Rarely is it a good thing for a soldier to volunteer for something.  It would seem that Project Bellwether fell into the category of "bad deal".  During this experiment, volunteer soldiers were brought into the open desert to let mosquitoes feast on them in the name of research.   The experiment was conducted in 2 parts, and we will briefly examine what happened in both parts of the project, and what the overall goals were for this project.

Here's a quick summation of the activities:

30 (in the first trial) or 100 (in the second trial) US Army soldiers were subjected to studies on mosquito biting patterns at Dugway Proving Grounds, Utah.  Troops from the 45th and 46th Chemical Companies (in the first trial) or the 2nd Chemical Battalion (in the second trial), dressed in the standard Army work uniform of fatigue trousers bloused into combat boots and fatigue caps, went out into the field during two periods- September and October of 1959, and September and October of 1960- and were subjected to bites from mosquitoes released from specially designed release mechanisms, along with some mosquito traps.

[NOTE: Both reports cited in this thread are declassified and released to the public.]


First, we will examine BELLWETHER-I.


From the project report, pp. 3-6:


"The primary objective of this test was to ascertain the effects of major meteorological parameters on the biting rates of starved, virgin female Aedes aegypti mosquitoes upon troops in the open. Other objectives were (1) to see whether this mosquito, a domesticated, house-loving tropical mosquito, could be effectively tested in hot, open, temperate desert terrain and (2) to ascertain whether traps could be used to replace human samplers.

A total of 52 field trials were conducted between 1 September and 8 October 1959. The basic trial design oonsisted of three 15-foot radius circles, located 1/8-mile apart on a orosswind line. Ten men were located equidistantly around the perimeter of one circle, and either ten guinea pig baited traps or ten nonbaited traps were placed around the perimeters of the other two circles. The volunteers all remained seated in these trials. One hundred vectors were released in the center of each circle, and sampling was conducted for 30 minutes. A mobile meteorological station was located 1/4-mile upwind of the center circle.

Using uninfected, virgin female A. aegypti mosquitoes in Phase A of BW 459, the results obtained within the ranges of conditions encompassed in these trials indicate that:

1. It is feasible to test this mosquito under hot, dry, desert conditions, at least for the initial primary time period, and to assess the effects of various meteorological variables upon biting activity.

2. Although many of these trials produced erratic and unpredictable results, it would appear from the analysis of these data that each of the meteorological variables studied--wind speed, temperature, relative humidity, and solar radiation--exert a significant influence on the biting activity of the A. aegypti mosquito, and all would have to be considered as important parameters in any model designad to predict biting activity. However, the effects of the latter three factors were manifested only in terms of interaction with wind speed and with each other; wind speed alone had a direct effect upon biting activity. Moreover, within the ranges of conditions encompassed in these trials, it appears that wind speed was the most important factor affecting biting activity.

3. An increase of 1 mile per hour in the ambient wind speed was associated with a decrease of approximately six bites in a 15-foot radius circle with 10 volunteers over a 30-minute time interval.

4. The data suggest that the previously determined lower temperature limit of 59 degrees F for vector biting activity of the non-cold resistant strain is placed too high; however, at these lower temperatures some other factor(s), at present unknown, produce erratic results.

5. A 20-minute sampling period is sufficient to encompass an average of 80 per cent of the expected initial primary biting activity.

6. Guinea pig-baited traps captured one and one-half times as many mosquitoes as did nonbaited traps.

7. Whereas, on the over-all average, a single mosquito in a baited trap was equivalent to two bites on a human, excessive intertrial variation precludes replacing human samplers with traps to determine biting rate activity.

8. With a vector/host ratio of 10:1, in only a very few trials did 100 per cent of the volunteers report bites. The mean percentage of test subjects bitten lay between 60 and 70 per cent.

9. The over-all average outdoor biting rate for this vector was 40 bites per 100 mosquitoes in the time period studied.

10. No evidence of crepuscular-period biting preference was obtained in these trials."




Now, this trial was not just set up in order to find out about what mosquitos do, but as a basis to try to find out a basis for entomological warfare (EW) studies and creation of EW munitions.  The report mentioned that a lot was known about how to calculate effects of chem or bio warfare, but that more research on EW effects was needed- including how mosquitos move, their biting habits, and how long they stay around to continue biting.

Some more excerpts:



"The biological effectiveness of laboratory-reared and laboratory infected anthropods in transmitting disease to human beings is well documented in the literature. One hundred and thirty human volunteers have been infected with yellow fever and over two hundred with dengue by laboratory-reared and laboratory-infected yellow fever mosquitoes. ... The literature dealing with accidental and intentional laboratory infection indicates that the deliberate employment of infected arthropod vectors against enemy targets holds great strategic potential, and limited, feasibility field tests with several vectors have borne out this possibility Concurrent with these feasibiliity  demonstration field trials, mass production techniques of rearing selected vectors, producing agent, and infecting vectors with agent have been developed.

There are other requirements, however, that ordinarily must be satisfied in accepting and standardizing a potential weapon system for use. Some requirements of obvious importance would be realistic predictions of target effectiveness and casualty rates upon which munition expenditure calculations ultimately would be based. These requirements are, of course, formidable for any of the more conventional CW or BW weapons, and the numerous factors that must be considered and quantitatively appraised to satisfy these requirements need not be mentioned here. Although many of these problems would be common to any weapon systen, there are others attendant to the evaluation of an entomological weapon that are inherently unique.

...

