Infestations and distribution of Rhopalosiphum padi (L.) on different varieties of barley and oats, effect of nitrogen fertilization and chemical control.

Infestations of Rhopalosiphumpadiwere compared between common varieties grown at nitrogen levels of 50, 100 and 150 kg/ha in 1986 and 1988, and controlled with dimethoate in 1988. Colonization of the aphid was slightly quicker on oats than on barley. The highest peak inside field cages was 305 aphids and in the open field 74 aphids per plant. On barley, the number of aphids was highest on var. Pomo and var. Pokko, six-row type varieties with the longest growth period. On oats the number of aphids remained lower on var. Nasta, an early maturing variety with strong straw, than on other cultivars. Excessive nitrogen either increased or decreased the peak number of aphids per plant, depending on the variety and the year. Over 40 % ofaphids on oats fed on the lower base of the plant at or under the soil surface. On barley, aphids lived slightlyhigher, 30 % in the stem and 30 % on leaves. Dimethoate spray controlled aphids on the top of the plant but not those on the base and lower leaves. Its efficacy against aphids on the upper stem and lower leaves on barley was decreased when an increased amount of nitrogen was used. The percentages of parasitized and diseasedaphids decreased after the use of dimethoate. Index words: barley, oat, Rhopalosiphum padi, bird cherry-oat aphid, varieties, nitrogen, distribution, chemical control, feeding site, dimethoate, parasitized, diseased


Introduction
Differences in Rhopalosiphum padi (L.) infestation between species of spring cereals and different varieties have been reported in Finland by Markkula and Roukka (1972) R. padi much less than other spring cereals, and aphids on wheat seldom need to be controlled in Finland. On the basis of experiments in cages, Rautapää (1968) has described changes in yield components on cereals infested with R. padi.
Nitrogen fertilization is known to promote vegetative growth and the ability of spring cereals to tiller, thereby creating new sites for R. padi to feed on. By extending the period of vegetative development, excessive nitrogen extends the time for R. padi to multiply. It has been claimed that excessive nitrogen increases the need for aphid control.
The behaviour of aphids, especially their choice of feeding site, the stage of plant growth and biochemical interaction between aphids and plants have been shown to be important to the growth of aphid populations and to yield losses (W ratten 1978, Leather 1982, Leszczynski 1985. Wiktelius (1987) has presented information about the feeding site of the bird cherry-oat aphid on spring barley. R. padi was found generally to favour the area of the plant that is under the soil surface. Differences in the abundance and behaviour of R. padi on different varieties of spring barley and oats grown at different levels of nitrogen fertilization were studied in order to evaluate the effect of variety and nitrogen supply to the rate of increase of the aphid population and to the need for and the efficacy of chemical control. The experiment was performed in a year with a low expected R. padi population, 1986, and in a year with an exceptionally high expected aphid population, 1988, in Finland.

Material and methods
The field experiments, carried out at the Agricultural Research Centre in Jokioinen, included ten varieties of barley and oats (Fig. 2) (subplots) and three levels of nitrogen fertilization (50, 100 and 150 kg nitrogen/ha) (plots). The soil type was clay. Primary nitrogen fertilization was given as a Normal-Y fertilizer containing 10.1°7o of ammonium-N and 5.9 % of nitrate-N. The excessive nitrogen to the levels of 100 kg and 150 kg nitrogen/ha was given as a salpetre fertilizer, called Oulun salpietari, containing 13.8 % ammonium-N and 13.7 % nitrate-N. The number of replications was three, devided into randomized plots and randomized subplots, 12.5 m 2 in size. Sowing dates were May the 19 ,h in 1986 and May the 16 lh in 1988. Seedlings emerged from ten to fifteen days after sowing. No aphid control was needed in 1986. In 1988, the crop was sprayed on the 7th June, a week from the first peak of migration and the time when the first migrants were observed in the crop and 3 days from the second, final peak of migration. Dimethoate, 0.5 1/ha in 200 litres of water, was sprayed with a normal farm sprayer. At spraying, most varieties were at the 2-4 leaf stage. Half of each plot was sprayed, thus forming subsubplots of 6.25 m 2.
In 1986, at the peak population, aphids were counted from 25 plants taken randomly from each subplot. In 1988 the population was observed more precisely. Migrants and exules of R. padi were first counted separately five days after the start of the spring migration, from a randomly chosen section of 0.5 m (about 25 plants) of a row, in the centre of each subsubplot. The second time, aphids were counted, one week after from the first count, from five 0.5 m long sections of rows (one of sections the same as first time). At the peak of the population (three weeks after first count), samples from one random row were taken to the laboratory and deepfrozen quickly, for observation later, in the autumn. In addition, during the first count, three plants were chosen from each unsprayed subsubplot, two with I-s exules and a third with s -l o 5-10 exules per plant. These plants were covered with an aerated PVC cage to prevent the aphids from escaping and to keep their enemies out. The number of aphids from these plants were counted again two weeks later, and three weeks later (the peak of population) the plants were taken from the soil to the laboratory. During the last count made from samples of plant rows or single plants, apterae and nymphs were counted separately from the base and upper stem from each leaf. The white area of the stem above the roots was considered to be the base of the plant. The height of the plant as well as the length of the flag leaf of the main shoot and of tillers were also recorded. During the sampling of the plants for the last count, number of coccinellids (larvae and adults) and syrphids (larvae) were counted. The number of parasitized and diseased aphids was recorded in the laboratory from samples.
Both experimental years were relatively warm. Daily precipitation and mean temperature are presented in Fig. 1.

