Morphological and molecular characterization of Pratylenchus species from Yam (Dioscorea spp.) in West Africa

Morphological characterization

Nematodes from 27 samples fixed in formalin were processed to anhydrous glycerin following the glycerin-ethanol method (Seinhorst, 1959) as modified by De Grisse (1969). Permanent slides were prepared and used to record morphometrics and morphological features (Castillo and Vovlas, 2007; Inserra et al., 2007) using an Olympus BX51 DIC microscope equipped with a Nikon digital camera. Additional morphological and morphometrical data were recorded from temporary slides made from DESS fixed specimens, before DNA extraction (see Table 1).

Molecular characterization

Following morphological identification, the same individual nematodes were picked from temporary slides and used for extraction of genomic DNA using a quick alkaline lysis protocol (Janssen et al., 2016). DNA was amplified by preparing 24 μl PCR master mix comprising 16 μl double sterilized distilled water, 2.5 μl 10x buffers, 2 μl MgCl2, 0.05 μl of dNTP (10 mM), 1 μl of reverse and forward primers, 0.05 μl of Toptaq and 2 μl of nematode template DNA. The primer set D2A (5’–ACA AGT ACC GTG AGG GAA AGT TG–3’) and D3B (5’–TCG GAA GGA ACC AGC TAC TA–3’) (Subbotin et al., 2006) was used for amplification of the D2-D3 expansion regions of 28 S rDNA gene and the cytochrome c oxidase subunit 1 (COI) gene fragment was amplified using the primer set JB3Prat (5’–TTT TTT GGG CAT CCT GAA GTC TAT–3’) and JB4Prat (5’–CCT ATT CTT AAA ACA TAA TGA AAA TG–3’) following DNA amplification profile described in Kolombia (2017).

PCR products were electrophoretically separated on a 1% agarose gel and stained with ethidium bromide. PCR products were purified using the Wizard® SV Gel and PCR Clean-Up System Kit (Promega, the Netherlands) as described in the manufacturer’s instructions and sequenced by Macrogen Inc. (the Netherlands) in forward and reverse directions. Consensus sequences were assembled using GENEIOUS 9.15 (Biomatters; http://www.geneious.com) and deposited in the NCBI GenBank (Table 1).

Phylogenetic analysis

Both D2-D3 of 28 S rDNA and COI of mtDNA sequence datasets were aligned using MUSCLE (Edgar, 2004) with default settings. Outgroup taxa of each dataset were chosen based on previously published data (Subbotin et al., 2008; Liu et al., 2016). The best fit models of DNA evolution were estimated using the program jModeltest 0.1.1 (Posada, 2008) under the Akaike information criterion (AIC). Bayesian phylogenetic analysis (BI) was undertaken using MrBayes 3.2.6 for 1 × 10⁶ generations with a general time-reversible model with a gamma distribution for the remaining sites (GTR + I + G), four runs, 20% burn-in, and subsampling frequency of 500 generations (Huelsenbeck and Ronquist, 2001) for both D2-D3 and COI.

Results

Occurrence and morphological characterization of Pratylenchus spp. from yam

From the rhizosphere, Pratylenchus spp. were detected in 48 samples (59%) collected in Ghana (Fig. 1A) and 45 samples (39%) in Nigeria (Fig. 1B). The density of Pratylenchus spp. from the rhizosphere varied from 2 to 704 individuals per 100 ml soil and roots in Ghana, and from 2 to 398 individuals in 100 ml of soil and roots in Nigeria. From 400 tubers examined, Pratylenchus spp. were recovered from just 6% of the 400 tuber peels (Figure 1C). Twenty-four tubers were infected with Pratylenchus spp., of which, 19 tubers (79%) also had galling and crazy root damage caused by the root-knot nematode (Meloidogyne spp.), 4 tubers (17%) showed dry rot symptoms caused by the yam nematode (Scutellonema bradys) while no symptoms were observed in one tuber, which had a density of 50 specimens of Pratylenchus brachyurus, per 5 g of yam peels (Figure 1D). Densities of Pratylenchus spp. were as higher as 340 nematodes in tubers with symptoms and up to 525 individuals per 5 g of yam peels in tubers with dry rot and galling, respectively. Twenty-eight populations from 12 yam tubers and 16 rhizosphere samples were studied using morphological and molecular data, which resulted in the identification of Pratylenchus brachyurus and P. hexincisus (Taylor and Jenkins, 1957) and P. zeae (Graham, 1951).

