The Distribution of the Midge C Riparius in a Polluted River Ystem and Its Environs 499 M

Air & Wat. Pollut. Int. J. Pergamon Press 1966. Vol. 10, pp. 757-768 Printed in Great Britain.

THE DISTRIBUTION OF THE MIDGE C RIPARIUS IN A POLLUTED RIVER YSTEM AND ITS ENVIRONS 499 M. A. LEARNER and R. W. EDWARDS

Water Pollution Research Laboratory, Ministry of Technology, Stevenage, England

(Received 15 June 1966)

Abstract—A survey was carried out in an area which had previously given rise to complaints of infestations of the midge, Chironomus riparius, a species generally associated with organically polluted water courses. Larval densities in the water courses of the area, where previously high numbers had been recorded, were quite low. C. riparius which had seven emergences between March and November, rarely constituted more than 30 per cent of the catch of flying insects on sticky traps. Other chironomids coming from neighbouring gravel pits were often present in numbers similar to C. riparius The species composition of the catches of sticky and suction traps was different—the sticky traps catching a higher proportion of chironomid midges. Significant aspects of the life-history of C. riparius are described. Three granular formulations of DDT were tested as larvicides in the laboratory. Mortalities of less than 70 per cent were attributed to the patchy distribution of DDT on the mud.

1. INTRODUCTION FOR SEVERAL years, swarms of non-biting midges, mainly Chironomus riparius Meigen gave rise to complaints in parts of the Colne Valley, near the Hertfordshire Main Drainage Authority's sewage works at Maple Lodge. Until recently the effluent from this works flowed down Flakes Channel (FIG. 1), where in 1954 and 1955 EDWARDS (1957) found larval densities exceeding 100,000/m' of river bed. Although further estimates of larval densities were made in Flakes Channel in 1962 when about 40,000/m' were present, it was not until 1965 that a field survey of the problem could be carried out. In 1964 discharge to Flakes Channel ceased and the effluent from the sewage works was discharged directly to the Grand Union Canal (FIG. 1). In view of the possible need to control midges chemically, a field experiment was also conducted in certain Hertfordshire ponds during 1961 and 1962 to test the efficacy of TDE, an organo-chlorine pesticide, in controlling the larvae of C. riparius and related species (EDWARDS et al., 1964).

2. BIOLOGY OF Chironomus Riparius C. riparius belongs to the sub-family Chironominae of the fly family Chironomidae in which there are about 400 known species in Britain. A useful review of those species found in water-supply reservoirs is given by MUNDIE (1956). As adult C. riparius midges do not bite and, as far as is known, do not transfer pathogens, complaints generally arise from their habit of collecting in large mating swarms under certain conditions. Swarms, whose development is largely determined by light intensity, temperature, and wind velocity (GinsoN, 1945) consist principally of males. Females appear to be attracted to the swarm by vision and the sound 757 758 M. A. LEARNER and R. W. EDWARDS produced by wing vibrations, and after coupling, pairs drop from the swarm on to near-by foliage (GissoN, 1945; COE et al. 1950). After mating the female does not return to the swarm but lays her eggs in a gelatinous mass which is attached to aquatic plants and other objects just below the water surface. It has been found that, under laboratory conditions, 90 per cent of the females lay their eggs between 2 and 4 days after emergence and that 10 per cent lay a second batch before dying. Each female lays about 600 eggs, which hatch in 4 to 5 days at 18° C. The larval stage has 4 instars, separated by moults, and overwintering occurs in the fourth larval instar. Pupation does not seem to occur at low temperatures. The larvae, which are detritus feeders, build tubes in mud, and fourth instar larvae penetrate to a depth of about 10 cm. The earlier larval instars and pupae are generally confined to the surface layers of mud.

R.COLNE

MAPLE LODGE SEWAGE WORKS WE I R---.

