Resistant accessions of wild Psidium spp. to Meloidogyne enterolobii and histological characterization of resistance

Authors


Abstract

Meloidogyne enterolobii has been reported in some states of Brazil and other countries causing severe damage on commercial guava (Psidium guajava). The use of resistant varieties is the most effective way to manage nematode parasitism. This study screened 51 accessions of Psidium spp. selected from the Psidium Germplasm Collection (Embrapa) to look for resistance against M. enterolobii. Six months after inoculation, nematode reproduction factor (RF) was used to assess resistance. The following species were resistant to Menterolobii: Pcattleianum (yellow guava), Pfriedrichsthalianum (Costa Rican guava), Acca sellowiana (feijoa) and Prufum (purple guava). All 43 wild accessions of Pguajava were susceptible, as well as three accessions of Pguineense (Brazilian guava), one of Pacutangulum (pear guava) and the susceptible control P. guajava cv. Paluma. When used as rootstocks under greenhouse conditions, Pcattleianum and Pfriedrichsthalianum were compatible with cv. Paluma; however, in greenhouse and field conditions only 50% of both scions survived. No apparent hypersensitive response (HR) was seen in the resistant guava Pcattleianum and Pfriedrichsthalianum. Juveniles were able to develop normal feeding sites similar to those in susceptible roots 6–13 days after inoculation (dai). From 27 to 32 dai, giant cell deterioration was observed and nematodes showed arrested development. The majority of nematodes failed to reach maturity and did not begin laying eggs in resistant roots. These results suggested that the induction of resistance is relatively late in this pathosystem.

Introduction

The root-knot nematode (RKN) Meloidogyne enterolobii (syn. M. mayaguensis) was described from a population sampled in China and isolated from a tree species (Enterolobium sp.). This nematode was also reported from other regions in China, mainly isolated from guava (Psidium guajava). The presence of this parasite has increasingly been detected worldwide from a wide range of hosts, including crops carrying resistance genes to major Meloidogyne spp. (Carneiro et al., 2006; Brito et al., 2007). Recently, Menterolobii was detected in two commercial greenhouses in tomato (Solanum lycopersicum cv. Maxifort) resistant to Meloidogyne spp. in Switzerland (Kiewnick et al., 2008). Considering the risk of introducing and disseminating this pathogen in the European region, Menterolobii was recently added to the 2012 EPPO alert list as a quarantine nematode.

Morphological identification demands considerable skills and can be unreliable as a result of significant intraspecific morphological variation in Meloidogyne spp. Because of its morphological resemblance to Mincognita and Marenaria when only the perineal patterns are considered (Carneiro et al., 2001; Brito et al., 2004), Menterolobii has been misidentified in a number of surveys. It is probable that the severe root-knot problem on guava in the Americas and the isolated cases in Africa and Asia involve only this particularly virulent species (Carneiro et al., 2012).

In Brazil, Menterolobii was originally detected on guava orchards in 2001 in Pernambuco and Bahia states (Carneiro et al., 2001). Since then, this nematode has been a matter of major concern in the country because it has spread rapidly, making the cultivation of guava unviable in heavily infested areas (Carneiro et al., 2007; Siqueira et al., 2009).

Typical symptoms of Menterolobii in guava are a strong tanning of leaves and branches, followed by general yellowing of plant aerial parts, culminating in leaf fall and plant death. Infected roots show multiple galls and generalized necrosis, as well as a drastic decrease in fine roots. Moderate infections are associated with common chlorosis, nutrient deficiency and reduced flowers and fruits. Meloidogyne enterolobii infection in guava is a complex disease, usually associated with soil fungi infestations (Gomes et al., 2010).

