 9 mm. (instead of 6 mm.) in inner diameter. This was to provide more room for group rearing. To check the growth of moulds in the rearing jars and cells, a solution of one part in a thousand terpinol in water was used as recommended by Edwards (1955) and employed by Marshall and Kevan (1962). In addition to laboratory studies some field observations were also made in selected sites as noted in the appropriate parts of the text. In connection with these field studies, soil temperatures were recorded

16 12 by means of a thermistor direct-recording thermometer. 1 For morphological studies, specimens were killed and fixed in Gisin's fluid C (Kevan, 1955). 2 For microscopical examination, specimens were mounted in a medium almost identical to Salmon's (195lb) PVA mounting medium A 2 3 and examined by means of a phase contrast microscope. The four species selected for study were not all reared in an identical manner. ISOTOMA NOTABILIS For the study of growth and development of instars, first instar individuals were separated on the day of hatching and reared to maturity at 17 C. They were then placed in groups of ten together in a cell. Small clutches of eggs laid by the females of these groups were then placed on the day of oviposition in 1 "Tele-thermometer"; Model 44 TE, Yellow Spring Instrument Company Incorporated, Yellow Spring, Ohio. 2 Ethyl alcohol (95%), 750 ml.; Rectified ether, 250 ml.; Glacial acetic acid, 30 ml.; Formalin (40% formaldehyde solution in water), 3 ml. 3 Polyvinyl alcohol {"Polyviol" W 28/20, Reference number , Wacker Chemie GmbH, MÜnchen 22, West Germany), 2.5 gm.; Lactophenol (50 gm. Phenol crystals in 50 ml. lactic acid), 30 ml.; Distilled water, 10 ml.

17 13 separate cells and reared at 14, 11, à, 6 and 4 c. The number of instars was studied up to the time of first oviposition by the resulting females. To observe the numbers of ovipositions and of eggs laid, adult females from severa! cells, kept at each of the above temperatures, were isolated after the first oviposition and placed in individual cells. Records of ten egg-laying females at each temperature were kept. FOLSOMIA SI~ULIS For observations on the life history, thirty individuals were selected at random from the rearing jars and reared individually in small cells. Twelve of these laid eggs which hatched in the rearing cells. In order to study the daily increase in length and the duratien of each stadium, a total of fifteen individuals was removed just after hatching from two of the cells and placed in further individual cells. In the remaining ten cells the young that hatched (from three to seven individuals in each) were allowed to live together at 24 c. (approximately) until they matured. Data regarding the total number of days required for postembryonic development were obtained from these groups. To study the effect of different temperatures,

18 14 twenty-one egg batches were taken at random from the communal rearing jars just after they were laid. Each batch was transferred to an individual rearing cell and then placed in a refrigerated incubator and kept at 22, 17 or 11 c., seven cells being kept at each temperature. From three to seven eggs hatched in each cell. Information on the number of instars before the first oviposition and the total number of days required for postembryonic development was recorded and the data compared with those obtained from the groups reared at 24 C. {above). A further batch of thirty eggs was kept at 4 C. and the incubation period observed. PSEUDOSINELLA PETTERSENI!liQ f.alba The number of moults and the duration of the various instars prior to oviposition were studied by isolating young individuals just after hat ching in groups of from 2 to 6 in small rearing cells..r.earing was carried out at 11, 17 and 24 c. At 11 and 24 c., 6 and 7 ce1ls respective1y were used for each species. At 17 c.' 7 ce1ls were used in the case of P.:eetterseni and 6 for f.alba. To observe the incubation period of each species at low temperatures, 30 eggs were placed in each of two rearing jars a few hours after they had been laid and then kept at 4 C.

19 15 III. ISOTOMA NOTABILIS SCHAFFEa, 1896 ( ISOTOlVliDAE) "

20 16 GENERAL REMARKS Isotoma notabilis was described from Europe by Schaffer (1896). From North America Folsom {1937} described Isotoma eunotabilis, but Stach (1947) states that this represents only a geographical race of I.~bilis, a species which probably occurs throughout the whole Holarctic region. Maynard {1951), however, continues to use the name I.eunotabilis for specimens collected from New York State, as do Bellinger (1954) for Connecticut, Wray and Knowlton (1956; 1961) for Idaho and Utah, and Wilkey (1959) for Californian material. The species also has the following additional synonyms: Isotoma menotabilis Borner, 1903, Isotoma delicatula Brown, 1929 and Parisotoma notabilis Bagnall, The material studied herein was identified as I.notabilis by Dr.W.R.Richards, Ottawa. The species has hitherto been recorded in Canada from Ontario by Macnamara (cf.mills, 1934} and from the British Columbia Alberta border (Rocky Mountains), the Yukon-Mackenzie border (Richardson Mountain), Mackenzie District (Reindeer Station, YellowKnife and Coppermine}, northern Manitoba (Churchill) by Hammer (1953}; it is known also from Alaska (Hammer, ~~.cit.). I.notabilis is a widely distributed species, known especially from the northern hemisphere. It probably originated in the north temperate zone {Stach,

21 ). Outside Continental North America it has been recorded under one name or another from Europe (Poland, Slovakia, Lithuania, White Russia, west and central Ukrain~ Crimea, Caucasus, Bulgaria, Hungary, Germany, Austria (including Styria), Italy and Sicily, Spain, England and Swedish Lapland), Iceland, Faroe Island, Greenland, New Zealand, South Australia and Costa Rica (see Stach, 1947; Hammer, 1953). It may be noted that Stach (.2.E,.Cit.) states that it is "knowntt from Asia, but his text makes it clear that this was a misprint for nunknownu. Christiansen (1958a), however, records it from Syria. Generally it is an inhabitant of litter, but it is also found in moist forest soil covered with moss, on old tree stumps, in moist humus soil and under flower pots in greenhouses, and in caves. At I~cdonald College, Ste. Anne de Bellevue, Quebec, it has been collected from mull- and mor-type soils in the Morgan Arboretum and from heaps of moist rotting leaves. So far, the biology of only three species of Isotoma {one of these since removed to another genus) has received more than casual attention. Handschin (1926) and Steinbeck (1931; 1939) give some information on the "glacier fleau,!.saltans (Nicolet), James {1933) has touched on a few aspects of the biology of Isotoma [= Isotomurus] palustris (M;ller), and Milne (1959; 1960) has given a few details concerning Isotoma viridis

22 18 Bourlet. Milne, however, was unable to rear the last species to maturity, so that few facts are known concerning its life history. Some further scattered information concerning Isotoma species is also known, and references will be found in Paclt (1956) who also reviews the physiological literature. Several species in addition to I.saltans, such as!.hiemalis Schott, I.kosiana Bagnall and I.nivalis Carl, are well known for their association with snow and ice, their mass occurrences sometimes being spectacular; Kos {1944) gives some information on I ~ Paclt ~! kosianaj, and Gisin {1960) notes that adults of I.nivalis are not found in summer. Poole {1959) has described the gut contents of I.notabilis, and Macnamara (1924) pointed out that I.viridis may be regarded as carnivorous since it can kill I.notabilis and devour it; I.macnamarai Folsom ~ I.grandiceps Reuter] is also carnivorous, and! sepulcralis Folsom is a carrion feeder (Macnamara, 1924). Poole <2E ~ ) also noted that I.viridis will feed on I.notabilis and Brown (1954) observed that it will eat nematodes. Schaller (1949), cited by Poole {1959), has also observed that, whereas Tomocerus flavescens (Tullberg) and Onychiurus armatus (Tullberg}, feed on the excrement of larger arthropods, Isotoma olivacea Tullberg is not coprophagous. Denis (1949) states that Isotoma [now Isotomuru~J palustris

23 19 will eat bacterial slime. Other observations on species of the genus {or closely related forms) include those of Folsom (1933), who refers to the economie importance of! Palustris and! [now Isotomode~ denisi Folsom, which damage the roots of sugar-cane, and, in the case of!.denisi, of pineapple also (there are also several other references in the literature to this - see Paclt, 1956), of Davies {1925) who associates I.viridis and!.palustris with damage to mangolds, and of Splendore (1912), who records the latter species damaging tobacco. Cassagnau (1956) refers to variation in chaetotaxy in!.propingua Axelson brought about by unfavorable temperatures, and Christiansen (195Sa) has pointed out certain variations in the abdominal setae of Syrian I.notabilis. DESCRIPTION OF ADULT Before the first moult the juveniles of!.notabilis are almost colorless, but during development a light grey color develops. Young adults are greyish or light bluish in color. Older adults become dark blue with a pale mid-dorsal line. The bodies of most specimens of all ages after the first moult glisten. The length of the largest specimen observed was 0.9 mm.; that of the smallest adult was 0.72 mm. Stach (1947) and Gisin (1960) give the maximum length as 1 mm.

24 20 The body of the adult is densely covered with moderately smooth setae. Stach (1947) describes a pair of longer, erect, dorsal bristles on each abdominal segment except the first, but specimens collected in southern Quebec have such bristles on the fifth and sixth segments only. Christiansen {1958a) also indicates the absence of long bristles on the second and third abdominal segments. On either si de of the head are two black patches each having four ocelli, the black pigment usually being denser around their circumferences. The post-antennal or gans are close to the eye-patches. They are broad and elliptical in shape and about as long as a whole eye-patch. The third antennal segment of I.notabilis is said to be a little longer than the second (Stach, 1947), but in specimens collected in southern Quebec the second antennal segment is slightly longer than the third. The claws of the legs are without inner and lateral teeth. The ratio between the lengths of the third and fourth abdominal terga is about 7:8 in material from southern Quebec, but appears to be subject to sorne variation (see below). The ventral tube is quite long and beset with many setae. The manubrium of the springing organ is shorter than the dentes and clothed on all

25 21 aides with many stiff setae. The mucro is usually tridentate with the apical tooth hooked, the antapical tooth suberect, smaller than the apical, and the third tooth proximal. The term "proximolateral", used for the third tooth by Mills (1934) and Maynard (1951), is misleading since a "proximolateral" tooth, would infer that there is one tooth on either side making four teeth in all, when only three are present. Borner (1901, ~~ also Stach, 1947) first pointed out a variation in the mucro of this species: "Mucro mit 3 sehr selten mit 4 Zahnen". instead of one. His variation, however, had two proximal teeth Similar variations in the mucro of other collembola, auch as Folsomia candida Willem, are " reported by Goto and Ogel (1961) and differences in the number of setae of the manubrium have been observed in Folsomia similis Bagnall by Sharma and Kevan (1962). Since the name I.notabilis is not in general use in North America, it might be appropriate to comment on the alleged differences between this species and l eunotabilis. According to Mills (1934), I.eunotabilis differs from I.notabilis in the shape of eyes, the size of the third and fourth abdominal segments, the proportions of the second and third antennal segments, and in the chaetotaxy of some abdominal segments. The second and third antennal segments in!.eunotabilis are said to be equal in length, whereas in I.notabilis there appears

26 22 to be considerable variation (see above). The lengths of the third and fourth abdominal segments are also alleged to differ between I.eunotabilis and I.notabilis. In the former it is said to be 56:66, whereas in the latter it is given as 46:47. Quebec material differs again {as above), but the ratio is fairly close to that given for I.eunotabilis. In view of the amount of variation, it would appear that Stach (1947) was justified in synonymising I.eunotabilis with I notabilis, but further investigation of biology and variation are desirable to settle the problem finally. FOOD Macnamara (1924) has divided the Collembola according to their feeding habits into two groups: the suctorial, which have no molar plate on the mandibles, and chewing forms, which have a well developed mandibular plate. The suctorial forms are regarded as being carnivorous and the chewing forms vegetarian. I.macna ~arai Folsom ~ I.grandiceps Reute~ is raptorial and devours prey, including other colembola, as does I.viridis (Macnamara, 1924; Denis, 1949). The latter has also been observed to be carnivorous in the laboratory during the present investigation. I.sepulcralis Folsom feeds on dead fish and differs from other carnivorous collembola in having a molar plate on its jaws

27 23 (Macnamara, 2E ~.). I.notabilis, however, is a vegetarian in the strict sense of Macnamara. I (now IsotomurusJ palustris feeds on bacteria (Denis, 1949), and I.saltans on coniferous pollen (Handschin, 1926). The latter seems to be a more general feeder and pollen constitutes only part of its diet (Steinbock, 1931}; the same is true of I kosi [= I.kosianaJ (Kos, 1944). According to Agrell (cited by Poole, 1959), soil collembo1a are unspecialized feeders. Sorne species can certainly thrive on food which presumably does not form their natural diet, such as yeast (Britt, 1951; Marshall and Kevan, 1962). In the laboratory I.notabilis thrives well on decaying leaf fragments and on yeast. It also feeds on fungi. Poole (1959) found that, out of ten ~.notabilis he examined, the gut contents of eight contained hyphae, spores being present in two specimens. Lignin or cellulose and mineral particles were also present in small but constant proportions. In the present study I.notabilis has been reared in the laboratory for about four months purely on Mucor mycelium. MATING During development it is difficult to distinguish between living males and females of I.notabilis. There seems to be no sexual dimorphism in this species. Only with mounted material can the sexes be different!-

28 24 ated by the form of genital aperture. Little is known regarding the mating behavior of collembola, but the subject has been briefly reviewed by Paclt (1956) and by Marshall and Kevan (1962). In I.notabilis no special behavior of any sort during the reproductive period has been observed. Despite careful observations, no deposition of stalked spermatophores, such as those described by Schaller (1954) for Orches ~lla, has been seen in I.notabilis. It is, however, possible that the males deposit spermatophores directly on the ground and that the females pick them up by means of their gonopores in the manner referred to by Schaller (22.cit.), for it has been observed severa! times that individuals will touch the substratum with the tips of their abdomens on occasions other than the deposition of faecal matter. It is perhaps significant that this activity increases during the reproductive period. Due to the form of gonopore it is likely that the female could pick up a spermatophore in this manner. PARTHENOGENESIS The subject of parthenogenesis in the Collembola has been briefly reviewed by Marshall and Kevan (1962) who found, 1ike Goto (1960a) and Goto and Ogel (1961), that Folsomia candida Willem could be exc1usive- 1y parthenogenetic. A number of eggs of 1.notabilis

