Identification of Planigale ingrami and Planigale tenuirostris from mandibles deposited in eastern barn owl (Tyto javanica delicatula) pellets

Introduction

In his landmark revision of the genus Planigale, Archer (1976) described Planigale ingrami (long-tailed planigale) and Planigale tenuirostris (narrow-nosed planigale) as often ‘not clear that they are separate species, (p. 358)’, on the basis of similarities in craniodental features. Subsequent genetic studies have clearly supported Planigale ingrami and Planigale tenuirostris as distinct (Painter et al. 1995; Blacket et al. 2000; Umbrello et al. 2023). Nevertheless, morphological similarities between the two often make confident identification difficult, especially when dealing with skeletal remains, which typically rely on intact skulls (via cranium depth:width ratios) (Blacket et al. 2008; Van Dyck et al. 2013). This presents a challenge for confident identification where skeletal material is incomplete, such as owl pellet content analysis, which is increasingly used as a detection tool to generate small mammal species richness estimates (Heisler et al. 2016; Schoenefuss et al. 2024; Charley et al. 2025).

Planigale craniodental material sourced from owl pellets mostly consists of mandibles, albeit well preserved, with the cranium typically crushed and fragmented. This currently limits confident identification of planigales in owl pellets to genus level (Charley et al. 2025). The inability to accurately attribute species-level identification to planigales found in owl pellets means that species richness estimates may be underestimated, and knowledge of planigale species distributions remains unclear. Here, with reference to registered, identified specimens held at Australian museums, we undertake an analysis of eastern barn owl (Tyto javanica delicatula) pellets by using a mandibular identification feature in an attempt to discriminate between Planigale ingrami and Planigale tenuirostris.

Materials and methods

This research was conducted under the auspices of Queensland Department of Environment, Science and Innovation (DESI) Permit P-PTUKI-100171210 and QUT Research Ethics Permit 5154.

Planigale mandibles assessed in the present study (n = 68) consisted of both Queensland and South Australian Museum reference specimen mandibles (total n = 30; P. ingrami n = 16; P. tenuirostris n = 14) for which correct identification could be strongly attributed by museum curators [i.e. with an intact skull (Blacket et al. 2008)] (see Appendix 1), and a representative subset of Planigale spp. mandibles (n = 38, designated as Unknown) collected from eastern barn owl pellet material taken from an abandoned wool shed at Toorak, north-western Queensland, Australia (21.03290 S, 141.79958 E). Toorak is located ~40 km south of Julia Creek and lies within the semi-arid region of the Mitchell Grass Downs bioregion (Fig. 1). For full details of the owl pellet collection and processing methodology, see Charley et al. (2025). Only left mandibles were assessed in the present study to ensure that each individual contributed only one set of measurements to the analysis. Only adult specimens were used, because some measurements could be affected by bone lengths. Adult specimens were designated by full emergence of the lower (M₄) molar tooth (Blacket et al. 2008).

Fig. 1.

Map of Australia with state and territory borders. Points represent planigale specimen collection locations used in the present study. Shaded areas represent the respective species distributions (generated from Marsh et al. 2022).

Planigale ingrami and Planigale tenuirostris have potentially overlapping distributions at Toorak (Baker and Gynther 2023). Two other species of planigale occur in Queensland, Planigale gilesi and Planigale maculata. Although these two species were unlikely to be present at the surveyed site, they are in any case relatively easily identified because the former has only two premolar teeth and the latter is notably larger in size than are congeners (Baker and Gynther 2023).

On the basis of expert recommendation (S. Van Dyck, pers. comm.), the angle the ascending ramus makes at the emergence of the lower molar tooth row (henceforth, LMR-AR angle) was suggested as a possible diagnostic feature for distinguishing between Planigale ingrami and Planigale tenuirostris. This feature was measured using a disconnected angle tool (using Zeiss labscope software ver. 4.3.1, https://www.zeiss.com/microscopy/en/products/software/zeiss-labscope.html, although any angle software would be applicable) using a Zeiss Stemi 305 stereo microscope with attached Axiocam 208 colour microscopy camera, placing the lower angle arm to intersect both the anterior molar emergence point on M₁ and the posterior molar emergence point on M₄, and the upper angle arm following the anterior border of the ascending ramus (Archer 1976) (Fig. 2).