In the São Francisco Valley, a major guava producing region in northeast Brazil, 70% reduction was reported in guava production within a 7 year period (2000–2006; Plantec/Codevasf, SGAN 601, Ed. Manoel Novaes, Brasília-DF, Brazil, personal communication). Moreover, as this is an important area for seedling production, M. enterolobii may have been spread to other regions throughout Brazil (Carneiro et al., 2007). According to Pereira et al. (2009), direct losses associated with guava infection by M. enterolobii in several states in Brazil might reach up to US$ 61 million, not to mention direct job losses as a result of decline in guava orchards. Taken together, these observations reveal that new control methods should be undertaken in order to reduce the major economic impact of this Meloidogyne-induced disease on guava. Some screenings for resistance against Meloidogyne spp. in Psidium spp. have been carried out, including one grafting assay; however, in the majority, either the nematode was not identified as M. enterolobii (Carneiro et al., 2012) or the rootstocks were misidentified. In a previous study, Pcattleianum was resistant (reproduction factor, RF, = 0) and Pfriedrichsthalianum moderately resistant (RF = 1·9) to M. enterolobii (Carneiro et al., 2007).

This work investigates accessions of guava and its wild relatives (Psidium spp.) to identify novel sources of resistance to Menterolobii, with the identification of high levels of resistance in Pcattleianum and Pfriedrichsthalianum. Histological observations revealed that the resistance is a late defence response that resulted in the degradation of feeding sites, and production of males, but not adult females or eggs.

Materials and methods

Accessions of Psidium spp.

Accessions of Psidium spp. used in this study were collected from different regions in Brazil and also from Costa Rica (Pfriedrichsthalianum). These accessions are deposited in the CENARGEN bank and are specified in Table 1. The commercial Pguajava cv. Paluma was used as a susceptible control.

Table 1. Response of guava accessions (Psidium guajava) and Psidium spp. to Meloidogyne enterolobii, 180 days after inoculation under controlled conditions
AccessionaRoot fresh weight (g)bGIcEMIdTNEe per g of rootsRFfReactiong
  1. Data were transformed as log (+ 1). Means followed by different letters are significantly different according to Scott–Knot's test (< 0·05). Coefficient of variation (CV) = 37%.

  2. a

    Brazilian states: DF, Distrito Federal; RS, Rio Grande do Sul; PR, Paraná; SP, São Paulo; SC, Santa Catarina; GO, Goiás; MG, Minas Gerais; PA, Pará; PE, Pernambuco.

  3. b

    Mean values (= 6) of fresh weight of roots (g).

  4. c

    Mean values (= 6) of gall index (GI) based on a 0–5 scale. 0: no galls, 1: 1–2 galls, 2: 3–10 galls, 3: 11–30 galls, 4: 31–100 galls, 5: >100 galls per root system (Taylor & Sasser, 1978).

  5. d

    Mean values (= 6) of egg mass index (EMI) based on a 0–5 scale. 0: no galls, 1: 1–2 galls, 2: 3–10 galls, 3: 11–30 galls, 4: 31–100 galls, 5: >100 egg masses per root system (Taylor & Sasser, 1978).

  6. e

    Mean values (= 6) of total number of eggs (TNE) per gram of roots.

  7. f

    Mean values (= 6) of reproduction factor (RF = final population/5000 eggs of M. enterolobii).

  8. g

    Reaction of inoculated plants. RF ≥ 1 = susceptible, S; RF < 1 = resistant, R.

