Experimental zone

Two experimental locations were established in the state of Mexico (Figure 1). Temamatla (19° 12’ N and 98° 52’ W), altitude of 2260 m above sea level, humidity of 45 - 70% and temperature of 17- 20 °C and the second location, Ayapango (19° 8’ N and 98° 48’ W), altitude of 2440 m above sea level, humidity of 50 - 90% and temperature of 15 - 18 °C. (^{SMN, 2012}; ^{INEGI, 2013}).

Figure 1: Geographic location of the experimental maize fields in the State of Mexico, during the 2010 spring-summer cycle. 

Genetic material

Three maize cultivars were used: the tri-linear HS2 hybrid (cross: CL12 × CL11 × CL7) from the Colegio de Postgraduados in Montecillo, Mexico; a simple-cross Triunfo hybrid (cross between lines L-10 and L-52) (^{Bravo-Quirino and Muñoz-González, 2003}) from the National Institute of Research in Forestry, Agriculture and Fishing (INIFAP); and the native Cacahuacintle from CHVM. The two improved cultivars and native maize were developed at altitudes of 2200 to 2750 meters above sea level.

Experimental units

A split-plot design with randomized complete blocks was used (Figure 2). The largest unit was represented by the maize cultivars, with subunits for each of the isolates. Each block consisted of 12 treatments with three replicates, resulting in a total of 36 experimental units. Each unit consisted of four 10-m-long rows 0.8 m apart and 0.2 m between plants. A fertilization formula of 160N-60P-00K was applied. Fieldwork was performed manually or with machinery, depending on the conditions in CHVM.

Figure 2: Model of the split-plot design with randomized complete blocks used. a,b,c means the P. agglomerans bacterial isolates. Ctrl means controls and colors means the cultivars. 

Origin of the inoculum source

The P. agglomerans bacterial isolates A, B and C were obtained from maize plants with symptoms of chlorotic streaks in the experimental f ields of the Colegio de Postgraduados. These isolates were previously identified by sequencing the 16S rRNA gene. Molecular analysis of sequences showed that they were 99% similar to the sequences for the P. agglomerans cluster, while the control strain of P. agglomerans ATCC 27155 showed 98-99% similarity. Sequences were deposited in GenBank (Accession Nos. EF050806 to EF050810). The isolates were maintained at -80 °C in a solution of 25% glycerol (^{Morales-Valenzuela et al., 2007}; ^{Silva-Rojas et al., 2007}).

Activation and inoculum preparation

To activate the bacteria, each isolate was seeded with the cross-streak method over CPG (Bioxon®, México) (10 g casamino acid, 10 g peptone, 10 g glucose and 18 g agar) and the cells were grown at 28 °C. After 48 h, pale yellow colonies characteristic of P. agglomerans were observed. The isolates of P. agglomerans were inoculated in chalk-agar culture medium (10 g yeast extract, 20 g glucose, 20 g calcium carbonate and 17 g agar). The isolates were incubated at 28 °C for 48 h. For each isolate, a bacterial suspension of 10⁸ CFU mL^-1 was prepared using sterile distilled water (^{Schaad et al., 2001}).

Inoculation of maize seedlings with P. agglomerans

For each experimental unit, 10 maize plants from the central rows were selected. P. agglomerans inoculation was performed on maize plants with six true leaves (phenological stage V3) (CIMMYT, 2012) 30 days after sowing by the injection method on the stem 2 cm above the ground (^{Hui-Ying et al., 2011}; ^{Lamka et al., 1991}). One mL of bacterial suspension with 10⁸ CFU was injected. As the negative control, 1 mL of sterile distilled water was used. At the end of the experiment, leaves with symptoms were collected for each strain. Koch’s postulates were validated in maize plants inoculated with P. agglomerans.

Measuring disease progress in maize

After harvesting and based on the presence of chlorotic streaks in inoculated maize leaves, the yield of each treatment was recorded by location, cultivar and isolates. For each treatment ear weight, width and length were evaluated, as well as grain number and weight.

Statistical analysis

The effect of the disease on yield was determined per treatment and analysed statistically with SAS V9.3 software (^{SAS, 2014}). The program design was developed for a split-plot experimental design (^{Littell et al., 2006}) with the following model:

where y_ijk is the response variable associated with level k of the factor in the small plot within the large plot in the j^th block that received the i^th level applied to the large plot, Blo_j is the effect of the j^th block, α_i is the main effect of the i^th level of the factor applied to the large plot (A), β_k is the main effect of the k^th level of the factor applied to the small plot (B), (αβ)_ik is the interaction between factor A and factor B, η_ij is a common random effect for all subplots in the large plot (i, j) and e_ijk is a random component particular to the subplot with the k^th level of the small plot in the largest plot (i, j). Significant differences were analysed with ANOVA and the Tukey test with P < 0.05.

