The surprising complexity and diversity of sperm storage structures across Galliformes

Abstract

In internal fertilisers, the precise timing of ovulation with the arrival of sperm at the site of fertilisation is essential for fertilisation success. In birds, mating is often not synchronised with ovulation, but instead females utilise specialised sperm storage tubules (SSTs) in the reproductive tract, which can ensure sperm are always available for fertilisation at the time of ovulation, whilst simultaneously providing a mechanism of post‐copulatory sexual selection. Despite the clear importance of SSTs for fertilisation success, we know little about the mechanisms involved in sperm acceptance, storage, and release. Furthermore, most research has been conducted on only a small number of species, based on which SSTs are usually assumed to look and function in the same way across all species. Here, we conduct a comparative exploration of SST morphology across 26 species of Galliformes. We show that SSTs, and the surrounding tissue, can vary significantly in morphology across species. We provide observational evidence that Galliformes exhibit at least 5 distinct categories of tubule types, including distinctive coiled and multi‐branched tubules, and describe 2 additional features of the surrounding tissue. We suggest functional explanations for variation in tubule morphology and propose next steps for future research. Our findings indicate that SSTs are likely to be far more variable than has previously been assumed, with potentially important consequences for our understanding of sperm storage in birds and post‐copulatory sexual selection in general.

Full text

Ecology and Evolution. 2024;14:e11585.  
|
1 of 17
https://doi.org/10.1002/ece3.11585
www.ecolevol.org
1 | INTRODUCTIO N
In internal fertilisers, successful fertilisation relies on the arrival of
sperm at the site of fertilisation at the precise time of ovulation. In
many species, ensuring sufficient sperm are available for fertilisa-
tion requires insemination to be precisely timed with the release
of the ovum. In others, insemination and ovulation are not syn-
chronised, sometimes occurring many days or even months apart
Received:8April2024
|
Revised:2 5May2024
|
Accepted :31May2024
DOI: 10.1002/ece3.115 85
NATURE NOTES
The surprising complexity and diversity of sperm storage
structures across Galliformes
Katherine Assersohn | J. Paul Richards | Nicola Hemmings
This is an op en access ar ticle under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium,
provide d the original work is properly cited.
© 2024 The Aut hor(s). Ecolog y and Evolution published by John Wiley & Sons Ltd.
School of Bioscience s, University of
Sheffield, Sheffield, UK
Correspondence
Katherine Assersohn and Nicola
Hemmings, School of Bioscience s,
University of Shef field, Sheffield S10 2TN ,
UK.
Email: kassersohn1@sheffield.ac.uk and
n.hemmings@sheffield.ac.uk
Funding information
Royal Society, Grant/Award Number:
DHF160200; Natural Environment
Research Council, Grant/Award Number:
NE/S00713X/1; University of Sheffield
Abstract
In internal fertilisers, the precise timing of ovulation with the arrival of sperm at the
site of fertilisation is essential for fertilisation success. In birds, mating is often not
synchronised with ovulation, but instead females utilise specialised sperm storage tu-
bules (SSTs) in the reproductive tract, which can ensure sperm are always available
for fertilisation at the time of ovulation, whilst simultaneously providing a mechanism
of post- copulatory sexual selection. Despite the clear importance of SSTs for fertili-
sation success, we know little about the mechanisms involved in sperm acceptance,
storage, and release. Furthermore, most research has been conducted on only a small
number of species, based on which SSTs are usually assumed to look and function in
the same way across all species. Here, we conduct a comparative exploration of SST
morphology across 26 species of Galliformes. We show that SSTs, and the surround-
ing tissue, can vary significantly in morphology across species. We provide observa-
tional evidence that Galliformes exhibit at least 5 distinct categories of tubule types,
including distinctive coiled and multi- branched tubules, and describe 2 additional fea-
tures of the surrounding tissue. We suggest functional explanations for variation in
tubule morphology and propose next steps for future research. Our findings indicate
that SSTs are likely to be far more variable than has previously been assumed, with
potentially important consequences for our understanding of sperm storage in birds
and post- copulatory sexual selection in general.
KEY WORDS
cryptic female choice, fertilit y, post- copulatory sexual selection, reproduction, sperm
selection, uterovaginal junction
TAXONOMY CLASSIFICATION
Evolutionary ecology, Zoology