Fig. 2.

The LMR-AR angle measurement for planigales assessed in the present study. QLD Museum specimen (a) JM18198, identified by museum curators as Planigale ingrami, and (b) JM2192, identified by museum curators as Planigale tenuirostris.

Analyses were performed in R (R Core Team 2024). A Shapiro–Wilk test was first conducted on the dataset to test for normality. Then, a one-way ANOVA was used to assess for differences in planigale angle measurements among the three assessed groups (Unknown, Planigale ingrami and Planigale tenuirostris). Tukey’s honest significant differences (TukeyHSD) test was then used to compare pairwise differences between the groups. Tukey’s HSD test is used when performing multiple simultaneous comparisons to ensure that the P-value achieves an 0.05 error rate (Agbangba et al. 2024).

Results and discussion

Complete data are presented in Appendix 1. Data were deemed to be normally distributed (Shapiro–Wilk normality test, W = 0.99, P-value = 0.82). ANOVA testing demonstrated a significant difference between species (d.f. = 2, f-value = 7.54, P-value = 1.14 × 10⁻³). Post hoc testing (TukeyHSD) demonstrated a significant difference between Planigale tenuirostris and Planigale ingrami museum specimens (diff = −4.45, lower = −7.57, upper = −1.33, adjusted P-value = 3.03 × 10⁻³) (Fig. 3). Overlap in the molar row-ascending ramus angle measurement between Planigale ingrami and Planigale tenuirostris was common (Table 1, Fig. 3), although there was clear separation in the upper and lower extremes for both species. Any angle measurement in the assessed museum registered mandibles that was ≥120° was attributable to Planigale ingrami, whereas any angle of ≤112° was attributable to Planigale tenuirostris.

Fig. 3.

Boxplot comparison of the LMR-AR angle for various Planigale ingrami and Planigale tenuirostris museum reference specimens and owl pellet collection mandibles (Unknown). Boxes represent the interquartile ranges, the vertical lines the median and the horizontal lines the total range.

There was a significant difference between Unknown and Planigale ingrami (diff = −3.66, lower = −6.20, upper = −1.12, adjusted P-value = 2.72 × 10⁻³), but no significant difference between Unknown and Planigale tenuirostris (diff = 0.79, lower = −1.87, upper = 3.45, adjusted P-value = 0.76). The measured angles for Planigale tenuirostris had a markedly larger range (105.5°–119.3°) than did those of Planigale ingrami (113.1°–121.9°). This range showed notable overlap with the Unknown sample (104.15°–121.32°), which may explain the lack of significant difference between Planigale tenuirostris and the Unknown samples (Fig. 3).

One unknown mandible specimen could be attributed to Planigale ingrami (≥120°), and 13 unknown mandible specimens could be attributed to Planigale tenuirostris (≤112°) by using the proposed LMR-AR angle. Despite this, the observed interquartile range (IQR) of both species was largely separate (Planigale tenuirostris 109.9°–115.6°; Planigale ingrami 115.1°–118.0°; Fig. 3), with 20 unknown specimens falling within the IQR of Planigale tenuirostris, and seven unknown specimens falling within the IQR of Planigale ingrami. Interestingly, one unknown specimen (104.15°) fell outside of the observed minimum of Planigale tenuirostris (105.5°), suggesting possible further variability in the measurement.

The Unknown sample measurements showed notable overlap with both Planigale ingrami and Planigale tenuirostris. This suggests the presence of both planigale species at Toorak, despite prior owl pellet collections from this location being assumed to consist solely of Planigale ingrami (Woolley 2010). Further field-based detection methodologies [pitfall trapping; Catling et al. (1997)] should be employed to corroborate this finding with live planigales at Toorak and surrounds. Voucher specimens would permit comparative examination of both DNA and a full suite of craniodental features.