29 25 were therefore separated on the second day after being laid and placed individually in rearing cells and incubated at 17 C. On hatching they developed normally and, after the fourth moult, some of the resulting females laid three to four eggs each, although ethers did not lay. The eggs, however, did not hatch. All the individuals were then placed together in one rearing cell. Further eggs were laid, and these proved to be viable, indicating that fertilization had occurred. After a short period of communal existence each individual was again separated from its fellows and the females continued to lay fertile eggs. Out of a batch of eleven eggs laid by a single female, four males and seven females were obtained (as determined by the subsequent microscopie examination of the mounted progeny}. Thus, although unmated females will oviposit, there is no evidence of parthenogenesis in this species. Dr.Christiansen (!a litt., 1961) notes that in preliminary investigations with two ether species of Isotoma, evidence of parthenogenesis is also lacking. OVIPOSITION Not a great deal is known concerning oviposition by species of Isotoma, but Handschin (1926} notes that the eggs of! saltans may be so numerous that stones upon which they were laid appeared to be covered

30 26 with a yellow dust. 0 Generally, at 17 C., I.notabilis females begin to lay eggs after the fourth moult. the small rearing cells they were prevented from descending below the surface to oviposit because of the rather smooth nature of the plaster of Paris and charcoal bases. They therefore laid eggs in small pits in the plaster, if these were present, or under leaf fragments or around other food material. Several times eggs were found glued to the smooth walls of the cells, which suggests that, in the field, the species may glue its eggs under debris, or on soil particles. I.notabilis does not hesitate to oviposit on or under a fine network of fungus mycelium. hatched. In Eggs laid in such situations also During oviposition the female lowers the posterior end of the abdomen and depresses the head. Eggs are laid singly but are heaped together. A batch of eggs takes only a moment to lay. If the female is disturbed, it does not resume laying at once, but delays until the next period of oviposition. Some fluid is secreted from the genital pore together with eggs, which enables freshly laid eggs to be glued together. Apparently no mixture is secreted from the anus as reported for Smynthurus viridis by MacLagan (1932).

31 27 Eggs are laid in small clutches and are at first smooth, spherical and colorless. The size of a normal egg is approximately 0.13 mm. Occasionally, at the first oviposition, some of the eggs are smaller than usual, and such eggs measure approximately 0.08 to mm. and usually fail to develop. This may be due to the relative immaturity of the individuals when the first batch of eggs is laid {MacLagan, 1932). When reared at 17 c., the average number of eggs laid at the first oviposition was just over seven, but the maximum number (17) was produced at the third oviposition. Four ovipositions were observed at intervals of approximately six days. At lower temperatures the total number of eggs laid and the number of ovipositions were progressively reduced until, at 4 C., an average of only three eggs were laid per female, there being but a single oviposition. Young adults survived for more than a month at 0 C., but laid no eggs. Lowering the temperature thus reduces the fecundity in I.notabilis {Table I). On the second day after being laid normal eggs incubated at 17 c., increased from 0.13 to 0.14 mm. in diameter and changed in color to a pale yellow. Further increase in size was followed on the third day when the

32 28 eggs reached a maximum aize of 0.16 mm. On the same day the chorion ruptured and divided into two equal parts. No subsequent increase in size occurred after the chorion had ruptured. The method of hatching is similar to that referred to by Marshall and Kevan (1962) for Folsomia candida. HATCHING At 17 c. the eggs required seven to eight days for hatching. With a decrease in temperature the time required for hatching was increased (see Table II). The percentage of eggs hatching decreased from 87 to 49 per cent. when the temperature was lowered from 17 to 6 C.; 0 figures for 11, 8 and 4 C. are unfortunately, not available. Lowering of temperature thus has the same effect on the eggs of I.notabilis as it has on those of Folsomia candida (Marshall and Kevan, 1962). In that species the incubation period increased and the percentage of eggs which hatched dropped from 88.9 to 52 per cent. when the temperature was lowered from 24 to 16 c. POSTEMBHYONIC DEVELOPMENT Juveniles of I.notabilis are almost pure white in color and measure about 0.32 mm. in length immediately

33 29 after hatching. Since collembola are ametabolous, their young have more or less the same body characters as the adults, but it may be noted that the eye patches have the same number of ocelli in all instars. At 17 C. the first moult takes place on the fourth day, at which time the general coloration changes from white to light blue. The intensity of the blue color increases with age and instar. The number of instars appears to be indefinite, but maturity is usually reached at the fifth instar. The duration of each instar increases as the temperature decreases (see Table III). The total number of days required to complete the life history from oviposition to oviposition at various temperatures is shown in Table IV. Only the females are considered since no method of determining the sexual maturity of males proved possible. The rate of growth of juveniles from hatching until maturity was studied and found to be very uniform. The daily increase in body-length of ten individuals is shown in Table V. The uniformity in the rate of growth and time of moulting in the earlier instars may be due to the uniform and adequate food supply. After egglaying had commenced, however, increase in length was greatly retarded. A single individual of I.notabilis in its fifth instar, which measured 0.72 mm. approximately seventeen days after hatching and which had already begun to lay eggs, when measured on the eighty-sixth day

34 30 after emergence from the egg, had attained a length of only 0.9 mm. MOULTING In I.notabilis, as in other collembola, moulting continues after the adult stage bas been reached, although little increase in size occurs in the later instars. Five instars appear to be normal before maturity (see page 29). The length of the time between each moult varies with the individual (see Table III). The moulting process in!.notabilis has been observed, especially on the smooth surface of the rearing cells. At first the swelling of the thorax and peristaltic movements of the body cause the integument to split along the mid-dorsal line of the head, thorax and first abdominal segment. Davis and Harris (1936) have also observed the splitting of the old skin along the mid-dorsal region of the thorax in Pseudosinella violenta (Folsom) f f.petterseni Borner]. During this process locomotion ceases. Peristaltic contractions then force the nota of the thorax through the split, and at the same time the head, which is held within the exuviae, is forced downward. This causes the intersegmenta! membranes to stretch, and simultaneously the antennae are laid under the body. Further peristaltic

35 31 contractions, originating at the tip of the abdomen, force the thorax and abdomen to arch upward. This causes the split to widen, and the animal then becomes free. The moulting process takes approximately five to seven minutes in the adults. After ecdysis there is a short pause before the newly moulted individual begins to walk about. Sometimes an individual may fail to free its furcula from the exuviae and it will then die. Several attempts on the part of the observer to free trapped individuals from such a predicament, or to disturb the process of moulting, all proved fatal. In a few individuals the intersegmental membranes did not return to their normal position after the moulting process was completed. Such individuals assumed an abnormal appearance, having the body a little longer than usual. They were unable to move and survived for only two or three days. PREDATORS Little or nothing is recorded regarding the predators which attack Isotoma species in particular, but Steinbock (1931) says that the phalangid, Parodiellus obliguus (Koch) is an enemy of I.saltans. From litter samples taken from the field, predatory mites belonging to three species of Pergamasus

36 32 (Parasitidae) and one of Hyoaspis (Laelaptidae) were obtained along with I.notabilis. The most abundant of these was Pergamasus crassipes (Linné), although all species would attack I.notabilis in laboratory cultures. Karg (1961) refers to this species as a predator of collembola. The number of P.crassipes obtained during June to September, 1961, from five samples for each month, is shown in Table VI. In the laboratory it was observed that P.crassiEes preferred I.notabilis to species of collembola belonging to other families (Poàuridae and Entomobryidae) when these were present in the same culture. Sometimes E.crassipes caught and ate young Entomobryidae, but the dark brown Hypogastrura armata Nicolet (Poduridae) were never eaten. Bellinger (1960) has in fact suggested that the pigmented podurids may be distasteful to other animals. Lipovsky {1951), who reared several species of Euschoengastia and Trombicula (Trombiculidae) exclusively upon the active forms of Sinella curviseta Brook (Entomobryidae) and their eggs, has also observed that Poduridae were never eaten by the nymphs or adults of these chigger mites. P.crassipes was maintained for three months exclusively on a diet of I.notabilis. This mite, when kept in captivity was observed to catch and devour, or

37 33 at least kill, an average of fourteen I.notabilis per day when reared individually in small cells with an abundant food supply (see Table VII). Twenty-five I.notabilis were given to each mite each day. FIELD OBSERVATIONS From laboratory rearing it was established that, at temperatures above 14 c., the life cycle of I.notabilis can be completed in a month or less (see Table IV). Soil temperatures at a depth of approximately 10 to 12 cm. at the sampling site in the Morgan Arboretum showed that the mean temperature was above 14 c. throughout June to September, 1961, so that it would appear that there could be at least four generations of I.notabilis during this period, and probably more in a full year. Sampling during June to September showed that the population was highest in the month of June when the species could complete its life history within a month. Contrary to expectations, however, there was apparently a decline in numbers in July which also continued into August, perhaps as a result of temperatures being too high for development {laboratory figures are not available for 20 C. or above}. If, however, the temperature in September were not to fall below 14 c., there might again be another generation by October {Table VI indicates a rise in the adult popula-

38 34 tion during September), otherwise the life-history of this generation would probably not be completed until after the winter. Eggs laid in October probably hatch before the temperature becomes too low for development since, even at 6 C., eggs take only just over five weeks to develop. It is therefore likely that young I.notabilis may survive the winter in southern Quebec and mature in spring after the melting of the snow, which would account for the comparatively high population early in the year. Twelve young adults of I.notabilis were in fact kept at 0 c. in the laboratory for more than a month and, of these, seven survived. V.G.Marshall (persona! communication) took soil samples under snow cover in the Morgan Arboretum during February, 1960, and found these to contain living collembola (although ~.notabilis was not positively identified), thus supporting the possibility that r.notabilis may survive the winter as juveniles or young adults. By migrating vertically downwards from the surface of the soil and away from the coldest layers these could escape from the most severe conditions. It should be noted, however, that the soil in winter may have a characteristic "winter fauna" composed of species with a law-temperature tolerance or requiring low temperature for development so

39 35 that confirmation is necessary for Quebec conditions. Bellinger (1954), however, has shown that the adults of this species will pass the winter in the soil in Connecticut. Vertical migration by collembola and mites in winter has also been observed by Wallwork (1959) in a hemlock-yellow birch stand at Imperial Lake, ~lichigan, and by Dr.N.R.Fraser (persona! communication) at Macdonald College, Quebec. Not a great deal is known about the population of the mite predator Pergamasus crassipes in the field, but Table VI indicates that this does not follow exactly the pattern of I.notabilis. Adults build up during the summer, reaching a peak in August and then fall off sharply in numbers with the advancement of cooler weather in September. COMPARISON OF LIFE-HISTOHIES OF ISOTOMA SPECIES As already indicated {page 17) information on the biology of Isotoma species has been published only for I.saltans, I. [= Isotomurus] palustris and J..viridis. Few details are known; the first of these is not known from Canada and will not be considered further. James (1933) found the eggs of!.palustris in the soil of strawberry beds and under damp pieces of wood. In rearing cells, he observed that they were laid

40 36 upon similar wood fragments, usually in batches, but sometimes scattered. The eggs are spherical in shape, light cream in color, and vary in diameter. They are laid 8-25 at a time and take days to hatch at 30 C. The!ife-cycle appears to take about a month at high temperatures and about seven instars have been recorded, but it is not known how many of these instars could be considered juvenile. According to Milne (1959; 1960), I.viridis lives in the upper humus layers under surface litter. The eggs are smooth, globular and pale red in color. They measure about 0.21 mm. in diameter and are laid in batches of In cultures they were laid on the surface of the medium. At 12 C. the eggs take 16 to 20 days to develop. Milne has also shown that raising the temperature by 12 C. from 12 to 24 C., more than halves the developmental time for the eggs of I.viridis {see Table VIII), thus agreeing in general with previous records for the development of the eggs of other collembola, such as those of Ripper (1930) for Hypogastrura!!nubrialis (Tullberg), and Lindenmann {1950) for Orchesella cincta (Linné). These authors also found that the life cycle (including incubation period) were shortened by raising the temperature. Strebel {1932), however, found that the incubation period for Hypogastrura purpurascens (Lubbock) was increased by raising

41 37 the temperature (see ~lilne, 1960). In the present experiments, the number of days required for hatching of eggs of I.notabilis and the total time for completing the life-cycle was considerably reduced with increase in temperature. At the temperatures studied, an increase of 6 C. resulted in the life-cycle being shortened by about half or more (see Tables II and IV). These results are thus in general agreement with the observations of Milne {1959; 1960} for!.viridis, although the effects of temperature change seem to be more pronounced. Milne, however, was unable to study the complete life cycle of I.viridis. A comparison between the life histories of I.notabilis and these two species is given in Table VIII. The differences observed are possibly due to differences in food requirements, size, and ability to withstand cold. In this latter regard it has been noted that!.notabilis can live for more than a month at 0 C. (see page 27), but it is a small species and can probably migrate downwards into the soil during unfavorable conditions more readily than can the other species. It is usually less than 1 mm. in length, which is half the size of! viridis and only one-third as large as ~.palustris. Similarly Tomocerus sp., which is very large by comparison, requires more space to manoeuvre in soil, and this may explain why its population in the

42 J$ month of October is low (Knight, 1961); it simply cannot descend to the warmer soil layers, and thus does not survive. Differences between the species of Isotoma may also lie in their vulnerability to attack by predators, but there is very little information on this aspect of their biology. We know, however, that I viridis may itself be a predator and that I.notabilis can surfer heavy predation by mites such as Pergamasus crassiees (see Table VII).