Figure 1:

Pratylenchus brachyurus was the most prevalent RLN species in Ghana and Nigeria, present in 11 of the 12 tubers used for species identification and 88% of Pratylenchus-positive rhizosphere samples. Twenty-five specimens per 5 g of yam peels of Pratylenchus hexincisus were recovered in just one tuber showing galls from Nigeria, and P. zeae was detected in 12% of the rhizosphere samples from Ghana (26 nem/100 ml soil) and Nigeria (3 nem/100 ml soil).

Systematics

Pratylenchus brachyurus Godfrey, 1929 Filipjev and Schuurmans Stekhoven, 1941 (Figures 2 and 3; Tables 2 and 3).

Table 2.

Measurements of thirteen Pratylenchus brachyurus populations from Ghana. All measurements are in μm and in the form: mean ± s.d. (range).

Molecular characterization of Pratylenchus spp. from yam

The D2-D3 of 28 S rDNA gene

The D2-D3 alignment included 80 Pratylenchus sequences, and two outgroup sequences. Thirteen new D2-D3 sequences were obtained in the present study. Following the numbering proposed by Subbotin et al. (2008), the BI tree contained five highly supported clades except for clade III (Figure 6).

Figure 6:

The sequences of P. hexincisus generated in this study formed a very well supported clade without internal resolution with P. hexincisus sequences from China (MT362902 and MT362903), P. hexincisus sensu Inserra et al., 2007 obtained from the type locality (DQ498832 and DQ498833), P. scribneri Steiner in Sherbakoff & Stanley, 1943 (EU130864, EU130865, JX047001 and KM094196) and P. agilis (Thorne and Malek, 1968) (EU130841). However, sequences of P. scribneri sensu Inserra et al. (2007) (DQ498830) and P. scribneri from California U47554 (Al-Banna et al., 1997) formed a separate clade. The intraspecific variation of our P. hexincisus populations was 1–2 bp (0.1–0.3%) and differed only 0–2 bp (0–0.3%) with P. hexincisus from the type location (Inserra et al., 2007) (DQ498832 and DQ498833) and 0–3 bp (0–0.4%) with P. agilis (EU130841) and 1–5 bp (0.1–0.6%) with P. scribneri (EU130864, EU130865, JX047001 and KM094196), while it was clearly different (14–17 bp, 2.5–5.7%) from P. scribneri sensu Inserra et al. (2007) (DQ498830).

Sequences of P. brachyurus from this study, together with P. brachyurus sequences from GenBank were grouped in a well-supported subclade C of the clade III. The intraspecific variation of P. brachyurus was 2–51 bp (0.3–6.6%) and nucleotide difference between P. brachyurus and the most similar sequence, P. penetrans, was 152–177 bp (19.5–23%).

Pratylenchus zeae sequences formed a well-supported clade together with P. zeae sequences from GenBank. The intraspecific sequence variation of P. zeae was 23–65 bp (3.2–9%) and the interspecific sequence difference with the closest related species, Pratylenchus sp. (JX261959), was 23–80 bp (3.2–11.1%).

The mitochondrial COI gene

The COI sequences alignment was 422 bp in length and included 58 sequences of Pratylenchus including eight newly generated sequences, and four outgroup taxa (Meloidogyne, Hirschmanniella, Pratylenchoides and Radopholus). The BI tree contained five highly supported clades following numbering proposed by Subbotin et al. (2008) (Figure 7).

