Introduction

Citing publications to substantiate claims or facts, or to show congruency or not with prior results, is a cornerstone of scientific writing. As part of this process, authors are responsible for ensuring that the references they select genuinely support the statement they have written. To do so requires authors to read the references being cited to verify the appropriateness of the paper. Inadequate verification of citations can lead to perpetuation of inaccuracies within the scientific literature, which can become an overwhelming influence given that it has been estimated that 70–90% of scientific citations are copied from reference lists of other papers rather than from the original paper (Simkin and Roychowdhury 2005). Such issues erode the integrity of scientific literature (Santini 2018).

Inaccurate citing is nothing new to scientific writing and has been the lament of numerous publications. Among the numerous issues are incorrect summaries of work (Hernandez and White 1989; Stordal 2009; van de Weert and Stella 2019), and the rise of a persistent incorrect description or understanding of a scenario, such as of the native ant extinctions in Madeira by Pheidole megacephala and Linepithema humile (Wetterer 2006). Here, we present an unusual scenario, whereby we agree that the statement being made is correct, but there appears to be no baseline publication that can be truly cited, either because it never existed, or that demonstrative publication has been lost to knowledge. This statement is that “red imported fire ant Solenopsis invicta workers do not have ovaries”. This statement is cited in many papers, and is a critical underlying basis of highly influential works on eusociality and inclusive fitness (e.g., Keller and Ross 1998). Here, we demonstrate that we cannot find a source publication that proves this statement, we demonstrate how publications are erroneously citing this statement, and finally we provide demonstrative proof of the statement, which to the best of our knowledge is the first time this has been published in the scientific literature.

Materials and methods

We first tried to find a publication that demonstrated S. invicta workers do not have ovaries. We anticipated that demonstration would be in the form of quantified dissections of workers finding a lack of ovarioles and potentially other structures associated with reproduction, such as in Bickford (1895, 1897). To do this, we conducted a literature search using the ISI Web of Science platform and Google Scholar using the following search terms: “ant”, “Solenopsis”, “invicta”, “ovar*”, “steril*”. We also conducted targeted searches of any references cited within publications stating that S. invicta or Solenopsis workers did not have ovaries and/or were sterile. After not finding any suitable references from those searches, we conducted another literature search for the Spanish words “hormiga”, “obrera”, and “ovarios”, on the basis that the native range of S. invicta is within countries that speak Spanish. After again not finding a suitable reference, we then repeated the search in German, Portuguese, and Dutch.

To demonstrate how publications were citing that S. invicta or Solenopsis workers did not have ovaries and/or were sterile, we generated a flow chart. We also compiled a list of papers that provided no citation to support the statement (Fig. 1). Note that these summaries were not intended to be comprehensive of all global literature, just representative to demonstrate the issue. These summaries were also restricted to literature written in English.

Fig. 1
figure 1

Flow diagrams of publications citing other publications to support the statement that Solenopsis or S. invicta workers do not have ovaries and/or are sterile. Arrows flow from the citing paper to the paper cited. Note that the top diagram also has an anti-clockwise circular date order of papers referring to other publications commencing with Cole 1986 and finishing with Lee et al. 2017. Page numbers are given when multiple statements are made in a publication with different references. “No reference” indicates that the publication did not provide a reference to support the statement. No data indicates that the statement was implicit within a phylogenetic figure, but no data were provided. “States nothing” indicates that there was nothing written in the publication specifically about any Solenopsis even though it was used as the supporting reference

Full size image

Upon not finding any publication that met the criteria of demonstrating that S. invicta workers do not have ovaries, we dissected S. invicta workers to quantify the presence/absence of reproductive structures, specifically ovarioles and spermatheca. Hand collections were made from four nests from throughout Brisbane. Most workers were preserved upon arrival at the Brisbane laboratory, but 25 random workers from each nest were kept alive for 48 h in Brisbane and fed pH-indicating dye to aid photography following dissection. When killed, all workers were preserved in 100% ethanol. A total of 87 workers were dissected, 25 using a Nikon SMZ25 stereoscopic microscope in Brisbane, and 62 using a Zeiss Stemi 2000-C stereoscopic microscope in Darwin.

Because S. invicta is polymorphic (Mirenda and Vinson, 1981), we dissected individuals that represented the full-size range of the individuals collected, in case a subset of workers had reproductive structures such as occurs in yellow crazy ant Anoplolepis gracilipes (Lee et al. 2017; Lenancker et al. 2021). To demonstrate the morphometric diversity we assessed, we measured standard morphometrics of the 62 specimens dissected in Darwin. First, the bodies of the ants (minus the abdomens) were glued to cardboard triangles as per standard ant curation. Then, using the Darwin microscope with an ocular ruler, the individuals were positioned so that their heads were perpendicular to the view, and their head length and width were measured (Fig. 2).

Fig. 2
figure 2

Head width and length measurements of workers assessed for the presence/absence of ovaries

Full size image

Results

We found many publications were being cited to reference the statement Solenopsis or S. invicta workers do not have ovaries and/or are sterile (Fig. 1), and numerous publications that provided no citation to support the statement (e.g., Aron et al. 1995; Gadagkar 2019; Manfredini et al. 2014; Villalta et al. 2018), but we found no publication that provided any data or evidence that S. invicta workers do not have ovaries. None of the 87 workers dissected contained ovarioles or spermatheca, but all other major abdominal organs were found (Fig. 3).

Fig. 3
figure 3

Dissected abdomens of Solenopsis invicta showing: the digestive track along with the posterior end of the crop (a–e, g–h); colored digestive track after feeding with pH-indicating dye (c); and a close-up of the posterior end of the crop showing the poison glad but no presence of spermatheca or ovarioles (f). CR  crop, MG midgut, HG  hindgut, PG  poison gland

Full size image

Discussion

This paper is the first known publication of quantified work that demonstrates S. invicta workers do not have ovaries. Our findings align with the often made statement that workers of this species are completely sterile, not even able to contribute trophic eggs to colony nutrition, as can occur in other ant species (Bourke 1988). We were certainly not the first people to dissect this species to conduct the same investigation though, as we had informal discussions with numerous researchers who had historically done so, but they had not published quantified findings. Potentially such a quantitative study has been conducted and published previously, but if so the knowledge of its existence appears to have been lost. That others have also been unable to find a baseline publication is clear from the large number of different and inappropriate papers being cited to support the statement. Indeed, even one of us (BDH) was found to have previously made this citation error within Lenancker et al. (2019). We acknowledge that the samples were obtained from the invasive range only, but we are unaware of any instance of a species having different organ composition between its native ant exotic range as has been shown for other aspects such as morphology, diet and colony structure (Wills et al. 2014).

Although our work focused solely on S. invicta, while reading literature associated with our searches, we noted that many other species and genera were stated as not having ovaries, but we did not come across the base research that proved those statements. Likely most of these statements are correct, but we foresee that the same referencing issue we have identified here may extend to many other ant species and genera. Moreover, while most are likely correct, some may well not be. We suggest to the scientific community that if their own literature searches fail to find a base publication that supports a statement, especially a broad statement used often in publications, there may be a general need for somebody to conduct a simple demonstrative investigation to provide an adequate reference. Such work could indeed be highly beneficial should the broad statement prove to be erroneous (Letrud and Hernes 2019).