FORMER SEWAGE EFFLUENT OUTFALL N GRAND UNION PRESENT CANAL OUTFALL F

1 0-0 SITES OF STICKY-TRAPS 0-0 MUD SAMPLING STATIONS GRAVEL PIT c0) LAKE I55m

GRAND UNION TROY ARM CANAL

R. COLNE FIG. 1. Map of the survey area showing trapping and coring stations.

Larvae are apparently resistant to low oxygen concentrations, although they are unable to withstand complete absence of dissolved oxygen for prolonged periods. During October and November 1965, there was little flow along Flakes Channel and there was no detectable dissolved oxygen in the water. The surface of the mud became covered with Beggiatoa, a sulphur bacterium, and Chironomus larvae left their tubes

Distribution of the Midge Chironomus riparius in a Polluted River System and its Environs 759 and moved to the air—water interface at the edges of the channel. The numbers of larvae at both stations in this channel (E and F) were extremely low at the end of this period of anaerobiosis. The pupal stage lasts a few days and until emergence the pupa remains within the mud. Before emergence the pupa rises to the water surface: the whole emergence process takes only a few minutes. Laboratory experiments by PHILLIPP (1938) suggested that C. thummi (C. riparius) emerges principally in the afternoon. On one occasion, in March 1966, the daily emergence rhythm was studied in Flakes Channel by collecting for 24 hr all the ascend- ing pupae, pupal exuviae (cast pupal cases), and adults caught in a plankton net fixed in mid-channel so that a quarter of the net opening was above the water surface. It will be seen from FIG. 2 that the ascent of pupae and subsequent emergence occurred between noon and 8 p.m., the main emergence being around 3 p.m. Pupal exuviae and adults were found in low numbers throughout the day but their presence does not provide evidence of emergence at this time.

14 12 ADULTS 10 8 6- E 4 2 13 0 0, 200 180 EXUVIAE •o 160 140 LI Z 120 100 80 60 o 40 in ct 20 Li 0 6 - PUPAE 4 - a - 1-1 o 7 1;I31 1 o 12 16 20 24 TIME (h)

Flo. 2. Daily emergence rhythm of C. riparius. Hours of darkness are indicated by shading.

FIGURE 3 summarizes the life-history of C. riparius and shows scale drawings of all the stages together with details of wing venation, adult body pigmentation, and the male reproductive apparatus, these latter features being used in identification (CoE et al., 1950). 760 M. A. LEARNER and R. W. EDWARDS

I M

ADULT

AIR PUPA EGG BATCH WATER LiiRVA

may// pAuD P‘LARVAe.

FIG. 3. Life-cycle of C. riparius (e) and details of importance in identification. (a) Adult male (dorsal view), (b) adult female, (c) wing, (d) male—tip of abdomen (dorsal view), (f) egg mass (g) larva-4th instar, (h) pupa. (a) and (b) after STRENZKE (1959) and (h) after JOHANNSEN (1937)

3. ESTIMATION OF DENSITIES OF CHIRONOMUS LARVAE AND OTHER AQUATIC INVERTEBRATES Larval densities within the mud were determined using a metal corer fitted with non-return valves in the lid (FIG. 4A). Mud samples were sifted through 18 and 72 mesh/in, sieves and the retained animals removed for counting and identification. Other macro-invertebrates, mainly tubificid worms and Asellus aquaticus, were also found. Initially, in February and early March 1965, three samples were taken every 100 yd (91.4 m) down the water courses within the study area—a total length of 4000 yd (3656 m) of channel, river, and canal. The cores were taken one near each bank and one in the middle of the water course. For routine monthly collections six stations were selected, and four cores, distributed across the water course, were taken at each station (FIG. 1).

Distribution of the Midge Chironomus riparius in a Polluted River System and its Environs 761

NON RETURN \. —VALVE

PROTECTIVE --COVER

RE MOVEABLE 14‘.. PLATE

FIG. 4. Diagrams of (A) corer and (B) sticky trap. FIGURE 5 shows the density of tubificid worms and of Chironomus and tanypodine larvae in the mud at all stations during 1965. Although there were increases in the density of Chironomus larvae in the canal downstream of the effluent (B and C) in the late summer, the average density remained below 10,000/m2 at all stations throughout the year. Such densities of chironomids are no higher than have been reported from some lakes and storage reservoirs. MUNDIE (1957) found 36,000 midge larvae/m2 during the winter in Kempton Park East Reservoir, and 10,000/m2 during May in Staines South Reservoir.

