Sticky Trap Estimates of Efficiency

Nzi Trap Efficiency in Canada - Initial Observations Mihok, S., Carlson, D.A. & Ndegwa, P.N. (2007) Tsetse and other biting fly responses to Nzi traps baited with octenol, phenols and acetone. Medical and Veterinary Entomology, 21, 1-15.
Pooled Estimates Individual estimates Assumptions Orientation behaviour
Sticky Triangle (Experimental Trap)		Control Trap (Plastic Sheeting Only)
During early August 2002, I attempted to estimate the number of biting flies attracted to the vicinity of a trap in a residential setting at my home in Russell, Ontario. I used a standard cotton drill Nzi trap, either unbaited or baited with a Biosensory octenol lure. I encircled the trap with a 1-m high barrier of Rentokil Fly Control Adhesive Film to form a transparent "sticky triangle" similar to "traps" I have set in Africa (here forming a continuous sheet). I counted the flies landing on the sticky film relative to those captured at a control trap nearby. To control for the modified appearance of the experimental trap, I attached plain plastic film to all blue surfaces of the control trap. The experiment was run for six days with collections at 3-hr intervals from 9 am to 9 pm. For the first four days, I baited the traps with octenol on alternate days. On the fifth and sixth days, I removed the trap from inside the sticky triangle to examine attraction to the sticky triangle alone. Attraction to the "transparent" sticky film is important for some species as the chemical adhesive absorbs ultraviolet light, and, along with the plastic base, is shiny in direct sunlight (polarizes light). For example, in Africa, a supposedly "transparent" and unbaited 1-m-sized sticky triangle can catch 1,000 Stomoxyinae (stable fly species) a day. In contrast, other biting flies (tabanids and tsetse) are not caught in large numbers, if odour attractants are not present. Unfortunately, weather was quite variable after the first day, so replication among days was not satisfactory. I also now suspect that I may have lowered trap efficiency by simply having the trap face East (based on the assumption that flies should first encounter the front of the trap when following the octenol odour plume upwind). In June, 2003, I tested East/West trap orientations and found a significantly higher catch of tabanids (+45%) with traps facing West (i.e. with the entrance facing the afternoon sun). I have found the same for stable flies in another formal experiment. The sticky material maintained its adhesiveness throughout. I kept it relatively clean by regular removal of flies. The only problem encountered was morning dew on the first two days. This made the sticky film opaque until about 9 am. Morning was a period of low fly activity, so catches at this time contributed little to daily totals. I collected useful data for stable flies (Stomoxys calcitrans) and the most common tabanid (Tabanus quinquevittatus). The landing behaviour (areas, surfaces, heights) of these species was similar to unpublished observations with ecologically similar species in Africa. For deer flies (Chrysops), the sticky setup was not appropriate for estimating trap efficiency (Assumptions).
Both traps were set facing east. Direct sunlight therefore fell on the SW back side of the setup during peak fly activity in the afternoon. The traps were in full sun nearly all day, with only minimal shade from nearby trees early in the morning, and late in the evening. Tabanids are thought to orient in flight relative to the sun, and possibly also to polarized light, but very little work has been done to analyze their behaviour in detail. At Russell, peak tabanid activity occurs between 3 and 6 pm from the results below. During this time period in July at this latitude, the sun is 45 degrees above the horizon at a compass angle 15 degrees south of due west at 4:17 pm (Daylight Saving Time, Eastern Time Zone). Similar calculations can be made on the web for any locality. See the web sites: Sustainable by Design & Time and Date
Pooled Estimates of Trap Efficiency (total catches on sticky triangle / total catches in control trap)
Species	Attractant	Efficiency	N - Sticky Catch
Stomoxys calcitrans	Octenol	6.7%	508
	Unbaited	29.7%	145
Males	Octenol	9.0%	268
	Unbaited	40.3%	77
Females	Octenol	4.2%	240
	Unbaited	17.6%	68
T. quinquevittatus	Octenol	7.0%	258
	Unbaited	0.0%	39
Weather varied substantially, resulting in some puzzling trends in efficiency estimates. It's hard to know how typical these estimates might be for other situations, given strong correlations between catches and weather indices (tabanids - temperature, stable flies - temperature, wind speed). Further details are tabulated in Individual Estimates. The apparent differences between male and female efficiency estimates for stable flies may account for the male-biased sex ratios (2.3 5: 1 up to 2007) at this locality in Nzi trap catches of S. calcitrans. Here, sticky catches show that females may be present in nearly equal numbers, but are not getting caught with equal efficiency. A male bias occurred in both trap catches (1.9 : 1) and sticky catches (3.4 : 1) in unbaited Nzi traps set near livestock in Alberta. I have also collected with Vavoua traps in residential settings in both Ontario and in Alberta and have found an even stronger male bias (10.6 : 1, N=545). *In Africa, there is no sex bias in catches of S. calcitrans* in Nzi traps under diverse conditions (0.95 : 1, N=1,630). These inconsistencies are puzzling.** Unfortunately, I have almost no examples of sticky catches of S. calcitrans from Africa, but I have considerable data for other African species. For three common stable flies (S. niger niger, S. niger bilineatus, S. taeniatus) there is an opposite trend, i.e. a modest female bias in sticky captures (0.61 : 1, N=11,936). A female bias is also typical of Nzi trap captures in Africa. During early work before the Nzi trap was developed, a female bias was found for many Stomoxyinae in Vavoua traps. This bias can be extreme seasonally, and may have some biological significance (Mihok et al., 1996)

