Introduction

Nematodes belonging to the Longidoridae family, together with Trichodoridae, are the only representatives of the class Enoplea that have a phytoparasitic habit. These nematodes are vermiform migratory ectoparasites that feed on the roots of plants, causing damage such as darkening and reduction in growth of the root system, deformations in the root apex (galls) and collapse of the cortex (Cohn 1970; Taylor and Brown 1997). However, their agricultural importance is augmented due to the capacity of some species of Longidorus, Paralongidorus and Xiphinema to transmit various Nepoviruses, which cause diseases in a wide range of fruit and vegetable plants (Taylor and Brown 1997).

In Brazil, studies on the diversity of Longidoridae are relatively scarce. The occurrence of these nematodes, especially Xiphinema spp., has been reported in association with various cultivated and wild plants, but in most of these reports the nematodes are generally identified only to genus level (Lehman et al. 1977; Moura and Almeida 1981; Sharma 1982; Rossi and Ferraz 2005; Castro et al. 2008). Out of 260 valid Xiphinema species (Decraemer and Geraert 2013), only 11.15 % have been reported in Brazil (Doucet et al. 1998; Siddiqi 2000; Oliveira et al. 2003) and five Xiphidorus species (Doucet et al. 1998; Oliveira et al. 2003). Among these species, Xiphinema index Thorne & Allen, 1950 and X. californicum Lamberti and Bleve-Zacheo 1979 are potential transmitters of viruses (Decraemer and Geraert 2013).

In contrast, approximately twice as many Xiphinema species are reported, for example, in South Africa (Lamberti et al. 2000; Heyns and Swart 2002), a country one-seventh the size of Brazil. This discrepancy reflects the lack of research in Brazil, which can be explained by the scarcity of taxonomists who are dedicated to characterizing and describing new species or even accurately identifying the species. Thus, reports on Longidoridae are undercounted not only in Brazil but throughout Latin America and reflect mainly the distribution and interests of individual nematologists, and not the biodiversity of the group (Doucet et al. 1998; Oliveira 2004).

Minas Gerais is one of the largest Brazilian states, with 586,528 km2 and occupying 7 % of the national territory. In this vast and diverse area, only a few surveys reported occurrence of Longidoridae, which were identified solely to genus level. Apart from isolated reports of new findings, only two national surveys of Longidorids have been published (Ferraz 1980a; Oliveira et al. 2003). These surveys included only a few samples from Minas Gerais, but the occurrence of X. elongatum Schuurmans Stekhoven & Teunissen, 1938, X. krugi Lordello, 1955, X. surinamense Loof and Maas 1972 and X. setariae/vulgare complex was reported. However, Ferraz (1980b) conducted a broader study in Minas Gerais to evaluate the occurrence of phytonematodes, and X. brevicolle Lordello & Costa, 1961, X. brasiliense Lordello 1951 and X. paritaliae Loof and Sharma 1979 were also recorded. Therefore, there is no reference to Longidoridae associated with a large number of plant species in the state of Minas Gerais.

In this work, we report results from a survey conducted in 12 municipalities in the state of Minas Gerais, followed by morphometric analysis, geographical distribution and identification of host plants in association with the species of Longidoridae.

Materials and Methods

Sampling

From 12 municipalities in Minas Gerais, 126 soil samples were collected from different types of vegetation, including forest, Cerrado and crops, from February to May 2010. The sampling points were geo-referenced and the plant species identified (Table 1). Each sample consisted of approximately 1.5–2.0 kg of soil collected at four soil cores drilled in the area covered by each plant canopy, at a depth of 0–30 cm. The soil was kept in a polyethylene bag within a polystyrene box and immediately transported to the laboratory and kept chilled at 4 °C.

Table 1 Information for the samples where Longidoridae were found associated with plants in municipalities of Minas Gerais State, Brazil
Full size table

Extracting, fixing and mounting nematodes

The nematodes were extracted from a sample of 1 kg that had been divided into four aliquots. These consisted of 250 cm3 of soil, processed sequentially, according to the modified method of decanting and sieving combined with modified Baermman (Ploeg and Brown 1997). Nematodes were examined under a stereoscopic microscope and longidorid specimens were removed for morphological and morphometrical studies. The longidorids were heat-killed at 60 °C, fixed in a 1 % triethanolamine and formalin (TAF) mixture, and processed to anhydrous glycerine using a slow method (Hooper 1986). Species identification and measurements were made using a high power microscope with the aid of a camera lucida.

Identification

To identify the Longidoridae species, taxonomic keys proposed by Loof and Luc (1990), Oliveira et al. (2003), Lamberti et al. (2004) and Oliveira and Neilson (2006) were used, as well as original descriptions of the species from the literature.