1 Guava (wild, red, Brasília, DF)37·3 b5·04·867 012·2 b499·9 aS
2 Guava (wild, white, Brasília, DF)22·2 c4·34·336 556·6 c162·3 cS
3 Guava (wild, red, Planaltina, DF)45·8 b5·05·070 538·6 b646·1 aS
4 Guava (wild, red, Gama, DF)55·6 b5·04·065 459·6 b727·9 aS
5 Guava (wild, white, Planaltina, DF)37·3 b4·54·511 569·8 d86·3 cS
6 Guava (wild, white, Gama, DF26·2 c5·04·783 511·4 a437·6 aS
7 Guava (wild, red, Sobradinho, DF42·4 b5·05·018 480·1 d156·7 cS
8 Guava (wild, white, Brasília, DF)48·7 b5·04·8 39 671·5 bc386·4 bS
9 Guava (wild, red, Brasília, DF)17·4 c5·05·0 18 646·2 cd 64·9 dS
10 Guava (wild, white, Sobradinho, DF)15·7 c4·34·314 012·7 d 44·0 dS
11 Guava (wild, red, Sobradinho, DF)70·8 a5·05·063 107·3 b893·6 aS
12 Guava (wild, red, Guará, DF)18·0 c4·24·226 913·6 c96·9 cS
13 Guava (wild, white, Guará, DF)43·3 b4·85·056 864·2 b492·4 aS
14 Guava (wild, white, Brasília, DF)30·3 b5·04·716 076·3 d97·4 cS
15 Guava (wild, red, Brasília, DF)32·0 b4·34·395 041·7 a608·3 aS
16 Guava (wild, red, Ceilândia, DF)74·8 a5·05·085 448·6 a639·2 aS
17 Guava (wild, white, Ceilândia, DF)24·7 c5·04·561 610·4 b304·4 bS
18 Guava (wild, red, Núcleo Band, DF)55·9 b5·05·017 101·9 d191·2 cS
19 Guava (wild, red, Brasília, DF)59·9 b4·84·834 731·9 c416·1 bS
20 Guava (wild, white, Brasília, DF)87·9 a4·75·020 300·8 c356·9 bS
21 P. cattleianum (yellow guava, Pelotas, RS)62·5 b0·00·00·0 e0·0 eR
22 Guava (wild, white, Brasília, DF)60·1 b5·05·019 966·7 c240·0 bS
23 Guava (wild, red, Brasília, DF32·5 b5·05·052 635·9 b342·1 bS
24 Guava (wild, red, Campo Morão, PR)30·8 b5·05·070 360·7 b433·4 bS
25 Guava (wild, red, Londrina, PR)43·9 b4·84·3 78 906·6 ab692·8 aS
26 Guava (wild, white, Piracicaba, SP)64·7 b5·05·025 849·2 c334·5 bS
27 Guava (wild, white, Campinas, SP)50·1 b5·05·055 364·8 b554·8 aS
28 Guava (wild, white, Limeira, SP)54·0 b5·05·052 716·1 b569·3 aS
29 Guava (wild, red, Limeira, SP)55·8 b5·05·051 366·0 b573·2 aS
30 Guava (wild white, Botucatu, SP)100·5 a4·54·327 347·7 c549·7 aS
31 Guava (wild, yellow, Botucatu, SP)76·1 a5·05·042 476·3 c646·5 aS
32 Guava (wild, white, Indaiatuba, SP)35·6 b5·04·756 354·6 b401·2 bS
33 Guava (wild, red, Indaiatuba, SP)94·0 a5·05·011 834·5 d222·5 cS
34 Guava (wild, red, Jaboticabal, SP)32·4 b5·05·050 220·6 b325·3 bS
35 Guava (wild, white, Jaboticabal, SP)22·3 c4·24·266 288·0 b295·6 bS
36 Guava (wild, white, Brasília, DF)12·7 c5·05·022 117·2 c 56·2 dS
37 Acca sellowiana (pineapple guava or feijoa, Capão Alto, SC)32·6 b0·00·00·0 e0·0 eR
38 Pguineense (Brazilian guava, Brasília, DF)44·2 b4·34·339 259·3 c 33·0 dS
39 Pguineense (Brazilian guava, Divinópolis, GO)61·4 b4·74·716 438·6 d201·9 bS
40 Pguineense (Brazilian guava, Planaltina, DF)29·7 b3·83·94365·8 d 25·9 dS
41 Guava (wild, white, GO)35·3 b5·05·058 633·9 b414·0 bS
42 Guava (wild, red, Divinópolis, GO)41·3 b5·05·0114 210·4 a943·4 aS
43 Pfriedrichsthalianum (Costa Rican wild guava, Turrialba, Costa Rica)24·1 c1·00·0337·0 e0·8 eR
44 Guava (wild, Brasília, DF)41·6 b2·92·92750·4 d 22·9 dS
45 Guava (wild, white, Brasília, DF)96·3 a5·05·019 044·7 c366·8 bS
46 Guava (wild, white, Brasília, DF)119·3 a5·05·020 106·2 c479·7 bS
47 Guava (wild, white, Brasília, DF)50·5 b5·05·025 656·8 c259·1 bS
48 Guava (wild, red, Cristalina, GO)58·8 b4·14·915 493·2 d182·2 bS
49 Guava (wild, white, Cristalina, GO)43·3 b4·94·922 525·0 c195·1 bS
50 P. rufum (purple guava, Paracatu, MG)13·5 c0·00·00·0 de0·0 eR
51 Pacutangulum (pear guava, Belém, PA)183·7 a5·05·09656·4 d177·4 cS
52 Control: guava cv. Paluma (Petrolina, PE)44·3 b5·05·019 633·8 c174·0 cS

