Glossina fuscipes fuscipes Newstead 1910 (Diptera, glossinidae): Colour Preferences, Host-Funding Strategies and Responsiveness to Baited Traps

Abstract

This study was conducted on Glossina fuscipes fuscipes, a riverine species of Central and Eastern Africa. It focuses on (i) colour preferences, (ii) landing and feeding behaviours on a host, the Nile monitor and (iii) on the trappability in odour- and colour-baited traps. Colour preferences were studied using electric screens in the field. Colours that were found most attractive were combined to make traps. These traps were tested as visual baits, compared to the standard biconical trap Challier Laveissiere (1977). The performance of the most successful trap type was analyzed in relation with fly sex and fly abundance. It was baited with zebu urine at changing age of storage to estimate an optimum age of urine for baiting G. f juscipes traps in field conditions. Zebu urine was also tested in conjunction with acetone. Whole body odours and urine of the monitor lizard were tested as attractants too. Royal blue relatively attracted more flies than black colour ( 1. 7x). The peony purple red colour stimulated landing significantly more than royal blue. The attractancy of the trap colours was a function ofreflectance in the blue (P < 0.01) and in the infrared (P < 0.05) ranges of the spectrum. However, very bright colours performed weakly both for attraction and for landing, suggesting that moderate chroma in the range 450-490 nm and in the near IR (750-1100 run) largely determine the attraction of this fly. A biconical trap with the purple red component inside, caught 1. 4 times more males and 1.63 times more females (P = 0.03; n = 21); an increase of up to 3 times at low abundance sites (catch of maximum 10 flies/trap/day) was obtained. Such a trap is advisable at low densities, in routine trapping on farms and immediately after the initial fast depletion of tsetse numbers during extermination campaigns with conventional tools. A blue/red target is also proposed instead of the usual black or blue targets. More flies approached the biconical trap in the light region 1,600-1,700 μE/m2/s; the response of the fly became photonegative when illumination was above 1700 μE/m2/s. The number of males attracted was associated with the presence of bushes and that of females was very dependent on illumination rather than vegetation. G. f fuscipes showed a significant preference for the neck of the monitor lizard (P < 0.05). When the Nile monitor was present in the odour chamber of a compartmentalized experimental cage~ traversed by a wind flow, 61.5 to 67.5% of the flies performed an active upwind flight towards the next compartment where further search for the host occurred. Out of 100 flies that took off from the release chamber towards the varan, approximately 50% did so by ranging upwind. In nature, flies tended to aggregate at sites where monitor lizards live (r = 0.67; P = 0.05 for females and r = 0.54; P = 0.03 for males). When acetone and 5 to 6 week-old cow urine were used together as a bait on the standard biconical trap, the highest catch indices occurred at sites with low abundance of tsetse (average catch 0-10 flies/day), 1.7 times for males and 2.3 for females (P < 0.05). At low abundance (0-10 flies/day) the catch index of colour or odour-baited traps was high. A strong point of inflection appeared at the abundance level of7 flies/day, where the treatment was just as good as the control trap. At high abundance (>20 flies/day) the catch index was very low (ca. 0.25).