Principal component analysis of Xiphinema krugi populations

Principal Components Analysis (PCA) was performed on the correlation matrix of the set of ten measurements [length of body (L), length of both odontostyle and odontophore, tail length, body diameter at anus, largest body diameter, position of the vulva in relation to the anterior end of the body expressed as a percentage of the length of the body (V %), body length divided by largest body diameter (a), tail length divided by body diameter at the anus (c’) and ratio of the length of the body and tail (c)] of 255 females taken of X. krugi populations (44 populations in total). In addition, an analysis of variance (ANOVA) was carried out to test statistical significance to groups observed on PCA plot. GenStat for Windows, 16thEdition (VSN International, Hemel Hempstead, U.K.) was used to perform the statistical analyses.

Results

Distribution and morphometry

Eight species of Xiphinema (X. brasiliense, X. diffusum Lamberti and Bleve-Zacheo 1979, X. elongatum, X. ensiculiferum (Cobb, 1913) Thorne, 1937, X. krugi, X. variegatum Siddiqi 2000, X. setariae/vulgare complex and X. surinamense and two species of Xiphidorus (Xiphidorus sp. and X. amazonensis Uesugi et al. 1985) were found associated with a broad range of plant species (Table 2), with the most frequent species being X. krugi (31.7 % of all samples) and X. variegatum (30.9 %). Two species, X. diffusum and X. amazonensis, were associated only with cultivated plants, present in the municipalities of Viçosa and Jaíba, respectively, while the other species were associated with cultivated plants and a wide range of naturally occurring arboreal species, in various localities (Table 2). On the other hand, X. variegatum, X. surinamense and Xiphidorus sp. were geographically restricted to natural forest vegetation (Atlantic Rainforest).

Table 2 Longidoridae species associated with different host plants in areas with natural and cultivated vegetation in the municipalities of Minas Gerais State, Brazil
Full size table

Based on the morphometric data of the identified species (Tables 3, 4, 5 and 6), individuals of the X. krugi populations varied sharply in body size (L) (averages ranging from 1.6 to 2.7 mm) and in the shape of the tail (Table 3; Fig. 1). The shape of the tail in all populations was typical of the species with a distinct ventral peg, but it showed a slight variation, where the female tail was hemispheroid with only a moderately developed bulge. Morphometric variation was also observed in X. elongatum with emphasis on the averages of body length L (2.4–3.8 μm), odontostyle (96–127 μm), and odontophore sizes (59–76 μm) (Table 5). The populations of X. ensiculiferum stood out for the mean values of L, which ranged from 2.7 to 2.8 mm, being higher than those reported in the literature (Table 4).

Table 3 Morphometric data of 11 representative populations of Xiphinema krugi present in municipalities of Minas Gerais State, Brazil. Data are expressed as mean ± standard deviation (range)
Full size table
Table 4 Morphometric data of Xiphinema variegatum, X. surinamense, X. diffusum and X. ensiculiferum present in municipalities of Minas Gerais State, Brazil. Data are expressed as mean ± standard deviation (range)
Full size table
Table 5 Morphometric data of Xiphinema elongatum and X. brasiliense present in municipalities of Minas Gerais State, Brazil. Data are expressed as mean ± standard deviation (range)
Full size table
Table 6 Morphometric data of Xiphidorus sp. and X. amazonensis present in municipalities of Minas Gerais State, Brazil. Data are expressed as mean ± standard deviation (range)
Full size table
Fig. 1
figure 1

Photomicrograph of the anterior and posterior region of females and males of Xiphinema krugi. a-b: females of X. krugi (sample 01, Lagoa Grande); c-f: female and male of X. krugi (sample 73, Viçosa); b, d and f: different tail shapes observed in populations of X. krugi. Bars correspond to 30 μm

Full size image

Principal component analysis (PCA) of X. krugi populations

A plot of the first two principal components (Fig. 2; Table 7) indicated that individuals taken from the Sample 125 tended to have larger scores than those from other habitats on the second principal component. The first and second principal components accounted for 32.8 and 25.7 % of the variation, respectively.

Fig. 2
figure 2

Principal Component scores plot of PC1 against PC2 for a PCA of ten measurements of the sampled Xiphinema krugi. The symbols are for the nematode Habitats: Sample 125 = open white circles; Cultivated = open white triangles; Cerrado = filled black circle; Atlantic Rainforest = plus symbol

Full size image
Table 7 Principal component loadings generated by PCA analysis. The characters are ranked according to the magnitude of loading in PC2
Full size table

An analysis of variance (ANOVA) model considering location (municipality) as a factor to test for differences among the four locations within the cultivated habitat showed highly significant differences amongst the locations (p < 0.001). Contrasts were also fitted to test the pairwise difference between Sample 125 and the other locations each of which was highly significant (p < 0.001).