Nematode inoculation

A population of M. enterolobii collected in Petrolina (Pernambuco state, Brazil) was used in this study because of its pathogenicity on commercial guava. The identification of species was done using the esterase (Est) phenotype (Carneiro & Almeida, 2001) and SCAR markers as described (Tigano et al., 2010). Prior to inoculation, the population was multiplied in tomato cv. Santa Clara for 3 months under greenhouse conditions. Eggs were extracted from infected roots using 0·5% NaOCl, according to Hussey & Barker (1973), using a blender instead of manual agitation. For histopathological studies, freshly hatched second-stage juveniles (J2) were collected using a modified Baermann funnel. In both cases, counting was done under a light microscope using Peters’ slides.

Nematode resistance in controlled conditions

Plants of each accession were grown in 5 L pots filled with a mixture (1:1) of autoclaved soil and Plantimax compost under greenhouse conditions. Seedlings of about 15–20 cm were inoculated with 5000 eggs of Menterolobii by pipetting the nematode suspension around the stem base. Plants were arranged in a completely randomized design with 52 treatments (accessions of wild Psidium and guava) and six replications. They were maintained under greenhouse conditions at c. 25–30°C and were watered and fertilized as required. Six months after inoculation, the root system was rinsed with tap water and weighed. Roots were stained with phloxine B and evaluated for gall and egg mass numbers (galling index, GI; egg mass index, EMI), using a 0–5 scale, where 0 = no galls or egg masses; 1 = 1–2; 2 = 3–10; 3 = 11–30; 4 = 31–100; and 5 = >100 galls or egg masses per root system (Taylor & Sasser, 1978).

Eggs were extracted using a modified extraction method according to Hussey & Barker (1973), using a blender instead of manual agitation and 1% NaOCl. Total egg number per plant was quantified under a light microscope using Peters’ slides. The reproduction factor (RF) was calculated as RF = FP/IP, where FP = final nematode population and IP = initial nematode population (IP = 5000). The average RF was transformed as log10 (x + 1), submitted to analysis of variance and the means separated using the Scott–Knot test at 5% confidence level. The accessions for which RF = 0 were considered immune, RF < 1 resistant (R), and those for which RF ≥ 1 were considered susceptible (S; Sasser et al., 1984).

Plant grafting in greenhouse and field conditions

Twelve-month-old plants of Pcattleianum, Pfriedrichsthalianum, Pguajava, Prufum and Asellowiana were grafted with Pguajava cv. Paluma using the budding technique. Successful grafted plants (Pcattleianum, Pfriedrichsthalianum and Pguajava) were planted in a field infested with Menterolobii. The plants were evaluated in May and September of 2009 and February, June and October of 2010. Plant height and diameter in the grafted region were evaluated.