Figure 7:

Sequences of P. hexincisus from yam formed a well-supported clade with P. hexincisus sequences from China, Italy and the USA and P. scribneri sequences from China and the USA, with P. loosi (PP 0.84) as sister species. The sequences of P. hexincisus generated in this study and P. hexincisus sequences from Italy (KY828322) and China (KY828321) and P. scribneri (MK877999: USA; MK878000: USA; MK878268: USA, KY424093: China; KY424090: China; KY424089: China; KX349425: China) were very similar 0–8 bp (0–1.93%). However, these sequences were different from the recently deposited P. hexincisus sequences from Wheat and Corn in the USA (MK877467, MK877469, MK877471, MK877482, MK877492) with 51–81 bp (19.8–21.1%). Sequences of the closest related species, P. loosi, differed 54–102 bp (19.1–24.5%).

Sequences of P. brachyurus from this study, together with other P. brachyurus sequences available in the NCBI GenBank database formed a well-supported subclade C of clade III, sister to P. oleae (clade IV) (Palomares-Rius et al., 2014). The intraspecific variation of P. brachyurus was 0–16 bp (0–4.1%) and the interspecific sequence difference between P. brachyurus and P. oleae was 78–81 bp (21.1–22%).

Pratylenchus zeae sequences formed a well-supported clade (VI) together with P. zeae sequences from GenBank. The intraspecific sequence variations of P. zeae were 0–37 bp (0–9.6%) and the interspecific sequence difference was 99–112 bp (25.9–28.6%) with P. parazeae, the closest related species.

Discussion

Prior to the current study, seven RLN species, i.e. P. brachyurus, P. crenatus (Loof, 1960), P. coffeae, P. loosi (Loof, 1960), P. sudanensis, P. pseudopratensis and P. zeae have been reported from yam rhizosphere and yam tubers (Caveness, 1967; Bridge, 1973, 1988; Coyne et al., 2003; Varghese and Mohandas, 2004; Bridge and Starr, 2007; Mudiope et al., 2007; Osei et al., 2015; Coyne et al., 2018). Using a combination of morphological and molecular identification, P. brachyurus and P. hexincisus were identified from yam tubers, while P. zeae was recovered from the yam rhizosphere only. Pratylenchus brachyurus, a cosmopolitan species, appears as the predominant species on yam in Nigeria and Ghana, which is in agreement with other studies that have reported P. brachyurus from yam in Nigeria and West Africa (Luc and de Guiran, 1960; Unny and Jerath, 1965; Caveness, 1967; Bridge, 1972, 1973). In this region, the polyphagous P. brachyurus has also been recorded as a pest of numerous crops (Miège, 1957; Luc and de Guiran, 1960; Bridge, 1973; Egunjobi, 1974; Egunjobi and Larinde, 1975; Guerout, 1975; Coyne et al., 1999; Castillo and Vovlas, 2007), including an interception from Colocasia sp. (another tuber crop) from Nigeria to China (Zhao et al., 2011). Also, it is known to affect plant growth and the yield of crops in West Africa, for instance on pineapple (Guerout, 1975) and cassava (De Guiran, 1965).

Pratylenchus species, and in particular P. coffeae are known to cause “dry rot” on yam tubers, a condition similar to that caused by S. bradys, based on what is known for P. coffeae and P. sudanensis (Bridge et al., 2005; Bridge and Starr, 2007; Coyne and Affokpon, 2018). However, symptoms of P. brachyurus or its effects on yam production are not well known, Given the predominance of P. brachyurus in yam tubers and yam rhizosphere, it appears that this species is a major RLN on yam in West Africa. However, more work is necessary to clearly establish the effect of this species on yam growth, yield and tuber quality. The ability of P. brachyurus to survive a long period without a host and its polyphagous nature, could make its management particularly difficult, without the use of resistant cultivars.

