A research programme in New Brunswick coordinating the use of radar, specially-instrumented aircraft, and observations from platforms extending above the forest canopy has established the massive scale and regularity of evening take-off flights by spruce budworm moths, Choristoneura fumiferana, and of subsequent nocturnal dispersal at levels high above the ground. A night-viewing telescope showed moths taking off from the tree crowns up until 2330 h, 2.5 h after sunset, while radar dot echoes of uniform strength and size began to appear each night at about the same time. Direct identification of radar echoes was provided by simultaneous catches of budworm moths taken in insect-collecting nets on Cessna 185 aircraft. Moths collected during emigration, displacement, and immigration contained a high proportion of egg-carrying females.Radar recorded the rates of climb and the altitudes reached by the flying moths and provided extensive data on their numbers, density, orientation, direction, speed, and duration of displacement. Airborne moths became concentrated in zones of wind convergence, and line-echoes from moths at wind-shift fronts were detected on radar at distances of 30 km. The time of passage of wind-shift fronts over a surface site was detectable by pilot-balloon observations and by sensing equipment on a 24-m meteorological tower. Wind-fields over New Brunswick, as found from a DC-3, showed a marked contrast between the high degree of wind uniformity on many evenings and convergent winds on other evenings. Attention is focussed on the potential effects of meso-scale sea breeze fronts, of which one example is presented, and of storm cells on moth concentration and dispersal.Through the use of radar, spruce-budworm moth dispersal has been viewed for the first time in its entirety and the integrated research programme has provided new data for better evaluation of the significance of moth dispersal in the initiation and spread of infestations.

Colour can potentially provide useful information for a variety of computer vision tasks such as image segmentation, image retrieval, object recognition and tracking. However, for it to be helpful in practice, colour must relate directly to the intrinsic properties of the imaged objects and be independent of imaging conditions such as scene illumination and the imaging device. To this end many invariant colour representations have been proposed in the literature. Unfortunately, recent work (Second Workshop on Content-based Multimedia Indexing) has shown that none of them provides good enough practical performance.In this paper we propose a new colour invariant image representation based on an existing grey-scale image enhancement technique: histogram equalisation. We show that provided the rank ordering of sensor responses are preserved across a change in imaging conditions (lighting or device) a histogram equalisation of each channel of a colour image renders it invariant to these conditions. We set out theoretical conditions under which rank ordering of sensor responses is preserved and we present empirical evidence which demonstrates that rank ordering is maintained in practice for a wide range of illuminants and imaging devices. Finally, we apply the method to an image indexing application and show that the method out performs all previous invariant representations, giving close to perfect illumination invariance and very good performance across a change in device.

A non-technical description is given of a new, powerful, low cost field system (Infra-red Active Determination of Insect Flight Trajectories or IRADIT) for detailed and automatic remote sensing studies of natural insect flight behaviour. The special requirements and difficulties of the detection problem are defined. A series of examples of field devices and techniques are presented to illustrate the key factors of the optical sensing and tracking of insects in flight. In the finally adopted IRADIT system, flying insects are differentially illuminated, under all natural light conditions, by an intense beam of pulsed near-infra-red radiation and detected using a shuttered image intensifier linked with a video camera operating at a rate of 50 frames per second. Immediate fully-automatic determination of the flight trajectories of several simultaneously detected insects was achieved, at this same high rate and in the presence of sky background photon noise, by processing the video signals with electronic circuits and a microcomputer. Flight trajectories are influenced by the local wind, whose vector must be subtracted to study insect flight behaviour. This was achieved by the use of a specially developed sensitive three-vector vane anemometer, providing digital data to the microcomputer at a minimum rate of 5 Hz. In tests of the prototype IRADIT-anemometer system in the field, insects with wing area of only 1-5 mm 2 and flying against the midday sky were readily tracked at ranges up to 15 m. A range of at least 100 m is expected for nocturnal moth tracking.

A new airborne radar technique has been developed and used during two seasons on three aircraft in Canada, to investigate the source areas, flight behaviour and areas of deposition of migrating insects. The radar measures, records and analyses more than 100 high resolution profiles of insect orientation and absolute density per second, at a spacing of less than one metre of aircraft track. During some 11000 km flown with this equipment in New Brunswick and neighbouring areas, synoptic and small-scale meteorological systems have been traversed and their effects on the flying insects (spruce budworm moths) have been measured.

Radar methods have been extended to measure the aerial density of small insects. Results obtained during an outbreak of the cereal aphid Metopolophium dirhodum (Walker) in south-eastern England were compared with simultaneous suction trap catches to study the sensitivity of trap effectiveness to windspeed. Two traps were studied: the Rothamsted Insect Survey trap (12-2-m) and a standard aerofoil trap. The Survey trap effectiveness is moderately sensitive to windspeed, decreasing exponentially by a factor of two for each 2-4 m/s (5 knots) of average windspeed. The two trap sensitivities did not differ significantly, but both results are very significantly different (P<0-001) from the published predictions, which were based upon a comparison of catches from suction traps and a combination of a rotary (whirligig) net and a tow net. These differences are discussed. The average catching rate is about 40% of that of an ideal trap. Seven-day catches could vary by a factor of 0-5-2-0 from average due to prolonged periods of extra strong or light winds. Systematic windspeed gradients can corrupt suction trap studies of insect dispersal in relation to vertical density profiles, diurnal flight patterns and geographical distribution. Absolute calibration of the aerofoil trap was achieved by using the remote-sensing IRADIT infra-red system to measure the aerial density of aphid-size insects near to the trap inlet in very light winds; the effectiveness was not statistically different from unity, and the Survey trap is expected to perform similarly. IntroductionThe Rothamsted Insect Survey (12-2-m) suction trap (Taylor & Palmer, 1972) has become a standard tool for monitoring the density of flying insects, particularly aphids. A network of traps, at present at 23 sites, has been managed by the Survey for 15 years in the United Kingdom, and in recent years similar traps have been operated throughout Europe (Taylor et ai, 1981). The Survey trap takes in air from a height of 12-2 m (40 ft) at the rate of approximately 0-8 m 3 /s, through a circular inlet pipe with a diameter of 0-254 m (10 in.) at a speed of about 16 m/s. Trapped insects are removed daily, the aphids identified and weekly bulletins distributed to the agricultural industry giving warnings of impending outbreaks (Woiwod et ai, 1984).The 'efficiency' of suction traps for quantifying the aerial density of passing insects has been assessed by Taylor (1962) by an indirect method. The insect catching rates of a variety of traps were measured relative to that of an aerofoil trap with an inlet diameter of