Having established that Sample 125 was significantly different from the other cultivated locations, these samples were now tested against those from the other habitats. For this ANOVA, the treatment factor was habitat where in addition to ‘Rainforest’, ‘Cerrado’ and ‘Cultivated’ we added Sample 125 as an additional level of the factor. The differences amongst these groups were also highly significant (p < 0.001). Contrasts for the pairwise comparison of Sample 125 against the other habitats also confirmed the statistical difference (p < 0.001).

Discussion

Previously, seven species of Xiphinema had been reported from Minas Gerais State, Brazil (X. elongatum, X. krugi, X. surinamense, X. setariae/vulgare, X. brevicolle, X. brasiliense and X. paritaliae) and only X. brevicolle and X. paritaliae were not recorded during the present study. Collectively, results from our study and the previous reports, showed that ten species of Xiphinema occurs in Minas Gerais, including X. diffusum, X. ensiculiferum and X. variegatum, which constitute their first report in the state. None of them is known to transmit viruses.

The majority of species reported in this study are known as of cosmopolitan distribution. Xiphinema diffusum has been recorded in many countries in Africa and Asia (Lamberti and Bleve-Zacheo 1979; Lamberti et al. 1991), in the USA (Robbins 1993), Easter Island, and Chile (Lamberti et al. 1991). The first report in Brazil was in 2003 (Oliveira et al. 2003). This nematode is one of Xiphinema americanum group species more widely distributed in the world (Lamberti et al. 1991).

Xiphinema brasiliense was originally described from the rhizosphere of potato plants collected from Sapecado (SP), southern Brazil (Lordello 1951) and subsequent studies have confirmed its widespread distribution on Brazilian soils (Loof and Sharma 1979; Lamberti et al. 1987; Germani 1989; Oliveira et al. 2003). Nevertheless, it was found only in Viçosa municipality associated Atlantic Rainforest vegetation. Likewise, Xiphinema elongatum is widespread throughout the Brazil (Ferraz 1980a; Lamberti et al. 1987; Costa Manso et al. 1994; Oliveira et al. 2003). In our survey, this species was found in association with Prunus domestica, Macadamia sp. and Cerrado vegetation.

Xiphinema ensiculiferum was found only in two samples collected in Viçosa municipality, this being the first record of the state of Minas Gerais. In Brazil, this nematode had been previously associated with banana in Bahia State (Sharma and Sher 1973; Sharma and Loof 1984), unidentified native plants in Maranhão State (Ferraz et al. 1989), from soil collected from the northern bank of the Rio Negro, in Amazonas State (Marais et al. 1995) and natural vegetation in Mato Grosso State (Oliveira et al. 2003).

Xiphinema vulgare is considered a junior synonym of X. setariae (Cohn and Sher 1972; Loof and Luc 1990; Heyns and Coomans 1991). However, Lamberti et al. (1995) rejected this synonymisation and retained both species as distinct and valid. Here, for clarity of nomenclature, these nematodes were referred as X. setariae/vulgare complex. In the present survey, specimens of the X. setariae/vulgare complex were found associated with Cerrado vegetation, Litchi chinensis, Araucaria angustifolia, Hibiscus rosa-sinensis and Saccharum officinarum.

In this study, X. surinamense was found only in the municipality of Araponga associated arboreal species in an Atlantic Rainforest preservation area. However, this nematode has been reported in different habitats in many Brazilian States (Loof and Sharma 1979; Ferraz 1980a; Lamberti et al. 1987; Oliveira et al. 2003). Also, X. variegatum was found only associated with plants from Atlantic Rainforest, and occurred in a mixed population with X. surinamense in a sample from Araponga municipality. Its morphological similarity with X. surinamense hampered the identification of these two species, which is based on slight morphometric differences.

Xiphinema krugi and X. variegatum were the most widespread species. The frequent occurrence of X. krugi diverges from that reported by Ferraz (1980b), in which X. brevicolle (probably, in fact, X. diffusum) was the most frequent species of Longidoridae in Minas Gerais. In a survey by Oliveira et al. (2003) in various Brazilian states, X. krugi was prevalent.

The constant association of X. variegatum with Atlantic Rainforest vegetation observed in this work corroborates the suggestion of Oliveira et al. (2003) that X. variegatum may be well adapted to tropical forest conditions.