Assumptions About How Flies Behave
Tabanus quinquevittatus - Many horse flies were well-distributed over the sticky triangle, as if they had flown into it, without seeing it. Some flies were clearly stuck down head first. When I was at the trap, horse flies flew too fast to be observed, but they typically hit the sticky triangle with a "thump". Hence, at least for this species, sticky catches appear to be a reasonable estimate of the numbers of flies orienting to the trap. But, from the results below with no trap inside the sticky triangle, this assumption is clearly not absolutely true, especially in the presence of octenol. Unfortunately, weather was too variable across days to make reliable, direct comparisons among treatments. Better insights into fly behaviour will require simultaneous variations on the experimental setup.
Mean *Daily Catches of T. quinquevittatus***
Sticky Setup	Control Trap	Sticky Triangle	Exptl Trap
Trap present - unbaited	0	20	0
Trap present - octenol	9	129	0
No trap - unbaited	0	12	--
No trap - octenol	8	36	--
For stable flies, there was considerable variation in captures from day to day. Modest numbers of stable flies landed on the sticky triangle in the absence of any other visual cue, with or without octenol present. As with T. quinquevittatus, on warm days, octenol likely increased the numbers of stable flies attracted to the area.
Mean Daily Catches of Stomoxys calcitrans
Sticky Setup	Control Trap	Sticky Triangle	Exptl Trap
Trap present - unbaited	22	73	0
Trap present - octenol	17	254	0
No trap - unbaited	7	21	--
No trap - octenol	11	22	--
*Chrysops aberrans* The deer flies C. aberrans and C. univittatus were observed on a few occasions near the trap. They circled and hovered at slow to moderate speed in the vicinity. They were also observed flying near the trap cone, or the gap between the sticky triangle and the trap. They always were found to be stuck down "feet first". Some were caught in the experimental trap, even though access below 1 m was blocked. Clearly, these data do not fit the assumptions of the experimental design. Hence, it would be inappropriate to estimate trap efficiency for deer flies (100+%). Surprisingly, the sticky triangle itself was unattractive to deer flies, although deer flies landed on it when a trap was present inside.
*Total Catches of Chrysops* (deer flies)**
Sticky Setup	Control Trap	Sticky Triangle	Exptl Trap
Trap present - unbaited	5	4	1
Trap present - octenol	21	21	6
No trap - unbaited	8	0	--
No trap - octenol	18	1	--
I followed up on these observations by suspending a vertical piece of blue cotton drill cloth (100 cm x 50 cm) between two poles, with a 25 cm gap to the ground. I then wrapped a horizontal 50 x 100 cm layer of sticky film between the top of the poles so that it covered the top 25 x 50 cm section of the blue cloth, and the area beside and above it. I set this up to monitor the landing and circling behaviour of deer flies at a height typical of a person's upper torso (125 to 175 cm), relative to what I thought would be an appropriate cloth silhouette of a person's body. I monitored catches for a total of six days (3 days with octenol and 3 days without) relative to two Nzi traps set nearby (one with octenol and one without each day). The sticky human model caught only 2 deer flies (landing on the blue area) relative to a total catch in the two Nzi traps of 159 deer flies. Hence, it appears as if Chrysops rarely investigates stationary, 2-dimensional objects covered with sticky film, even at heights above one metre. Movement or 3-dimensional structure may be critical to entice flies to investigate. These quirks of behaviour are documented in the literature and have been exploited in a commercial product: the Trolling Deer Fly Trap (PDF).

