Genetic dissimilarity for resistance to Mononychellus tanajoa ( bondar ) ( Acari , Tetranychidae ) among domesticated and wild Manihot species

The aim of this study was to evaluate the genetic dissimilarity among wild and domesticated species of Manihot for resistance to cassava green mite during the insect life cycle. Nine accessions of wild Manihot species, M. esculenta ssp. flabellifolia, M. esculenta ssp. peruviana, and M. carthaginensis ssp. glaziovii, and two clones of M. esculenta (Cigana Preta and Sacaí) were evaluated under laboratory conditions at 25 ± 1°C, 70 ± 10% RH, and a 12-h photophase. Daily observations during the mite life cycle stages (larva-adult) were recorded. The data were subjected to an analysis of variance, a Scott-Knott test (5%), and Singh criterion, cluster, and principal component analyses. The larval-adult period ranged from 5.53 to 7.01 days: the longest period was observed on an M. glaziovii accession (GLA-19-DF) and the shortest on an M. flabellifolia accession (FLA-025V). The UPGMA method allowed the division of the genotypes into six groups, with the greatest distance between the FLA-025V and GLA-19-DF accessions. The first two main components explained 77.50% of the total accumulated variation. The association of the longest cycle duration of M. tanajoa with the lowest larval-adult viability suggests that GLA-19-DF is less favorable to mite development compared to the other accessions. Significant variability among the genotypes was observed.


Introduction
In Brazil, cassava is cultivated in all biomes due to the great diversity of varieties adapted to different ecosystems, resulting in a great genetic diversity among species.This diversity is due to natural selection and is favored by cross-pollination, a high degree of heterozygosity, and abrupt fruit dehiscence, leading to a plethora of new genotypes (FUKUDA;SILVA 2002).
Historically, cassava is a culture characterized by low supplies of inputs and agrochemicals with a high tolerance to drought, which can remain in the soil until its consumption, playing an important role in feeding the population (CAMARGO, 2003).However, due to its long growing cycle, the crop is subject to attack by a wide range of feeding arthropods.It is estimated that there are over 200 mite species associated with cassava in the Americas among which the green mite Mononychellus tanajoa (Bondar, 1938) (Acari, Tetranychidae) is considered a major pest affecting cassava cultivation in northeastern Brazil, particularly in the semiarid region (FUKUDA, 2006).
Green mite infestations develop largely under conditions of high temperature and low relative humidity, beginning at the top of the plants, attacking the buds and young leaves.Such infestation affects the formation of leaves, which become reduced in severely attacked plants and exhibit deformation, with the shortening of internodes, potentially leading to the death of the branch apices.Losses are estimated to be 10-50%, depending on the intensity of infestation (MORAES; FLECHTMANN, 2008).
The potential use of wild species has recently been reported in studies conducted at the International Center for Tropical Agriculture (CIAT).The results of these experiments include the following: 1) interspecific hybrids of M. esculenta ssp.flabellifolia show moderate to high levels of resistance to mite, whitefly, and scale insects (BURBANO et al., 2007); 2) an interspecific hybrid between M. esculenta and M. walkerae Croizat exhibits reduced post-harvest physiological spoilage (CIAT, 2003); 3) M. glaziovii Muell, M. catingae Ule, and M. carthaginensis Mull.Arg. are adapted to semiarid conditions and are potential sources of genes for drought tolerance; 4) the roots of interspecific hybrids of M. esculenta ssp.flabellifolia contain a high protein content (AKINBO et al., 2012a); 5) M. flabellifolia accessions and interspecific crosses of M. esculenta ssp.flabellifolia exhibit different levels of resistance to the whitefly Aleurotrachelus socialis Bondar (CARABALI et al. 2010); and 6) resistance to A. socialis can be introgressed in interspecific hybrids of commercial cassava (AKINBO et al., 2012b).
Considering the characteristics of agronomical interest found in wild and domesticated cassava species, it is important to select cassava genotypes that are resistant to green mite for use in integrated mite control programs (ARGOLO et al., 2005).
This study was aimed at assessing the genetic dissimilarity among wild and domesticated Manihot species for resistance to M. tanajoa during the life cycle of the insect (larva-adult period).

