Blue sticky traps are more efficient for the monitoring of Lygus rugulipennis ( Heteroptera , Miridae ) than yellow sticky traps

Jarmo K. Holopainen MTT Agrifood Research Finland, Plant Production Research, Plant Protection, FIN-31600 Jokioinen, Finland. Current address: Department of Ecology and Environmental Science, University of Kuopio, PO Box 1627, FIN-70211 Kuopio, Finland, e-mail: jarmo.holopainen@uku.fi Sakari Raiskio MTT Agrifood Research Finland, Plant Production Research, Plant Protection, FIN-31600 Jokioinen, Finland Anu Wulff Department of Ecology and Environmental Science, University of Kuopio, PO Box 1627, FIN-70211 Kuopio, Finland Kari Tiilikkala MTT Agrifood Research Finland, Plant Production Research, Plant Protection, FIN-31600 Jokioinen, Finland


Introduction
In host plant detection by insects visual stimuli play an important role, especially among gener-alist insects which have host plants with variable morphological characteristics.The physical stimulus perceived by the insect is affected by brightness (intensity of reflected light), hue (dominant wavelength of reflected light) and Holopainen, J.K. et al. Lygus spp.monitoring with blue sticky traps saturation (spectral purity of reflected light) (Prokopy and Owen 1983).Relative to other colours, the yellow colour is known to attract many herbivorous insects (e.g.Johnson andMueller 1988, Kostal andFinch 1996) and therefore yellow water pans and sticky traps are usually used for insect monitoring.In cabbage root flies the preference for yellow traps compared to blue and violet trap was innate even in young flies (Kostal and Finch 1996).
Several species in the genus Lygus have hundreds of host plant species and they are significant pests of many important crop plants (Holopainen andVaris 1991, Young 1986).The most important species in Europe is L. rugulipennis, which is a pest on crop plants as well as forest trees in nurseries (Holopainen 1986) like other Lygus species in Northern America (Schowalter and Stein 1987).The movement of Lygus bugs to the crops from their hibernation sites (Varis 1995) or from other cultivated plants (Fleisher et al. 1988) is easily monitored with sweep net sampling (Varis 1995).However, sweep netting is time-consuming, may harm emerging seedlings and does not catch samples properly from small seedlings, which are often most severely damaged by Lygus spp.(Fye 1984, Holopainen 1986).
For the development of efficient monitoring and forecasting systems of Lygus populations, the low catches of bugs with yellow sticky traps have been a problem.Especially, in organic farming, monitoring of L. rugulipennis is needed for timing different plant protection measures.The use of virgin Lygus female (Slymaker andTugwell 1984, Holopainen andRikala 1991) as a bait in yellow sticky traps has increased catching efficiency of yellow sticky traps significantly, but this method is too laborious for routine use.Therefore, we tested attractivity of blue sticky traps for monitoring purposes of L. rugulipennis and other Lygus spp.damaging especially small carrot seedlings.

Traps
The traps were disposable 190 mm by 290 mm cardboards with sticky upper surface, yellow or blue in colour.The blue sticky traps we used are especially aimed to attract thrips species on Saintpaulia and Frankliniella occidentalis, while the yellow colour is expected to attract all other flying insects.Also the sticky glue of the traps lures certain insects (Catch-it TM , Silva miljö AB, Knäred, Sweden, http://www.silvandersson.se).
Traps were formed into a vertically positioned cylinder, fixed in a wooden stick at the level of carrot foliage top, being about 10 cm to 30 cm from soil surface depending on the height of foliage.Blue and yellow traps were placed in a row in alternating order, 10 m from each other.Traps, replaced with new ones at one-week intervals, were collected, covered with transparent plastic film and stored in a cold room before insect identification and counting.

Reflectance of traps
Spectrum of the sun and reflectance of sunlight from the traps were measured with a Macam 9910 spectrophotometer (Macam Photometrics Ltd, Livingston, Scotland) on a sunny morning between 9.30 a.m. and 10.00 a.m. in late April.Trap surfaces were at right-angle to the sun during measurements.

Fields
Research was carried out in one carrot and one grassland field in 1997, six carrot, five swede, five cabbage, three potato, two spinach, and one red beet field in 1998.In 1999 traps were tested in two carrot fields and a cabbage field, and in 2000 in two and in 2001 one carrot field, respectively.The size of the fields varied between 0.5 and 5 ha.Traps were arranged in one row, ex-Vol. 10 (2001): 277-284.cept in two rows per field in 2000 and two rows per plot in 2001.The distance of the traps in row, parallel to the field edge, was 10 meters in each plot.The trap row was placed on the side of field, which was on the potential entering direction of the bugs.Usually traps were on the side that was closest to the forest.
In 1997 two blue and two yellow traps per field per week were used in carrot and one blue and one yellow trap on grassland for four weeks.In 1998 several crop plants were studied using three blue and four yellow traps per field for eight weeks.In 1999 three blue and three yellow traps per week were used in carrot for seven weeks.In 2000 6 blue and 6 yellow traps in two rows were used in both localities (Jokioinen 3.5 ha field and Kokemäki, 0.4 ha field) for nine weeks, and the positions of traps were reversed (yellow vs. blue) each week to reduce any bias due to positional effects.In 2001 a ten week investigation was conducted on a 3.6 ha carrot field in Jokioinen with four replicate plots each having two yellow and two blue traps in 10 m distances from each other.
Trap catches of Lygus bugs from blue and yellow traps from 1997 to 2000 were analysed using weekly catch of each trap as a true replicate.Non parametric tests (Mann-Whitney U) were used for comparisons, since data were not normally distributed because of small number of observations.Data from replicated plot experiment in 2001 were analysed with t-test using mean plot catches.SPSS for Windows statistical package was used for analyses.