43 39 IV. FOLSOMIA SIMILIS BAGNALL, 1939 {ISOTOMIDAE)

44 40 GENERAL REMARKS Folsomia similis was described by Bagnall (1939) from England, and recorded by Stach (1947) from Poland, by Gisin (1949) from Switzerland, and by Sharma and Kevan (1962) from eastern Canada. It has been found in garden soil, under flower pots, in vineyard soil, and in mull soil under maple trees (Acer saccharum L.). It is primarily a soil-inhabiting species and probably widely distributed in the Holarctic region. It was transferred to the genus Listeria by Bagnall (1949), but it appears to have no other synonym. Stach (1947) refers to certain morphologieal differences between specimens of this species from Poland and those described by Bagnall (1939) from England, and Sharma and Kevan {1962) briefly indicate variation in the number of ventral setae on the manubrium... Goto and Ogel (1961) also mention structural variation in the mucro of the related species, F.candida Willem, associated with parthenogenesis. The importance of such variation for taxonomie work are considerable and an account of their occurrence in F. similis is given below. As far as the author is aware, the biology of but one species of Folsomia,!.candida Willem, has been studied in any detail. Milne (1959; 1960} provides some information regarding this species, and Marshall and

45 41 Kevan (1962) give further details regarding development under different conditions. They also refer to partheno- genesis (see also Goto, l960a; Goto and Ogel, 1961), kinds of food consumed, and the possible importance for survival, during times of temporary flooding, of the ability of the eggs to hatch on or under water. Torne (1961) has recently presented some interesting resulta on populations of this species reared in the laboratory under different conditions of substrate and temperature. Other observations on Folsomia include those of Folsom (1933) that some species, such as F.fimentaria {Linné) and F.nivalis (Packard}, are often so abundant on the surface of well water as to be a nuisance, and that the former may be injurious to the roots of sugar-cane. Potatoes, carrots and lily-bulbs are also attacked by the same species (Krjukova, 1932; Tomaszewski, 1949). l fimentarioides (Axelson) is also injurious to root crops (Womersley, 1939). Denis (1949) reports that!.fimentaria will feed on pollen. Poole (1959) indicates that the diet of!.quadrioculata (Tullberg) differs for different individuals. The only parasites recorded for Folsomia appear to be micro-organisms originally thought to be sporozoa but which are actually lower fungi (Thor, 1930). In the present study, the biology of! similis

46 42 has been shown to differ from that of F.candida, for which independant observations confirm the resulta obtained by Milne (1959; 1960) and by Marshall and Kevan (1962). DESCRIPTION OF ADULT Qeneral Folsomia similis is greyish-white in color with scattered black spots. The latter are present on all the body segments, but are more abundant on the head and the second and third segments of the thorax. The body is densely clothed with short simple setae, irregularly arranged. According to Stach (1947) the setae occur in irregular rows of 5 or 6 on the first and second abdominal terga and of 6 or 7 on the third abdominal tergum. However, specimens from southern Quebec differ from Stach's Polish material in having 4 or 5, 5 or 6 and 5 to 7 rows of setae on the first, second and tbird abdominal terga respectively. Also the arrangement of the last row of setae on all these segments was always found to be comparatively regular as compared with the condition described for Polish specimens. The last abdominal tergum bears approximately 15 to 20 stiff bristles. The post-antennal organ, which is situated on

47 43 each side of the head below the eyes is narrowly elliptical having a narrow margin all round. In the majority of specimens it is slightly curved. In most of the specimens from southern Quebec, the organ has a transverse septum near the middle (Fig.l}. Stach (1947} states that (in Polish material} there is a slight tendency for the post-antennal organ to be divided in half. Irregular denticulations usually about 7-20 in number are present on the inner posterior margin of postantennal organ of almost all (SS%) of southern Quebec specimens (see Fig.l Band C}, but these are not referred toby Stach (1947). There is a great deal of variation in the number of the denticles, but in only three specimens were they absent. Occasionally a few denticles are also present on the anterior margin {Fig.l C}. The variations were first detected by Dr.W.R.rlichards, Entomological Research Institute, Ottawa, when the specimens were sent to him for identification (personal communication, 1960). The second thoracic segment is a little longer than the third. The first abdominal segment is smaller than the second and the third which are of nearly equal size. The fourth, fifth and sixth terga are ankylosed and together are about twice the length of the second abdominal tergum. There is a faint intersegmental suture between the fourth and fifth but this does not

48 44 reach the lateral margins. The manubrium of the springing organ is shorter than the dentes. According to Stach (1947) it bears eight ventral setae, but Gisin (1960) reports 3+3 or 4+4 setae on the manubrium. Specimens from southern Quebee, however, show great variation in the number and arrangement of setae which vary in number from 4 to 17 {see page 47). The dens has narrow crenulations on its dorsal surface; on its ventral side there are fifteen short setae. The mucro is always bidentate and was not found to vary as it may do in parthenogenetic strains of Folsomia candida (Goto and Ogel, 1961). The anal opening is terminal and triradiate. On viewing it from the ventral aspect, a median anterior and two posterio-lateral slits may be seen. Posteriorly are about nine long stiff setae forming a transverse row between the posterio-lateral slits; anteriorly are about four small setae on either side of the median slit (Fig.2). The anal slits appear to be guarded by three integumental flaps. Mature specimens usually measure between 0.95 and 1.3 mm. in length. The maximum length was found to be 1.45 mm., as recorded also by Gisin (1960).

49 45 Genital Apertures The male genital plate has proved useful in the taxonomie study of certain species of collembola. Stach (1947} has published diagrams for the males of several species, Mills and Richards (1953) give good descriptions and diagrams for both male and female genital apertures in Uzelia hansoni Mills and Richards (Isotomidae), Christiansen (195Sb) has shown the importance of the male genital plate in taxonomie work on Entomobryidae, and Gisin (1960) describes the male and female genital apertures of Onychiurus silvarius Gisin (Onychiuridae). The male genital aperture in f.similis is situated on a small specialized genital plate on the ventral side of the fifth abdominal segment behind the manubrium. The aperture is guarded by two longitudinal lips, each having two sclerites (genital dises). These latter do not appear to have been previously described for any collembola. Each lip bears three small setae, two anterior and one posterior, just proximad and caudad to the anterior and posterior genital dises respectively. There are fourteen to fifteen small guard-setae round the margin of the genital plate {Fig.J). The female genital aperture in! similis is transverse having an anterior and posterior lip. Each

50 46 lip bears a pair of small, widely spaced setae (Fig.4). A similar type of genital aperture may also be seen in F.candida and Isotoma notabilis Schaffer (Isotomidae), and has been reported by Davis and Harris (1936) for Pseudosinella violenta Folsom [: P.petterseni Borne~ {Entomobryidae), and by Gisin (1960) for Onxchiurus silvarius Gisin (Onychiuridae). The female genital opening in the various species examined is very uniform (although the number of setae on the lips is variable in most). taxonomie work. Because of this it has not been found useful in The structure of the female genital aperture appears to be associated with the method of picking up the spermatophore (page 24). yariation in the Setae of ~ Manubrium Variations in the mucro and body setae in co11embo1a are not unusua1. For examp1e Borner (1901) and Christiansen (1958a) have described sorne variations in Isotoma notab11is, and Cassagnau (1956) refers to changes brought about by high temperatures in the same genus. Sharma and Kevan (1962) refer briefly to variation in manubrial setae in f.similis, and Goto and n Oge1's {1961} investigation of the mucro of Fo1somia candida has suggested that such variations may cause confusion in the correct identification of sorne species of Fo1somia. Of 188 specimens of F.simi1is examined, 60 had

51 47 eight setae on the manubrium, the "normal" number given by Stach (1947), and a further 29 had seven. When the setae number 4+4 (formula of Gisin, 1960), their arrangement is as follows: one seta near the base of each dens, a pair on each side a little in front of these, and one seta a little beyond this pair (Fig.5A). In twenty-two specimens the setae were arranged in the above manner but, in the remaining thirty-eight specimens having eight setae, there was no definite arrangement. In the three specimens having a setal count of less than seven, one had the setae arranged 2+2 {Fig.5C) and in two they were grouped 3+3, there being one seta in each row in all the cases. In 76 specimens the setae numbered between 9 and 12. Twenty specimens had more than twelve setae. Table II shows the number of manubrial setae present in all the specimens examined. Figure 5 shows the setal arrangement in sorne of these. It is clear from the above observations that in Folsomia species the number and arrangement of setae on the ventral side of the manubrium have little or no taxonomie value. FOOD Torne (1961) observes that the nutritional requirements of Folsomia candida are very sma11. Marshall

52 and Kevan (1962) found that this species thrives well on commercial baker's yeast alone and can live and reproduce satisfactorily on a diet of decaying sugar-maple and American elm leaves (!.9..!!: saccharum L. and Ulmus ~mericana L.) and that it shows preference for these species. Their findings were subsequently confirmed independently by Sharrna (unpublished) using a culture of F.candida from the same source as the F.similis herein discussed (see page 10). F.candida can also be reared on bracken (Pteris) spores (Milne, 1959; 1960). Poole (1959) has indicated that individual feeding preferences may exist within a single species of the genus: he found that sorne specimens of f.guadrioculata (Tullberg) feed for preference on fungi, whereas others prefer tissues of the higher plants. Little was known about the food of F.similis, but it soon became apparent that this species differa in its feeding habits from F.candida taken from the same soil samples. At the beginning of the experimenta an attempt was made to rear F.similis on decaying sugarmaple and elm leaves (as with F.candida) but, although they remained alive for three months on this diet, no increase in population occurred during this period. Ordinary commercial baker's yeast was then added to the rearing jars, and, within a fortnight, several batches of eggs had been laid. It was found that f.similis

53 49 could not be reared very successfully on yeast alone, nor was it observed to feed on the moulds which were allowed to flourish after the yeast was added to the rearing cells. If, however, a piece of decaying maple leaf was introduced into rearing cells containing only yeast (and moulds) as food, satisfactory results were obtained, and the Collembola developed normally. The difference in feeding habits between F.similis and f.candida thus helps to explain their existence together in the same habitat. EGGS Folsomia similis lays its eggs in batches of from 3 to 14 together. The eggs are glued together by means of a sticky substance in a manner similar to that reported for F.candida by Marshall and Kevan (1962) and for Isotoma notabilis (see page 26). The substance appears to be secreted at the time of oviposition and is apparent when an egg-mass is picked up with a fine needle. Generally the eggs were found beneath pieces of leaves in the rearing jars, but they have also been observed on the surface of the rearing cells. Macnamara (1919) has noted that oviposition by collembola in general apparently takes place only in the dark and this

54 50 appears to be true for F.similis. Batches of eggs, laid by several individuals, were usually found very close to each other or even in aggregate heaps in the culture jars, and it appears that F.similis females prefer to lay their eggs where they or other members of their group have previously oviposited, although sorne scattered egg batches may also be found. With f.candida, Milne (1959; 1960) and Marshall and Kevan (1962) also observed aggregations of large numbers of eggs, and Macnamara (1919) states, for collembola generally, "that masses often contain from 50 to loo eggs and bulk much larger than the body of the insect, but are usually all stuck together with the appearance of being the product of one female". Such aggregations of eggs may help to explain the patchy distribution of many collembola in the soil. Poole (1961) has given three possible explanations for such aggregations, the first of which is that individuals may not wander far from the original egg clusters as suggested by Murphy (1953) 4. This may be true to some extent for species such as f.candida and f.similis where the young are round together but it certainly does not explain the vast aggregations of adults which have 4 Poole (1961) cites Murphy (1955} in Kevan, D.K.McE. (Ed.) "Soil Zoology", but this is-rncorrect.

55 51 been observed for species such as Hypogastrura armata (Nicolet) and Tullbergia krausbaueri (Borner) which do not deposit in large masses. Freshly laid eggs of F.similis are spherical, soft and pure white in color, but within ten hours the color of the chorion begins to change to light brown. By the following day they become quite dark. With the change in the color of the chorion, gas bubbles appear around the egg, but as the chorion darkens the gas bubbles begin to disappear and by the third day after oviposition they vanish. The eggs of F.similis are therefore similar in their behavior to those of l ~ ~ida {Marshall and Kevan, 1962). At oviposition the egg of! similis is about O.lOà mm. in diameter, but by the second day it expands to about mm. This may be due to the absorption of water. Similar increase in egg size after laying is reported by Britt (1951) for Achorutes [= HypogastruraJ armatus and by Marshall and Kevan (1962) for F.candida. On the third to fourth day the chorion ruptures and no subsequent increase in the size occurs. Unlike! candida, there is no evidence of parthenogenesis in!-similis.