Location of bioassays
The study was performed at the Laboratory of Entomology -Embrapa Cassava and Fruits, Cruz das Almas, Bahia State, Brazil.

Mites and Manihot accessions
Cassava green mite specimens (M.tanajoa) were obtained from colonies established in the cassava clone "Cigana Preta" maintained in a greenhouse.The wild Manihot genotypes are part of the Work Collection of Wild Manihot Species -Embrapa Cassava and Fruits, comprising 628 accessions belonging to 28 species, and were installed in the field in 2005.This study evaluated nine accessions belonging to three wild cassava species, M. esculenta ssp.flabellifolia (FLA-025V, FLA-026V, and FLA-029V), M. esculenta ssp.peruviana (PER-006V, PER-007V, and PER-010V), and M. carthaginensis ssp.glaziovii (GLA-03-DF, GLA-10-DF, and GLA-19-DF), and clones of a domesticated species (M.esculenta), Cigana Preta (BGM 0116) and Sacaí (BGM 0384), all from the experimental area belonging to Embrapa Cassava and Fruits.These genotypes were selected according to characteristics of interest: M. esculenta ssp.flabellifolia is the ancestor of M. esculenta and provides resistance to such pests as mites; M. esculenta ssp.peruviana is close to M. flabellifolia; M. carthaginensis ssp.glaziovii presents drought tolerance; and the M. esculenta clone Cigana Preta is considered susceptible and Sacaí resistant to M. tanajoa.

Mite eggs and bioassays
To obtain eggs, eighty female M. tanajoa obtained from maintenance breeding were allowed to lay eggs in the lobes of fully developed young leaves of the tested accessions.The lobes were positioned with the top (adaxial surface) in contact with nylon foam (1 cm thick) that was moistened with distilled water and surrounded with strips of cotton wool inside trays (20 x 15 x 5 cm).The trays were sealed with transparent polyvinylchloride film (PVC), which was perforated to allow aeration.After 24 hours, the females were removed, and the eggs were allowed to develop on the lobes until hatching.After hatching, the larvae were isolated on leaf disks (2.5 cm in diameter) of each accession tested; the disks were deposited on nylon foam (1 cm thick) moistened with distilled water inside Petri dishes (14 cm in diameter x 2 cm deep) and sealed with PVC film perforated to allow aeration, according to the methodology of Noronha et al. (1995).The Petri dishes were maintained in a BOD-type incubator at 25 ± 1°C, a relative humidity of 70 ± 10%, and a photoperiod of 12 hours.Except when they were in quiescent stage (protochrysalis, deutochrysalis, and teleiochrysalis stages), the mites were transferred to new disks every two days with the aid of a stereoscopic microscope and a brush with fine bristles.The development of M. tanajoa was followed to adulthood, with daily observations of the larval, protochrysalis, protonymph, deutochrysalis, deutonymph, teleiochrysalis, and adult stages to verify the succession to the next stage, as characterized by the presence of exuviae.A completely randomized design with 30 replicates per genotype was used.The first 30 hatched larvae were captured.Each plot consisted of one mite.

Statistical analysis
The data were subjected to an analysis of variance, and the means of the genotypes were grouped by the Scott-Knott test (SCOTT; KNOTT, 1974) at 5% significance using SISVAR software (FERREIRA, 2011).Cluster and principal components analyses were performed.For the cluster analysis, the mean Euclidean distance was considered as a dissimilarity measure.The hierarchical groupings from the dissimilarity matrix were obtained by the Unweighted Pair Group Method with Arithmetic Mean (Upgma) method (SNEATH;SOKAL, 1973).The Singh (1981) criterion was also used to quantify the relative contribution of the duration of the life cycle stages for the genetic dissimilarity analyses performed using the Genes (CRUZ, 2001) and Statistica 7.1 software (STATSOFT, 2005).

