This research reports, for the first time, validation of microscopic identification keys of single infectious ovine GIN larvae from faecal samples (containing multiple species) through molecular analysis. The authors verified the identity of each extracted larvae, offering a unique perspective into the accuracy of existing morphometric and morphological identification keys. Besides, the novelty of this research can be found in the enhancement of such keys from a practical point of view, providing the most reliable traits for correct L3 identification.
Overall, microscopic identification of ovine GIN infectious larvae is best achieved through a combination of both morphology and morphometry. As the size characteristics of different species/genera of GIN larvae overlap, measurements alone are insufficient for accurate identification. Shape of head, tail and sheathed tail are key additional characteristics for the identification of GIN L3 [20]. Our findings, as well as current scientific literature, stress the importance of examining L3 in their totality for accurate identification and thus not solely relying on one characteristic in particular [7, 12].
The authors have found, in accordance with Bowman [35], that the easiest way to identify individual infectious GIN larvae of sheep is firstly based on their sheathed tail length, dividing encountered larvae into three main groups: short (A), Trichostrongylus spp. and T. circumcincta; medium (B), Cooperia spp. and H. contortus; and long (C), C. ovina and Oesophagostomum spp. (Fig. 1). Although morphometric parameters for each group are defined here, with minimal practice, larvae can easily be classified within their respective groups without the performance of microscopic measurement. Further identification of the GIN species/genera within each group is based on microscopic measurements as well as additional morphological characteristics. A global decision tree (based on our findings as well as current scientific literature) for the identification of the infectious larvae of the six common species/genera of ovine GIN is given in Fig. 3.
Gastrointestinal nematode species included within the first group (A, short sheathed tail) are challenging to differentiate as morphological differences between Trichostrongylus spp. (Additional file 2: Figure S1) and T. circumcincta (Additional file 2: Figure S2) are quite small [12, 17, 19, 22, 25]. In particular, infectious T. axei and T. circumcincta larvae are very similar, as T. axei larvae do not have tubercles on their caudal extremity [22]. Given this fact, misidentification is common and, ultimately, three larvae morphologically classified as being T. circumcincta within this research were genetically identified as T. axei.
Despite the frequent overlap in total body length between the larvae of the common GIN species in sheep, a rudimentary distinction between Trichostrongylus spp. and T. circumcincta can be made based on overall size (Fig. 2c). As previously reported [12, 19, 20, 22, 24, 25], and confirmed in this research, full body length of T. circumcincta is commonly greater compared to those of Trichostrongylus spp. (Trichostrongylus spp., 709.43 ± 30.43 µm; T. circumcincta, 795.32 ± 25.84 µm). As a general rule, a cut-off of 720 µm can be used for the identification of T. circumcincta [25]. Nevertheless, as different species of GIN require (slightly) different developmental conditions (e.g., temperature and moisture) [7, 17, 23], any specific faecal culture protocol could favour the development of one species over another. Although unlikely (based on the continuous reports of size difference between the two genera in question), any size difference here might thus have been the result of more favourable culture conditions for T. circumcincta.
Next, distinction between Trichostrongylus spp. and T. circumcincta larvae can be made based on the presence of an inflexion (“shoulder”) at the base of the cranial extremity of T. circumcincta (Fig. 4), as first described by Lancaster and Hong [37]. In instances where this inflexion can clearly be noted, this morphological characteristic offers an easy distinction between the two types of larvae. However, the authors agree with Roeber and Kahn [7] that this morphological feature is rather subjective and thus easily missed. Besides this, the head of infectious T. circumcincta larvae is rather flat and square compared to the rounded shape of Trichostrongylus spp. [20], although most likely only noticeable for the experienced parasitologist.
Lastly, most species within the genus Trichostrongylus (except T. axei) have tubercles (Fig. 5) on their caudal extremity (tail), while these are absent in all other infectious GIN larvae of sheep, including T. circumcincta [12, 19, 22, 24]. Usually, these tubercles are readily visible when examining ensheathed larvae under high magnification (×400), and for T. colubriformis and T. vitrinus, 1–3 tubercles can be noted [22]. Even though this trait cannot be applied for definite differentiation between the two genera within group A, visible tubercles on the caudal extremity allow the larvae to be classified under Trichostrongylus spp.