56 52 HATCHING The process of hatching in collembola differa from that of most insects. Since the embryos are not equipped with any special "egg burster", nor do they gnaw a hale with the mandibles, they rupture the egg shell solely by their body movements. On examining the egg of f.similis in water under the microscope just prior to hatching, the body movements are clearly visible. The young collembolan, after rupturing the anterior end of the shell with its head quickly extricates itself and becomes free. Sometimes the anterior end of the egg shell may remain on the head like a cap and, in such cases, despite repeated attempts to free itself, the young collembolan fails to do so and after 24 hours or less it will die. On several occasions an attempt was made very carefully to remove the egg shell fragments from the head by means of a fine needle, but the slightest injury at this stage caused death and such attempts were seldom successful. At 22 and 24 c., eggs took from 9 to 10 days to hatch. No significant difference in the length of the incubation period for the two temperature regimes was observed, but the percent age of eggs hatching at 22 C. was a little greater than at 24 c. ( 83.3% as against 79.2%). Eggs incubated at 17 C. took 12 to 13

57 53 days to hatch and the percentage hatching was again slightly less than at 22 C. (76%), but, when the temperature was lowered to 11 C., less than half of the eggs hatched and the time taken for development increased to about one month. At 4 C. a few eggs hatched (the exact percentage was not recorded) after about two and a half months (see Table X). The optimum temperature for development for F.similis would thus appear to be 22 C. Increased mortality in the egg stage is small when the temperature is raised by a few degrees, but large when there is a considerable decrease in temperature. Marshall and Kevan {1962) observed that, in F.candida, the percentage of eggs hatching was slightly (though not significantly) higher at 24 c. than at 22 c. (88.9% as against 87.8%) and significantly lower at 18 C. The optimum developmental temperature for - - ~~~~ F.similis is thus a little lower than for F.candida which also appears to be able to tolerate somewhat higher temperatures. POSTEMBRYONIC DEVELOPMENT First instar F.similis are pure white in color and measure about 0.32 mm. in length upon hatching. Black pigment spots first appear on the thoracic seg-

58 54 ments after the first moult and not until later instars on the other body segments. From a study of fifteen individuals reared in 0 isolation the daily rate of growth at 24 C. was found to be mostly rather uniform until a length of approximately 0.94 mm. was attained on or after 34 to 39 days (average 36.6) from hatching {see Table XI). In six of the fifteen individuals, however, the rate of growth was much slower and less uniform. One of these six took 50 and another 60 days to attain a length of 0.9 mm., while the remaining four never reached this size although moulting continued irregularly until they died. Irregularities in growth rates are also recorded for Isotoma [= IsotomurusJ palustris (MÜller) by James (1933), who says that, "after hatching, the larvae moulted at irregular intervals, requiring about one month to reach stage No.7. Following this stage, for sorne unknown cause the larvae showed no appreciable increase in length. They appeared active enough in feeding and continued to moult regularly, but never reached the average length (2 mm.) of adults." The majority of the individuals of F.similis referred to above, in addition to showing a fairly uniform size increase, also moulted fairly regularly at intervals of between 6 and 13 days, but in a few instances considerable differences in the time elapsing

59 55 between successive moults were found. In four individuals the fourth instar lasted 16 or 18 days and, in another two, 23 days passed between the fifth and sixth instars (see Table XI). At lower temperatures there is also much variation in the duration of various stadia (see below}. The variations referred to may possibly be due to artificial surroundings and unnatural diet. Wigglesworth (1953) and others have indicated that the level of nutrition may cause great variations in the number of moults and period of maturation in insects. In addition to the observations on isolated individuals reared at 24 C., data were obtained on postembryonic development for groups (of three to seven at a time) reared together at various temperatures. indicated on page 13, ten such groups were reared at room temperature (approximately 24 C.) and seven each at 22, 17 and 11 C. Table XII summarizes the resulta obtained up to the end of the sixth instar. It will be seen that at 24 C. the rate of development was much the same for the group rearing as for individuals reared singly, except that there was greater uniformity. The later instars took about twice as long to moult as did 0 the earliest ones. At 22 C. the rate of development was more rapid, particularly in the later instars; As

60 indeed at this temperature there was a tendency to mature more rapidly, and oviposition by several individuals occurred before the fifth moult (sixth instar), whereas at 24 C. no oviposition occurred before the sixth moult (seventh instar). At 17 C. the average rate of development was somewhat similar to that at 24 c., if anything slightly faster, but there was much more individual variation. 0 At 11 C. the average rate of development was much slower and even more erratic. The young hatching from eggs kept at 4 C. (see page 53) were not included in these experimenta, but little or no development occurred at this temperature before they died accidently. As indicated above, the females, at least, of!.similis do not all become mature in the same instar. A few reared at 22 C. began to oviposit in their sixth instar, but most did not do so until the seventh. other temperature conditions egg-laying also occurred most commonly after the sixth moult {i.e. in seventh instar), but sorne individuals moulted more frequently before oviposition. For example in one of the groups reared at 24 C. the first oviposition did not occur Under until after the tenth moult (the other three individuals in the group did not lay; they were probably males). two other groups (one of three and the other of four individuals) the first oviposition did not occur until In

61 57 after the eighth moult (apparently only one individual in each group was a female). Table XIII summarizes the number of days from hatching to oviposition for females reared at different temperatures. The females concerned were members of the groups referred to in the preceding paragraph. Only those females that began to lay after the sixth moult are included. It is clear from the table that the optimum temperature for maturation is 22 C. at which females will begin to oviposit about 40 days after hatching. A few degrees above or below this temperature appreciably prolong the process, and halving the temperature more than doubles the average time taken for females to reach maturity. COlV1PAiiiSON OF LIFE-HIS'l'ORIES OF FOLSOMIA SPECIES Comparing the biological observations for!.candida given by ~lilne (1959; 1960), Goto (1960a), Goto and Ogel (1961) and Marshall and Kevan (1962), with those now available for!.similis, it seems that the two species differ in respect of food requirements (see page 48), low temperature embryonic development, rate of postembryonic development and the occurrence of parthenogenesis. Information on the biology of the two species is summarized in Table XIV. Marshall and Kevan (2E.cit.) have concluded

62 that 22 C. is optimum for deve1opment of the eggs of F.candida, and that, at this temperature, the number of days between successive ovipositions is minimal. For f.simi1is, 22 C. also seems to be optimum, for, at this temperature, the eggs deve1op as rapid1y as at any other investigated, a higher proportion hatch (see Table X), and the postembryonic and maturation period is shortest {see Tables XI, XII and XIII). The number of days between successive ovipositions in F.similis has not, however, been determined. The rate of embryonic development in F.similis is simi1ar to that of!.candida at 0 0 higher (22-24 C.), and faster at 1ower (4-5 C.}, temperatures. It is perhaps general1y more rapid also at intermediate temperatures a1though Milne (1959; 1960) 0 found that at 12 C. the eggs of!.candida hatched after on1y days as compared with and days for 16 and 18 C. respectively given by Marshall and Kevan (1962). At 11 C. F.similis took twice as long to hatch as did F.candida at 12 c. (Milne's data). There may be sorne discrepancy in Milne's figures since Marshall and Kevan's experiments were more comprehensive. The former author was, however, usïng European materia1 and the latter authors North American. It may also be noted that an increase in the length of the incubation period with an increase in temperature was observed by Strebel (1932, - ~.Milne,

63 ; 1960) for Hypogastrura purpurascens (Lubbock) which took 19 days to hatch at 10 c., and 2$ days at 22 C., but there is no evidence in any species for the bimodal curve suggested by the figures for F.candida given by I~lne, on the one hand, and Marshall and Kevan on the other. F.similis shows a distinct correlation between temperature and rate of postembryonic development: the latter decreases as the former rises. In respect of low temperatures a female of r.similis thrived well for over two months at 0 C. and laid nine eggs. No observation for F.candida at such a low temperature has been recorded. F.similis eggs kept at 0 c. did not hatch after two months at this temperature. They were subsequently killed accidently by desiccation so that it was not possible to discover if they would have hatched had the temperature been raised subsequently. Postembryonic development in E similis is much slower than in!.candida (see Table XIV}. At 24 c. the latter species usually completes its life cycle in about three weeks, while the former takes about two months at 0 the same temperature although much less at 22 c. At 0 12 c. F.candida requires days, but at almost the same temperature (11 C.} more than thrice this period is needed by E similis. It has been estimated by Marshall and Kevan (1962) that! candida may conceivably pass through as many as twelve generations in a year in

64 60 the greenhouse, but it is doubtful if, in F.similis, there would be likely to be more than six at most. There is no evidence of parthenogenesis in f.simikis. A sample of adults obtained from the stock laboratory cultures yielded 85 specimens in which the genital aperture was visible under the microscope. these, 26 were males and 59 females. Of F.candida on the... other hand, is commonly parthenogenetic (Goto, 1960a; ~ Goto and Ogel, 1961; Y~rshall and Kevan, 1962), although Goto (1960a) says that in wild populations in England, males and females occur in approximately equal numbers. Marshall and Kevan (1962), who discuss the parthenogenesis in collembola generally, found no males in their laboratory cultures of f.candida and the evidence for the existence of males in wild populations in southern Quebec is still lacking. Parthenogenesis enables a species to reproduce more rapidly so that the rate of multiplication by F.candida may further surpass that of F.similis. When it is also realized that f.candida may lay as many as 36 eggs at a time, as compared with a maximum of 14 recorded for F.similis, the difference in reproductive rate between the two species can be very considerable. Another important aspect of parthenogenesis in ri [.candida has also been discussed by Goto and Ogel (1961}

65 61 who concluded that whole populations exhibiting some special morphological variation {in this case the form of mucro) could be built up as a result and that this might lead to incorrect specifie diagnosis. With F.similis (since it is not parthenogenetic} there is less likelihood of this happening. Considerable morphological variation in fact occurs in the number and arrangement of the ventral setae of the manubrium and in the post-antennal organ of F.similis from southern Quebec but, although there may be several differences from European populations, these do not seem to be constant and there appears to be insufficient morphological stability to necessitate regarding the European and North American populations as distinct, even subspecially.

66 62.. V. PSEUDOSINELLA PETTErtSENI BORNER, 1901 AND PSEUDOSINELLA ALBA (PACKARD, 1873) (ENTOMOBRYIDAE)

67 63 GENERAL REMAR.KS fseudosinella petterseni BÔrner, 1901 Pseudosinella petterseni was first described from Europe by Borner (1901). Guthrie (1903) and Folsom (1924) redescribed the same species from Minnesota and from Massachusetts, New York, Indiana and Illinois under the names Cyphodeirus albinus and Lepidocyrtus violentus respectively. Mills (1934} transferred the latter to the genus Pseudosinella and recorded it - as P.violenta (Folsom) - from Iowa. Later, Maynard (1951) synonymized it with P.petterseni. Another synonym: Pseudosinella hofti (Schaffer, 1896} is given by Denis (1949), but Gisin (1960) lisuhofti as a synonym of Sinella coeca ( Schott, 1896). P.petterseni is a very widely distributed species. In Europe it has been recorded from Ireland, Spain, Denmark, Germany, Switzerland, and Czechoslovakia (Gisin, 1960); in North America it has been recorded in the United States from the various states already ment ioned and from Connecticut, North Caro lina, 1'-üssouri, Tennessee, Louisiana, Texas, Utah and Washington (Maynard, 1951; Bellinger, 1954; Christiansen, 1960) and, in Canada, from Ontario (Chatham, Belleville and Kingston) (James, 1933), although not hitherto from Quebec. In Central America it has been recorded from Costa Rica (Denis, 1931)

68 64 P.petterseni belongs to the soil fauna and is common under logs or stones; it is also abundant in greenhouses. Like other collembola with a dense scaly covering, this species may be found associated with soil that is rather dry (Folsom, 1924). Folsom further states that, "it occurs often in company with ants of various species, especially under stones embedded in the ground". Christiansen (1960) recorded it from caves in Texas and Iowa. At ~~cdonald College, Ste. Anne de Bellevue, Quebec, it has been collected from the thick litter layer and mull-type soil under sugar-maple trees (Acer ~accharum L.) in the Morgan Arboretum. Some published information on the biology of the species is available. Davis and Harris (1936), using the name P.violenta, have described the effect of 0 humidity on its life stages at 25, 30, 34 and 36.7 C., and Bellinger (1954) has studied its vertical distribution in Connecticut soil. Spencer and Stracener (1929; 1930) have indicated that Lepidocyrtus violentus [= P.petterseniJ makes pits in the roots of sugar-cane and causes damage by gnawing off the lateral roots, retarding cane-sprout growth and causing a marked reduction in the percentage of sucrose. Flor (1930) also refers to the species as being associated with root-rot of sugar-cane. Ingram (1931) observed that it will graze on the root-hairs of the same crop, and Folsom

69 (1933) has reported it damaging the roots of beets. Pseudosinella alba (Packard, 1873) Packard (1873) described this species (in the genus Lepidocyrtus) from Massachusetts. Elsewhere in North America, it has been recorded, as b.alba, for New York by Folsom (1928), and for Ontario (Chatham, Belleville, Grimsby and Crieff Hills) by James (1933). Schaffer (1900) referred to European material of this species as Sira {Pseudosinella) alba, and Borner (1901) used the name Pseudosinella alba. Mills (1934) was responsible for applying the latter to earlier North American records of the species, and reported it from Minnesota, Ohio and Tennessee, as we11 as from the two states a1ready mentioned; Bellinger (1954) notes that it occurs in Connecticut. Like ~.petterseni, ~.alba is thus widely distributed in North America, although not previously recorded from Quebec. It also occurs throughout most of Europe, except for the arctic regions, and is known from Australia (cf.gisin, 1944; 1960). Tullbergia ocellata Lie-Pettersen, 1896, is listed as a synonym by Gisin {1960). P.alba is a common species occurring under stones on sandy soil (Maynard, 1951) and in the soil under conifer stands (Bellinger, 1954). Christiansen (1960) records it from caves in Minnesota and Tennessee.