Colonization of plants by R. padi
In 1988, the first migrants were observed in the plants on the I st 1 st of May, three days from the first peak of the spring migration. (Spring migration peaked second time on the st h5 th of May.) The total mean number of spring migrants that infested oats (2.25 aphids/plant) was double the number of migrants on barley (1.0 aphids/plant) (P<0.05). However, there was no difference between the incidence of infested plants of barley and oats. No significant differences were found between the number of spring migrants on the different barley and oats varieties, either. Incidence of migrants on all barley and oat varieties was significantly lower at the level of 50 kg nitrogen/ha than at the higher levels of nitrogen, the respective means being 17 + 6 and 26 ± 5 % of plants, P<0.05.
The number of first exules after the first days of spring migration was highest with the medium amount of nitrogen fertilization on oats (mean at the level of 100 kg/ha 2.5 aphids/plant, total mean 1.76 aphids/plant) (PcO.Ol). No other differences were found.  (Fig. 2). There was a significant difference (P< 0.01) in number of aphids between two groups of varieties. The group of varieties having the highest number of aphids included: Pomo (1), Pokko (3), J01439 (5), Eero (7), Puhti (11), Virma (Hja 75430) (14) and Titus (Tiitus) (16). The second group of varieties, which had the lowest number of aphids, included: Agneta (4), Arra (6), HlOl4 (9), Veli (12), Pol (13), Stil (Karhu) (19) and Nasta (20). No difference was found between the pooled results for barley and oats or between the pooled results for different levels of nitrogen fertilization. The peak number of aphids was, on average, 2.5 times higher in cages than in the surrounding open field.
In the open field in 1988, significant differences were found in aphid numbers per plant between varieties Pomo, Agneta and Ida at the lowest level of nitrogen fertilization (Fig. 3).
But at higher amounts of nitrogen fertilization, the differences were not as clear. (Porno and Agneta are six-rowed and Ida two-rowed varieties). On var. Ida only, the number of aphids increased continually with increasing amounts of nitrogen (Fig. 3). With the same variety, the number of (subsidiary) tillers more than doubled, from 0.5 to 1.2 tillers/plant, as the nitrogen level was increased from 50 kg/ha to 150 kg/ha (P<0.05). On var. Porno, the number of tillers first increased from a mean of 1.3 to 1.8, but it then decreased to 0.9 as the amount of nitrogen was increased from 50 kg/ha to 100 kg/ha and 150 kg/ha. On var.
Agneta, however, the number of tillers remained unaffected. When the aphid numbers per shoot (main shoot + tillers) were compared only the number of aphids on cv. Porno at the lowest nitrogen level (32.5) was significantly higher than the numbers of aphids on other varieties (mean 20.75 respectively) (PCO.01). On oats, the number of aphids on var. Puhti tended to be higher than on var. Nasta (Fig. 3). The number of aphids per plant remained almost unaffected as the amount of nitrogen was increased, but the number of aphids per shoot (main shoot + tillers) de-creased slightly. With var. Puhti, the decrease was from 34.3 to 20.2 aphids/shoot between 50 and 100 kg nitrogen/ha, and with var. Titus (Tiitus) from 27.4 to 15.8 aphids/shoot between 50 and 150 kg nitrogen/ha (P<0.05). With Nasta, the number of aphids per shoot (mean 22.4) remained unaffected.
A similar tendency in the results for aphids on various varieties and with different levels of nitrogen was found in 1986, the year with a lower aphid population. On Pomo, the number of aphids (9.3/plant) was significantly higher than on other varieties at the lowest level of nitrogen (total mean 4.7/plant) (P<0.05). At the level of 100kg/ha varieties Porno, Pokko and Eero had significantly higher numbers of aphids (10.6, 9.2 and 8.7 aphids/plant) than other varieties: Kalle, Agneta and Arra (mean 3.3 aphids/plant) (PC0.05). The number of aphids on var. Ida was 7.6, on average. Already in 1986, the absolute maximum (14.2 aphids/plant), significantly higher than that of any other variety (P<0.05), was found on ti, Titus and Nasta) fertilized with different levels of nitrogen in the field. Significant differences (PC0.05) between varieties at each level of nitrogen indicated by capitals, and significant differences (PC0.05) between levels of nitrogen within each variety indicated by small letters.
Pokko at the highest level of nitrogen.
On oats in 1986, the mean number of aphids on var. Nasta at the two highest level of nitrogen (2.9 aphids/plant) and on var. Pol at the medium level of nitrogen (2.2 aphids/ plant) was significantly lower than on other varieties (mean 6.5 aphids/plant at nitrogen level of 100 kg/ha and 5.9 aphids/plant at nitrogen level of 150 kg/ha) (P<0.05).
Distribution of R. padi on barley and oat plants On oats, 44 % of aphids fed on the base of plant and 29 % fed on the upper stem. On barley, 19 % fed on the base, 32 % on the stem and up to 34 % on the lowest leaves (Table 1) (Fig. 4). There were only slight differences in the distribution between levels of nitrogen (Table 1) and varieties. The relative number of aphids on the strong and short straw of Ida (27-41 %) was significantly higher than that of Porno (22 -32 %) (P<0.05) Leaves of var. Porno were broader and attracted relatively higher number of aphids.
In the comparison between the main shoot and tillers, significantly more aphids (P<0.05) were found on the upper stem of the main shoot than on the stem of tillers, and in tillers the number of aphids on leaves was significantly higher. In dimethoate sprayed areas, the mean number of parasitized aphids on barley was 0.06 (0.4 %), and the mean number diseased aphids was 0.76 (4.6 %). The respective numbers on oats were 0.06 (0.3°/o) and 0.7/plant (2.8 %).