Histopathological observations

Plantlets of Pcattleianum and Pfriedrichsthalianum accessions and the susceptible control Pguajava cv. Paluma were grown in 3 L pots containing washed sterilized sand and were fertilized weekly. Plantlets were inoculated with 20 000 J2 of Menterolobii per plant. Six plantlets of each accession and three for the susceptible control per time point were carefully removed from cups at 3, 6, 12, 16, 18, 22, 28, 34 and 45 days after inoculation (dai), and their roots rinsed with water. Some of the roots from susceptible and resistant plants were stained with acid fuchsin to observe J2 penetration, localization and subsequent development within the roots. After staining, root segments were observed under a stereomicroscope and those parts that showed nematode infection were mounted on a slide for observation under a light microscope (Axiophoto Zeiss).

Subsamples of roots were embedded in resin to produce thin sections. Root fragments showing galls/swellings or without symptoms were excised under a stereomicroscope, fixed, and embedded in Technovit 7100 epoxy resin (Kulzer Friedrichsdorf) as described by Pegard et al. (2005) and according to the manufacturer's recommendations. Unstained root sections were mounted on glass slides and fluorescence was observed after UV excitation (UV filter set A2 Zeiss 02; 488002-0000). The same sections were subsequently stained (1 min at 60°C) with 0·5% toluidine blue in 0·1 m sodium phosphate buffer, pH 5·5 and observed using a light microscope.

Results

Characterization of nematode resistance in controlled conditions

The RF values from the species Pcattleianum, Pfriedrichsthalianum, Prufum and Asellowiana indicated that they were resistant (RF < 1) to Menterolobii (Table 1). All wild accessions of Pguajava as well as the commercial guava cv. Paluma, Pacutangulum and Pguineense were susceptible (RF > 1; Table 1). Some wild guava (accessions 9, 10, 36 and 44) showed significantly smaller RF than other Pguajava, and similar to accessions 38 and 40 of Pguineense.

Evaluation of plant grafting in greenhouse and field conditions

All 10 plants of Pfriedrichsthalianum (Costa Rican guava) and nine out of 10 plants of Pcattleianum (yellow guava) were compatible with guava cv. Paluma, using the budding technique, in greenhouse conditions. None of the 10 grafted plants of Prufum or Asellowiana were compatible.

All successfully grafted plants were planted in the field in May 2009, in an area naturally infested with Menterolobii (Luziânia, GO, Brazil), for evaluation of their development. The evaluations of plant height were done in May and September 2009, and in February, June and October 2010.

Based on data collected up to September 2009, it was observed that c. 50% of guava trees grafted on Pfriedrichsthalianum and Pcattleianum died under field conditions. The guava trees grafted on Pfriedrichsthalianum were vigorous, producing buds, flowers and fruits from June 2010. Plants grafted on Pcattleianum showed an incompatibility reaction in the field (enlargement in the grafted region) observed in September 2009. Consequently, all nine plants planted in the field had died by June 2010 (Table 2). Control plants (guava grafted on guava) were taller than plants grafted on Pfriedrichsthalianum (Table 2). Unfortunately, during the rainy season of October 2010, cattle gained access into the experimental plot and grazed guava shoots, reducing their height (Table 2).

Table 2. Comparison of grafted plant heights (cm) under field conditions
Psidium speciesaTime of year
May 2009Sept 2009Feb 2010Jun 2010Oct 2010
  1. a

    Rootstocks were grafted with P. guajava cv. Paluma. Plant height was measured 90 days after planting in the field.

  2. b

    Mean (= 9–10) of plant heights (cm) ± SE.

  3. c

    Plant height was affected by cattle grazing.

  4. d

    Scions died.