2 of 17
|
ASSERSOHN et al.
(Birkhead, 1992; Birkhead & d el Nevo, 1987; Hatch, 1983; He mmings
& Birkhead, 2020; Wanless & Harris, 1986). To compensate for this,
females may store sperm within the reproductive tract, maintaining
it in a viable state to be released at ovulation (Bakst, 2011).
Female sperm storage is a widespread phenomenon, occur-
ring in many invertebrates and all major ver tebrate groups (Holt &
Fazeli, 2016; Shankar et al., 2022). In species exhibiting sperm stor-
age (as opposed to just sperm longevity), females can store sperm
for a few hours to months or even years (Holt & Fazeli, 2016; Levine
et al., 2021; Orr & Zuk, 2012). This is achieved through providing
an environment favourable for sperm sur vival, often in the form
of specialised morphological structures e.g., seminal receptacles,
(gastropods and arthropods), spermathecae (many invertebrates),
tubules (birds, reptiles, fish and amphibians), or a combination of
both spermathecae and seminal receptacles (Drosophila melanogas-
ter) (reviewed in; Holt, 2011 and Orr & Brennan, 2015). Sperm stor-
age structures not only guard against a lack of sperm at fertilisation
but may also allow females to preferentially store/release sperm
from preferred males, thereby providing a mechanism of postcop-
ulatory sexual selection (i.e., cr yptic female choice) (Bakst, 2011;
Eberhard, 1996;Mendoncaetal.,2019; Sasanami, 2017).
Sperm storage has been particularly well- studied in birds, be-
ginning with speculation in the early- 1900s that viable sperm
were capable of surviving in the oviduct long after insemination
(Payne, 1914). It was not until the mid- 1900s that the first special-
ised storage sites, then termed ‘sperm nests’, were identified and
suggested to be responsible for sustained fertility in female birds
(Van Drimmelen, 1946). ‘Sperm nests’ were initially described as
being located within shallow crypt s of the infundibulum (the site
of fertilisation), but in a series of histological studies, sperm stor-
age was later suggested to occur principally at the utero- vaginal
junction (UVJ), a small non- distinct section of the vagina bordering
the uterus (Bobr, Lorenz, & Ogasawara, 196 4; Bobr, Ogasawara, &
Lorenz, 1964; Verma & Cherms, 1964, 1965). Today, these structures
are commonly known as ‘sperm storage tubules’ (SSTs), occurring
primarily in the UVJ, with evidence for infundibular storage sites
remaining equivocal (but still widely cited) (Assersohn et al., 2021;
Bakst, 2011; Sasanami, 2017).
Although commonly defined as ‘simple tubular invaginations’,
SSTs are increasingly appreciated as being highly specialised, with
an expanding body of evidence to suggest that the molecular and
biochemical processes occurring within SSTs are complex and
highly regulated (Bakst, 2011; Freedman et al., 2001; Hemmings
et al., 2015; Holm et al., 2000; Khillare et al., 2018; Mendonca
et al., 2019; Sasanami, 2017 ). The vagina is t ypically hostile to sperm:
only 1% of inseminated sperm make it into storage (Bakst, 2011),
and it is thought to be a selective process that probably ensures
sperm with atypical morpholog y and physiology are inhibited
from participating in fertilisation (Bakst, 1994; Bobr, Lorenz, &
Ogasawara, 1964; Khillare et al., 2018; Ogasawara et al., 1966).
Once accepted into SSTs, numerous compounds are produced that
likely act to suppress sperm motility and provide protection from
structural damage and oxidative stress (Bakst & Bauchan, 2015;
Freedman et al., 2001; Holm et al., 2000; Huang et al., 2016, 2017;
Khillare et al., 2018;Matsuzaki etal., 2015;Mendoncaetal.,2019;
Sasanami, 2017). At ovulation, sperm are re- mobilised and released
from storage (Hiyama et al., 2014; Ito et al., 2011). In birds, there
isaverynarrow window of time (around 15 min in domesticfowl;
Bakst, 1994) between ovum release and the laying down of the
outer perivitelline layer (a matrix of glycoproteins that surrounds
the ovum and blocks fur ther penetration by sperm; Hemmings &
Birkhead, 2015). The timing of acceptance and release of sperm
is therefore highly regulated, probably through fine hormonal and
possibly nervous control (Hemmings et al., 2015). Furthermore, 3D
imaging of SSTs in zebra finches (Taeniopygia guttata) has identi-
fied gate- like constric ted openings that may act to limit the ability
of sperm swimming below a certain velocity to enter storage, pro-
viding an additional mechanism of selec tion for high- quality sperm
(Hiyama et al., 2014; Ito et al., 2011;Mendoncaetal.,2019). Despite
these advances, we still lack a comprehensive understanding of
the mechanisms by which SSTs accept, store/maintain and release
sperm (Khillare et al., 2018; Sasanami, 2017; Shankar et al., 2022).
We have long known that even among commercial birds selected for
consistent and high fertility, sperm storage duration varies greatly.
Forexample, sperm arestored forjust5–10 daysin Japanesequail
(Coturnix japonica)butupto15 weeksinturkeys(Meleagris gallopavo)
(Birkhead&Møller,1990 ; Sasanami, 2017). However, the causes of
such variation are not known or generally discussed in the context of
SST diversity across avian species.
Perhaps an even more fundamental barrier to our understand-
ing of sperm storage in birds is our lack of knowledge of inter-
specific variation in SST morphology. Sperm are some of the
most diverse cells in the animal kingdom (Pitnick et al., 2009), and
so it stands to reason that the structures that selectively store
them might also vary considerably in both structure and function
(Cramer et al., 2023). Sperm length has been found to correlate
negatively with SST number and positively with SST length in
passerines(Briskie&Montgomerie,1993), suggesting that the co-
evolutionary dynamics between male and female post- copulatory
sexual selection may be a driving fac tor in the evolution of sperm
morphology (Kustra & Alonzo, 2023). Sperm morphology and
female sperm storage organ morphology have been shown to
correlate in other taxa as well (mainly invertebrates; Hig ginson
et al., 2012; Miller& Pitnick,20 02). Within- population variation
in sperm storage organ morphology and function has also been
documented in Drosophila (Lüpold et al., 2013). However, the gen-
eral assumption in birds is that SSTs always look – and function
in – the same way as those observed in the species in which SSTs
have been most well studied (e.g., domestic chickens (Gallus gallus
domesticus), turkey, Japanese quail, and the zebra finch). In studies
that have identified SSTs, there is also inconsistency in how quan-
titative variables such as SST number and length are measured,
making cross study comparison difficult. For example, there is no
consensus on whether branched tubules should be considered
(and measured as) one tubule with great total length, or many dis-
tinct shorter tubules. An important step in determining the degree

