Patterns of asexual reproduction of the soybean aphid, Aphis glycines (Matsumura), with and without the secondary symbionts Wolbachia and Arsenophonus, on susceptible and resistant soybean genotypes

Abstract

Plant breeding is used to develop crops with host resistance to aphids, however, virulent biotypes often develop that overcome host resistance genes. We tested whether the symbionts, Arsenophonus (A) and Wolbachia (W), affect virulence and fecundity in soybean aphid biotypes Bt1 and Bt3 cultured on whole plants and detached leaves of three resistant, Rag1, Rag2 and Rag1 + 2, and one susceptible, W82, soybean genotypes. Whole plants and individual aphid experiments of A. glycines with and without Arsenophonus and Wolbachia did not show differences in overall fecundity. Differences were observed in peak fecundity, first day of deposition, and day of maximum nymph deposition of individual aphids on detached leaves. Bt3 had higher fecundity than Bt1 on detached leaves of all plant genotypes regardless of bacterial profile. Symbionts did not affect peak fecundity of Bt1 but increased it in Bt3 (A+W+) and all Bt3 strains began to deposit nymphs earlier than the Bt1 (A+W-). Arsenophonus in Bt1 delayed the first day of nymph deposition in comparison to aposymbiotic Bt1 except when reared on Rag1 + 2. For the Bt1 and Bt3 strains, symbionts did not result in a significant difference in the day they deposited the maximum number of nymphs nor was there a difference in survival or variability in number of nymphs deposited. Variability of number of aphids deposited was higher in aphids feeding on resistant plant genotypes. The impact of Arsenophonus on soybean aphid patterns of fecundity was dependent on the aphid biotype and plant genotype. Wolbachia alone had no detectable impact but may have contributed to the increased fecundity of Bt3 (A+W+). An individual based model, using data from the detached leaves experiment and with intraspecific competition removed, found patterns similar to those observed in the greenhouse and growth chamber experiments including a significant interaction between soybean genotype and aphid strain. Combining individual data with the individual based model of population growth isolated the impact of fecundity and host resistance from intraspecific competition and host health. Changes to patterns of fecundity, influenced by the composition and concentration of symbionts, may contribute to competitive interactions among aphid genotypes and influence selection on virulent aphid populations.

Keywords: Arsenophonus; Hamiltonella; Wolbachia; reproduction; resistant soybean varieties; soybean aphid; symbionts.

Figure 1

Both aphid strain and soybean genotype affected aphid population size in whole plant experiments. Average number of aphids per plant for 5 clonal strains of aphids [Bt1(A+W−), Bt1(AW−), Bt3(A+W+), Bt3(A−W+), Bt3(A−W−)], grown on caged whole plants in a greenhouse (A) and an environmental chamber (B) on 4 genotypes of soybean (W82, Rag1, Rag2, Rag1 + 2). Counts were made at the end of the experiment. Data were analyzed with a general model using a negative binomial distribution and a log link function. There was a significant interaction between aphid strain and soybean genotype for the greenhouse (df = 12, X² = 22.53, p = 0.0320) and environmental chamber experiments (df = 12, X² = 39.49, p < 0.0001). Means with the same letter are not significantly different using Tukey–Kramer post hoc comparisons (p < 0.05). The study included three replicates of each aphid strain/soybean variety combination for both experiments.

Figure 2

Pattern of cumulative and daily aphid nymph deposition by individual aphids on detached leaves. Average number (A) and cumulative number (B) of aphids deposited per day by strain for individual aphids grown on detached leaves of four soybean genotypes (W82, Rag1, Rag2, and Rag1 + 2). Error bars are standard errors of the average number of aphid nymph deposited. Characterizations of the fecundity curves are described in the text.