Results
Reflectances from yellow and blue traps were 25% and 12%, respectively, of the irradiation from the sun (Fig. 1) in the integrated wavelengths 280-700 nm (UV-radiation and photosynthetically active visible light (PAR)).Blue traps reflected from blue to red light the peak being at blue (460 nm).Yellow traps had a great-er overall reflectance than blue traps reflecting mainly wavelengths from yellowish green (550 nm) to red (750 nm) the peak being at 665 nm.Reflectance in the UV-radiation area (< 400 nm) was very low in both trap types.
The dominant Lygus species in trap catches was L. rugulipennis (Popp.) in most crops.L. kalmi (L.), which is a specialist feeding on Apiaceae (Linnavuori 1966), was found in carrot fields in 1999 and 2000.In 1997, in the carrot field, weekly mean catch of L. rugulipennis in blue traps 3.75 ± 1.13 (SE) was significantly (n = 8, Mann-Whitney U = 1, P < 0.001) higher than in yellow traps 0.13 ± 0.12.The accumulated total catch in blue traps during the monitoring period was 29-times compared to yellow traps (Fig. 2a).In the grassland fields, in 1997, weekly mean catch of L. rugulipennis in blue traps 3.25 ± 2.00 was only marginally (n = 4, U = 2.5, P = 0.089) higher than in yellow traps, 0.25 ± 0.30 individuals.In 1998, the total catch of L. rugulipennis (and possibly the density also) in different field crops was so random and low that statistical significant differences were not found (Table 1).

Discussion
In the years when L. rugulipennis population densities are high enough to cause extensive damage to cultivated plants, yellow traps do not indicate bug densities efficiently (Holopainen and Rikala 1990) (Prokopy et al. 1979) and yellow water-pans traps to red and green traps (Landis and Fox 1972).The variation of colour detection among insect families is poorly known, but e.g.some Heteroptera species do not perceive orange colours (Hénaut et al. 1999).Blue sticky traps have been observed to be efficient in the monitoring of some thrips species in greenhouses (Gillespie and Vernon 1990).In natural ecosystems Thysanoptera is the only insect order that is more efficiently caught by blue than yellow traps (Ho-  .Holopainen, J.K. et al. Lygus spp. monitoring with blue sticky traps back et al. 1999).Hoback et al. (1999) demonstrated that attractiveness of blue and yellow traps could be variable inside one insect genus as in Frankliniella spp.thrips.Our reflectance analysis demonstrated that blue traps are more efficiently reflecting short blue light waves (400-500 nm) than yellow traps, which are more efficiently reflecting in the green area (500-560 nm).This might be one explanation for the higher catch of blue traps for L. rugulipennis, since this species is not an absolute herbivore.The omnivorous diet could include insect eggs and larvae (Varis 1972), and possibly green colour of the plant foliage is not as attractive to this species as to other species in the genus Lygus.It is interesting to note, that Frankliniella occidentalis is also an omnivore (Agrawal et al. 1999) and is also trapped in greater numbers on blue traps than on yellow (Gillespie and Vernon 1990).
Colour preference of Lygus bugs might be different also in different crop types.Prokopy et al. (1979) found that for L. lineolaris that damage flowers in apple trees, bright blue sticky traps were not as attractive as non-ultraviolet-light reflecting white, yellow or plexiglas rectangles, hung vertically at about 0.7 m above ground.We did not find yellow traps to catch L. rugulipennis more efficiently than blue traps in any of the studied fields.
Blue traps are probably most effective for Lygus monitoring on crops where they damage plants on seedling stage, and the foliage of young plants do not shade the soil surface e.g.carrot, conifer seedlings, transplanted cabbage seed-lings.Kostal and Finch (1996) found that the cabbage root fly Delia radicum (L.), avoided landing on blue traps when background was green, while brown background resulted in the relative increase of catches in the non-preferred violet and blue traps.Prokopy and Owens (1978) and Prokopy et al. (1979) found that white sticky traps were more efficient than yellow sticky traps to collect L. lineolaris.Therefore efficiency of white sticky traps to collect L. rugulipennis should be tested and compared to blue sticky traps.However, at the moment blue traps are significantly more efficient than yellow traps for L. rugulipennis monitoring.
Since L. rugulipennis frequently damage the apical meristem in the production of vegetable seedlings during spring in greenhouses, in addition to the thrips monitoring, the use of blue traps for timing of chemical control of Lygus bugs is recommendable.The blue sticky traps seem also to be suitable monitoring tool for the forecast system of L. rugulipennis that is under development in MTT Agrifood Research Finland.Vegetable growers make observations and countings of bug densities in their fields and send the information by the short message service of their mobile phones to the central database of MTT.As a feedback the growers receive on the display screen of they mobile phone the latest forecast of bug densities in their area.

Fig. 1 .
Fig. 1.Radiation (W/m 2 ) spectrum of the sun and reflectances of yellow and blue sticky traps in UV area (250-400 nm) and visible light area (400-750 nm).

Table 1 .
Weekly numbers of Lygus rugulipennis per trap in yellow and blue traps (n is same for both colors) in different field crops (21 fields) in 1998.