|
3 of 17
ASSERSOHN et al.
of variation in SST form and function across species, and how
this correlates with post- copulatory processes, will be to develop
more consistent and reproducible descriptions and measures of
SST traits.
Here, we report and present the discover y of remarkable varia-
tion in mor phological structure of SS Ts across species of Galliformes.
We also present methods for dissecting and examining the folds of
the utero- vaginal junc tion in birds, suggest criteria for categorising
tubule morphology based on our observations, and discuss the po-
tential for future research in Galliformes and across birds in general.
2 | MATERIALS AND METHODS
2.1 | Animals
Galliformes are a very large, diverse, and well- studied bird group in
which body size and mating system vary greatly, and samples are
easily accessible, making them ideal for exploring interspecific vari-
ation in SST morpholog y. We collected the whole oviduct of a single
female from 28 different species of Galliformes. Dead birds were
sourced in 2016 from breeders in the process of disposing of excess
stock, anddissections tookplaceon sitewithin30 min ofeachbird
being killed. Females were included in the study only if they were
in breeding condition (confirmed by the presence of an ovum in the
oviduct and/or a hierarchy of developed ova in the ovary), since the
avian oviduct is known to regress in size outside of the reproductive
period.Malesandfemaleswerehoused together,allowed free ac-
cess to each other, and all copulations were natural. Females were
dissectedatmostwithin1–2 daysofcommencingegglaying,andall
females were confirmed to have laid fertile eggs (demonstrating that
they had recently accepted and stored sperm).
The oviduct was removed int act (including the cloaca and ovary),
unravelled, stripped of connective tissue, and briefly cleaned in
phosphate- buffered saline (PBS). The oviduct was then pinned out
lengthways in a long, shallow wax- based tray, where it was photo-
graphed and measured. The length of each individual section of the
oviduct , as well as its entire length, was measured. We also recorded
the wet mass of the oviduct after briefly dabbing off excess liquid
with absorbent tissue. Once all measurements were complete, the
oviduct was transferred into a deeper tray, pinned so that each sec-
tion was straight, and submerged in 10% formalin solution to fix the
tissue.Afteratleast48 hinfixative,asegmentofthevaginacontain-
ing the UVJ was then cut away from the rest of the oviduc t (Figure 1;
step 1; Briskie & Birkhead, 1993). Since the UVJ has previously been
reported to be located at the uterus end of the vagina, the segment
was always cut at the beginning of the uterus. However, due to the
use of vaginal tissue for another study, the degree to which the seg-
ment extended along the vagina varied from 19 to 30% of the total
vaginal length. Upon microscopic examination of each sample, the
start of the UVJ segment was designated as the point at which SST
first appeared, and the end of the UVJ segment was where uter-
ine tissue started. Occasionally, it appeared that the distribution of
tubules continued slightly further down the vagina than the length
of our sample allowed us to visualise; however, this was uncommon
(only occurring in 4 species; California quail (Callipepla californica),
European quail (Coturnix coturnix), mountain quail (Oreortyx pictus),
and Temminck's tragopan (Tragopan temminckii)), and the dissected
segments were expected to contain the majority (if not all) of the
SSTs in the sample.
2.2 | Dissection
Samples were cut longitudinally with micro- dissecting scissors,
pinned open with butterfly specimen pins on Sylgard™ 184 gel
(Sigma- Aldrich), and covered with phosphate buffered saline (PBS),
to reveal the internal luminal mucosa and mucosal folds of the va-
gina (Figure 1; step 2). For each sample, 2 folds were dissected.
Folds were examined under a Leica MZ75 dissecting microscope
and dissected by cutting the entire length of one side of the ‘valley’
(as opposed to the crest) of a single fold along the entire leng th of
the sample (Figure 1; step 3) (Briskie & Birkhead, 1993). Care was
taken to remove only the mucosa, peeling it away from the underly-
ing connective tissue (the lamina propria), but avoiding cut s to the
underlying tissue. Before cutting along the remaining side, the fold
was gently prised open along the length. Opening the fold out whilst
it was still attached to the sample along one side had the benefit
of providing leverage for gently teasing apar t the ‘crest’ of the fold.
Flattening the crest was necessar y to allow the fold to lie smoothly
once mounted, but was much more difficult to achieve once the fold
was fully dissected. If necessary, additional cuts were made to re-
move connec ting tissue whilst ensuring the integrity of the fold was
maintained and the SSTs were not damaged. The entire length of
the remaining attached side was then cut longitudinally, and the fold
gently removed from the underlying tissue. The fold was then incu-
batedfor5–10 minwith10–20 μL of Hoechst 33342 dye (the exact
volume depended on how much was needed to fully submerge the
fold). Hoechst was used to help visualise sperm in storage, which
aided in identifying SSTs, and had the added benefit of clarifying the
edges of the SSTs against the surrounding tissue. The fold was then
placed onto a slide (lamina propria side down) and gently opened to
lie flat. Once flattened along the entire length, a coverslip was added
withFluoromount-G™ AqueousMounting Medium(Sigma-Aldrich)
and left to dry. Dried samples were sealed with transparent nail var-
nish and kept in the dark until imaging (Figure 1; step 4).
2.3 | Morphological observations from dissection
We observed a remarkable amount of variation in the thickness,
texture, integrity, and overall appearance of vaginal tissue between
species. The visual presentation of the folds of the vagina varied
between species, with some species having large and very distinct
folds whilst in others the individual folds were less pronounced.
Some species displayed obvious pigmentation in the vagina, and

4 of 17
|
ASSERSOHN et al.
in many cases, tubules were visible by eye at this stage (Figure 1;
step 5). There was always an obvious difference between vaginal
and uterine tissue, though in some species this difference was more
pronounced than others. Uterine tissue was always to some degree
thicker, darker in appearance and with larger/taller folds relative to
the vagina.
Samples also clearly varied in their response to preservation in
formalin: While some samples remained very well preserved, others
had become fragile and difficult to dissect. In the case of 2 species,
Javan peahen (Pavo muticus), and Grey partridge (Perdix perdix), sam-
ples were so fragile they were impossible to dissect. Consequently,
we removed these species from the sample pool, reducing the sam-
ple size to 26 species (52 folds). This difference in response to pres-
ervation, ease of dissection and mounting suggests there are some
innate differences in tissue structure between species; however,
most samples remained in suitable condition for dissection, despite
FIGURE 1 Schematicillustratingtheavianoviduc tandovary,andthemethodsofUVJdissection,folddissectionandSSTcategorisation.
Within the oviduct, the utero- vaginal junction (UVJ) (purple segment) – the site of primary sperm storage, was dissected and pinned open to
expose the internal folds. A single fold was then dissected along its length and stained with Hoechst fluorescent dye. The fold was then laid
flat and fixed onto a slide before being imaged along its entire length using fluorescent microscopy. SSTs are typically visible along the centre
of UVJ folds. Schematic drawn in Adobe Illustrator (V.28.4.1).