STATION A 1000 ' - 0.000 0 w) -Q J. 13 .. _.. _.... SR ,31 P— 5000 STATION B 0.000 c F, - 4 000 - t 0,000 .. o c 3000 HI RONOMUS .1% - 30,000 Li F ./ . 2000 TA NY POD I NAE ,, 20.000 • TUBIFICIDAE .. 1 000 , r / 0,000 1 RI N P' O -I " STATIONN C c2 E 1000 . 0,000 4-'i P P Ea fl P PP PIL uj 0 - O Fi 1 - > 4000 . STATION D ,..' L.. J 40.000 g •a 3000 _ 30,000 n 1 r - 2000 . 1 20.000 3 1000 0,000 _N P„, P P P PM PI STATION E 0 3000 30.000 1I 2000 1 20,000 1000 , 10,000 I L IL -9 •1 E I' E PO P 0 STAT 10 N F 1000 - 0,000 0 IF1 ., 2 4 I _R P P 0 Feb. Mac Apr. May Jun. July. Aug Sept Oct. Nov. Dec. FIG. 5. Densities of Chironomus, Tanypodinae and Tubificidae in mud, 1965. P indicates the presence of tubificids in small numbers. 762 M. A. LEARNER and R. W. EDWARDS There was a fairly even distribution of Chironomus larvae across the bed of Flakes Channel (E and F) and Troy Arm (D), but in the Canal, just downstream of the effluent, the greatest numbers were found near the bank on which the effluent outfall was sited. Tanypodine larvae, which are generally carnivorous and have been reported to feed on larvae of other chironomid species (MuNDtE, 1957) occurred principally in the Canal upstream of the effluent (A) and in Flakes Channel (E and F) where densities up to 750/m2 were found. Tubificid worms were fairly abundant at all stations and particularly so in Flakes 2 Channel (E and F) where densities reached 10,000/m . BRINKHURST and KENNEDY (1965) found evidence that in one river the presence of Chironomus larvae adversely affected tubificid populations: there is no evidence of such competition in the present study. In April 1965, mud samples were taken at all stations (A—F) for organic carbon analysis (TABLE 1). While there was an increase in organic carbon content downstream of the effluent, it is similar to values recorded for muds taken from other rivers in Hertfordshire (EDwARDs and ROLLEY, 1965). There was a correlation between the average density of larvae at each station and the organic carbon content of the mud (r = 0.88 : P <0.01); a similar relation has been reported by WENE (1940).

TABLE 1. ORGANIC CARBON CONTENT OF DEPOSITS IN APRIL AND AVERAGE NUMBERS OF LARVAE OF C. riparius FROM APRIL TO DECEMBER 1965

Organic C Average Station (%) average of larval 3 samples (Nos./m2) A 6.6 137 B 10.8 1480 C 8.7 370 D 10.9 1309 E 8.0 810 F 5.8 516

Each time mud cores were taken the dissolved-oxygen content of the overlying water was measured. Except in the Canal upstream of the effluent (A), dissolved-oxygen concentrations were generally below 6.5 mg/1 and at stations E and F fell to nil in October and November.

4. DENSITIES OF ADULT MIDGES • Emergence cages (EDwARDs et al., 1964), which either float on the water surface or are submerged, could not be used in parts of the study area because of boat traffic, so counts were made of the numbers of aquatic insects alighting on sticky traps placed near the water courses. Traps, each consisting of four Perspex sheets (15 x 10 cm) mounted vertically at right angles on a 2-m pole (FIG. 4B), were placed at the six stations shown in FIG. 1. The Perspex sheets were coated in a proprietary grease and were removed weekly for examination. Trapped insects were removed from the sheets by soaking in methanol. Results of this trapping programme are shown in FIG. 6. •

Distribution of the Midge Chironomus riparius in a Polluted River System and its Environs 763 Although it is claimed that the aerial density of insects can be determined from sticky-trap catches, provided the wind velocity is known (JOHNSON, 1950), the present authors have used the data only to give an indication of changes in density at different sites throughout the year. A suction-trap of the Johnson—Taylor type (TAYLOR, 1951) was installed at Station 1 so that sticky-trap catches could be compared against a standard device which draws air through a gauze cone at a constant rate. FIGURE 7 shows this comparison for the twelve weeks during 1965 when suction-trap catches were analysed. Although there is a correlation between catches of C. riparius (r = 0.73; P<0.01), there is no significant correlation for total insect catches. In view of the poor