Results and discussion
The biological life cycle (larva-adult period) of M. tanajoa was grouped according to the means of the Scott-Knott test at 5% significance (Table 1).The egg stage lasted for five days and did not vary between the accessions.The larvae used in the study were of the same age.A variation from 1.07 to 1.53 days was observed for the larval stage, whereas the protonymph stage ranged from 0.96 to 1.29 days, with the formation of only one group, which did not statistically differ from Manihot accessions studied in the two cycle stages.In contrast, two groups were distinguished for the other stages.
For the protochrysalis stage, the variation ranged from 0.98 to 1.62 days, with significant differences between the accessions.The longest mite protochrysalis, protonymph, and larva-adult stage durations were found for the accessions of M. peruviana species; the shortest protochrysalis stage was observed on the M. carthaginensis ssp.glaziovii accessions.
For the deutochrysalis and deutonymph phases, the variation ranged from 0.93 to 1.28 days and 0.90 to 1.27 days, respectively.The M. peruviana accessions and M. esculenta clones Cigana Preta and Sacaí of were among the group showing the shortest deutochrysalis stage, whereas the M. esculenta clones were among the group showing the longest deutonymph stage.The teleiochrysalis stage varied from 0.64 to 1.09 days, with the M. flabellifolia accessions and M. esculenta clones exhibiting the longest durations and the M. peruviana accessions and two M. glaziovii accessions exhibiting the shortest duration.
The larva-adult period ranged from 5.53 to 7.01 days.The group that showed the longest larva-adult period was formed by FLA-026V, GLA-03-DF, GLA-019-DF, PER-006V, PER-007V, PER-010V, Cigana Preta, and Sacaí.The accessions that formed the group with the shortest larva-adult period were FLA-025V, GLA-10, and DF-FLA-029V.The M. esculenta clones were included in the same group during all stages of the mite life cycle.
The viability of the immature stages of the cassava green mite varied among the Manihot species.The survival rate ranged from 76.67 to 100%, with the lowest viability being observed on accession GLA-19-DF (76.67%) of M. carthaginensis ssp.glaziovii.The viability was greater than 80% on the accessions of M. esculenta ssp.flabellifolia, and the survival rate was greater than 90% on GLA-03-DF and GLA-10-DF.Mortality was observed during the immature mite stage on the M. peruviana accessions and Cigana Preta and Sacaí clones (Table 1).These data show that there is genetic dissimilarity between the tested accessions, as there were significant differences in the green mite biological cycle among different genotypes of the same species.Argolo et al. (2005) evaluated the resistance of cassava varieties (BGM 0876-Pretinha, BGM 0080-Engana Ladrão, BGM 0537-Do Céu, and BGM 0384-Sacai) selected as tolerant to M. tanajoa for the semiarid conditions of northeastern Brazil for varieties and found that the larva-adult period ranged from 7.39 to 9.43 days.The mite life cycle for clone Sacaí was 9.28 days, which was higher than that obtained in the present work for the same clone.However, there was a difference among the individual stages, with the protonymph, deutochrysalis, and deutonymph stages being longer in our study.These data showed that the duration of each stage does not interfere with the duration of the next stage or, consequently, in the development of cassava green mites.
The present study indicates that the characteristics having the greatest influence on mite development are related to the genetics of host species because the bioassays were performed under conditions of identical temperature, relative humidity, and availability of water and nutrients.It also indicates different types of antibiosis of resistance levels of these accessions, which should be considered as an indicator of pest resistance.Burbano et al. (2007) evaluated and compared two Brazilian wild Manihot species (M.flabellifolia and M. peruviana) to commercial M. esculenta genotypes to determine their resistance potential to M. tanajoa and found variable degrees of mite damage.The M. flabellifolia genotypes were significantly different from M. esculenta but similar to the M. peruviana genotypes, which showed lower levels of damage.