|
5 of 17
ASSERSOHN et al.
there being an approximate 7- year gap between tissue fixation and
sample dissection.
2.4 | Imaging
Slides were examined under fluorescent and brightfield light on a
Leica DMLB compound microscope with an LEJ ebq100 fluores-
cence source. Images were captured with Infinity Analyse software
via a Lumenera Infinity3 USB camera. Images were taken along
the entire length of each fold, following the centre line of the fold
(Figure 1; step 5), at 50× magnification, with one image directly bor-
dering the next with no overlap and no missing sections. Depending
on the individual sample, images were taken either using brightfield,
darkfield, or fluorescent illumination, or a combination, in whichever
way produced the highest quality image for that sample. In many
cases, images of one field of view were taken repeatedly through
multiple planes to avoid missing hidden structures where possible.
Image scale bars were produced in ImageJ (Schneider et al., 2012),
and contrast and brightness were adjusted in Adobe Photoshop
(version 25. 5) to aid visualisation where necessary. For some struc-
tures of note, additional images were taken at either 100 or 20 0×
magnification.
2.5 | Tubule categorisation
On microscopic examination of each image, careful notes of visible
structures were made that were subsequently assessed to create
distinc t categories that could be used to define consistent features
(Figure 1; step 6). An initial criterion for categorising structure type
was then created, which was used to categorise the visible struc-
tures within each image across all samples. A second observer then
used the categorisation criteria to categorise a subset of images: one
image per sample (two samples per species; 52 images in total) was
chosen for re- categorisation, at random (using a random number
generator), from a pool of images that only contained visible struc-
tures (i.e., excluding images in which no features were observed).
Repeatability of categorisation was then calculated in R (R Core
Tea m, 2023) using the package rptR (Stoffel et al., 2017). A binary
family was used for each model, with tubule category as the re-
sponse variable and sample ID as the random effect. For categories
with lower repeatability, adjustment s to the categorisation criteria
were made to improve reproducibility. We then used these revised
categorisation criteria to create a flow chart (also reproduced as a di-
chotomous key in Appendix S1 ) that could be used and built upon by
future researchers to categorise the structures they observe in their
own samples, with the aim of encouraging a consistent and definitive
use of terminology throughout the literature. We acknowledge that
a greater number of distinct structures are likely to exist across spe-
cies and encourage future work to critically examine the structures
observed and consider whether they fit into the existing categories
presented here, or whether new categorisations are necessary.
To aid in categorising tubule location, we considered the samples
as being divided into 3 sections, each containing an equal number of
images: C (the most distal section, i.e., closest to the cloaca and bor-
deringtherestofthevagina),M(thecentralthirdofthesample),and
U (the most proximal section that borders the uterus; Briskie, 1996).
If the sample contained an uneven number of images making equal
thirdsimpossible,theMsectionwasexpandedtoaccommodatethe
additional images, with the C and U sections always containing the
same number of images and so being propor tionally the same size.
Note that we always consider the first image in C to be the star t of
the UVJ region of the vagina and will contain the first appearance
of tubules.
3 | RESULTS
3.1 | Tubule location within the UVJ
Tubule location varied between species; however, in the majorit y of
species (20; 76.9% of species), tubules were found in all three re-
gionsofthesample(C,M,andU)inatleastoneofthesamplefolds.
In 13 species (50% of species), tubules were only found in the C and
M(but notthe U)regionsinatleastone ofthesamplefolds,but of
these, 6 species never had tubules in the U region of either fold:
harlequin quail (Coturnix delegorguei), Japanese quail, mountain quail,
Swinhoe's pheasant (Lophura swinhoii), silver pheasant (Lophura
nycthemera), and wild turkey. There was only 1 species (California
quail)inwhichtubuleswereseenonlyintheCregion(andnottheM
and U), but this was only the case for one fold of the sample. Taken
together, tubules were commonly found across the entire length of
the sample, but in half of all species, at least one fold sample did not
have tubules that directly bordered the uterus.
3.2 | Presentation of fold tissue
Variation in the texture of vaginal and uterine tissue between spe-
cies was apparent upon microscopic examination: In some cases, tis-
sue was thin, and tubules were clear, and in others, the tissue was
thick and layered, with tubules sometimes being embedded mak-
ing them more difficult to identify, even when examining through
multiple planes of focus. Sperm was not always observed in stor-
age in every sample, possibly due to lower retention of sperm, fewer
copulations, or fewer sperm being transferred during copulation to
begin with. In samples that contained stored sperm, the distribution
of sperm was uneven throughout the samples, with most tubules
being free of sperm.
3.3 | Channels
We repeatedly observed ‘channel- like’ structures (here- on referred
to as channels) that presented as grooves along a single fold and

6 of 17
|
ASSERSOHN et al.
often extended along a large portion of the fold (see Figure 2a where
no channels are present, relative to Figure 2b–f where channels can
be seen). Although occasionally shor ter and more ‘funnel- like’ in ap-
pearance (Figure 2e,f, seen in; black fr ancolin (Francolinus francolinus),
grey junglefowl (Gallus sonneratii), golden pheasant (Chrysolophus pic-
tus), Lady Amherst's pheasant (Chrysolophus amherstiae), mountain
quail, red legged partridge (Alectoris rufa), roul roul partridge (Rollulus
rouloul), and Temminck's tragopan), these structures were distinct
in presentation from tubules but contained clear thickened tissue
on either side (channel walls) with a ‘lumen- like’ depression along
the centre. Channels were sometimes seen beginning at the start
of the sample (the C region) and ending at the presence of tubules
(Figure 2c), and they either extended strictly in parallel or merge and
split into a series of interconnected pathways. They often (but not
always) traversed more than one field of view (Figure 2c,d). In some
cases, channels appeared further along the sample and were occa-
sionally seen to end direc tly in a tubule (Figure 2c–f). Whilst in many
cases channels appear as ‘open- top’ grooves (e.g., Figure 2b,e,f), it
was not always clear from 2D imaging whether in some cases they
may be enclosed tubes (e.g., Figure 2c,d). Ultimately, future work
exploring the 3D ultrastructure of channels will be needed to de-
termine this.
Channels were observed to some degree in most species ex-
amined (24 species (92.3% of all species) and 46 folds (88.5% of
fold samples); Figure 3). There were only 2 species where channels
were obser ved in one fold but not the other (black francolin and
Madagascarpartridge(Margaroperdix madagarensis)), and only 2 spe-
cies where no channels were observed in either fold of the sample
FIGURE 2 (a)Anexampleofthe
typical presentation of UVJ tissue that
does not contain channels or SSTs,
from European quail (Coturnix coturnix)
at 50× magnification. (b) Channel-
like structures of the UVJ tissue from
helmeted guineafowl (Numidea meleagris)
at 50× magnification. (c and d) Channel-
like structures of the UVJ tissue in grey
francolin (Francolinus pondicerianus) at
50× magnification. Each image crosses
3 fields of view. Channels are very
long and end in a region populated by
tubules. In (c), you can see a large globular
(agglomerate) branched tubule (ag) and
channels that appear to end directly in
a tubule with obvious stored sperm (sp).
(d) Particularly clear example of a long
channel ending in a forked tubule. (e)
Slightly different presentation of channel
tissue in mountain quail (Oreortyx pictus)
at 100× magnification. Channels are
shorter in length, funnel- like and end
directly in tubules. (f) Example of channels
apparently ending in a highly branched
tubule in mountain quail at 200×
magnification. Stored sperm is visible
in more than one branch. Arrows point
to representative (but not all) examples
of channel walls (cw), tubule entrances
(te), stored sperm (lighter/brighter blue)
(sp), and agglomerate tubules (ag) (where
relevant). Images are all orientated with
the cloaca closest to the left side and the
uterus closest to the right.