P. friedrichsthalianum102·5 ± 11·3b108·5 ± 12·9116·7 ± 13·6118·8 ± 14·282·7 ± 4·2c
Pcattleianum97·2 ± 6·2098·0 ± 6·6112·5 ± 8·5dd
P. guajava 110·0 ± 1·15111·0 ± 1·2113·2 ± 1·5158·8 ± 1·2122·3 ± 2·2c

Histopathological observations of Psidium spp. inoculated with M. enterolobii

The susceptible Pguajava cv. Paluma was chosen for the histopathological observations of the compatible reaction (Fig. 1). In the susceptible host, examination of acid fuchsin-stained guava roots and toluidine blue-stained sections showed that J2 were able to penetrate the root tips, migrate along with the sieve elements, and develop normally after having initiated the differentiation of feeding sites (Fig. 1a–c). At 3 dai, numerous J2 were observed localized in the subapical meristem of the roots (Fig. 1a), and J2 were present in the root cortex, probably migrating towards the vascular cylinder; at this stage, cell wall damage (Fig. 1b) was frequently seen. At 6–10 dai, numerous juveniles were observed within the vascular cylinder, and feeding juveniles were visible inside the vascular cylinder close to giant cells (Fig. 1c). At 17–23 dai, numerous J3/J4 had established feeding sites with 6–14 well-formed giant cells containing numerous nuclei, dense cytoplasm, and small vacuoles (Fig. 1d,e). At this time, enlargement of the vascular cylinder region took place, and large galls were apparent. Giant cells displayed thickened cell walls (Fig. 1d–f). At 27 dai, adult females were observed (Fig. 1f) associated with severe injuries to the surrounding cells, disruption of the root cortex and accumulation of egg masses at the root surface (Fig. 1g).

Figure 1.

Root sections of Psidium guajava cv. Paluma (susceptible) infested with Meloidogyne enterolobii. (a) Roots showing J2 at the penetration stage; (b, c) giant cells indicating the differentiation of feeding sites close to the nematode; (d, e) giant cells in different stages of formation along with J3 and J4; (f) an adult female close to giant cells; (g) red-stained eggs; (h) large number of completely formed giant cells. vc, vascular cylinder; gc, giant cells; N, nematodes; E, eggs. Sections visualized after staining with toluidine blue (b, c, d, e, f, h) and acid fuchsin (a, g).

The resistant accessions of Pcattleianum and Pfriedrichsthalianum were used to histologically analyse defence responses of the plant. Because results were very similar, only P. cattleianum is shown (Fig. 2). In both accessions, nematode penetration did not seem to be affected, and a lot of J2s could penetrate the resistant guava accessions, as in the susceptible one (data not shown). Microscopic examination of acid fuchsin-stained roots and observations of toluidine blue-stained sections showed that J2 were localized in the epidermis and cortex at 3–6 dai, indicating that penetration and colonization were in progress (Fig. 2a–c).

Figure 2.

Root sections of Psidium cattleianum (resistant) infested with Meloidogyne enterolobii. (a) Numerous J2 nematodes at the penetration stage; (b) numerous blue fluorescent J2 nematodes in the vascular cylinder; (c,d) giant cells organized around J3/J4; (e) giant cells with degraded cytoplasm; (f) J4 with arrested development; (g, h) males inside J4, and male with fluorescent blue staining. vc, vascular cylinder; co, cortex; gc, giant cells; N, nematodes. Sections visualized under UV light (b, h); sections stained with toluidine blue (c, d, e, g) and acid fuchsin (a, f).

The two resistant Psidium spp. did not show any HR response to nematode penetration at the early infection stage (Fig. 2b). When roots were examined at 13–17 dai, nematodes were seen close to feeding sites in which the giant cells displayed similar features to those in susceptible roots (Fig. 2d). The first evident differences between the guava cv. Paluma and the resistant Psidium spp. were detected at 15–23 dai, when the giant cells adjacent to the nematode in resistant roots had some larger vacuoles that took up almost the entire volume of the cells, whereas the giant cells in the susceptible roots had uniformly dense cytoplasm with less vacuolation (Fig. 2e–h). Prior to 13 dai, some nematodes were developing normally in the resistant accessions based on observations of their size, shape and appearance of internal contents. However, nematodes associated with resistant roots at 22–27 dai were arrested in their development (J4; Fig. 2f) in parallel with the deterioration of giant cells. At this stage, most of the giant cells appeared to be on the verge of collapse and devoid of any visible cytoplasmic content (Fig. 2e), whereas in susceptible roots, robust giant cell complexes with dense cytoplasm and thick cell walls were present (Fig. 1f,h). At 22–31 dai (Fig. 2e,g,h) several males were observed inside the J4 and none were observed to have reached to a mature female stage. The presence of putatively phenolic compounds (blue fluorescent material) was visible inside the body of J4 females and males (Fig. 2h).