70 66 At Ste. Anne de Bellevue, Quebec, it has been extracted in large numbers from the same habitat as f.petterseni, many of the soil samples containing both species together. Little seems to be known of the biology of ~.alba, although Bellinger (1954} has studied its vertical distribution in Connecticut along with that of P.petterseni. No economie importance has so far been attributed to this species. DESCRIPTION OF ADULT The ~ genital plate The general morphology of adult ~.petterseni and P.!12! has been adequately deseribed (Packard, 1873; Borner, 1901; Folsom, 1928; l~lls, 1934; Gisin, 1960; Christiansen, 1960), but the male genital plate in the Entomobryidae has, until recently, been largely ignored, presumably because of the difficulty in examining its structure; it is very small and complex and is obscured by a dense body covering of large setae and scales. These facts, combined with the scarcity of males in many species, have resulted in almost total ignorance of the organ in this family {Christiansen, 195Sb). Christiansen (! ~ ) has, however, recently given a general description of the male genital plates in various genera of Entomobryidae, although he does not refer specifically

71 67 to t.petterseni or to f.alba. Davis and Harris (1936) state that the male genital aperture of the former species (they use the name P.violenta} is situated on a setiferous, biscuit-shaped papilla, but give no other details. The female genital pore is merely a transverse slit guarded by anterior and posterior lips, devoid of setae; it is virtually identical in both the species studied. Pseudosinella petterseni In P.petterseni, as in other Entomobryidae, the male genital plate, on which is situated the genital aperture, is located just behind the manubrium on the ventral side of the fifth abdominal segment. The plate is transverse and somewhat kidney-shaped, having a keyhole-like notch in the middle of the posterior margin, and at the anterior end of this notch lies the gonopore (Fig.6(d)). Smooth, acuminate setae are more or less uniformly arranged in two concentric series on the genital plate: an inner series of surround the gonopore and an outer series of 8-10 are placed submarginally (Fig.6(e,f)). The genital plate measures 35 x 20 ~. Pseudosinella alba In t.alba the subgenital plate is transversely oval, rather than reniform, but bas a similar keyhole-

72 68 shaped notch in the middle of the posterior margin to accommodate the gonopore. Numerous setae are arranged in three to four irregular concentric series between the gonopore and the margin of the plate. The latter measures 30 x 20 p. (Fig.7(d)}. The male genital plates of the two species differ in shape, size and number of setae and are thus quite distinct. SPERMATOPHORE As far as is known, males of arthropleone collembola deposit their spermatophores on the substrate. Schal1er (1954), for Orchesella villosa (Geoffroy), describes the structure of the stalked spermatophore after deposition. As far as the writer is aware, the method of formation of the spermatophore and its position in the ejaculatory duct (the term used by Lubbock (1873)) are still unknown. Pseudosinella petterseni To ensure that males had insufficient time to discharge their spermatophores before they were ki11ed, a rearing jar containing living f.petterseni was suddenly inverted over a petri-dish containing Gisin's fixative (see page 12). In one PVA-mounted specimen a structure, possibly a spermatophore, was clear1y visible in the ejaculatory duct (Fig.6}. This structure con-

73 69 sisted of two parts: a somewhat dome-shaped head 10~. in length and a tube 120 ~. long. The tube tapered posteriorly and ended bluntly. In the specimen referred to, the blunt end of the tube lay just at the genital opening. It would seem that this male was about to deposit a spermatophore just before it was killed. There is, however, sorne doubt whether the structure observed was, in fact, a spermatophore since Schaller (in litt., 1962} is of the opinion that these are not properly formed until they are extruded, at least in those species with which he has worked. On being deposited, the spermatophore of Orchesella villosa, as described by Schaller (1954), differa particularly from the structure described above for f.petterseni in the shape of the head, which is round instead of dome-shaped. Since the spermatophores of collembola are very minute, they are difficult to detect in rearing jars, and those of f.petterseni were, in fact, never encountered. Pseudosinella alba In several specimens of f.alba a similar spermatophore-like structure was discovered in the same manner as for P.petterseni. Its exact position before deposition was quite apparent from laterally mounted specimens and agreed with the above description. The

74 70 head of the structure was, however, rather larger than in P.petterseni (15 ~.long) and differed in shape, being somewhat drop-shaped. The tube a1so differed in being shorter (only 70~.) and broader at the anterior end, but it a1so tapered posteriorly and ended bluntly at the genital opening (Fig.?). As with f.petterseni, spermatophores of P.alba were never seen in the rearing jars. fseudosinella petterseni FOOD... P.petterseni seems to be a rather unspecialized feeder. Ingram (1931) reared this species on newly eut sugar-cane stalks and observed that a population of ten increased to two hundred and fifty in seven weeks. Davis and Harris {1936) have stated that their original co1ony of ten, when fed on soaked corn (maize), multip1ied to 3,447 in a period of eight and a half weeks. In a second experiment, where four parched peanuts served as the only source of food, these authors found that individuals mu1tiplied to 3,554 in seven and a half weeks. peanuts were first pitted and then consumed. The In the present study t.petterseni was reared in the laboratory on commercial baker's yeast, on which it flourished quite satisfactorily at various tempera-

75 71 tures. It was found, however, that the species was repelled by fungal mycelium and it avoided any part of the food so contaminated. This was also observed by Davis and Harris (1936). Pseudosinella alba P.alba was reared with moderate success on dead and partly decomposed leaves of American elm (Ulmus ~mericana L.) and sugar-maple leaves (Acer saccharum L.). On the addition of yeast to the diet, however, the colonies multiplied much more rapidly, but with the yeast alone no satisfactory increase in numbers occurred. E alba also avoided fungal mycelium in the same way as did P.petterseni. It was also noted that, in rearing cella, P.alba will eat the eggs of its own species. On one occasion, during rearing in small cells, one of three adults present in a cell, laid six eggs. On the following day two of the adults were seen apparently feeding on these eggs, and after half an hour pits were observed on the tops of four eggs. The pits appeared to have been chewed out by the collembola. On the third day, two of the eggs had vanished from the cell, and on the fourth day after oviposition the remaining four eggs had also disappeared. The usual food (yeast and partly decayed elm leaves) was present at all times. Similar

76 72 occurrences were observed in other rearing cells, confirming the egg-eating habit of this species. Macnamara (1924), Brown (1954) and others have mentioned various carnivorous feeding habits among collembola - on dead animal tissues, on other living collembola and on nematodes - but, so far, none has been reported feeding on eggs of any kind, let alone those of its own species. EGGS Pseudosinella petterseni Eggs were laid singly and scattered in the crevices (if these were present) of the plaster or on the smooth walls of the rearing cells. On deposition they were covered with a sticky fluid from the anus (Davis and Harris, 1936}. At room temperature {24 C. approx.), the number of eggs laid by a single female varied from six to fifteen at a time. Each female oviposited several times during its life, but after the third oviposition no further records were kept. Davis and Harris (1936), however, report the total number of eggs laid by one female throughout its lifetime as 45. The number of eggs laid per oviposition decreased with the temperature (see Table XV). occurred. At 4 c. no oviposition At oviposition the eggs were spherical, smooth

77 73 and white and measured 0.16 mm. in diameter. On the third day of incubation at 24 c. the diameter increased to mm. and hair-like outgrowths began to appear on the chorion. On the fourth day the 'hairy' covering became quite distinct. Such hair-like structures on the eggs of certain collembola were observed well over a century ago by Nicolet {1842) who states that, "dans ce dernier cas les oeufs sont parfois velus; garnis de poils longs et serrés". No splitting of the chorion such as that reported for other species of collembola (cf.marshall and Kevan, 1962) has been observed in this species. There was no further increase in size. Davis and Harris (1936) record that the eggs of this species hatched on the eighth day of incubation at 25 C., and comparable results were obtained during the present study. At 24 C. eggs took 7-9 days to hatch. A slightly lower percentage, however, hatched at 24 C. than was reported by Davis and Harris (2E.cit.) for temperatures only one degree higher (88.3 as against 92%). A significant drop in the percentage of hatching 0 was observed at lower temperatures, and at 11 C., only one-third of the eggs hatched. No hatching occurred at 4 C. although considerable numbers of eggs were exposed to this temperature for up to three months. The optimum incubation temperature for this species appears to be 30 C., at which all the eggs hatched in the minimum time (Table XVI}.

78 74 The chorion of eggs which failed to hatch became brown in color. Pseudosinella alba The eggs of P.alba are laid in a similar scattered manner to those of P.petterseni. At 24 C. females laid 6-16 eggs at a time. at ether temperatures are shown in Table XV. The numbers deposited The number of ovipositions and the total number of eggs per female were not determined. The eggs of P.alba when laid were similar to those of ~.petterseni, but slightly smaller; they measured 0.14 mm. in diameter. On the second day of incubation at 24 C. they increased to 0.16 mm., and on the third day the chorion became 'hairy' as in E.petterseni. Again no subsequent increase in size was observed. 0 At 24 C. eggs took 7-10 days to hatch. The incubation period increased as the temperature decreased and was days at 11 C. When thirty eggs were kept at 4 c. none hatched after being exposed to this temperature for three months. affected the percentage of eggs hatching. Lowering the temperature also There was a decrease of 24.1 in the percentage of eggs hatching, 0 when the temperature was lowered from 24 to 11 C. (Table XVI}. Optimum conditions for incubation were not determined in this species, but temperature requirements

79 75 are apparently similar to those of f.petterseni, although lower temperatures appear to be somewhat less lethal. Eggs kept at 24 C. which failed to hatch turned violet ten days after oviposition and a similar color change in dead eggs was also observed at other temperatures. The color of the chorion in the dead eggs of the two species thus shows a marked difference. POST~1BRYONIC DEVELOP~ŒNT Pseudosinella petterseni 0 Young E.petterseni reared at 24 C. averaged 0.41 mm. in length on hatching, their mean antennal length being 0.14 mm., an average ratio of 2.87:1 (i.e. nearly 3:1). Increases in body and antennal lengths were relatively uniform until the fourth moult. After the fifth and sixth moults the increase in the length of the antennae was significantly greater in relation to the body length. Adult females at first oviposition averaged 1.17 mm. in length, the mean length of their antennae being 0.48 mm., an average ratio of 2.34:1 (i.e. less than 2t:l) - see Table XVII. Such changes in the length of the antennae relative to the body are well known in collembola. Folsom (1919), for example, states that, in Tomocerus vulgaris (Tullberg), "as the individual grows there are changes in the relative lengths of the segments of the body and those

80 76 of the appendages". Juveniles of E.petterseni are white in color without trace of pigment. No change in color takes place throughout life. The largest individual measured was 1.41 mm. lvlills ( 1934) and Iviaynard ( 1951), however, record a maximum length of 1.5 mm. for North America; Gisin (1960) cites 1.0 mm. for European material. Since it is difficult to determine when male collembola are sexually mature, the period required for postembryonic development was determined for females only, measuring the time from the date of hatching until the first oviposition occurred. Males possibly mature a little sooner than do females. At 24 C. there were found to be five to six moults before the first oviposition. Davis and Harris {1936), using Dyar's factor also calculated that there were six instars at 30 C. before the first oviposition. The later stadia (5 and 6) took longer to complete than did earlier ones (see Table XVIII), probably due to their increasing sexual maturity. Pseudosinella alba Young individuals of P.~ measured 0.45 mm. in length on hatching. In the first instar the young are dull white in color, but in the second they become dirty white and shining. The color becomes darker grey

81 77 to blackish in successive instars. The length of the females at first oviposition averaged 0.91 mm. The longest individual measured was 1.0 mm., agreeing with the size given by Mills (1934), Maynard {1951) and Gisin (1960). As in î petterseni there are five to six moults at 24 C. before the first oviposition. The duration of each stadium, however, was more uniform since the earlier instars developed more slowly (see Table XVIII). Five individuals out of twenty-four reared at 24 C. oviposited in the sixth instar and eight in the seventh. The remaining individuals did not oviposit and were probably males EFFECT OF T~WERATURE ON POSTm~RYONIC STAGES Pseudosinella petterseni Temperature had a marked effect, not only on the incubation period and percentage of eggs hatching (see above), but also on surviva1, oviposition and postembryonic development. Ten adult specimens were kept first at 4 C. for one week and then at 0 C., but within a period of thirty days at the latter temperature all had died. At 4 C. adults survived for more than two months but they did not oviposit.

82 The duration of the postembryonic development decreased as the temperature increased {see Tables XVIII and XIX). At 24 c., 27 to 29 days were required before the first oviposition occurred, but at 30 C. Davis and Harris (1936} have stated that only loi days were needed to complete postembryonic development (data for other temperatures are not given by these authors). At 11 C. it was found that between 87 and 112 days were necessary, the average being nearly loo days. The above observations indicate that f.petterseni prefers warm temperatures. The optimum for this species is probably 30 C. because at this temperature, in the laboratory, the embryonic and postembryonic development was found to be the most rapid, although the information regarding the latter is not available for higher temperatures. At 30 C. the entire life-cycle can be completed in as little as l6à days. At 24 C. it averages 35i days and at 11 c. the average figure is days. P.petterseni does not seem to be able to survive freezing, but it can tolerate low temperatures (at!east down to 4 C.- see above). Bellinger (1954}, in Connecticut, obtained specimens of adult E pettersen~ in soil samples taken about a foot below the surface, suggesting that this species can probably survive low temperatures in nature by migrating to the deeper layers of the soil. There are, in fact, indications that the

83 79 species survives the Canadian winter in this manner. Pseudosinella alba As in f.petterseni, temperature had a marked effect, not only on the incubation period and percentage of egg hatching, but also on oviposition, survival and postembryonic development. Tolerance of low temperatures by adults is greater in P.alba than in f.petter ~eni. Of ten individuals kept first at 4 C. for about 0 a week and then at 0 G., only three died within a period of two months at the latter temperature (when the experiment was concluded). individuals, however. No eggs were laid by these The duration of the postembryonic period increased with a decrease in temperature for this species also (Tables XVIII and XIX). At 24 C. a rather longer time elapsed than in P.petterseni before the first oviposition. The majority of the 13 individuals referred to on page 77 laid eggs about 37 days after hatching, that is, after a considerably longer period than in f.petterseni. Two of the individuals which first laid eggs in the sixth instar required only 32 days, however, and one of those which did not oviposit until the seventh instar required as much as 45 days to mature. At 17 c. P.alba again required rather longer to mature than did P.petterseni (an average of 59.2 as

84 80 0 against 48.6 days), but at 11 C. development was a little less retarded than in that species (an average of 88.7 as against 99.7 days- see Table XIX). thus appears to be a slower maturing species than f ~ f.petterseni, but one which is less adversely affected 0 by lower temperatures, for, although at 24 C. the average time needed to complete the life-cycle was nearly 46, as against 35~ days, at 11 C. the number of days required was only 120, as against 134 for P.petterseni. FIELD OBSERVATIONS Bel1inger {1954), working in Connecticut, reported both P.petterseni and f.alba in his samp1es, but he found that the two species occurred in different areas. P.alba was round on1y under stands of White and Red pines and Hem1ock; P.petterseni was round in uncu1- tivated areas with some young \Vhite pines. In the Morgan Arboretum, Ste. Anne de Bellevue, Quebec, however, both species occurred together in the soil investigated. Neither species proved to be very abundant as soil arthropods go, but from the 20 samp1es taken between June and September, 1961, 47 adult P.petterseni and 152 P.~ were obtained (Table XX). Further samp1ing wou1d be necessary for statistica1 analysis, but it is interesting to note that there appeared to be a