|
7 of 17
ASSERSOHN et al.
(European quail and Chinese quail (Excalfactoria chinensis)). The cat-
egorisation of channel tissue was significantly repeatable between
observers (RLinkscale = 0.99,CI = 0.99,1.00,p < .0001).
3.4 | Transitional tissue
Across most species, we observed structures that we could neither
conclusively define as tubule tissue nor vaginal or uterine tissue.
These structures always directly preceded the uterus, and often
looked similar in morphology to tubules seen further down the sam-
ple, but in this case were either poorly defined, and/or much smaller
and increasing in density until merging into uterine tissue. We refer
to these structures as ‘transitional tissue’, as they appear transitional
between regular tubules and uterine tissue. In many cases, when
viewed at high magnification, very small tubule- like entrances could
be seen (see Figure 4c,d for a comparison of transitional tissue at
low and high magnification). Transitional tissue was not considered
to be tubule tissue as we did not deem it likely to be functioning in
sperm storage.
Transitional tissue was never seen to contain stored sperm and
was obser ved to some degree in 22 species (84.6% of species) and
40 folds (76.9% of fold samples) with only 4 species not display-
ing transitional tissue in either fold of the sample (bobwhite quail
(Colinus virginianus), European quail, helmeted guineafowl (Numidia
meleagris), and white eared pheasant (Crossoptilon crossoptilon))
(Figure 3). The distinction between tubules approaching the U re-
gion of the UVJ and early transitional tissue was sometimes difficult
and required a degree of subjectivity; however, the categorisation of
transitional tissue was significantly repeatable between observers
(RLinkscale = 0.97,CI = 0.98,1.00,p < .0001).
3.5 | Tubule categorisation
In addition to transitional tissue and channel tissue, we found five
distinct categories of tubules.
1. Straight unbranched tubules – the simplest structure, with
a clear single lumen. They may bend but they do not coil
or branch. At least some proportion of tubules within every
sample was straight unbranched, making it the most common
type of tubule (Figure 3). The categorisation of straight un-
branched tubules was highly repeatable between observers
FIGURE 3 Phylogenyforthe26speciesexamined,andthetubuleandtissuetypesobservedwithinthosespecies.Cellsmarkedwith‘X’
indicate that this species exhibited this tubule or tissue type to some degree in at least one fold of the sample. Galliformes phylogeny with
time- calibrated branch lengths was obtained from Stein et al. (2015), which was trimmed using the R package treeplyr (Harmon, 2023).

8 of 17
|
ASSERSOHN et al.
FIGURE 4 (a)Exampleoftypical(non-transitional)tissueattheuterusendofthevaginainblackfrancolin(Francolinusfrancolinus)at
100× magnification. (b) Transitional tissue in Japanese quail (Coturnix japonica) at 50× magnification. T issue appears like poorly defined
tubuletissue,muchsmallerthantubulesandinplacesdifficulttodistinguishfromsurroundingtissue.(c)Mountainquail(Oreortyx pictus)
at 50× magnification. Distinct clusters of agglomerate- like tissue that is distinct from uterine tissue but with a poorly defined ‘fluffy’
appearance. When viewed under high magnification, looks like smaller less well- defined agglomerate tissue (see d for an example). (d)
TransitionaltissuefromMountainquailat20 0× magnification. (e) Transitional tissue from Harlequin quail (Coturnix delegorguei) at 50×
magnification. Tissue appears like extremely densely packed tiny tubules, distinct from storage tubules in being overall significantly smaller,
but with tubule like entrances only seen when examined under high magnification. (f) Transitional tissue in Temminck's tragopan (Trago pan
temminckii) at 100× magnification, showing a clear transition from tubule- like (but poorly defined) tissue in the vagina (to the left) to dense
uterine tissue (on the right). Arrows point to representative (but not all) examples of tubule- like entrances (tle). Images are all orientated with
the cloaca closest to the left side and the uterus closest to the right.
FIGURE 5 (a)Exampleofstraightunbranchedtubulesinbamboopartridge(Bambusicola thoracicus) at 200× magnification; (b) example
of straight unbranched tubules in common pheasant (Phasianus colchicus) at 100× magnification; (c) example of straight branched tubules
in mountain quail (Oreortyx pictus) at 200× magnification; (d) example of a branched tubule that is also coiled, in bobwhite quail (Colinus
virginianus) at 200× magnification. It is common for coiled tubules to also be branched; (e) An example of a large number of coiled tubules in
bobwhite quail at 100× magnification; (f) An example of a branched tubule where one branch is coiled but not the other, in mountain quail at
200× magnification; (g) An example of a mini- tubule and a regular sized tubule in the same field of view in wild turkey (Meleagris gallopavo)
at 100× magnification. The mini- tubule has visible stored sperm; (h) An example of many mini- tubules in close proximity to one another in
bobwhite quail at 100× magnification. Some tubules are so short in length they appear as little more than an entrance; (i) An example of
agglomerate tubules in black francolin (Francolinus francolinus) at 50× magnification. Tubules are so highly branched they appear as globular
rather than tubular structures; (j) An example of agglomerate tubules in grey junglefowl (Gallus sonneratii) at 100× magnification. Tubules
are so highly branched they form some unique and interesting shapes. Arrows point to representative (but not all) examples of the tubule
lumen (lu), the tubule epithelium (ep), tubule entrances (te), and stored sperm (lighter/brighter blue) (sp). Images are all orientated with the
cloaca closest to the left side and the uterus closest to the right. Image title abbreviations are as follows: Straight unbranched – SU; straight
branched–SB;coiled–Coi;Mini-tubules–Min;Agglomeratetubules–Agg.

|
9 of 17
ASSERSOHN et al.
(RLinkscale = 0.80, CI = 0.40, 0.99, p = <.0001). Examples can be
seen in Figures 2, 5, and 6.
2. Straight branched tubules – generally possessing between 1
and 3 long branches (though occasionally more), without coils or
twist s. As with straight unbranched tubules, branches may bend.
These are commonly described in the literature and were present
to some degree in 8 species (30.8% of species). For each species
that contained straight branched tubules, they were common
and found in both folds of the sample (Figure 3). The categorisa-
tion of straight branched tubules was highly repeatable between