Pratylenchus zeae, retrieved only from the yam rhizosphere of one sample in Ghana and one in Nigeria, is a commonly occurring species on other crops in West Africa (Fortuner, 1976; Plowright and Hunt, 1994; Coyne et al., 1996; Castillo and Vovlas, 2007). Pratylenchus zeae was reported on yam in Nigeria (Bridge, 1973) but has never been reported on yam rhizosphere in Ghana. Its absence from tuber tissue, however, indicates that yam tubers may not support P. zeae and that its occurrence in this case may be related to other plant species occurring together with the sampled yam.

Pratylenchus coffeae, one of the major plant-parasitic nematodes of yam in the Americas and the Pacific Islands was not recorded in any of the samples collected from Ghana and Nigeria. A similar observation was reported by Kwoseh et al. (2005) in Ghana. This remarkable absence from P. coffeae supports the statement of Duncan and Moens (2013) that “P. coffeae is a pest of yam, interestingly, not in Africa”, despite being present on other crops in both localities (Duncan et al., 1999; Pourjam et al., 1999; Speijer et al., 2001; Bridge et al., 2005; Kwoseh et al., 2005; Coyne and Affokpon, 2018). Although Osei et al. (2015) recorded P. coffeae on yam in Ghana, but its identity was not ascertained by molecular method.

Traditional taxonomy can have serious limitations for differentiating species of Pratylenchus (Luc, 1987; Subbotin et al., 2008). In the current study, however, populations of P. brachyurus were, despite a remarkable intraspecific variation, relatively easily identified based on morphology and morphometrics, including the number of lip annuli (2), stylet length (17–21 µm), vulva position (77–88%), and a bluntly rounded tail, which while highly variable was never conically pointing, posteriorly confirmed by molecular data. Our observations agree with the descriptions provided by Corbett (1976) and Castillo and Vovlas (2007), including its well-known intraspecific variation on the tail, lips and knobs shape (Roman and Hirschmann, 1969; Corbett, 1976; Tarjan and Frederick, 1978; Payan, 1989). However, the presence of a developed sperm-filled spermatheca in 2 of the 108 analysed specimens was observed for the first time.

In this study, P. hexincisus was recorded from yam for the first time, although just from one sample. The relatively high number of specimens retrieved from yam peels unequivocally demonstrates its association with the tuber. Therefore, infection studies to prove Koch’s postulates are required to demonstrate that P. hexincisus is a pest of yam. Pratylenchus hexincisus recorded from Benue State, Nigeria, is morphologically very similar to P. scribneri, which has been reported on maize in the neighbouring Western region of Nigeria (Anonymous, 1975). Molecular researches are necessary to establish if both species represent two different species or are conspecific.