Chlorotic streak symptoms on maize plant

Figure 4 shows the cumulative incidence of plants with chlorotic streaks by location after 7 days of inoculation, the percentage of infected plants was less than 15%. For both microenvironments, the largest number of plants with chlorotic streaks was observed between 7 and 14 days after inoculation. The percentage of plants with chlorotic streaks was 92% in Temamatla and 70% in Ayapango. Both microenvironments showed significant differences among isolates (P < 0.0001) and over time (P < 0.0001). Differences in incidence were caused by different isolates: in Ayapango isolate A was highly virulent, and in Temamatla isolate B was more virulent.

Figure 4: Advance of lesion in leaves of maize plants caused by inoculation of the bacteria in two locations of the Central High Valleys of Mexico. a: Ayapango and b: Temamatla. ◊: incidence; □: average temperature; Δ: average relative humidity. 

Maize yield in plants with chlorotic streaks

Symptoms of chlorotic streaks were observed in 60% of inoculated plants (Figure 5). In Ayapango, the number of plants with fruit was lower in all the cultivars and for all the isolates (P < 0.0001). In Temamatla; all cultivars were highly susceptible, but to different isolates: HS2 to P. agglomerans isolate A; Triunfo to P. agglomerans isolate A and B and Cacahuacintle to P. agglomerans isolates B and C (Figure 5 and Table 2).

Figure 5: Percent of plants with chlorotic streaks and fruit development in two locations A (Ayapango) and T (Temamatla) and three isolate a, b, c. 

By location, in Ayapango, the cultivar and isolate effects on the variables cob length, width and weight were significant. Grain number and weight were influenced by the interaction cultivar * isolate. In Temamatla, the cultivar-isolate interaction affected all the yield variables (Table 2).

Average size of ears on plants with chlorotic streak was smaller, relative to the control: width 5.2 ± 0.57 cm, length 14.80 ± 3.07 cm and weight 175.73 ± 62.65 g. Plants without inoculation with P. agglomerans (controls) had ears 5.32 ± 0.46 cm wide, 15.77 ± 2.51 cm long, weighing 195.20 ± 55.29 g. The average values of grain number and weight, were 405.82 ± 116.38 and 136.38 ± 50.27g, significantly lower than the control: 443.12 ± 110.21 and 162.21 ± 49.26 g.

As shown in Table 3, the cultivar-isolate interaction was significant in all variables. In both locations, cultivar HS2 and Cacahuacintle had the smallest dimensions. Ear width was smallest in Ayapango with the interaction location-isolate C and in Temamatla with the cultivar Triunfo and isolate A, and in both locations the interaction Cacahuacintle-isolate C was significant.

The Central Valleys of the State of Mexico are notable for their production of maize (500 000 ha), 80% of which is rainfed. Climate in this region is temperate and humid, with an average annual temperature of 16.5 °C, relative humidity of 85%, and annual rainfall of 800-900 mm, at an altitude of 2200 meters above sea level (^{SMN, 2012}). These conditions are favourable for maize cultivation but allow the development of bacterial diseases, such as chlorotic streaks on maize leaves caused by P. agglomerans.

In this research, during the period of evaluation of chlorotic streaks, average humidity and temperature in Ayapango and Temamatla were 45% and 85% and 16-19 °C, respectively. Under these conditions, the incidence of maize plants with chlorotic streaks increased gradually in the three cultivars of maize (Figure 3). Differences between both locations (Temamatla and Ayapango) during the evaluation were found. In Temamatla the humidity was lower than Ayapango (70 and 90% respectively) and the temperature was 2 °C higher in Temamatla. These findings probably had influence in the incidence of plants with chlorotic streaks. Probably the bacteria requires a lower humidity to express its pathogenicity.

^{Morales-Valenzuela et al. (2007)} found similar results when they inoculated maize and sorghum seedlings with P. agglomerans under greenhouse conditions. ^{Frederickson et al. (1997)} found that at 25 °C and 95% relative humidity, different isolates of P. agglomerans caused necrosis on the leaves and edges of pearl millet (Pennisetum glaucum (L.) R. Br.).

Similarly, ^{Sabaratnam and Beattie (2001)} reported that when maize plants are grown in relative humidity below 95%, colonisation of P. agglomerans in leaf tissue increases. However, when maize plants are exposed to 45% relative humidity, colonisation decreases drastically.

In the two locations, maize yield was affected by P. agglomerans when relative humidity was 85%. Differences of ± 3 °C in temperature caused lower yield compared with the controls. The affinity of P. agglomerans to the cultivar Cacahuacintle is probably due to the high concentration of carbohydrates, of which more than 73% is amylopectin (^{Carballo et al., 2010}), in contrast to cultivar HS2 and Triunfo. Loss of grain weight of the cultivar Cacahuacintle reduces productivity by 1.9 Mg ha^-1 meaning economic repercussions for the producers (^{Téllez-Silva et al., 2016}).