Figure 3

Aphid strain and soybean genotype affected number of aphid nymphs deposited on detached leaves. Average number of nymphs deposited per aphid over a 14-day period for 5 clonal strains of aphids [Bt1(A + W−), Bt1(A−W−), Bt3(A + W+), Bt3(A−W−), Bt3(A−W−)] cultured on detached leaves of 4 soybean genotypes (W82, Rag1, Rag2, Rag1 + 2). Data presented by soybean genotype (A) and aphid strain (B). Data were analyzed with a general model using a negative binomial distribution and a log link function. Soybean variety was a significant factor in the number of nymphs deposited per aphid (X² = 48.33, df = 3, p < 0.0001) as was aphid strain (X² = 15.21, df = 4, p = 0.0043). Means with the same letter are not significantly different using Tukey–Kramer post hoc comparisons (p < 0.05). The study included 12 replicates of each aphid strain/soybean genotype combination except for Bt3 (A−W−) on W82 soybean which had 13 replicates.

Figure 4

Aphid strain and soybean genotype affected maximum number of aphid nymphs deposited on detached leaves. Average number of the maximum number of nymphs deposited by an individual aphid in a single day: by genotype (A) and by aphid strain (B). There were significant differences among genotypes (df = 3, X² = 92.83, p < 0.001) and aphid strains (df = 4, X² = 46.79, p < 0.0001). Using Bonferroni adjusted tests of specified contrasts of genotypes and strains (α = 0.0063 see methods for more information), aphids had a higher maximum number of nymphs deposited in a day on the susceptible genotype (W82) than the other genotypes. Bt1(A+W−) was significantly lower than Bt3(A+W+) but not from Bt1(A−W−). Bt3(A+W+) was significantly higher than Bt3(A−W+) and Bt3(A−W−).

Figure 5

Aphid strain affected the day maximum number of aphid nymphs were deposited on detached leaves. Average day that the maximum number of nymphs were deposited by five aphid strains on four soybean genotypes (df = 4, X² = 14.13, p = 0.0069). Error bars are standard errors of the means. Means with the same letter are not significantly different using Tukey–Kramer post hoc comparisons (p < 0.05). Aphids that died before depositing nymphs were not included in the analysis. See Table 1 for sample size.

Figure 6

Resistant soybean genotypes decreased aphid survival on detached leaves. Percent aphid survival at day 14, cultured on leaves of four varieties of soybean (W82, Rag1, Rag2, Rag1 + 2). The percentages shown were calculated by combining all five strains of aphid used in the study [Bt1(A+W−), Bt1(A−W−), Bt3(A+W+), Bt3(A−W+), Bt3(A−W−)]. Asterisks indicate a significant difference (p < 0.05) from the susceptible W82 soybean genotype. Aphids survived significantly better (p < 0.05) on the W82 variety of soybean than on genotypes with Rag1 or Rag1 + 2 (X² = 15.3293, df = 3, p = 0.0016). There was not a significant difference between W82 and Rag2.

Figure 7

Resistant soybean genotypes increased variability of total fecundity per aphid on detached leaves. Standard deviation of the total number of nymphs deposited by aphids that survived to the end of the experiment (18 days). The standard deviation was calculated across all five strains for each plant genotype. There is a significant difference in the variation of nymphs deposited using model II ANOVA testing for significant difference of variation (df = 3, p < 0.0001). Asterisks indicate a significant difference (p < 0.05) from the W82 soybean genotype. W82 genotype was significantly less variable that Rag1, Rag2 or Rag1 + 2.

Figure 8

Simulations using individual leaf data found significant interactions similar to those in whole plant experiments. Average number of aphids per plant from simulations of whole plant experiment using individual leaf data. The study simulated five clonal strains of aphids [Bt1(A+W−), Bt1(A−W−), Bt3(A+W+), Bt3(A−W+), Bt3(A−W−)], grown on four genotypes of soybean (W82, Rag1, Rag2, Rag1 + 2) in each of 1,000 simulated experiments. The average represents the average of the average number of aphids per plant and error bars represent the average standard error. Using the same statistical method as used in the whole plant experiment, there was a significant interaction effect in more than 99% of the model runs.

Figure 9

Proportion of model runs with significant differences (p < 0.05) among Bt1 (top) and Bt3 (bottom) biotype strains for all soybean varieties. Differences outside the diagonal, comparing results on the same soybean variety, are grayed out. Comparisons with a larger proportion of significant differences are purple/pink, those with a lower proportion are pale blue. Differences of 100% only occurred outside the diagonal and were left white.