Discussion

Considering the difficulty in identifying Menterolobii by just the perineal pattern (Carneiro et al., 2001; Brito et al., 2004), it is possible that Menterolobii from guava has been misidentified in several countries, because the identification of root-knot nematodes in this plant species has been based only on morphology and differential host data (Villota & Agudelo, 1997; Crozzoli & Casassa, 1998). Using isozyme and RAPD markers for Meloidogyne spp. identification, only Menterolobii was detected in guava in Brazil (Siqueira et al., 2009). Carneiro et al. (2012) found that Pguajava was a good host for Menterolobii and a non-host for Marenaria, Mincognita and Mjavanica. Rossi et al. (2002) observed similar results when three commercial guava cultivars were inoculated with Mincognita race 2 and Mjavanica.

Susceptibility of Pguajava to Menterolobii observed in this study was also reported by other authors for Meloidogyne spp. (Villota & Agudelo, 1997; Lee et al., 1998; Maranhão et al., 2001; Carneiro et al., 2007; Almeida et al., 2009; Gonzalez-Gaona et al., 2010; Castro et al., 2012). Casassa et al. (1997) evaluated seven Pguajava accessions 6 months after inoculation with 5000 eggs of M. incognita. Resistance was observed in Pguajava accession S3 (RF = 0·2). In this study, only the accession 44 of wild guava (Olhos d’água Park) showed good results (GI = EMI = 2·9 and RF = 22·9) when compared with other evaluated wild Pguajava. However, more studies under field conditions are necessary to confirm these results. Milan (2007, 2010) also detected resistance in guava accessions from Malaysia. In the first study, Pguajava B-12 was classified as resistant to Mincognita (RF = 0·88) and in the second report this author verified that three accessions of Pguajava (K-10, A-06 and J-16) were resistant to Mincognita (GI < 2); however, M. enterolobii may have been the species challenged.

Another important aspect is the difficulty of identifying species of Psidium. In some studies, Pfriedrichsthalianum was identified incorrectly as ‘Brazilian guava’, although it is not native to Brazil. The Brazilian guava was identified as Pguineense (C. Proença, Universidade de Brasília, Brazil, personal communication).

In the present study, Pcattleianum, Pfriedrichsthalianum, Prufum and Asellowiana were resistant to Menterolobii. Previous analysis using Psidium spp. accessions confirmed the resistance of Pcattleianum and Pfriedrichsthalianum (Cuadra & Quincosa, 1982; Casassa et al., 1997, 1998; Villota & Agudelo, 1997; Maranhão et al., 2003; Carneiro et al., 2007; Milan, 2007; Bogantes-Arias & Mora-Newcomer, 2010). Two accessions of Pguineense were classified as susceptible to Menterolobii, but having low RF (25·9 and 33) when compared to that of the control guava cv. Paluma (RF = 174) and accession 39 of Pguineense (RF = 201·9). Variation in susceptibility of Pguineense was seen following inoculation with Menterolobii (Maranhão et al., 2003; Costa et al., 2012). These accessions were classified from moderately resistant to highly susceptible (RF = 4·76–38·89) and resistant to susceptible (RF = 0–3·47), respectively. The unique Pguineense accession tested was evaluated as immune (Lee et al., 1998).

