Islands off southern Australia once harboured three subspecies of the mainland emu (Dromaius novaehollandiae), the smaller Tasmanian emu (D. n. diemenensis), and two dwarf emus, King Island emu (D. n. minor) and Kangaroo Island emu (D. n. baudinianus), which all became extinct rapidly after discovery by human settlers. Little was recorded about their life histories and only a few historical museum specimens exist, including a number of complete eggs from Tasmania and a unique egg from Kangaroo Island. Here we present a detailed analysis of eggs of dwarf emus, including a first record of an almost complete specimen from King Island. Our results show that despite the reduction in size of all island emus, especially the King Island emu that averaged 44% smaller than mainland birds, the egg remained similar-sized in linear measurements, but less in volume and mass, and had slightly thinner eggshell. We provide possible reasons why these phenomena occurred.

Samples

    We used 38 intact eggs of mainland emu Dromaius n. novaehollandiae in our analysis, including the subspecies, D. n. woodwardi of Northern Australia, a taxon no longer considered valid [1]. We used six intact eggs of Tasmanian emu D. n. diemenensis, two held at NHMUK and four at TMAG. A unique egg of Kangaroo Island emu and purchased by the NHMUK, was found hanging in a sealer’s hut prior to 1834 [2]. One of us (CR) excavated a unique, almost complete egg of King Island emu at Yellow Rock River, and found with an associated skeleton of King Island emu, plus other individual emu bones and eggshell. Modern eggshell measurements of D .n. novaehollandiae were obtained from recently broken eggs held at the NHMUK, whereas 19th century subfossil eggshell was collected in a dune deposit in central Australia. All subfossil eggshell came from sand dune deposits on King Island, as did Kangaroo Island eggshell, with the latter material showing slightly more signs of erosion. No Tasmanian eggshell was available for this study.

Measurements

     For intact eggs, all measurements were taken using dial calipers and rounded to the nearest 0.1 mm. Total length of complete eggs was taken along the long axis, and greatest width along the short axis at the widest medial point. Egg volume from museum eggs, including extinct taxa, used scaled photographic images of the eggs, and measured using a code of a custom mathematical model of egg-shape based on an ellipse deformed with a distribution function, as described in [3]. Eggshell thickness was measured using a micrometer rounded to the nearest 0.1 mm, with recent museum and subfossil eggshell fragments measured away from eroded or broken edges. Eggshell thickness remains constant in museum specimens after collection and drying [4]; therefore, comparative eggshell thickness measurements of subfossil fragments and museum intact eggs are considered reliable. For intact egg thickness, a measurement was taken in the blowhole region of D. n. novaehollandiae, D. n. diemenensis and D. baudinianus eggs, with the technique described in [4]. Measurements proved difficult for the Kangaroo Island and Tasmanian eggs, because of the small size of the blowhole and fragile nature. 

Statistics

Mainland and dwarf island emu body masses estimates were calculated using algorithms that based on femora total length (mm) and mid-shaft width (mm) (table S3), based on [5-8]. Data for extant mainland emu taken from [8]; dwarf island emu from [8-9, this paper].  Measurement values are given as mean (minimum-maximum) values. x̄ = the calculated mass and estimate range for the mean.

     A t-test was used to show any significance difference between mainland Dromaius n. novaehollandiae versus Tasmanian D. n. diemenensis for intact emu egg width, using a 2-tailed distribution and two-sampled unequal variance of total length, total width and volume of intact eggs. D. n. diemenensis showed only significant difference from D. n. novaehollandiae in volume (tables S4, S5, S6 and S7), which confirms our data. As unique specimens represent D. n. baudinianus and D. n. minor, they are not included in this statistical study, but the linear measurements show that D. n. baudinianus nestle within mainland emu size and D. n. minor is slightly smaller in all dimensions. T-test analysis of eggshell thickness showed no statistical difference with results significantly higher than the p-value 0.05, so are not considered an evolutionary factor.