10 of 17
|
ASSERSOHN et al.
observers (RLinkscale = 0.93,CI = 0.95,1.0 0,p = <.0001). Examples
can be seen in Figures 2, 5 and 7.
3. Coiled tubules – highly coiled or twisted, in some cases small,
‘blobby’ and unbranched, whilst in others they appeared longer
and branched. Coiled tubules were un common, appearing in just 3
species (11.5%) (Figure 3). For each species that contained coiled
tubules, they were common and found in both folds of the sam-
ple. In some cases, coiling was so extreme the tubule appeared
‘corkscrewed’ in shape. The categorisation of coiled tubules was
highly repeatable between observers (RLinkscale = 0.79, CI = 0 .8 9,
1.00, p = .0 02).ExamplescanbeseeninFigure 5.
4. Mini-tubules–canbedefinedasbeing<30% the size of the larg-
est tubule in the sample, and despite being very small relative to
the longest tubules, often contained stored sperm. Due to the
large number and consistency in length of mini- tubules, we are
confident that we were obser ving their full length in most cases.
Occasionally, mini- tubules appeared as barely longer than the
tubuleentrance.Mini-tubules wereverycommon, appearingto
some degree in every species and in 50 folds (96.2% of fold sam-
ples) (Figure 3).Mini-tubulescouldappearalongtheentirelength
of the sample but were particularly common towards the U region
as the length of tubules generally declined closer to the uterus.
Despite being small in length, the lumen and epithelial wall were
similar in thickness to other tubules within the sample, and the
tubule entrance was similar in size to other tubules in the sam-
ple. This distinguishes them from tubule- like transitional tissue
which are significantly smaller (e.g. Figure 4e). The categorisa-
tion of mini- tubules was highly repeatable between observers
(RLinkscale = 0.99,CI = 0.98,1.00,p = <.0001). Examples can be seen
in Figure 5.
5. Agglomerate tubules – tubules were highly branched and ap-
peared almost globular or star- like in shape, sometimes appearing
as dense clusters of tubules with multiple entrances. It was often
unclear whether these clusters shared a lumen or were in fact
FIGURE 6 (a)Exampleoftubule/swith‘expanded’entrances,inhelmetedguineafowl(Numidia meleagris) at 100× magnification, with
tubule/s that appear to have very wide entrances. It is not clear whether this is two overlapping tubules or a single tubule with multiple
entrances. (b) Examples of ‘bulbous’ t ypes in Chinese quail (Excalfactoria chinensis) at 50× magnification, with 2 tubules that appear engorged
relative to the other tubules in the sample. One tubule has obvious stored sperm. (c) Examples of ‘bulbous’ types in Swinhoe's pheasant
(Lophura swinhoii) at 20 0×magnification,withasmallernumberofstoredsperm.(d)Examplesof‘bulbous’typesinMadagascarpartridge
(Margaroperdix madagarensis) at 50× magnification with clear stored sperm. Arrows point to representative (but not all) examples of tubule
entrances (te), stored sperm (lighter/brighter blue) (sp) and some non- tissue debris that has picked up the dye (de). Images are all orientated
with the cloaca closest to the left side and the uterus closest to the right. Image title abbreviations are as follows: Bulbous – Bul; expanded –
Exp.
FIGURE 7 Categorisationflowchar tfordeterminingtubuletypeinGalliformes.Imageswithinthisflowcharthavebeencroppedand
expanded for easy visualisation, but are taken from the following species: mini- tubules – bobwhite quail (Colinus virginianus) at 100×
magnification; transitional tissue (top) – Harlequin quail (Coturnix delegorguei) at 50× magnification; channel tissue – grey francolin
(Francolinus pondicerianus) at 100× magnification; straight branched – Reeves pheasant (Syrmaticus reevesii) at 100× magnification; straight
unbranched – common pheasant (Phasianus colchicus) at 100× magnification; transitional tissue (bottom) – black francolin (Francolinus
francolinus) at 50× magnification; agglomerate – black francolin at 100× magnification; coiled – bobwhite quail at 100× magnification. This
flowchar t has also been provided in a dichotomous key format in Appendix S1 , which may prove useful for practical applications.

|
11 of 17
ASSERSOHN et al.
distinc t clusters of small agglomerate tubules. Transitional tissue
may often appear agglomerate but has a less distinct form and ap-
pears only towards the U region of the sample (e.g. Figure 2c,d).
Agglomerate tubules were found in 9 species (34.6% of species)
and 15 folds (28.8% of fold samples) (Figure 3). There were 3 spe-
cies in which agglomerate tubules were found in one fold of the