The largest distance was observed between FLA-025V and GLA-19-FD, i.e., genetically, the greatest dissimilarity was between the genotype of Group 6 and the genotype of Group 1.The M. flabellifolia genotypes showed different behavior for the duration of the mite life cycle, and each accession was classified into a different group.However, there was similarity between the accessions in each group, indicating genotypic and/or phenotypic proximity among them, whereas dissimilarity was observed between the groups.
The feature that most contributed to the assessment of the genetic diversity according to the Singh (1981) criterion was the larva-adult period, at 57.30%.
Alves et al. ( 2011) assessed 14 accessions of six wild Manihot species with regard to biological aspects of M. tanajoa and found that the life cycle (larva-adult period) ranged from 5.76 to 8.18 days.The accessions of wild M. anomala species formed the group showing the longest mite development period, whereas M. peruviana and M. flabellifolia accessions showed shorter larva-adult periods.
The first two principal components explained 77.50% of the total accumulated variation.This value is considered satisfactory because the principal components explained approximately 80% of the variance found in the set of characters analyzed, enabling the formation of groups among the accessions and the construction of a scatter plot based on the diversity observed (Figure 1 and 2).
The first principal component explained 52.50% of the total variance, and the phases that most contributed to explaining this variability were the teleiochrysalis, deutonymph, and protonymph stages, as these variables showed higher weighting coefficients.The second principal component explained 25.00%, with the larval-adult and larva stages contributing the most (Figure 2).The species distribution in the scatter plot (Figure 2) confirmed the greatest distance observed in the dendrogram (Figure 1) between FLA-025V and the other accessions.This result allowed the identification of the greatest genetic dissimilarity among individuals, in addition to grouping two M. flabellifolia accessions with M. esculenta and confirming the proximity of these accessions.To verify the cassava ancestry, Carvalho (2005) studied 27 Manihot esculenta ssp.flabellifolia populations using two types of genetic markers (the glyceraldehyde 3-phosphate dehydrogenase [G3PDH] gene and five microsatellite loci) and confirmed that cassava was originally domesticated from M. esculenta ssp.flabellifolia.
There is complementarity between the two multivariate techniques used.In the cluster analysis (Figure 1) and principal component analysis (Figure 2), the GLA-19-DF and FLA-025V accessions showed no similarity with the other accessions, as they were allocated to different groups.A similar result was found for the M. peruviana and M. carthaginensis ssp.glaziovii accessions (GLA-10-DF and GLA-03-DF).

Conclusion
The time in days of a biological cycle phase (larva-adult period) has little influence on the duration of subsequent stages.
The longest larval to adult duration and lowest viability of the immature stage of M. tanajoa suggest that the GLA-19-DF accession is less susceptible to mite development when compared to the other genotypes tested.
The different behavior of the mite life cycle on accessions of the same species indicates that genetic differences exist between accessions of the same Manihot species.
The accessions of M. esculenta ssp.peruviana species are genetically close.
There is genetic variability among wild and domesticated Manihot species with regard to resistance to Mononychellus tanajoa.

Figure 1 .
Figure 1.Dissimilarity dendrogram based on the mean Euclidean distance and UPGMA clustering method of the M. tanajoa life cycle (larva-adult) on 11 Manihot accessions.

Figure 2 .
Figure 2. Graphic dispersion of the scores of 11 Manihot accessions in relation to principal components 1 and 2.

Table 1 .
Duration (X ± SE) and total viability (%) of stages of the life cycle (larva-adult period) of M. tanajoa as a function of the Manihot accession at a temperature of 25 ± 1°C, relative humidity of 70 ± 10%, and photoperiod of 12 hours.Cruz das Almas, Bahia State, Brazil, 2011.