85 81 tendency for P.alba to be more abundant and to be the commoner species under cooler conditions. OTHEH OBSERVATIONS P.petterseni, although devoid of eyes, is very sensitive to light. The laboratory cultures became extremely agitated and tried to escape from lighted areas when the lid of the rearing jar was removed under artificial light. Negative phototaxis is strong and is doubtless the principal factor in compelling the species to seek out its appropriate habitat. P.petterseni is also very sensitive as regards oviposition. A single female apparently would not oviposit more than twice or thrice in the same rearing jar (although this was as muchas 6.0 cm. in diameter), no matter how long it was allowed to live there. On being transferred to a new rearing jar, however, the same female would oviposit three or four days later. It might be possible that the continuous presence of yeast in the original jar and its fermentation inhibit further oviposition ~~~~~ COMPAHISON OF THE LIFE-HISTORIES OF P.PETTERSENI AND P.ALBA -- - P.petterseni and P.alba appear to differ to sorne extent in their food and habitat preferences, in

86 82 their tolerance of lower temperatures and in their rate of embryonic and postembryonic development. The differences and similarities in their life histories are summarized in Table XXI. The principal differences seem to be that, while ~-Eetterseni can thrive satisfactorily on vegetable matter or yeast alone, E.alba cannet do so, that E petterseni develops more rapidly at higher temperatures but E.alba (with the possible exception of the last pre-adult instar) is less retarded at lower ones, and that P.petterseni cannet tolerate freezing, whereas P.~ can do so. The percentage of eggs hatching at lower temperatures (11 C.) is 25 per cent. more in E.alba than in P.petterseni. A reduction in temperature from 24 to 11 c. more than doubles the period of postembryonic development in P.alba, but more than triples it in E.petterseni. This, as much as habitat, may explain the greater abundance of E.alba in south-eastern Quebec as compared with more southerly stations. From laboratory results and records of field temperatures it seems likely that, in south-east Quebec, both species of Pseudosinella may have approximately three generations a year. Eggs laid in September by females of either species could hatch in October, but the progeny are not likely to mature before the melting of the snow in the following spring. Between May and September there may conceivably be two more generations of

87 83 each species, P.petterseni overtaking the more coldtolerant f..alba as the summer temperatures increase, and then in turn being overtaken by the latter species with the approach of cooler conditions.

88 84 VI. CONCLUSION

89 GENERAL DISCUSSION The morphological variations observed in lsotoma notabilis and Folsomia similis are of considerable taxonomie importance. In view of these variations it seems that Stach (1947) was correct in synonymizing I.eunotabilis, recorded from North America, with!.~bi!!! from Europe. Hammer (1953), in fact, records I.notabilis from various parts of Canada, although, in the United States, the name! eunotabilis is still in use (Wray and Knowlton, 1961}. Only further biological investigation can finally settle the problem of synonymy. In f.similis it has been observed that the number of manubrial setae, a diagnostic character for Folsomia species, is unreliable because this number is quite variable. The presence of denticles in the inner margin of the post-antennal organ in f.similis and variation in their number also have been observed, and it is concluded that, although North American specimens of F.similis seem to differ to some extent from their European counterparts, they probably do not constitute a distinct subspecies. The structure of the male genital plate was found to be of taxonomie significance, not only at the family level as indicated by Christiansen (1958b), but also at the species level. The genital plates of Pseudosinella petterseni and P.~ (Entomobryidae), on

90 86 the one hand, and of Folsomia similis (Isotomidae), on the other, differ considerably from each other, as might be expected, but at the specifie level also, those of f.petterseni and P.alba differ significantly, if to a lesser degree. The female genital aperture, however, in all of the four species studied, was found to be very similar and not of diagnostic value; it is suggested that its form is such as to be of help in picking up the spermatophore. The form and position of structures observed in the ejaculatory ducts of f.petterseni and P.alba (Figs.6 & 7) suggest these to be spermatophores, although Schaller (in litt., 1962) has expressed the opinion that the spermatophores in collembola are not properly formed until they are extruded - at least in the species he has studied. Further histological studies may help in elucidating this. All the four species of collembola studied herein may live for sorne time on a diet of yeast alone, but F.similis and P.alba, unlike I.notabilis and f.petterseni, did not form thriving colonies under such conditions. Britt (1951} and Marshall and Kevan {1962) also obtained satisfactory colonies of Achorutes armatus and Folsomia candida respectively using only yeast as food. Only the addition of partly decayed leaves of

91 87 sugar maple or American elm (Acer saccharum L. and Ulmus americana L.) to the diet enabled F.similis and P.alba to multiply satisfactorily in the laboratory. I.nota bilis readily fed on fungus (also reported by Poole, 1959), but the other remaining three species did not, and declined other food if it were covered by fungus. Davis and Harris (1936) also observed this in the case of P.violenta ~ P.petterseniJ. There are thus significant differences in the feeding habits of various species of collembola. Poole (! ~ } found that food preferences even extended to individuals. P.alba was seen feeding on the eggs of its own species, but there seem to have been no previous reports of collembola feeding on any kind of egg. No evidence of parthenogenesis was found in any of the four species studied although this is not rare in collembola and occurs in Folsomia candida. I.notabilis and F.similis lay their eggs singly but place them in groups. This has also been observed for other species by Britt (1951), Milne (1959; 1960) and Marshall and Kevan (1962). P.petterseni and and P.alba, however, lay their eggs in a scattered manner. Freshly laid eggs of I.notabilis and F.similis were smaller in size than those of P.petterseni and P.alba. The eggs were smooth when they were laid in all four species, but, in the course of development, two

92 88 different phenomena occurred depending on the family to which the species belonged. In I.notabilis and! similis {Isotomidae) the chorion of the egg ruptured on the third or fourth day of oviposition, as was also observed for f.candida by ~~rshall and Kevan (1962). No rupturing of the chorion occurred in P.petterseni and P.alba {Entomobryidae) and, instead, a 'hairy' covering appeared on the eggs after the third day of oviposition. Nicolet (1842) had also reported such outgrowths on the eggs of certain species of collembola. The incubation period for the eggs of.[.simili!' f.petterseni and f.alba is 9 to 10, 7 to 9 and 7 to 10 days respectively at room temperature. That is, their rates of development are very similar. The eggs of I.notabilis were not reared at such a high temperature, but at 17 C. they hatched on the 7th or 8th day after oviposition, whereas those of the other three species required 12 to 13, 12 to 15 and 12 to 14 respectively. Thus I.notabilis develops more rapidly at somewhat lower temperature than the ether species which all have comparable rates of development. At this temperature also the percentage of eggs hatching is higher for l notabilis than for the other three species. At intermediate temperatures the percentage of eggs hatching was less than at higher temperatures for all the four 0 species. At 4 C. the eggs of f.similis also took a

93 longer time to hatch than those of I.notabilis, while the two remaining species did not oviposit at this temperature and the eggs placed at this temperature for 92 days did not hatch. I.notabilis eggs are thus clearly more adapted to lower temperatures than are those of the other species. F.similis, however, will oviposit at lower teznperatures than will I.notabilis. Generally there are 4 or 5 moults before oviposition in I.notabilis, but the other three species did not oviposit before the 5th or 6th moult. During postembryonic development in P.Eetterseni the ratio between the body length and length of antennae in the earlier instars differs from that in the later ones. Such differences in ratio have also been reported by Folsom (1919) for Tomocerus vulgaris and appear to be fairly general. In all four species, temperature had a similar marked effect on the postembryonic development to that on incubation period and percentage of eggs hatching. All took a longer time to mature at lower temperatures, thus agreeing with the results reported by ~lilne (1959; 1960) for other species of collembola. ~.petterseni did not survive at 0 c. for more than 30 days, but the remaining three species were able

94 90 to do so. It thus seems to be the least adapted to Canadian conditions and was, in fact, found to be less common than f.alba in similar situations in southwestern Quebec. I.notabilis has been recorded from colder parts of the world, such as Iceland and northern parts of Canada, than have the species of Pseudosinella, and Bellinger {1954) collected specimens during the winter in Connecticut soil, which is in keeping with the present conclusions that I.notabilis can tolerate lower temperatures than these species. r.similis has been shown to be able to lay eggs at 0 C. and to survive for more than 30 days at this temperature, suggesting that adult females remain active and even reproduce during the winter in eastern Canada. It is possible that in Canada P.Eetterseni and E.alba adults escape the winter cold by migrating deep into the soil to the warmer layers since in Connecticut, Bellinger {1954) has obtained specimens of these two species a foot below the surface. su~~ry General Four species of collembola, Isotoma notabilis, Folsomia similis, Pseudosinella petterseni and P.alba were selected from among those extracted from samples of mull-type soil taken under sugar-maple trees {~ saccharum L.)in the Morgan Arboretum, Macdonald College,

95 91 Ste. Anne de Bellevue, Quebec. Three of the species were also taken in litter samples from the same location, but f.similis, a species new to North America, was found only in the soil below the litter layer. It was, however, also taken in pot-soil from the College greenhouses. The same culture methods proved successful for all the species studied although there were differences in food requirements. species: The following summarizes the findings for each Isotoma notabilis Isotoma notabilis is widely distributed in the world, but probably best known from Europe. What is apparently the same species has generally been recorded from North America as I.eunotabilis. The species is subject to considerable morphological variations in Quebec populations. It is fungus feeder, but it will also feed on dead organic matter and yeast. Females will lay eggs without fertilization after four or five moults, but such eggs do not hatch. There is no evidence of parthenogenesis in this species. At 17 C. as many as là eggs were laid by one female and at least four ovipositions per female have

96 92 been observed. 0 but at 4 An egg-batch often contains 7 to $ eggs, C. an average of only three eggs were laid and only a single oviposition occurred. 0 At 17 C., eggs took an average of 7.4 days to hatch, but the mean numo 0 ber of days needed at 4 C. was At 17 C. an average of 24.7 days were required to complete the life history, while at 4 C. the mean number of days required was It has been estimated that in southern Quebec there could be at least four generations of!.notabilis in a year. I.notabilis appears to suffer heavy predation by mites such as Pergamasus crassipes {Linné), which, in the laboratory, can catch and devour or kill an average of fourteen I.notabilis per day. The life history of I.notabilis has been compared with those of certain other species of Isotoma and it is found that I.notabilis differa in size, food, and its greater ability to withstand cold. Folsomia similis Folsomia similis appears to be a widely distributed Holarctic species; it was reported for the first time from North America during the course of the present investigation. Most specimens from southern Quebec differ from those recorded from Europe in having denticles on the inner margin of the post-antennal organ

97 93 and also in having a more variable number and arrangement of the ventral setae of the manubrium. The importance of this variation in taxonomie work has been discussed. Each lip of the male genital aperture bears two genital 'dises' which have not previously been described for collembola. Yeast and fungi or decaying leaves of sugarmaple (!_Ç,!ê.!:, saccharum 1.) or American elm ( Ulmus americana L.) alone did not provide an adequate diet for rearing F.similis, but when yeast and decaying leaves were mixed, they proved satisfactory. aggregated in heaps. Eggs are laid up to 14 in a batch and usually They take from 9 to 10 days at 22 C. or 24 c., and from 72 to $0 days at 4 C. to 0 hatch. The greatest percentage hatch at 22 c. Postembryonic development is most rapid at 0 22 C., at which temperature an average of about 40 days is required from hatching until oviposition. At 24 C. and 17 C. development is a little slower. At 11 c. the life-cycle takes more than $0 days to complete. Maturity is usually reached by females after the sixth moult, but at 22 C. this occasionally occurs after the fifth moult. of F.candida. The biology of F.similis is compared with that The optimum temperature for development

98 94 0 of both species appears to be 22 G., but! similis seems to thrive better than r.candida at lower temperatures. Pseudosinella petterseni and f.alba Pseudosinella petterseni and P.alba are two soil- and litter-inhabiting collembola commonly found under stones and logs, in caves and elsewhere in many parts of the world. Both species have been extracted together from soil and litter samples in southern Quebec. The male genital plate of each species has been described for the first time; that of each species is characteristic in shape and chaetotaxy. A spermatophorelike structure is described in the ejaculatory duct of both species; these structures are similar in the two species but differ considerably in detail. According to the literature, ~.petterseni an unspecialized feeder. reared on yeast alone. is In the laboratory it has been For f.~, however, partly decayed leaves had to be added to the diet in order to establish a thriving colony. of its own species. P.alba will eat the eggs In both species the eggs are smooth and white and are laid in a scattered manner. After a few days they become covered with a 'hairyt coating. Eggs of f.petterseni develop more rapidly at higher, and more

99 95 slowly at lower, temperatures than do those of E ~; in the latter species also, a higher proportion hatch at lower temperatures. In each species there are five or six moults before maturity is reached. The average life cycle of E.petterseni is completed in 35i days at 24 C. and days at ll C.; that of P.alba is completed in 46 days 0 0 at 24 C. and 120 days at 11 C. Adults of both species survive at 4 C. but do not reproduce; at 0 C. P.alba can survive, but P.petterseni can not. There are probably about three generations of each species per year in south-western Quebec.