12 of 17
|
ASSERSOHN et al.
sample but not the other (Madagascar partridge, mountain quail,
and red legged partridge). The categorisation of agglomerate tu-
bules was highly repeatable between observers (RLinkscale = 0.93,
CI = 0.95,1.00,p = <.0001). E xamples can be seen in Figures 5 and
7.
In addition to the 5 categories above, we noticed several sam-
ples contained tubules with interesting features that were not con-
sistent or common enough to be considered separate categories, but
are of note and worth discussing: first , we observed some tubules
with unusually large/wide or ‘expanded’ entrances (Figure 6a) in hel-
meted guineafowl, Lady Amherst's pheasant and Reeve's pheasant
(Syrmaticus reevesii); and second, we observed some tubules that
were bulbous or inflated along their length (but had normal sized
entrances) compared to the rest of the tubules in the C region of
Chinese quail,Madagascar partridgeandSwinhoe'spheasant sam-
ples (Figures 3 and 6b–d).
3.6 | Categorisation flow chart
We created a flowchart for categorising tubule type based on our
observations across species (Figure 7; and reproduced as a dichoto-
mous key in the supplementary material). We found this process ap-
propriate for categorising the tubules found in the species examined
here, with the quality of images produced from our methodology,
but recognise that there are likely to be more categories of tubules
across other species/taxa and encourage researchers exploring tu-
bule diversity to expand on our categories as necessary. Figure 3
provides a summary of the categories assigned to each species.
4 | DISCUSSION
Avian SSTs are commonly described as ‘simple tubular invaginations’,
located within a small strip of vagina tissue bordering the uterus,
known as the UVJ (Bakst, 20 11; Sasanami, 2017). They are widely
assumed to look – and function in – the same way across all birds,
according to the most frequently examined species: the domestic
chicken, turkey, Japanese quail, and zebra finches. Here, we provide
evidence that Galliformes – a large group of ground- dwelling birds
– exhibit striking and surprising variation in SST morphology across
species. We propose the variety of tubules we observed across 26
species can be partitioned into 5 distinct categories: (1) Straight un-
branched tubules; (2) straight branched tubules; (3) coiled tubules;
(4) mini- tubules; and (5) agglomerate tubules. We also show that
above and beyond the apparent diversity in SST morphology, the
surrounding tissue of the UVJ region appears to vary between spe-
cies in some respects and includes at least 2 additional features: (1)
channels and (2) transitional tissue, the function (if any) of which is
currently unknown but warrants further investigation. Additionally,
whether tubule function varies by category is unknown, but may
have important consequences for our understanding of sperm stor-
age and post- copulatory sexual selection in general.
In most samples, tubules were found throughout the length of
the UVJ. However, in half of all species, tubules did not always ex-
tend to the distal region of the sample that borders the uterus. The
region directly preceding the uterus was often populated by what
we refer to here as ‘transitional tissue’. Transitional tissue could not
be defined as tubule tissue, surrounding vaginal tissue, or uterine
tissue. It of ten appeared to contain smaller or poorly defined forms
of the tubules seen further down the sample (Figure 4). While many
of the structures observed in transitional tissue appeared to have
entrances (when viewed under high magnification), due to their
small size and the fact they were never obser ved to contain stored
sperm, it seems unlikely they function as sperm storage structures.
Previous work in yellow- headed blackbirds (Xanthocephalus xantho-
cephalus) has shown that SSTs vary in length and likelihood to store
sperm across regions of the UVJ (Briskie, 1996); it is possible that
the transitional tissue observed here represents the smaller and less
functional tubules found in that study. Another possibility is that
transitional tissue represents a transitional/developmental state
between the surrounding tissue and fully formed and functional
storage structures. We currently do not know the processes of SST
ontogeny, or fully understand how tubule morphology changes in
response to the regression and subsequent regrowth of the oviduct
in seasonally reproducing birds. Further work is needed to elucidate
the precise ultrastructure and functional significance of transitional
tissue.
Across the 26 Galliformes species examined here, we observed
channel- like depressions within the surrounding tissue of the UVJ in
most samples. Channels varied in appearance between species, but
were generally either very long, extending multiple fields of view, or
were shor t and funnel- like. Long channels (e.g. Figure 2b–d) often
ended in a region populated by SSTs (e.g. Figure 2c), or in some cases
ended directly in a tubule (Figure 2d). Shorter funnel- like channels
(e.g. Figure 2e,f) were most often observed at the start of the sam-
ple (i.e. the C end of the UVJ) and also frequently ended directly in
a tubule.
It is difficult to discern the precise orientation of structures from
a 2D image, and we cannot rule out the possibility that in some cases
tubules may be stacked directly on top of channels rather than being
connected; however, this seems unlikely to account for all observa-
tions, particularly for funnel- like channels that were frequent and
consistent in appearance. Further investigation is certainly war-
ranted to confirm the presence and 3D ultrastructure of channels,
and channel- tubule connections, and to explore their functional
significance. However, it seems likely that these channels are some-
how involved in sperm transport or storage (at least to some de-
gree). One convincing possibility is that channels simply provide a
more direct route to the region of the vagina populated by tubules,
orenhanceswimmingefficiency(Magdanzetal.,2015). The precise
mechanism of sperm transport through the vagina is not known: in
addition to sperm motility (Allen & Grigg, 1957), it has been sug-
gested that some additional mechanisms may exist including ciliary
movements of the surface epithelium of the vagina, contractions of
the oviduct (possibly in combination with ciliary movement), and/

|
13 of 17
ASSERSOHN et al.
or chemotactic guidance (Orr & Brennan, 2015; Sasanami, 2017). It
is possible that in species that have them, channels aid in the rapid
passage of sperm, and/or provide some protection from the envi-
ronment of the vagina (e.g. the anti- sperm immune response, vaginal
sperm ejection, and mechanical flushes), while shorter funnel- like
channels may simply direc t sperm into tubules once they reach the
site of storage. Alternatively, close association with the vaginal mu-
cosal epithelium might increase sperm contact with immune cells
(Yoshimura et al., 1997 ), providing an opportunity for cryptic female
choice to be intensified.
Sperm swimming mechanics are known to be strongly influ-
enced by their interaction with surfaces (Denissenko et al., 2012).
In mammals, sperm are thought to travel through micro- channels in
the reproductive tract, whereby motile sperm preferentially swim
near the boundary of channel walls and this may act as a guiding
mechanism towards the egg (Denissenko et al., 2012; Magdanz
et al., 2015). Fur thermore, there is some evidence that suggests that
boundary- following behaviour may be associated with higher sperm
DNA integrity in humans (Eamer et al., 2016), and that channels may
elicit intensified competition between sperm (Zaferani et al., 2019).
Whether the channels we observe here are analogous to the micro-
channels of the mammalian female reproductive tract are unclear,
but this possibility has potential implications for our understanding
of sperm transport and post- copulatory sexual selection within the
avian vagina and warrant s further investigation.
On the other hand, if tubules are directly connected to a channel,
then sperm boundary following behaviour (Denissenko et al., 2012)
may make it more difficult for sperm to exit storage without having
to first travel back down towards the channel entrance. This would
also be the c ase if long channels exist as enclosed tubes rather than
open grooves (which is yet to be determined). This possibility raises
the intriguing prospect that SSTs may not always function to store
the fertilising subset of sperm, but instead may act as sperm ‘bins’
that inhibit – rather than facilitate – sperm participation in fer til-
isation. Sperm stored in SSTs are usually assumed to make up the
fertilising subset of sperm because past studies in turkeys have
shown that the number of sperm residing in SSTs is strongly cor-
related with the number that reach the ovum (Bakst, 2 011; Brillard &
Bakst, 1990). However, evidence that stored sperm always form the
fertilising subset is lacking for other species, and so it may be worth
revisiting this assumption. There is also evidence from other taxa
that females can directly differentially remove sperm from storage,
for example through dif ferential ejection (aka sperm ‘dumping’), as
seen in Drosophila (Snook & Hosken, 2004).
While the vagina is expec ted to be the major site of sperm
selection in birds (Bakst, 2011; Orr & Brennan, 2015; Steele &
Wishart, 1996), SS Ts are now also predicted to play an imp ortant role
in sperm selection by way of non- random acceptance or release of
sperm. For example, sperm filling rate into SSTs has been found to be
unevenly distributed across the UVJ (Briskie, 1996; Ito et al., 2011;
Sasanami et al., 2015); and in zebra finches, SSTs have been found to
vary widely in diameter, possess constricted entrances that may act
asa controlledbarrierto sperm entry (Mendoncaet al.,2019), and
differentially store sperm from different inseminations (Hemmings &
Birkhead, 2017 ). Sperm dumping and selective sperm displacement
in other taxa (Barnett et al., 1995; Gasparini et al., 2018; Lüpold
et al., 2012; Snook & Hosken, 2004) also suppor ts the role of stor-
age structures in sperm selection. In bats, species that store sperm
also appear to produce more sperm, and testis size is correlated with
sperm storage duration suggesting that longer sperm storage is as-
sociated with increased sperm competition (Orr & Brennan, 2015;
Orr & Zuk, 2013). The large variation in SST morpholog y observed
in the current study may lend support to the hypothesis that SSTs
exhibit functional differences that could influence the outcome of
sperm competition. In accordance with this, we also observed a
non- uniform distribution of stored sperm among tubules, suggest-
ing individual SSTs may vary in their ability to accept, store and re-
lease sperm. Sperm were also commonly observed distributed along
the entire length of the tubules, rather than being congregated just
within the blind- end; however, we did not notice any patterns in
sperm distribution across species or SST categor y.
Theoretically, tubule shape and length may influence the speed
or order in which sperm are released from storage. Briskie (1996)
found that in yellow- headed blackbirds, uterine- end SSTs were
smaller and later to mature but released a greater number of sperm
during the egg- laying period, suggesting that smaller tubules may
accept sperm later, but release them more quickly, than longer tu-
bules. Accordingly, sperm stored in the shorter mini- tubules we ob-
serve here (e.g. Figure 4g), may exit storage more quickly than sperm
stored at the end of longer or more highly branched tubules (e.g.,
Figure 5c). Branch length and distance from the SST entrance will
presumably influence how easily/fast sperm are released, potentially
providing a mechanism of control over the order of sperm release
from storage. Variation in tubule shape within samples may there-
fore also help to reduce mating order effects, by creating variation
in ‘the playing field’, in which sperm from different males are stored
in a variet y of tubules of varying leng th, branch number, and com-
plexity, and with a varying advantage/disadvantage over the timing
of release (Hemmings & Birkhead, 2017).
We observe d that agglome rate tubules a re often extr emely highly
branched, complex, and diverse in morphology (e.g. Figure 5i,j). How
this structural diversity might translate to variation in SST function
is unclear and introduces several questions. For example, do all
branches function in the same way or are there selective features
that allow branches to var y in their ability to accept sperm? Does
mating order influence which branch sperm are accepted into? Do
the number of branches of a tubule influence the ease with which
sperm can exit the tubule upon release? Given the degree of varia-
tion in SST structures observed here, and the highly diverse nature
of sperm cells (Pitnick et al., 2009), a logical line of questioning would
be to assess whether SST morphology and complexity correlate with
sperm morphology, sperm storage duration, and/or sexual selection
intensity (which is typically correlated with mating system) across
species. One possibility is that highly branched or complex tubule
structures introduce variation in the ease with which sperm can
enter or reach the most protective regions of the storage tubule.