STATION! 400 o---e TOTAL INSECTS o---o TOTALCHIRONOMIDAE 300 ..--• TOTAL CHIRONOMUS RPARIUS 200 - -a 100 • 0 I- 300 STATION 2 200 - • 100 O 0 STATION 3 .t 100 ... "litlarrtf8t1010W La _ 0 S TIO 4 u • .' „....A16:„...._,A„. _ ._ _ STATION 5 _ _ co 100 ... -. ..- • 0 300 _ ST ION 6 ZOO 100 0 ... _uuln ■ •-...reia■miam•A-urome■Ala.- _ April May June July Aug. Sept. Oct. Nor. Dec. TIME OF YEAR FIG. 6. Weekly insect catches on sticky traps, 1965. correlation between numbers of insects caught by the two types of trap and the difference in composition of the insect catches from these traps (TABLE 2) only data from the sticky traps will be considered further. These have been selected because sticky traps were installed at several stations and caught a higher proportion of C. riparius than the suction trap. Furthermore they are likely to be used on similar surveys because of their low cost and simplicity of operation. FIGURE 6 shows the numbers of C. riparius, all chironomids, and all insects caught on sticky traps during 1965. Numbers of C. riparius were very low during the spring and early summer but increased rapidly in August at all stations; emergences did not decline until mid-November. This seasonal increase in catches confirms observations on larval density shown in FIG. 5. From April until September, C. riparius generally constituted less than 10 per cent, and rarely more than 30 per cent, of the total insect catch. Only at the beginning and end of their emergence season, March and November, when most other insects were not emerging, did this species constitute more than 50 per cent of the total catch. The ratio of male to female C. riparius caught on the sticky traps was between 2 : 1 and 3 : 1. This might be expected since males are in flight for much longer 764 M. A. LEARNER and R. W. EDWARDS periods than females. However, results from the suction trap generally gave a sex ratio much nearer unity. The numbers of C. riparius caught on all sticky traps throughout 1965 are shown in FIG. 8. These data were analysed by the method of HARDING (1949) to determine the number of generations giving rise to the polymodal distribution shown in FIG. 8 and it was concluded that there were seven emergences during the year. The period

;2-80 400 TOTAL INSECTS CL g60

s rt, 40 200 X

>- 20

0 0 200 400 5000 10000 SUCTION-TRAP (Nos/trap week) FIG. 7. Comparison of the numbers of insects caught with suction and sticky traps at Station 1 during 12 weeks in 1965.

TABLE 2. COMPOSMON OF INSECT CATCHES FROM SUCTION AND STICKY TRAPS AT STATION 1 DURING 6 WEEKS IN 1965

Percentage composition Insect group Sticky trap Suction trap C. riparius 12.2 4.8 Other Chironomus species 4-2 1.0 Other Chironomidae 20.0 6.9 Culicidae 0.1 1.3 Ceratopogonidae F7 0.6 Psychodidae 17.4 48.6 Trichoptera 0.1 0-2 Tipulidae (+ Trichoceridae) 0.2 0.3 Other insects 44-1 36.3 between emergence peaks, determined during the present study and one carried out in the same area during 1954, is related to water temperature (FIG. 9). The relation is not a precise one and at the moment development times cannot be predicted with great accuracy. The proportion of chironomids in the total insect catch was frequently between 50 and 60 per cent but was occasionally above 80 per cent. These very high proportions were due to the appearance of large numbers of Tanytarsus spp. and of orthocladinines, particularly Cricotopus spp. Tanytarsus sylvaticus, which was numerous during April and early May, is particularly associated with fairly barren lakes and reservoirs. MUNDIE (1957) recorded T. sylvaticus in the Kempton Park East Reservoir where a single emergence occurred in March and April. In the present investigation larvae of

Distribution of the Midge Chironomus riparius in a Polluted River System and its Environs 765 this species were not found in mud cores within the study area, and as pupal exuviae were found on a near-by gravel pit it seems likely that this species emerges from this and other pits. Other Tanytarsus species appeared during the year, particularly T. mancus, and these too probably came from gravel pits. Adults of certain Cricotopus

350-

a. f! 300 -

LT1 z 250- 0

150 -

u 100- 'I) 0

50-

Mar April May June July Auo. Sept. Oct. Nov. Dec. TIME OF YEAR FIG. 8. Emergence of C. riparius in 1965. Peaks indicated by encircled figures.

x-1955

1 0-1965 '20

0 w 15 cc

Euc 10 0.