Two species of Xiphidorus (X. amazonensis and Xiphidorus sp.) were found in this survey, and constitute the first records of this genus in Minas Gerais. Xiphidorus amazonensis was originally found in a lowland area on the island of Xiborena, near Manaus (AM), but it occurs in Venezuela associated with various cultivated plants, such as fruit trees, gramineous species, sugarcane and okra (Uesugi et al. 1985). In Minas Gerais state, it was only found associated with sugarcane. The constant association of Xiphidorus spp. with cultivated plants, especially sugarcane, suggests that these nematodes can cause damage to crops; however, further studies on its pathogenicity should be done in order to provide more evidence. The absence of Longidorus species in this study suggests that this genus is not widely distributed in Brazil, contradictory to what is reported by Doucet et al. (1998), but in agreement with Oliveira et al. (2003).

Based on morphology, the species X. diffusum and X. brevicolle are distinguished by minimal morphological and morphometrical differences. For many years, X. brevicolle was thought to be distributed worldwide. However, the study of Lamberti et al. (1991), which involved populations previously identified as X. diffusum and X. brevicolle, concluded that X. brevicolle distribution is restricted to South America, and that most reports of this species in the world should be considered as X. diffusum. One of the populations analyzed by Lamberti et al. (1991) was from Viçosa (MG), previously identified as X. brevicolle. Curiously, this population grouped with various other populations of X. diffusum, which indicates that it was erroneously identified. Thus, it is probable that X. diffusum was already present in MG, but that it had been confused with X. brevicolle and reported as such.

Some of the populations studied, such as X. elongatum, X. ensiculiferum and X. krugi, showed variability in some morphometric characters evaluated. The means for L in populations of X. krugi were higher than those presented by Luc and Hunt (1978). In the populations of X. ensiculiferum, the variation in the means of L (2.7–2.8 mm) shows a contrast, for example, with what was presented by Southey and Luc (1973), with mean L equal to 1.95 (1.78–2.15 μm) for the species. In X. elongatum this variation was observed in the body length L (2.4–3.8 μm) and the size of the odontostyle (96–127 μm) and odontophore (59–76 μm). These values are higher than those reported by Oliveira et al. (2003): 2.1 mm, 93 and 59 μm for the same characters.

Although species in this study were identified using taxonomic keys based on morphological characters, such identification has become a difficult task, because some of these species are morphologically similar, with very small morphometric differences. This difficulty was noted especially in the identification of X. variegatum and X. surinamense. These two species have very similar morphology, with a pseudomonodephic female reproductive system, a short hemispheric tail, and a rounded labial region continuous with the body outline (Loof and Maas 1972; Siddiqi 2000). Despite these similarities, the two species differ in the body size, with X. surinamense (>2.5 mm) being larger than X. variegatum (<2.4 mm) (Oliveira et al. 2003). Thus, determining the limits of morphometric variability within a species becomes necessary moreover that identification of Longidoridae is based both on morphological and morphometric characters. Thus, the difficulties to secure identification, opposite the morphologic and morphometric variability observed in longidorids, can be remedied with the application of molecular methods complementary to classical taxonomy.

Xiphidorus sp. stood out, mainly because of the rounded labial region which is continuous with the rest of the body, a character that is distinct from other species belonging to the genus. This, associated with the morphometric measurements, confirms it to be a new species for science.

The morphotype (X. krugi) found in sample 125 was characterized by larger values of most characters, particularly ‘a’ and ‘c’ apart from lower values of both body diameter characters. A great variation in several morphological and morphometric characters in X. krugi is well documented in Luc and Hunt (1978). Likewise, the study of 14 X. krugi populations reported by Oliveira et al. (2006), encapsulating both molecular and classical taxonomic data, has demonstrated the possibility that in fact X. krugi is a species complex comprised of four distinct genotypes and/or cryptic species that have a morphological basis. Future studies on X. krugi populations from Minas Gerais, such as the contemporary molecular analysis used by Oliveira et al. (2006) may provide more conclusive evidence of the observed morphometric variability and to assist with the resolution of taxonomic controversies such as X. krugi and putative evolutionary studies.

Xiphinema spp. and Xiphidorus spp. were associated with 49 different plant species, including cultivated plants, natural vegetation from the Atlantic Rainforest and Cerrado, as well as with unidentified plants. The results presented here expand significantly the knowledge on the diversity of family Longidoridae in soils of Minas Gerais soils. Xiphinema diffusum, X. variegatum and X. ensiculiferum and species from the genus Xiphidorus represent new records for Minas Gerais. Additionally, the morphometric variability observed in populations of some Longidoridae species increased the range of character variation for these species.