Unfortunately, no definite identification of the respective Trichostrongylus species can be achieved through microscopic analysis of their infective larvae. McMurtry [22] originally proposed these to be distinguishable based on detailed assessment of the tubercles seen at high magnification after exsheathment of said larvae, but this was later drawn into question due to significant overlap between the number of tubercles between species [7].
In this study, two larvae morphologically identified as T. circumcincta turned out to belong to the genus Cooperia. Revision of the morphometric characteristics of said larvae led to the conclusion that a mistake was made when classifying these within their respective preliminary groups. For both specimens, a sheathed tail length exceeding 55 µm was recorded. Within the morphological keys used as a base for this research, only Zajac and Conboy [20] report the upper range of the sheathed tail length of T. circumcincta to exceed this value, while other references state the upper limit of the sheathed tail length of the L3 of this species as not exceeding 46 µm [12, 19, 27], corresponding to our data. Furthermore, the misidentification which occurred here most likely resulted from the fact that both T. circumcincta and C. curticei have very similar full body lengths. Statistical analysis showed no significant difference between the average full body length of these two larval types. Besides, considerable overlap was shown between them (Fig. 2c, T. circumcincta: 795.32 ± 25.84 µm, C. curticei: 813.50 ± 19.35 µm). Lastly, a case can be made that the other morphologically defining characteristics of both these larval types are rather subjective and/or easily missed (e.g., cranial inflexion or the presence of refractile bodies).
Because Cooperia spp. was represented only by C. curticei larvae within this research, discussion of group B pertains solely to said species and H. contortus, and no conclusions were drawn regarding the other ovine Cooperia species.
Like the genera in group A, a rudimentary distinction between infectious C. curticei (Additional file 2: Figure S3) and H. contortus (Additional file 2: Figure S4) larvae in sheep can be made based on full body length measurements. Besides the fact that Cooperia spp. L3 are commonly reported to be larger than those of H. contortus [12, 19, 20, 24], a significant difference in mean body length (Fig. 2c, H. contortus: 738.89 ± 14.48 µm, C. curticei: 813.50 ± 19.35 µm) was found here. When in doubt, the authors recommend a cut-off of 790 µm for the identification of C. curticei [first quartile (Q1) of full body length measurements of C. curticei = 791.84 µm].
More notably, sheathed tail morphometry was found to be a reliable trait for the distinction between H. contortus and C. curticei infectious larvae. A significant difference was found between the sheathed tail lengths of these two types of larvae, and no overlap was demonstrated (Fig. 2b, H. contortus: 81.46 ± 2.99 µm, C. curticei: 58.75 ± 2.29 µm). This being said, data of the present research might be subject to a sampling bias, as larvae were selected based on specific morphometric characteristics. Larvae (within group B) with a sheathed tail clearly exceeding 65 µm were identified as H. contortus and those with a sheathed tail shorter than 65 µm as Cooperia spp. For this reason, execution of similar research on blindly sampled larvae might generate different results where considerable overlap can be seen and/or with no significant difference in sheathed tail length, as is commonly reported [12, 19, 20, 24]. Nevertheless, the identities of only two of the larvae microscopically classified within group B turned out to be incorrect. In both cases, H. contortus larvae were misidentified as Cooperia spp. Revision of the morphometric data of said larvae showed these two to have sheathed tails exceeding 65 µm (the only two within the Cooperia spp. dataset), thus providing additional credibility to the above-mentioned morphometric key.
The authors would like to underline that, although sheathed tail length was found to be a good differentiating trait between C. curticei and H. contortus, this is most likely not the case for all Cooperia spp. For example, Cooperia oncophora is reported to have a sheathed tail very similar in size to that of H. contortus [13, 19, 24], and thus definite differentiation in mixed Cooperia samples or samples where C. oncophora and H. contortus can be found side by side might not be as straightforward if only morphology is used.