5 0 10 20 30 40 50 60 TIME REQUIRED FOR COMPLETION OF ONE GENERATION (Days)

FIG. 9. Relation between temperature and generation time for C. riparius. species were numerous at Stations 1, 2, and 3 in April and May and again at Station 1 in September and October. The larvae were probably associated with weeds developing in parts of Flakes Channel and gravel pits.

5. CHIRONOMID CONTROL

MUNDIE (1956) concluded that if any measure of lasting control is to be obtained operations must be directed against the aquatic stages. Several biological methods and chemical agents have been suggested (EDwARDs et al. 1964) but the most effective control has been achieved with the use of organo-chlorine insecticides—these corn- 766 M. A. LEARNER and R. W. EDWARDS pounds sometimes have unfortunate side effects however (HUNT and BISCHOFF, 1960). A settleable powder formulation of an organo-chlorine insecticide (TDE) has been used successfully in static bodies of water (EDwARDs et al. 1964), but in flowing waters (LEWIS, 1957), where coarser formulations are required for more rapid settlement, adequate control has not been achieved. In the present investigation, the toxicity of three granular formulations of DDT to larvae of C. riparius was determined in laboratory tests. The formulations consisted of 5 per cent active insecticide (about 70 per cent pp'-DDT, almost 30 per cent pp'-TDE plus op'-DDT, and a trace of pp'-DDE) and a heavy inert carrier. Larvae were intro- duced to aquaria containing a layer of mud, at a density of 20,000/m2. After one day, the formulations were applied at rates equivalent to 1.1 kg/hectare, and nine days later mortalities were determined (TABLE 3). It seems likely that inadequate dispersion of

TABLE 3. PROPERTIES OF 3 EXPERIMENTAL DDT FORMULATIONS AND LABORATORY TESTS ON LARVAL MORTALITY

Rate of settlement Disintegration Particle (time in sec for % Kill in water size 95% of formulation of C. riparius Formulation (hr) to sink 1 m) larvae

1 <2 600-1200 40 63 2 <24 600-1200 45 40 3 <24 250-350 130 67

18 16 14 12 NO LARVAE 10

6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34

NUMBER OF SAMPLES 14 LARVA.E PRESENT 12 10 8 6 4 2 0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32 34 WEIGHT OF DDT+T DE IN CORE (y9) FIG. 10. Distribution of DDT and TDE (Formulation 3) in mud, determined from analysis of cores (surface area 1-33 cm2). Distribution of the Midge Chironomus riparius in a Polluted River System and its Environs 767 the formulations on the mud surface was responsible for the limited mortalities since of the formulations which were most effective one had the smallest particle size and the other disintegrated most easily (TABLE 3). If formulations were applied to a river at the start of the larval overwintering period, it is likely that higher mortalities would be achieved because of the considerably longer exposure period than was provided in laboratory tests. FIGURE 10 shows the distribution of DDT + TDE in the mud at the end of the test period after the application of Formulation 3 to two tanks, one without and the other containing larvae. The distribution was determined by taking 36 mud cores (1-33 cm' surface area) from each tank. The distribution of pesticide was very uneven and as the area of each core contained on average about 16 granules of the formulation, a more normal distribution was expected. Over 30 per cent of the cores contained less than 2pg DDT + TDE. The presence of larvae within the mud did not facilitate a more even distribution of DDT + TDE on the mud surface. The overall recovery of pesticide from the tanks was quite low, only about 50 per cent of that added being recovered within the mud and larvae. The overlying water contained negligible quantities of insecticide. Dead larvae contained on average 0-021 jig DDT + TDE and those re- maining alive at the end of the test contained 0.019 jig—the difference is not statistically significant.