The results of successful grafting are in accordance with those obtained by other authors (Villavicencio et al., 1995; Marin et al., 2000; Bogantes-Arias & Mora-Newcomer, 2010). Vegetative growth was significantly higher in the guava rootstock treatment when compared with Pfriedrichsthalianum rootstock. A similar observation was made by Bogantes-Arias & Mora-Newcomer (2010) who reported that in the field, guava cv. Paluma grafted on guava rootstock showed more vigorous growth than guava grafted on Costa Rican rootstock, which clearly showed a strong dwarfing effect, whereas Brazilian guava rootstocks (probably Pguineense) exhibited an intermediate effect on vegetative vigour. Number and weight of fruits was higher when cultivated guava was used as a rootstock compared to wild guava rootstocks; no differences were observed between two wild guava rootstocks (Bogantes-Arias & Mora-Newcomer, 2010). Recently, interspecific hybrids between Pguajava × Pguineense selected for resistance to Menterolobii showed high compatibility with guava cv. Paluma, indicating that this strategy may be relevant in controlling Menterolobii (Costa et al., 2012).

Host resistance mechanisms act to prevent Meloidogyne development in different plants. The hypersensitive response (HR) is the most common, manifesting either as a pre- or post-infection event, associated with a rapid host cell death surrounding initial infection sites by the pathogen. As a result, the pathogen is arrested and its development is partially or completely inhibited (Williamson & Hussey, 1996; Williamson & Kumar, 2006). Hypersensitive-like reactions have been observed in Mi-mediated resistance in tomato (Williamson & Hussey, 1996), Mex-1 in coffee (Anthony et al., 2005), Me7 in pepper (Pegard et al., 2005) and on wild peanut–Marenaria (Proite et al., 2008), Vitis spp.–Marenaria (Anwar & McKenry, 2002) and coffee–Mincognita (Albuquerque et al., 2010) interactions.

In the resistant accessions Pcattleianum and Pfriedrichsthalianum, root penetration by the nematode was not affected, and similar numbers of J2s could be observed in the susceptible guava cv. Paluma and resistant accessions. Similar observations were reported for the moderately resistant cotton cv. Clevewilt accession (McClure et al., 1974), in the highly resistant accession M-315 RNR (Jenkins et al., 1995), and in a number of other resistant accessions (Faske & Starr, 2009).

Interestingly, in the current study there was no evidence of early HR in Pcattleianum and Pfriedrichsthalianum. In fact, the nematode was able to initiate and maintain apparently healthy giant cells in resistant roots for about 2 weeks before visible signs of deterioration occurred, leading to giant cell collapse. A similar mechanism was observed in cowpea roots containing the Rk gene (Das et al., 2008). Other authors also reported a delayed resistance against RKN in tobacco (Powell, 1962) and in pepper HDA330 (Bleve-Zacheo et al., 1998), in which a late HR was seen in developed giant cells. In the present study, no HR-like phenotype was observed in the resistant guava, even during the late stage of infection. During this time, nematodes were able to feed and develop normally until reaching the J4 stage. Infections associated with normal feeding site development, without any HR, in resistant plants have also been reported in nematode interactions with cotton, Cucumis and cowpea (McClure et al., 1974; Walters et al., 2006; Das et al., 2008).

Male sex conversion when juveniles cannot establish appropriate feeding sites is common when nutritional and environmental conditions are not favourable to nematode development (Fassuliotis, 1970; Williamson & Hussey, 1996; Pofu & Mashela, 2011). In this study, the presence of males can be explained by the fact that alteration of giant cells strongly reduced their metabolic activity, providing suboptimal nutrition for female nematode development up to 28 dai. No males were observed in the proximity of roots or in the soil, indicating that they died inside the roots.

This study contributes to the provision of a strong foundation for the development of resistant guava rootstocks in Brazil and other regions, to be used in the control of Menterolobii under field conditions. The microscopic observations suggest that resistance of wild Psidium accessions results from late defence responses with no apparent HR, but severe alteration of feeding sites associated with a lack of female development and egg production.

Acknowledgements

This work was funded by CNPq (Conselho Nacional de Desenvolvimento Científico e Tecnológico), MAPA (Ministério da Agricultura, Pecuária e Abastecimento) and Embrapa Cenargen (Embrapa Recursos Genéticos e Biotecnologia): project 64. V. R. C. thanks CAPES for his scholarship.

Ancillary