Efficacy of dimethoate application
The distribution of R. padi on plants greatly affected the efficacy of dimethoate. The ef--39  ficacy was excellent against aphids at the top of the plant, but nonexistent against the aphids feeding on the base (Table 2). This result was clearly reflected in the total efficacy of dimethoate spray, which on barley, at the lowest level of nitrogen fertilization, was reasonable but on oats was unacceptable. Increased nitrogen affected the efficacy of dimethoate on barley (Table 2), which was due to the decreased efficacy against aphids on the upper stem and lower leaves. On sprayed plants, aphids concentrated more on the base and the lowest leaves (Table 3). On one sprayed variety, Porno, up to 71 % of aphids remained on the base of plant.

Correlations between R. padi population growth and plant growth
The number of tillers of oats was decreased by the aphids feeding the base of plant, regression equation: number of tillers = 2.16 0.017 ± 0.007 x number of aphids in the base, R 2 = 0.26, Pc0.05. The significant differences (PcO.Ol) between varieties and positive effect of excessive nitrogen in the number of tillers became clear in sprayed subsubplots, only. This was mainly due to an unusual result obtained with var. Puhti, where spraying caused a decrease in the mean number of tillers/plant; at the two lowest levels of nitrogen the decrease was from 1.0 and 0.88 to 0.30 and 0.41 (secondary) tillers/plant (PcO.Ol). On barley aphids had no effect on number of tillers. Spraying did not affect significantly the number of tillers of barley, either.
In the cages, a slight positive correlation was found between adults and nymphs on the base and upper parts of tillers and the height of the main shoot (r = 0.17, P<0.01) and the length of the flag leaf in centimetres (r = 0.175, PcO.Ol). In the field a positive correlation was found between the number of aphids on the base of tillers, or the num-ber of aphids on the stem, and the height of the tiller (aphids on base r = 0.06, aphids on stem r = 0.105, PcO.001). Similarly, a slight positive correlation was found between the number of aphids on the base or the stem of oats (r = 0.09, PcO.001) or aphids in the whole plant of barley (r = 0.16, PcO.Ol) and the length of flag leaf. A negative correlation was found between the number of aphids on the second to the fourth leaf and the height of the tiller of barley and oats (barley r= -0.096, Pc0.05, oats r = -0.176, PcO.001) and the length of flag leaf of oats (r = -0.066, P C 0.05). The tiller height of both cereal species, mean 33.9 cm for barley and 31.4 cm for oats, was not affected by excessive nitrogen. The length of the flag leaf of barley, mean 13.9cm, was increased (from 13 to 15 cm) by excessive nitrogen, but the mean length of the flag leaf of oats, 16.8 cm, was not affected.
Spraying had no significant effect on these parameters.
In the unsprayed subsubplots, there was a negative correlation between the yield of bar- Pomo, Agneta and Ida was 4445 kg/ha (unsprayed), and increased significantly with the increase of nitrogen fertilization from 50 kg/ ha to 100 kg/ha (PcO.Ol). The mean yield of oat varieties was 3320 kg/ha (unsprayed), and it also increased with an increased amount of nitrogen (P<0.05).
On barley, a variable yield increase 9 % on Pomo and 4°/o on Agneta and 3 % on Ida was obtained by a dimethoate spray at the lowest level of nitrogen, but at highest level of nitrogen, dimethoate caused a greater yield loss, with Porno 9.4 % and Ida 6.6 %.
On oats, the mean yield increase due to dimethoate spraying was 11.6 %, but the effect of dimethoate was highly variable (from slight negative to a positive effect of 50 %).