100 96 VII. REFERENCES

101 97 Bagnall,.R..S Notes on British Collembola III (con'd.). Ent.mon.Mag.75: Notes on British Collembola. Ent. mon.mag.$5:51, Bellinger,P.F Studies of soil fauna with special reference to the Collembola. Conn. Agric.Exp.Sta.Bull.5B3: Possible adaptations in Poduroid collembola. Ent.News.71: Borner,C Zur Kenntnis der Apterygoten-Fauna von Bremen und der Nachbardistrikte. Abh.naturwiss.Ver.Bremen. 17: Britt,N.W Observations on the life history of the Collembolan Achorutes armatus. Trans.Amer.microsc.Soc. 70: Brown,W.L Collembola feeding upon nematodes. Ecology.35:421. Carpentier,F Quelques remarques concernant la morphologie thoracique des Collemboles (Aptérygotes). Bull.Ann. Soc.ent.Belg.83: A propos des endosternites du thorax des Collemboles (Aptérygotes). Bull.Ann.Soc.ent.Belg. 85:41-52.

102 Cassagnau,P L'influence de la température sur la morphologie d'hypogastrura purpurascens (Lubbeek), Collembola Podurmorphe. C.R.Acad.Sci. Paris.240: L'influence de la température sur l'apparition du 'genre' Spinisotoma (Collembola:Isotomidae). C.h.Acad.Sci.Paris.242: Christiansen,K. 1958a. The Collembola of Lebanon and Western Syria. Psyche.65: Bb. The entomobryiform male genital plate. Proc.Iowa Acad.Sci. 65: The Genus Pseudosinella (Collembola:Entomobryidae) in caves of the United States. Psyche. 67:1-25. Davidson, J The influence of temperature on the incubation period of the eggs of Sminthurus viridis. Austral.J.exp.biol.med.Sci. 8: a. Resistance of the eggs of collembola ta drought conditions. Nature, London.l29:867.

103 99 Davidson,J. 1932b. Factors affecting oviposition of Sminthurus viridis. Austral.J.exp. biol.med.sci.lo:l c. On the viabili ty of the eggs of Sminthurus viridis in relation to their environment. Austral.J.exp. biol.med.sci.l0: a. Effects of rain fall-evaporation on insects inhabiting the sail surface. Nature, London.l31: b. The environmental factors affecting the development of the eggs of Sminthurus viridis. Austral.J.exp.biol. med.sci.l1:9-23. Davies,W.M Investigations of spring-tails attacking mangolds. J.Min.Agric. 32: On the tracheal System of collembola, with special reference to that of Sminthurus viridis Lubbock [sic] Quart.J.micros.Sci.71: On the economie status and bionomics of Sminthurus viridis. Bull.ent.Res. 18: Swarming of collembola in England. Nature, London.l30:94.

104 100 Davis,li. and Harris,H.M The biology of Pseudosinella violenta (Folsom), with some effects of temperture and humidity on its life stages (Collembola:Entomobryidae). Iowa St.Coll.J.l0: Denis,J.li Collemboles de Costa Rica avec une contribution au species de l'ordre. Boll.Lab.Zool.gen.Portici.25: Sous-classe des Aptérygotes (Aptérygogenea,Braner,l885,Apterygota, Lang,l889):Anatomie-Biologie Systématique. In Grassé (Ed.) Traite de Zoologie, Paris. 9: Edwards,C.A.T Simple technique for rearing Collembola, Symphyla and other small soil-inhabiting arthropods. In Kevan,D.K.McE.(Ed.}, Sail Zoology. London, New York, Toronto: Flor,H.H Factors influencing the severity of the root-rot troubles of sugarcane. Louisiana Bull.212:1-40. Folsom,J.W Symposium on "The life cycle in insects". I.Apterygota.Ann.ent. Soc.Amer.13:

105 101 Folsom,J.W New Species of collembola from New York State. Amer.Mus.Nov.lOS:l-12. Gisin,H. Goto,H.E A list of insects of New York State. Orders Thysanura and Collembola.Mem. Cornell.Univ.agric.Expt.Sta.lOl:ll-17. The economie importance of Collembola. J.econ.Ent.26: Nearctic Collembola or springtails of the family Isotomidae. Proc.u.s. nat.mus.l68:1-144 Hilfstabellen zum Bestimmen der holarktischen Collembolen. Verh. naturforsch.ges.basel.55: ~otes sur les Collemboles avec description de quatorze espèces et d'un genre nouveau. Mitt.schweiz. ent.ges.22: Collembolenfauna Europas. Genève: 312 pp. 1960a. Facultative Parthenogenesis in Collembola (Insecta). Nature,London. 188: b. Simple technique for the rearing of collembola and a note on the use of a fungistic substance in the cultures. Ent.mon.~~g.96:

106 102 " Goto,H.E. and Ogel,S Variations in the mucro of Folsomia {Co1lembo1a:Isotomidae}. Entomo1ogist.94: Guthrie,J.E The Co1lembo1a of Mînnesota. Rep. Haarl~v,N. Hammer,M. Handschin,E. Hilton,W.A. geo1.nat.hist.surv.minnesota.(zoo1.) 4: A new modification of the Tu11gren apparatus. J.Anim.Eco1.16: A modified Tu11gren apparatus. In Kevan,D.K.McE.(Ed.) Soi1 Zoology. London, New York, Toronto: Investigations on the Microfauna of Hoffmann,R.W Holdaway,F.G Northern Canada.Part II:Col1embo1a. Acta Arctica.6:10à pp. n Col1embola - Springschwanze. Biol. Tiere Deutschlands. 25 {20): The nervous system of Nearura g! gantea Tullberg. 6: J.Ent.Zoo Co11embo1a.Nervous system and sense organs. J.Ent.Zoo1.28: Zur Kenntnis der Entwick1ungsgeschichte der Co1lembolen. 37: Zoo1.Anz. The bionomics of Sminthurus viridis Linné, or the South Australian Lucerne flea.pamph1.commonw.counc. Sci.industr.Res.4:1-23.

107 103 Hubber,I Color as an index to the relative humidity of plaster of Paris culture jars. Proc.ent.Soc.Wash. 60: Imms,A.D A General Textbook of Entomology (revised O.W.rtichards and d.g. Davies). Methuen.London (9th Ed.): x+l-886. Ingram,J.W Soil animals attacking sugar-cane. J.econ.Ent.24: James,H.G Collembola of the Toronto region with notes on the biology of Isotoma palustris Mueller. Trans. Canad.Inst.l9: Janin, M Contribution à l'étude du thorax des Collemboles. C.R.Acad.Sci. Paris.225: Karg,V.W Okologische Untersuchungen von edaphischen Gamasiden (Acarina, Parasitiformes). Pedobiologia. 1: Kevan,D.K.McE A method of preparing permanent fluid mounts of small organisms, especially Collembola, developed by Dr.Ekkehard von Torne(Vienna) and Dr.Hermann Gisin {Geneva). In Kevan,D.K.McE.(Ed.) Soil Zoology London,New York,Toronto:

108 104 Knight,C.B. Kos,F. Krjukova,F.A The Tomocerinae(Collembola)in old field stands of North Carolina. Ecology.42: Iz biologije in ekologije triglavskih Izotomid. Prirodosl. Izvest.l:5-22 {cf.paclt,l956) Podury v kacestve vreditelej ovoséej i mery borby s nimi. Izvest.Leningr.Inst.Borby Vredit.Chozj.3: (~. Paclt, 1956). Kubiëna, Animal activity in soil as a decisive factor in establishment of humus forma. In Kevan, D.K.McE.(Ed.).Soil Zoology. London,New York,Toronto:?J-82. Lemoine, V De l'acte génital probable observé chez le Sminthurus fuscus. Congr.Assoc.fr.Avanc.Sci. La uochelle:9{18à2) {cf. Lindenmann,l950). Lie-Pettersen,O.J Biologisches Über norwegische Collembolen. Bergens Mus.Aarb. 1899{7):1-12. Lindemnann, W Untersuchungen zur postembryonalen Entwicklung schweizerischer Orchessellen.rtev.suisse Zool.57:

109 105 Lipovsky,L.G Co11embo1a as food for chiggers (Acarina,Trombicu1idae). J.Parasito1.37 : Lubbeek, Sir J l Ionograph of the Co1lembola and Thysanura. London:x+276pp. +7S pl. IvlacLagan, D.S An ecological study of the "Lucerne flea" (Sm!nthurus viridis Linné). Bull.ent.Res.32: , Macnamara,C Popular and Practical Entomology (Remarks on Co11embo1a). Canad. Ent.51:73-à0, , The food of Co1lembola. Canad.Ent. 56: Marsha1l,V.G. and Kevan,D.K.McE Preliminary oblv'iaynard servations on the bio1ogy of Fo1somia candida Wi11em, (Co11embola:Isotornidae).Canad. Ent.(in press)., E A A monograph of the Collembo1a or springtail insects of New York State. Ithaca,New York:ll3 pp. IV!i lls, H.B A monograph of the Collembola of Iowa.Mon.Div.industr.Sci.Iowa St. Coll.3:xii+l43 PP Mills, H.B., and Richards,'iJ.iL Collembola from arctic and boreal Canada.J.Kansas ent.soc.26:53-59.

110 106 Milne,S Eco1ogy of Collembola.Unpublished Ph.D.Thesis, University of Glasgow Studies on the life histories of various species of Arthopleone Collembola. Proc.R.ent.Soc.London (A}, 35: Murphy,P.W The biology of forest soils with special reference to the mesofauna or meiofauna. J.soil sei. 4(2): Murphy,P.W. and Doncaster,C. C A culture method for soil meiofauna and its application to the study of nematode predators. Nematologia, 2: Nicolet,H Recherches pour servir à l'histoire des Podrelles Nouv.Mém.Soc.helvet. Sci.nat.6(3):1-88. Packard,A.S.l873. Synopsis of the Thysanura of Essex County, Massachusetts. Ann.Rep.Peabody Acad.Sci.5: Paclt,J Biologie der primâr flügellosen Insekten. Jena:VIII 258 pp. Park,O A notable Aggregation of Collembola. Ann.ent.Soc.Amer.42:7-9. Philiptschenko,J Die Embryonalentwicklung von Isotoma cinerea.(~.johannsen, O.A. and Butt,F.H.,l941 Embryology of Insects and Myriapods.New York;London).

111 107 Poole,T.B Studies on food of Collembola in a Douglas fir Plantation. Proc.Zool. Soc.London,l32: An ecological study of the Collembo1a in a coniferous forest soil. Pedobiologia,l:l ri.ipper,w Champignon-Springschwanze,Biologie und Bekampfung von Hypogastrura manubria1is Tullb.Z.angew.Ent.l6: Salmon,J.T. 195la. Keys and bib1iography to the Collembola. Zool.Publ.Victoria Univ.Coll. Wellington,8:l l95lb. Polyvinyl alcoho1 as a mounting medium in microscopy. Microscope,London. March-Apr.l95l,sep., 4 pp Keys and bibliography to the Collembola.First Supplement. Zool.Publ. Victoria Uni v.coll. v-iellington.20: l-3 5. Schaffer,c Die Collembola der Umgebung von Hamburg und benachbarter Ge bi ete. llllitt. naturh.mus.hamburg,l3: (cf. Stach, 1947)., Die arktischen und subarktischen Collembola. Fauna arct.l:

112 108 Schaller,F Biologische Beobachtungen an humusbildenden Bodentieren inbesondere an Collembolen. Zool.Jb.(Syst.) 78: (cf.Poole,l959) Die "Copula" der Collembola (Springschwanze) Naturwissenschaften. 39: Untersuchungen zur Fortpflanzungsbiologie arthropleoner Collembolen. Z.Morphol.Okol.Tiere,41: Die indirekte Spermatophoren-Ubertragung und ihre Problem. Forsch. Fortschr.28: Sharma,G.D. and Kevan,D.K.McE Folsomia similis Bagnall,l939(Collembola:Isotomidae),apparently new to North America.Ent.News,73: Spencer,H. and Stracener,C.L Soil animals injurious to sugar-cane roots.ann.ent.soc.amer.22: Recent experiments with soil anima1s attacking roots of sugar-cane. J.econ.Ent.23: Splendore,A Collembolo dannoso ai semenzai di tabacco;isotomurus Ealustris (MÜll.) Born.var.maculatus Schaeff.Boll.tecn. Ist.sperim.Coltivaz.Tabacchi,ll:l47-15l(cf.Paclt,l956).

113 109 Stach,J The Apterygotan fauna of Poland in relation to the wordl [sic} -fauna of this group of insects (Family Isotomidae). Acta.monogr.Mus.Hist. nat.krakow,l:l-488. Steinbock,O Zur Lebensweise einiger Tiere des Ewigschneegebietes. Z.Morphol.Okol. Tiere,20: (cf.Paclt,l956) Der Gletscherfloh. Z.dtsch.Alpenver.70: (cf.Paclt,l956). Strebel,O Biologische und physiologische Untersuchungen an Hypogastrura purpurascens und Sminthurus niger {Apt.Coll.). Zool.Anz.84: Thor,S. (~.Paclt,l956) Beitrage zur Biologie,Oekologie und Physiologie einheimischer Collembolen. Z.Morph.Okol.Tiere,41: {cf.Vdlne,l960) Biologische Studien an einheimischen Collembolen.III.Zur Biologie qes Tomocerus vulgaris Tullb. Konowia 17: Beitrage zur Kenntnis der invertebraten Fauna von Svalbard.Str. Svalbard Ishavet,27:1-156 (cf. Paclt,1956).

114 llo Tomaszewski,W. Torne,E.von Uchida,H. and Abukawa,T Uchida,H. and Chiba,S Uchida,H. and Chiba,S Wallwork,J.A. Wigg1esworth,V.B Apterygota.In Sorauer,Handbuch der Pflanzenkrankheiten. Aufl.5. Berlin and Hamburg,4(1) : Okologische Experimente mit Folsomia candida (Collembola). Pedobiologia,l:l On the breeding habits and structure of eggs of Tomocerus minutus Tullberg{Ins. Collem.), [rn Japanese] Zool. ~~g.tokyo,65: Studies on Development of Tomocerus minutus Tullberg (Ins.Collem.).I.On the postembryonic development [In Japanese). Zool.lvlag.,Tokyo,67: Studies on the development of Tomocerus minutus Tull- berg(ins.collem.).ii.on the postembryonic development(later larval stages) [In Japanes~ Zoo.Mag.,Tokyo,68: The distribution and dynamics of some forest soil mites. Ecology,40: The princip1es of insect physiology(2nd.ed.).london:viii+434pp.