One particular differentiating characteristic between H. contortus and Cooperia spp. (and all other GIN L3 for that matter) is commonly reported in the current scientific literature. To this extent, larvae belonging to Cooperia spp. are reported to have a pair of oval refractile bodies in their cranial extremities (head) [12, 19, 24]. However, refractile bodies were not clearly observed in any of the Cooperia spp. samples within this research. This is in accord with the findings of Dikmans and Andrews [24], who examined pure strains of C. oncophora and C. curticei in sheep and reported refractile bodies to be clearly visible in infectious C. oncophora larvae but much less so in C. curticei [24]. Consequently, it can be concluded that this trait is not suitable for the definite differentiation of the larvae within group B.
Finally, as mentioned for the L3 classified to have short sheathed tails, the shape of the head of infectious larvae within group B is different as well. Haemonchus contortus larvae are reported to have a typical “bullet-shaped” head, while Cooperia spp. larvae have a more square/rounded head [12, 19, 20]. Yet again, this is most likely only noticeable for a trained parasitologist.
The identification of the species/genera within the third sheathed tail group (group C) is generally problematic [7, 12, 19]. Microscopic differentiation between C. ovina (Additional file 2: Figure S5) and Oesophagostomum spp. is most commonly suggested through the counting and the shape of the larvae’s intestinal cells [7], though various sources report different cell counts with significant overlap [12]. An intestinal cell count of 18–22 triangular cells for Oesophagostomum spp. and 28–32 rectangular cells for C. ovina larvae is believed to be accurate at this time [12]. Nevertheless, intestinal cells can only be clearly perceived in freshly hatched larvae [19], making this characteristic unreliable for samples that have been stored for some time, as was the case here. The number and shape of the intestinal cells of all examined larvae (regardless of the species/genus) were rarely evident within this research.
For the above-mentioned reasons, differentiation between C. ovina and Oesophagostomum spp. is frequently based on other characteristics, such as a difference in sheathed tail and full body length [12, 19, 20, 24], but, as is the case for most ovine GIN infectious larvae, considerable overlap is reported between these two types of larvae [12, 19, 20, 24]. More importantly, current scientific literature reports the sheathed tail filament length of these larvae to be a reliable trait for identification [12]. Oesophagostomum spp. L3 are described as having very long filaments constituting up to 70% of the sheathed tail, while for C. ovina, filaments typically not exceeding 25% are reported. This being said, on top of being difficult to recognize, no clear description of the transition of sheathed tail to filament exists, making these measurements quite subjective [12, 19].
Genetic analysis unfortunately revealed no Oesophagostomum spp. larvae to be present within our samples, and all nine successfully sequenced larvae morphologically identified as potentially representing this genus were wrongly identified, and in fact were C. ovina. Misidentification of Oesophagostomum spp. larvae here most likely resulted from the unreliable differentiating morphological and morphometric traits explained above. As neither the number and shape of intestinal cells nor sheathed tail filament length could be utilized, identification of Oesophagostomum spp. was based mainly on sheathed tail length, and the nine larvae in question thus represent the larvae with the longest sheathed tail within our database. Nevertheless, no larvae with a sheathed tail exceeding 160 µm were observed.
However, as Oesophagostomum spp. is rarely encountered in Sardinia [1], the identification of the larvae within this genus was uncertain from the start, and the larvae in question were included solely in an effort to include Oesophagostomum spp. within this research. A more sensitive diagnostic technique like deep amplicon sequencing could possibly have detected the presence of Oesophagostomum spp. within our samples, as was the case for Cooperia spp. in a Canadian study where this particular genus of GIN is uncommon [17].
Finally, the misidentification of Oesophagostomum spp. larvae led to the occurrence of additional mistakes within our database. More specifically, underestimation of the upper limit of the sheathed tail length of C. ovina resulted in an underestimation of the lower limit of the sheathed tail length of said species as well. Consequently, six larvae having a sheathed tail length between 80 and 100 µm were wrongly identified as C. ovina while actually being H. contortus. However, current scientific literature does report the upper limit of the sheathed tail length of H. contortus to be 78–82 µm [12, 19, 20, 27], clarifying these mistakes.