Reproductive performance of fiel&caught Glossina pa//idipes maintained on different host bloods

flics were significantlg different from goat and waterbuck-fed flies. The highest number of puparia produ- ced per 90 females was by rabbit-fed flies (83) whereas tbe lowest was by goat-fed (60) flies. Mortality was high (84-99 %) irrespectire of the group.


INTRODUCTION
Though successful colonization of the tsetse fly, G/ossina pallidipes has been reported (10,15,17,19,21), most populations have been reared with difficulty. For example, LANGLEY (15) was unsuccessful in rearing a Zimbabwe strain of G. pallidipes . Similarly, the colony of G. palhcfipes maintained by the ICIPE that originated from the Lambwe Valley eventually died out from viral infection. Most of these colonies have been maintained on rabbits, rather than on more preferred natural hosts. CURTIS and JORDAN (3) found that goat-fed Glossina austeni survived slightly better than rabbit-fed flies, however, the fertility of rabbit-fed flies was consistently higher. A very high fertility was also reported by ITARD et a/. (8) for Glossina tachinoides fed on rabbit blood.
Many species of animals are rarely fed on by the tsetse, for example zebra, wildbeest, waterbuck, impala, etc. (11). In the wild, G. palhdipes feeds mainly on bushbuck, buffalo, warthog and bushpig (11). It is not clear whether G. pallidipes cari be raised successfully on preferred host bloods. The objective of the study was to compare the reproductive performance of field-caught G. pallidipes maintained on various host bloods.

1, The International
Centre of Insect Physiology and Ecology, POB 30772, Nairobi, Kenya.

MATERIALS AND METHODS
G. palkfipes was trapped at Aitong near the Maasai Mara reserve in Kenya. Non-teneral females were selected and fed once on rabbits in the field at capture (males were discarded). Males were not used in this study since the females were assumed to have mated in the wild and therefore inseminated. Previous studies in Kenya show that most nonteneral females were already mated when caught in traps (2). Female flies were held in PCV cages (21 x 16 x 18 cm) with black nylon netting, and transported by air to Nairobi within 48 h, where they were maintained at 25 i 1 "C, with a IightIdark cycle of 12/12 h and 70-80 % relative humidity. They were allowed 5 days to stabilize while being fed on their respective host bloods. For each of the five host bloods, there were 6 cages, each containing 15 flies. Each group was membrane-fed on fresh defibrinitated blood (9) of either rabbit, buffalo, goat, eland or waterbuck on altemate days for 41 days. MADUBUNYI (18) showed that faily feeding had no advantage over feeding on alternate days. Daily mortality, puparia production, weight of puparia and number of abortions were recorded. Any fly which died during the experiment was immediately dissected in order to determine its mating status. All flies were dissected at the end of the experiment.

RESULTS
Mot-tality was very high (84-99 %) at the end of the study period, irrespective of the group ( fig. 1). Buffalo-fed flics had a high early mortality (30 and 77 % after one and two weeks, respectively) ; rabbit-fed flies survived best to the end of the study. Abortions were rare after the flies had stabilized in the laboratory for 5 days (~11 per group, table 1). Rabbit-fed flies produced the highest number of pupae (83). For statistical analysis (x2 tests), expected pupal production was calculated from the cumulative number of days females survived ,divided by the average expected cycle length (10 days). There was a significant heterogeneity (pc 0.05) in the pupal output among host bloods, with rabbit blood still being best even after weighting for mortality (table 1). The number of puparia per female reproductive cycle was about 1/2 of the expected output based on a 1 0-day reproductive cycle. Mean puparial weights were above 30 mg in all groups (table II). However, mean puparial weight for flies fed on rabbit blood was higher (37.2 mg + 0.85 SE) than for flies fed on buffalo, waterbuck, eland or goat blood. Mean puparial weights differed significantly (pc 0.05) between rabbit, eland and buffalo-fed flies, versus goat and waterbuck-fed flies.

DISCUSSION
Although tsetse are selective in their choice of hosts (except for the palpalis group) (1 l), the present study shows that they cari be maintained on less preferred hosts. This is in agreement with BUXTON (l), MOL00 et a/. (20) and LANGLEY (14). It is now well known that Glossina morsitans morsitans, Glossina palpalis, Glossina tachinoides, etc. cari be reared easily on less prefereed hosts and as such very little importance was given to them in this study.
Mortality of field-caught G. pallidipes brought to the laboratory for this study was uniformly high. This high mor-tality was probably due to the flies either rejecting captivity or adapting to a different feeding regime rather than infection picked up from the blood meal (blood meals were uninfected). ITARD and BAUER (9) stated that tsetse are adapted to feeding on certain host animals and would not necessarily change this habit unless a period of adaptation is provided. This fact was clearly demonstrated by FILLEDIER and BAUER (5) who successfully reared Glossina morsitans submorsitans in the laboratory after an initial period of difficulty. In this study, rabbit-fed flies survived best. Thus, failure to maintain certain populations of G. pallidipes successfully on rabbits in the laboratory is not due to a defi-~ ciency in the quality of the blood.
The authors also found that flies maintained on rabbit blood tended to have heavier puparia. This may be a reflection of the nutritional quality of rabbit blood. KABAYO and LANGLEY (13) found that dietary lipids play a vital role in supporting reproduction in the tsetse, e.g., serum albumin and lipoproteins are vital for ovarian growth. LANGLEY and PIMLE,Y (16) also found that a cow blood diet reduced the extent to which the lipids were utilized for larval growth, thus causing an altered secretory activity in the female uterine gland. There may therefore be some disadvantages of feeding on certain host bloods. For example, tsetse allowed to feed on crocodile produced 5 larvae but those on goat, 141 (1). WETZEL and LUGER (23) also obtained better results with porcine blood than equine or bovine blood. Similar results were reported by WILLET (24) and FOSTER (6).
The number of puparia produced by flies fed on rabbit blood was also higher than flies fed on any other host blood. These results for G. pallidipes contrast with those for G. m. 'tiorsitans, where MOL00 et a/. (20) found only marginal differences among flies fed on various host bloods. LANGLEY (14), also working on G. m. morsitans, repot-ted that impala blood (a very rare host) is asmdigestible as that of preferred hosts. The so-called preferred hosts may be those that the flies have adapted to feed on because they are present in the flies' locality. However, if these hosts are removed, it may lead to a disappearance of tsetse from the area (4, 7, 11). This was also the case in Nagupande, ex Rhodesia, an area well populated with game of many species, where G. morsitans was practically eliminated by removal of the preferred hosts only (22).