6. SUMMARY

1. A survey was carried out in an area which had previously given rise to complaints of infestations of the midge, Chironomus riparius. Significant aspects of the life-history of this species are described. 2. Larval densities in the water course, where previously high numbers had been recorded, were quite low. 3. C. riparius rarely constituted more than 30 per cent of the catch of flying insects on sticky traps. Other chironomids coming from neighbouring bodies of water, mainly gravel pits, were often present in numbers similar to C. riparius. 4. Analysis of trap catches suggested that there were seven emergences, between March and November, during 1965. The generation time was related to temperature. 5. Insect catches at one station, where a suction and sticky trap were installed, were poorly correlated and the species composition of the catches from these traps were quite different—the sticky trap catching a higher proportion of chironomid midges. 6. Three granular formulations of DDT were tested in the laboratory. Larval mortalities of less than 70 per cent were attributed to the patchy distribution of DDT on the mud.

Acknowledgements—The authors wish to thank riparian owners who gave permission for the erection of traps and the Physical Chemistry Division of the Woodstock Agricultural Research Centre, Shell Research Ltd., for supplying formulations of DDT. The Laboratory of the Government Chemist analysed mud and water samples for organo-chlorine residues. Mr. P. J. MAIUS helped with field sampling. Crown copyright. Reproduced by permission of the Controller of H.M. Stationery Office. 768 M. A. LEARNER and R. W. EDWARDS REFERENCES BIUNKHURST R. 0. and KENNEDY C. R. (1965) Studies on the biology of the Tubfficidae (Annelida, Oligochaeta) in a polluted stream. J. Anim. EcoL 34, 429-443. COE R. L., FREEMAN P. and MATTINGLY P. F. (1950) Handbooks for the Identification of British Insects No. 9 Diptera, Nematocera. 216 pp. Royal Entomological Society, London. EDWARDS R. W. (1957) Vernal sloughing of sludge deposits in a sewage effluent channel. Nature, Lond. 180, 100. EDWARDS R. W., EGAN H. A., LEARNER M. A. and MARIS P. J. (1964) The control of chironomid larvae in ponds using TDE (DDD) J. app!. EcoL 1, 97-117. EDWARDS R. W. and ROLLEY H. L. J. (1965) Oxygen consumption of river muds. J. Ecol. 53, 1-19. GmsoN N. H. E. (1945) On the mating swarms of certain Chironomidae (Diptera). Trans. R. ent. Soc., Lond. 95, 263-294. HARDING J. P. (1949) The use of probability paper for the graphical analysis of polymodal frequency distributions. J. mar. biol. Ass. U.K. 28, 141-153. HUNT G. E. and BISCHOFF A. I. (1960) Inimical effects on wildlife of periodic DDD applications to Clear Lake. Calif. Fish Game 46, 91-106. JOHANNSEN 0. A. (1937) Aquatic Diptera. Part IV. Chironomidae: Subfamily Chironominae. Cornell University Agricultural Experiment Station. 50 pp. JOHNSON C. G. (1950) The comparison of suction-trap, sticky trap and tow-net for the quantitative sampling of small air borne insects. Ann. app!. Biol. 37,268-285. LEWIS D. J. (1957) Observations on Chironomidae at Khartoum. Bull. ent. Res. 48, 155-184. MUNDTE J. H. (1956) The biology of flies associated with water supply. Instn pub!. Hlth Engrs J. 55, 178-193. MUNDIE J. H. (1957) The ecology of Chironomidae in storage reservoirs. Trans. R. ent. Soc., Lond. 109, 149-232. PHILLIP P. v. (1938) Experimentelle Studien zur Okologie von Chironomus thummi Kieffer. Zoo!. Anz. 122, 237-245. STRENZKE K. (1959) Revision der Gattung: Chironomus Meigen. 1. Die Imagines von 15 nord- deutschen Arten und Unterarten. Arch. HydrobioL 56, 1-42. TAYLOR L. R. (1951) An improved suction-trap for insects. Ann. app!. Biol. 38, 582-591. WEE, G. (1940) The soil as an ecological factor in the abundance of aquatic chironomid larvae. 4. Ohio J. Sci. 40, 193-199.