Discussion
A higher incidence of migrants at the higher levels of nitrogen referred to the effect of an intense green colour of an excessively fertilized crop. But, the observation that a higher number of spring migrants landed on oats than on barley was a surprise, on the basis of information about the poor ability of R. padi to select host plant presented by Ahman et ai. (1985). However, this result agrees with the findings reported by Weibull (1987). The higher number of exule nymphs at the medium level of nitrogen was partly resulted from incidence of migrants, but might also indicate some minor changes in the biochemical balance of free and structural amino acids in plants (Niraz et al. 1985).
The unstricted population growth of R. padi in cages was comparable to the results presented by Rautapää (1968). Great difference between population growth in cages and the open field may partially result from caging out facultative predators such as carabid beetles, and of the specific enemies coccinellids, the number of which increased explosively until the end of June in the open field. The numbers of syrphids and hymenopterid parasites, on the contrary, were very low. Another possible reason for the excessive population growth could be a great improvement in conditions inside cages, such as increased humidity.
The differences in the number of R. padi between barley varieties were in agreement in the two experimental years, as were those for cages and the open field. The differences were obvious in a year of lower aphid population, but during higher pressure resulting from a larger population, the differences were not so clear. During an aphid outbreak, there is no way chemical control can be avoided by the selection of a specific commercial variety. On oats, which is favoured more by R. padi, the differences between cultivars seem to be minor in the field, even at a low level of aphid infestation.
The decrease in the number of aphids on oats at the higher levels of nitrogen is in agreement with the results of Weibull (1987). He reported that extra nitrogen fertilization resulted in a decrease in the total content of free amino acids in oats at the beginning of stem elongation. Weibull increased the nitrogen level from 50 to 100 kg/ha and had one variety of barley and oats. Differences in the reactions of different varieties are thus possible. Another way nitrogen could have an effect on the aphid population would be by causing changes in the number of tillers. Barley seemed to be sensitive to this. In our experiment, unfortunately, it was not evaluated how much nitrogen was actually used by plants in the prevailing dry conditions. The feeding area of R. padi was shown to be an important factor in the efficacy of pest control application; this finding partly explains many of the complaints about the efficacy of dimethoate presented in 1988. The phytotoxic effect of dimethoate on barley became obvious, as the efficacy at highest level of nitrogen was minimal. A slightly earlier application, better directed at migrating aphids, might have given better efficacy. The decreased tillering of var. Puhti in the area where aphids became controlled on the upper parts of plant indicates a change in apical dominance. Feeding on the upper area of the main shoot possibly resulted in a decrease of apical dominance there, and tillering was promoted. When controled, feeding stopped right at the beginning, apical dominance remained effective. This issue has been discussed by Harris (1974).
The percentage of parasitized aphids was low, and the percentage of diseased aphids fairly high compared to counts made in Sweden in 1980-83 (Wiktelius and Ekbom 1985). The decrease in parasitized aphids due to dimethoate spraying was expected, but the decrease in the percentage of diseased aphids was more surprising. However, the effect of insecticides in eliminating sensitivity to disease in the rape blossom beetle, Meligethes aeneus F., has been discussed by Hokkanen et ai. (1988). The highest proportion of diseased aphids was found in oats on a variety that generally supported the lowest populations of R. padi. This might be an indication of weakened condition and induced sensitivity of aphids on that variety. Differences in disease abundance may also indicate differences in humidity within the crop, crop structure affecting the penetration of fungus spores or minor differences in facultative microflora on the plant surface.
Vigorous plants had the highest number of R. padi, especially in cages, as also reported by Honek (1985). However, there was a slight negative correlation between the height of plants and the length of the flag leaf of oats and the number of aphids on leaves in the middle part of plant; similar results were reported for winter wheat by Havlickova (1984). The correlations referred to the behaviour of aphids in the plant when the population peak is approaching, but for practice the results had just a value of curiosity. Aphids on the base of plants seemed to be important in connection to yield loss, the mechanism most possibly being a decrease in the number of grains. As the plant becomes older and the base becomes dry, R. padi typically moves to higher parts of plant. There R. padi could be expected to affect grain quality, grain weight and protein content, as reported with Metopolophium dirhodum Wlk. (W ratten 1978). These changes were demonstrated by Rautapää (1968) in caged plants, but at a very high population level. After moving to upper parts of plants, R. padi in the field very quickly started the summer migration. Injuries on plants in this stage last a few days only, and are certainly of minor importance.