Orientation of Flies to the Sticky Triangle + Trap

The distribution of flies on the three sides of the sticky triangle provided some useful insights into fly behaviour around Nzi traps. At Russell, prevailing winds during the summer biting fly season are mostly from the west. The trap inside the sticky triangle was therefore set facing east, so that flies would approach the entrance when flying upwind along an odour plume. The three sides of the sticky triangle therefore faced east, southwest and northwest.

The east side was in sun only in the morning at a time of low fly activity. The northwest side was in sun only in the evening at a time of waning fly activity. The southwest side was in direct sun during the critical time when flies were active (noon till 6 pm). The three sides also differed in terms of the background visual cues when a trap was present. The front contained 1.5 square metres of blue cloth with some contrasting black also visible. The other two sides each contained 1.0 square metres of blue and black cloth.

If flies do not react to winds/octenol, or to trap visual cues, then we would expect an even distribution with 33% of the captures on each side of the sticky triangle. Below are the total catches with and without a trap/octenol present, with my interpretation of what these results signify.

Orientation of Stomoxys

With no trap or octenol present, stable flies slightly favored the sunny side (SW). Addition of octenol resulted in a much stronger bias towards the sunny side. With a visual cue present (the trap) and no octenol, captures were almost even on all three sides. Addition of octenol resulted in a small shift in captures towards the downwind side (E).

Tabanus quinquevittatus Orientation

The sample sizes for Tabanus quinquevittatus when a trap was not present were not large (N=36 with octenol, N=12 unbaited). These results have therefore not been interpreted. With a trap present and no octenol, there were more flies captured on the sunny SW side. With octenol present, there was a clear shift in captures to the downwind side (E).

Hence, for stable flies and horse flies, there was a change in catch distribution indicative of probable upwind flight in reaction to an odour plume (octenol), so long as a visual cue (the trap) was present. This conclusion is also supported by an incremental shift in fly catches towards the eastern side of the sticky triangle during the course of the day (as the octenol evaporation rate presumably rose with rising air temperature). Catches on the eastern side increased progressively for each fly (12 pm to 6 pm collections: 36-42-50% for S. calcitrans, and 39-42-47% for T. quinquevittatus).

Chrysops Orientation

Deer flies (Chrysops) were not captured in large numbers on the sticky film. For the small sample of 42 flies captured when octenol and a trap were present, there was a very small bias towards capture on the downwind side (E).

My interpretation of these statistics is that trap orientation may not be absolutely critical, as flies clearly investigate the trap from various angles during the course of the day. Tradeoffs in efficiency between having traps face downwind, versus having traps face the afternoon sun, are not easy to predict.