The observed morphology and morphometrics of P. hexincisus agreed with the original description (Taylor and Jenkins, 1957), although variability in the number of lines in the lateral field, with four to six lines have been observed. This variation is known for P. brachyurus (Castillo and Vovlas, 2007) as well as other members of the genus Pratylenchus, for example four to seven lines in the lateral field have been reported in P. neglectus (Rensch, 1924) Filipjev and Schuurmans Stekhoven, 1941 (Corbett and Clark, 1983) and four to six lines in P. scribneri (Roman and Hirschmann, 1969; Loof, 1985; Inserra et al., 2007). Originally described from corn in Maryland, USA (Taylor and Jenkins, 1957), P. hexincisus was distinguished as a new species, separate from P. scribneri, by the presence of 6 lines in the lateral field, its smaller size and the fact that no spermatheca was observed. However, morphological studies of both species have revealed high morphological similarities, including the presence of empty spermatheca and variation in the number of lines in the lateral field (Roman and Hirschmann, 1969; Corbett and Clark, 1983; Castillo and Vovlas, 2007; Inserra et al., 2007; Ozbayrak et al., 2019). Yet, a comprehensive investigation by Inserra et al., (2007), including P. hexincisus from the type locality (DQ498832-33) and a reference population of P. scribneri (DQ498830), showed a molecular distinction between P. hexincisus and P. scribneri, based on the D2-D3 region of the 28 S rDNA. Moreover, this reference P. scribneri population (DQ498830) also formed a distinct clade with P. scribneri (U47551) reported by Al-Banna et al. (1997). Furthermore, the authors also highlighted morphological characters that could discriminate both species, including the presence of a rectangular-elongated spermatheca in P. hexincisus versus rounded in P. scribneri. However, although the D2-D3 sequences of the isolates of yam were virtually identical to the sequences of P. hexincisus (Inserra et al., 2007), our population did not show an elongated spermatheca but always a rounded spermatheca. On the other hand, our population showed crenated outer incisures of the lateral field (Figure 4) as described in the original description of P. hexincisus (Taylor and Jenkins, 1957). The number of lateral lines appeared to be less discriminative to distinguish P. hexincisus and P. scribneri as six lateral lines were also observed in P. scribneri (Inserra et al., 2007). Similar variability on the lateral lines was observed in P. brachyurus reported in this study indicating that caution is needed when using this character to discriminate species in the genus Pratylenchus.

Although the D2-D3 sequences of P. hexincisus in our study were similar with those of P. scribneri from Imperial Valley, California and Vero Beach, Florida (EU130864-65) the uncertainty of the identification of the isolates was already mentioned by Subbotin et al. (2008) and these sequences may therefore represent P. hexincisus.

Augmenting the problem to delimitate both species, the COI sequences P. hexincisus (MK877467, MK877469, MK877485, MK877471, MK877482, MK877492) in the study reported by Ozbayrak et al. (2019) were clearly different from P. hexincisus from yam when assessed in the current study. Remarkably, our P. hexincisus COI sequences were similar to the sequences of P. scribneri (MK877999, MK878000, MK878268) Ozbayrak et al. (2019). Also, P. scribneri D2-D3 and COI sequences from China (JX047001, KM094196, KX349425) were similar to our P. hexincisus sequences and these Chinese D2-D3 sequences are also similar to those from the P. hexincisus type locality (Inserra et al., 2007). The unclear identity of the COI sequences could be resolved if the materials from Al-Banna et al. (1997) and Inserra et al. (2007) could be linked to COI sequences, i. e. if the P. hexincisus and P. scribneri COI sequences sensu Ozbayrak et al. (2019) agree with the identity of the material from Al-Banna et al. (1997) and Inserra et al. (2007). However, this is likely not the case since, according to the supplementary D2D3 tree in Ozbayrak et al. (2019), their P. scribneri sequences are different to the P. scribneri (DQ498830) from Inserra et al. (2007).

In summary, P. hexincisus and P. scribneri have similar, indeed overlapping, morphometric characteristics and shared morphological characters, leading to a confuse and difficult identification. Hence, a topotype population of P. scribneri is needed to solve the identity and validity of P. scribneri and P. hexincisus, as suggested by Inserra et al. (2007) and Subbotin et al. (2008). Likewise, the D2-D3 sequence of P. agilis Thorne and Malek, 1968 is also similar to P. hexincisus sensu Inserra et al. (2007), as provided and mentioned by Subbotin et al. (2008). Loof (1978) had already doubted the validity of P. agilis and the species was considered as species inquerendae (Frederick and Tarjan, 1989). This was confirmed by ITS sequences and isozyme analysis. Pratylenchus agilis was proposed as a junior synonym of P. scribneri (Hernández et al., 2000), although Waeyenberge et al. (2000) indicated differences between P. scribneri and P. agilis with respect to ITS-rDNA length and the RFLPs.

Evidently, additional morphological and molecular characterizations are required to further analyses the species group of P. scribneri, P. hexincisus and P. agilis.