115 111 wilkey,r.f Preliminary list of the Collembola of California. Bull.Calif.Dep.Agric. 48: Willem, V Recherches sur les collemboles et les Th~sanoures. Mém.Acad.Sci.Belg. 58: Womersley,H Primitive insects of South Australia. Silverfish,springtai1s and their al1ies.adelaide:322 pp. Wray,D.L Swarming of Gollembola in North Carolina. J.econ.Ent.38:500. Wray,D.L. and Knowlton,G.F A preliminary list of Collembola of Idaho. Gt.Basin Nat.l6: Co11embo1a from Rodent nests. Ent. News,72:24S-251.

116 VIII. TABLES

117 Table I. r~.verage number of eggs per fema1e laid by ten I.notabilis at various temperatures. (Figures in parentheses give maxima and minima. ) Oviposition number Temperature ( 0 c.) Average number of eggs (5-9) 1 (3-4) (3-5) (3-4) {3-5) ( 1-4) ( 9-13) (6-10) {5-6) (2-4) {2-4) (15-18) (6-14) (3-4) ( 9-13)

118 Table II. Days required for hatching (means in parentheses) and percentage of eggs hatched for I.notabilis at various temperatures. Temperature ( 0 c.) à 6 4 Da ys là He qui red ( 7.4) (10.5) (19.0) (30.5) {35.7) (53.6) Percent age 87 68?? 49? hat ching

119 Table III. Duration in days (means in parentheses) for instar of I.notabilis at various temperatures. Instar Temperature ( c.) Da vs (4.0) (4.0) (7.0) ( 10.5) (12.6} (25.5) ( 4.3) (4.5) {8.7) (10.6) (12.6) ( 23.5) ( 4.2) (4.3) (7.2) ( 8. 5) ( 12.8) {28.0) { 4-4} (3.7) (7.7) {11.7) ( 13.1) {25.9) 5? 3-5 x {3.8) {10.6) (11.3) (20.6) x Number of days before oviposition, 1-2 (av. 1.4).

120 Table IV. Duration in days of life-history of I.notabilis females at various temperatures. (Means in parentheses) Temperature ( 0 c.) Da ys (24.7) (30.4) (52. 2) (81.0) (99.2) (171.7)

121 Table V. Body length (in millimetres) of ten I.notabilis during first nineteen days Measurements x's indicate days after hatching. after moulting are underlined. of egg-laying by females. No.of da ys Individuals after hat ching A B c D E F G H I J Body length (mm.) Q.40 ~ 0.40 ~ ~ 0.32 ~ ~!hll Q.JU ~ Q.: Q.: O.?t x x o. z:t! o. 1'

122 Table VI. Numbers of I.notabilis and Pergamasus crassipes extracted from five samples per month during June to Septembert (Diameter and depth of samples, 6.1 and 12.7 in. respectively) 1v1onth lvlean soil I.notabilis P.crassiEes temperature (0 c.) Young Adults Total Adults June July August September à

123 Table VII. Daily laboratory record of I.notabilis killed by ten Pergamus crassipes during a period of ten days. Day Indi vi dual I li tes A B c D E F G H I J Number of I.notabilis killed $ $ $ l$ $ Total kil1ed per mite Average number kil1ed 14.1 per mite per day

124 Table VIII. Comparison of life-histories of Isotomurus palustris (from James, 1933), Isotoma viridis {from ~lilne, 1959; 1960) and Isotoma Q2tabilis. Habitat Foocl I.palustris Under stones and dead wood near water, or on floating dead wood. Bacterial slime (Denis, 1949). Maximum size 3 {mm.) Color Variable, usually brownish yellow. Number of Not known. ovipositions per female Number of 8-25 at 30 C. eggs per bat ch Size of eggs (mm.) I.viridis Upper humus layers and surface litter. Damp rotting organic matter; soil animals (e.g.nematodes and I.notabilis}. Variable, usually reddish or yellow. Not known (temperature not given) I.notabilis Moss and forest litter; forest humus and upper soil layers. Damp rotting organic matter, but mainly fungi. Dark blue. Four at 17 C. (less at lower temperatures 1 see 'l'able I) 7-8 at 17 c. (less at lower temperatures) Incubation at 30 C. period(days) Resistance Not observed. of eggs to lower temperatures Size of 0.27 first instar (mm.) Number of 7 { approximately) instars bef ore oviposition Total length. l month(approximately) of but no oviposition life-history observed to confirm this at 24 C at 120 C. No hatching at 50 c. Not known. Not known at 17 C. (for other temperatures see Table II. Hatch at 4 c. (51-55 days to hatch) days at 17 C. (for other temperatures see Table IV.) 0

125 Table IX. Number of ventral setae on the manubrium of 188 specimens of Folsomia similis from a laboratory culture derived from southern Quebec stock. No. of setae No. of specimens 4 l ll

126 Table X. Percentage of Folsomia similis eggs hatching and incubation period at various temperatures (means in parentheses}. Temperature 0 c Percentage hatching (very few) Days incubation required (9.2) (9.3) (12.3) (30.2) {77.1)

127 Table XI. Growth of 15 Folsomia similis individua1s measured daily from the date of hatching (length of all individuals on hatching, 0.32 mm.), and the number r- Instar of days passed in each instar. Individua1s A B c D E F G H I J K 1 M N 0 Length in milimetres at the end of each instar 1 0.4à à à à 0.4à à à 0.75 O.àl 0.67 o.à3 0.7à O.àl à o.à O.à9 o.à o.à t o.6f o.9e ~ Number of days spent in each instar l à à 7 à 7 à à à à à à 9 9 ll à 9 à à à là ll ll là x Died before the moult z Lived but no moult observed; no increase in 1ength for 6-à days

128 Table XII. Number of days passed in each of the first six instars by Folsomia similis reared at various temperatures (means in parentheses), based on groups of 3 to 7 individuals of mixed sexes reared together in small cells; 10 such groups 0 at 24 C.; 7 at each of the lower temperatures. Instar Temperature ( 0 c.) Da ys (6.6) ( 5. 8) (6.5) (14.7) (7.7) (5.3) (8.9) ( 12. 7) (8.9} (5.6} ( 8.3) ( 14. 2) (10.1} (6.9) (9.4} (11.5) ( 12.1) (7.2} (8.4) (14.2} ( 12. 7) (7.8) (11.2) ( 16. 2}

129 Table XIII. Total number of days (means in parentheses) from hatching until first oviposition in Folsomia similis reared at various temperatures, based on the same groups as in Table XII. (Only females which began to lay after the sixth moult are included.) Temperature 0 c Da ys (59.1) (40.1) (53. 9) (g4.9) i

130 Table XIV. Comparison of!ife-histories of Folsomia candida (mostly from lfûlne, 1959; marked with an x - and from Marshall and Kevan, 1962) and F.similis. Habitat Food Size of adult (mm.) F.candida Flower pots;in vegetable mould on humus;among dead leaves;in caves; in upper true soil layers. Yeast;fungal hyphae;damp rotting leaves;bracken spores~ av. F.similis Flower pots;in vineyard soil;in soil under trees(acer saccharum 1.) - Yeast;fungus and damp rotting leaves;requires a mixture for satisfactory oviposition Color of adult wnite Parthenogenesis Common (males unknown in sorne Number of eggs in a bat ch Size of egg wh en laid(mm.) Incubation period in days 26 c. 24 c. 22 c. 20 c. 18 c. 17 c. 16 c. 12 c. 11 c. 50 c. 40 c. P ercentage eggs hat ching instances) 3-20 (leaf diet) (yeast diet) 9-36 x (bracken~spore diet) x 9-10 lg x x Greyish white with black pigment spots. Unknown 3-14 (mixed diet) c. 22 c. 18 c. 17 c. 16 c. 11 c. 40 c very few Size of first instar (mm.) Days required by females from hatching to oviposition 24 c. 22 c. 17 c. 12 c. 11 c. Low temperature tolerance ~ x x x No records of survival below 5 c Female can thrive and lay eggs ' at 0 C. l Gisin (1960) gives a size range of mm., but the maximum given seems excessively large. 2 M"l. 0 1 ne g1ves mm. but this may be after sorne swe 11 1ng has occurred.

131 Table xv. Number of eggs laid per oviposition by females of Pseudosinella petterseni and f.alba (number of females is 15, 13, 10 and 10 of the former and 13, 10, 13 and 10 of the latter species at 24, 17, 11 and 4 c., respectively). Species Temperature { 0 c.) No.of eggs f.petterseni maximum minimum mean E ill.! maximum minimum mean

132 Table XVI. Incubation period in days {means in parentheses) and percentage hatched for eggs of Pseudosinella petterseni and ~.alba reared at different temperatures. (Number of eggs of former spe~ies incubated, 224, 206, 54 and 30, and of latter, 114, 80, 48 and 30, at 24, 17, 11 and 4 c. respectively; figures for temperatures above 24 C. from Davis and Harris (1936)) Temperature ( 0 c.) Days required for hatching Percent age eggs hatched ~.petterseni P.alba P.petterseni ~.alba {7.6) (8.3) lf (12.2 { 13.03) (34.0) (31.2) No hatching after 92 days No hatching after 92 days

133 Table XVII. Length (in mm.) of body (b) and antennae (a) of seven individua1s (A-G) of Pseudosine1la petterseni, reared at 24 C. (Measurements for first instar were on hatching; those for other instars, immediately after ecdysis) - Instar Individuals Average ratio A B c D E F G b:a Length (mm.) 1 b a b a b :1 a (2.53:1-3.30:1) 4 b O.àl a b a b a :1 {2.20:1-7 b :1) a

134 Table XVIII. Duration in days (means in parentheses) instars of Pseudosinella petterseni and ~.alba. for each of the first six (Number of individuals involved of the former, 28, 22 and 20, and of the latter, 24, là and 23, at 24, 17 and 11 C. respectively.) Species f.;eetterseni f!!!2! Temperature ( 0 c.) Instar Number of days Number of days (3.8) (6.8) (15.6) ( 5.2) {7.2) (11.6) {3. 7) (6.2) (15.4) (5.5) (6.7) (11.0) (3.8) (5.6) { 13.1) (6.1) (7.4) ( 10.8) (4.5) (9.1) (15.6) (6.0) {8.5) (12.0) s (5.5) (12.5) ( 16.5) (6.3) (11.2) {16.0) (5.3) (10.9) ( 13.0) (6.5) {14.7) ( 16.1} 11

135 Table XIX. Total number of days (means in parentheses) from hatching until first oviposition in Pseudosinella petterseni and P.alba. (Number of individuals of the former species, 15, 13 and 10, and of the latter, 13,10 and 13,at 24,17 and 11 C.,respectively) Temperature ( C.) Species Number of days f. :eetterseni 10.5x (27.9) (48.6) (99.7) f..alba ~ {37.5) (59.2) {88.7) x From Davis and Harris (1936) z A single individual which never oviposited is inc1uded here since it was the only one which reached the seventh instar in so short a time.

136 Table XX. Numbers of adult Pseudosinel1a petterseni and t.alba extracted from five samp1es per month during June to September, 1961 (for details of samp1ing, see text). Mon th Mean soil. Numbers extracted temperature (0 c.) f..petterseni 1 P.alba June Ju1y August September Totals

137 Table XXI. Comparison of life-histories of fseudosinella petterseni and P.alba. (Figures in parentheses are averages} P.petterseni Size of adult (mm.} Color of adult Habitat Food Number of eggs per oviposition (24 C.) Size of egg (mm.} Incubation period at 24 C. (days) Incubation period at 11 C. (days} Percentage hatching at 240 C. Percentage hatching at 110 C. Size of young on hatching (mm.) Number of moults before oviposition Duration of first instar at 24 C. (days) Duration of first instar at llo C. (days) Duration of sixth instar at 240 C. (days) Duration of sixth instar at 110 C. (days) Duration of postembryonic development at 24 C. { days) Duration of postembryonic development at lloc.(days) Average time to complete life-cycle at 24 C.ldays) Average time to complete life-cycle at 110 C.l da ys) Percentage survival of adults after 30 days at oo C. (10 individua1s in each case) Dull white Under logs or stones, in mull soil and litter, in greenhouses, in caves and in nests of ants. A wide range of living and decaying vegetable matter; yeast alone in the laboratory {7.6) {34.0) (3.8) (15.6) 4-8 (5.3) ( 13.9) (27.9) (99.7) Variable, usually shining dirty white, grey or black Under stones on sandy soil and surface litter under conifer and maple stands, and in caves. Decaying leaves with yeast; possibly sorne animal matter since eggs of own species may be eaten (8.3) (31.2) (5.2) (11.6) 5-10 (6.5) (16.1) (37.5) (88.7)

138 IX. FIGURES

139 A B c ~ -1 Fig. l. Variation in post-antennal organ of Folsomia similis A. Without denticles on the inner margins B. With seven denticles C. With many denticles (number variable) Note the septum dividing the organ into two halves; this septum is rarely absent in Quebec specimens.

140 F i g. 2 Fig. 3 Fig. 2. Anal aperture of Folsomia similis Fig. 3. Male genital plate of Folsomia similis a. Genital aperture b. Lips c. Genital dise Fig. 4. Female genital aperture of Folsomia similis

141 Dens A B c D E F Fiq 5 Fig. 5. Sorne variation in the number and arrangement of ventral setae on the manubrium of Folsomia similis A. 8(4+4) setae ("normal" and regular) B. 8(4+4) setae (irregular) C. 4( 2+2) setae D. 11(7+4) setae E. 14(4+10) setae F. 17(12+5) setae