For EW weapons, comparable field testing techniques have not been developed. The agent is carried to the human target by the target-seeking entomological vector and consequently is rather independent of the meteorological diffusion processes. Therefore. the dissemination of arthropod vectors from an EW munition results in behavior quite unique from other CB weapons. This uniqueness is basically associated with three broad but mutually dependent vector characteristics, all of which are related to the potential transmission of the disease agent to a target population. These characteristics are (1) movement and distribution, (2) biting habits, and (3) survival or persistency of effect.

Whereas aerosols, once released, are subject solely to the effect of the various meteorological parameters, arthropod behavior is quite different. The vectors move actively in all directions, and this movement is erratic both in time and space. Temporatures, seldom limiting in aerosol work, have a sharp lower cutoff in entomological applications and the biting rate curves sharply downward as this cutoff is approached. Variation in the relative attractiveness of different human individuals to mosquitoes is a recognized but little studied prohlem. Further, a diurnal biting cycle is reported for Aedes aegypti, biting peaks occuring in early morning or evening. Although they may bite at anytime, especially if they have gained entry into lighted inhabited buildings, biting has been reported to approach zero during the night with those mosquitoes that remain outside of buildings. A female mosquito may bite and probe many times on one or more individuals in order to gain a full blood meal; however, once she has accomplished this meal, she is lost to the primary target effect until she has deposited her eggs. This involves an average of 3 to 4 days, but after oviposition is completed, she is again ready to continue the primary target effects.

...

Project BELLWETHER I,  BW 459, is a preliminary step in evaluating some of these problems with the A. egypti mosquito, and it is an approach to the over-all problem of developing field assessment techniques; i.e. techniques whereby valid data for the development of Munitions Expenditure Tables can eventuallv be obtained. The results of these trials will also be used in designing subsequent tests under BELLWETHER-II where some of the above cited problems, as well as others, will be further investigatad and elucidated."



On the plus side for us, some mosquito behavior characteristics were noted:



"On hot, sunny days, it was observed that most of the vectors moved onto the shady side of the person to hover, land, or bite. On windy days, they moved to the lee side and on cooler days, they were observed to move up inside of the sleeves before biting.

Most of the men had observed, at one time or another, that released vectors would fly towards them and then pass by and fly to another person without stopping.

The volunteers that had functioned the test fixtures observed that the vectors were attracted by movement--the movement involved in releasing the vectors resulted in many mosquitoes moving to that man. Other personnel observed that the vectors appeared to be attracted by sound--the more garrulous men getting more bites.

Several of the men noticed that some of the vectors would become so engorged with blood that they apparently could not fly, but instead dropped to the ground and crawled away."

Reference:

Outdoor Mosquito Biting Activity Studies, Project BELLWETHER-I, BW-459.  Technical Report DPGR-259, US Army Chemical Corps Research and Development Command, December 1960.

Available at:

http://www.dtic.mil/dtic/tr/fulltext/u2/596046.pdf



Yet, that was not the end of the testing.  There was one more event to complete: BELLWETHER-II.  One year later, from 6 September to 20 October 1960, more experiments were conducted, with larger numbers of mosquitoes, released at variable distances to see how far they would travel to get a meal, adding movement to the test subjects' activities instead of just sitting, and seeing if structures would affect mosquito behavior.  The report mentions that most of the subjects were new to the experiments, but some had participated in the 1959 trials.  Researchers also tried to see how long the mosquitoes would last in the wild, with inconclusive results.  Since I have gone on long enough with this post already, I will just post the digest of the second report.



"The objectives of BELLWETHER-II, using releases of uninfected, starved, virgin female Aedes aegypti mosquitoes, were, in part:

1. To evaluate the effects of varying the host distance, the host concentration, and the vector/host ratio:

2. To determine the effect of the presence or absence of overt movement of the human samplers upon the outdoor biting rate; and

3. To investigate methods of placement of human samplers in open terrain and within built-up areas.

Inherent throughout the entire test was the development and improvement of a basic entomological field test technology.

A total of 14 field trials were conducted in the period extending from 6 September to 20 October 1960. Up to 100 assigned military personnel were used as samplers in each trial, and grid arrays involving dispersal distances much greater than those involved in BELLWETHER-I were incorporated in this test design.

From the data generated in this test, and under the specific conditions encountered, it is concluded that:

1. In a 30-minute sampling period, there was no significant difference in vector biting activity at distances up to 100 feet from the release point, but maximum biting activity occurred at distances of less than 200 feet.

2. Intervening hosts did not interfere with either the vectors' outward spread or biting activity.

3. No conclusive findings were generated as to the effect of host concentration.

4. When the number of vectors was increased by a factor of 10, approximately 10 times as many bites were received and the proportion of hosts bitten was increased an average of 36 per cent.

5. Vector biting activity showed a tendency to be highest when the hosts were alternately in motion and then motionless for recurring 5-minute periods, and to be lowest when the hosts moved continuously.

6. Hosts located near buildings were subjected to significantly greater vector biting activity than were hosts located in open areas.

7. Vectors did not tend to distribute themselves evenly throughout an isolated built-up area, and, further, they did not tend to redistribute themselves evenly during the interims between host occupations.

8. No conclusive findings were generated as to the optimum sampling duration.

9. No evidence of crepuscular-period biting preference was obtained in these trials.

10. No conclusive findings were generated concerning the average longevity of this species when exposed to ambient desert conditions."



Reference:

Entomological Field Test Technology, BELLWETHER-II, BIO 531. Technical Report DPGR 293.   US Army Chemical Corps Research and Development Command, December 1961.

Available at:

http://documents.theblackvault.com/documents/biological/EntomologicalFieldTest.pdf