14 of 17
|
ASSERSOHN et al.
For example, if more vigorous sperm are better able to reach more
distal branches, they may be better protected from further selective
processes such as sperm ejection or the anti- sperm response within
the vagina. Exploring differential storage of sperm of varying quality
across a variety of tubule types may shed light on whether tubule
morphology is linked to sperm selection.
In the case of coiled tubules (e.g., Figure 5d–f), the tightness of
coils or their length may provide an even more extreme degree of
manipulation over the ability of sperm to enter storage, or the tim-
ing and speed of sperm release. The structure of coiled tubules has
some interesting parallels with the coiled vaginas of some species
of waterfowl, where the coiled shape of the reproductive tract has
a sexually antagonistic co- evolutionary relationship with the coiled
penis of the male (which is coiled in the opposite direction to the
vagina, making intromission more difficult) (Brennan et al., 2010).
In addition to their spiral structure, the vaginas of these species
also have multiple blind- ended ‘pouches’, in which the penis can be
directed to act as a mechanism of female control over sperm use
following forced copulations (Brennan et al., 2010). Figure 2f pres-
ents a striking example of a tubule in which one branch (notably, the
branch closest to the tubule entrance) is coiled whilst the other is
not. One possibility is that coiled tubules function in a similar way to
the blind- ended and spiral- shaped vaginas of ducks, possibly provid-
ing a sperm ‘bin’ by which a proportion of sperm are prevented from
being released, or released quickly enough, to par ticipate in sperm
competition.
Another comparison can be drawn with the uterine muscular
coiling (UMC)of viperidsnakes. UMCinvolvesa contractionof the
innermost layers of the UVJ, which causes the tissue to form a coiled
shape (not dissimilar to the coiled appearance of tubules seen in
Figure 5d–f ) (Muniz-Da-Silva et al ., 2020). It is thought t hat UMC
may function as a mechanism of sperm storage by maintaining the
position and viability of sperm. There is evidence that SSTs in tur-
keys are innervated, and individual SSTs house F- actin rich terminal
webs in the epithelium that exhibit a coiled appearance, suggesting
they are capable of contraction (Freedman et al., 2001). Nerve fibres
have also been observed in close association with SSTs in the al-
pine accentor (Prunella collaris) (Chiba & Nakamura, 2001), although
contractile elements were not found in this case. In support of the
contractile potential of SSTs, a contraction- like change to SST mor-
phology was observed in Japanese quail following an injection of
progesterone (which is thought to be one factor that triggers sperm
release from tubules) (Ito et al., 2011), suggesting that individual
SSTs are capable of contraction and relaxation under presumably
fine temp oral control. It may therefore be possible that tubule coiling
is a plastic mechanism (under neural or hormonal control) for main-
taining and protecting sperm in storage. Interestingly, it is reported
thatUMCinsnakesbecomeslessvisibleintissuepreservedin10%
formalin relative to fresh tissue. It may therefore be worth exam-
ining the morphology of SSTs across birds using fresh, rather than
preser ved tissue, and across different stages of reproduction (e.g.,
before, during, and after sperm acceptance and release). Exploring
SST morphology in fresh tissues will ensure that any features lost as
a result of the preservation process are uncovered. Coiled tubules
have in fact been obser ved at least once before, in a study examining
the morphology of SSTs in the American kestrel (Falco sparverius)
(Bakst & Bird, 1987). Their shape was only alluded to briefly and their
significance was not discussed at the time nor (to our knowledge)
since. We believe our images provide the first evidence for such
structures in Galliformes. While we have hypothesised several pos-
sible functional explanations for coiled SSTs, further work is needed
to uncover the significance of these structures for our understand-
ing of SST function and post- copulatory sperm storage and selection
in general.
Finally, we observed two additional features that were uncom-
mon and inconsistent within samples, suggesting they were not
appropriate for inclusion as separate categories. (1) Expanded en-
trances – these unusual tubules were observed to have abnormally
wide and funnel- like entrances. SSTs have been shown to have
constri cted gate-like entr ances (Mendo nca et al., 2019), and so it
could be that these tubules were simply in the processes of relaxing
following sperm release or prior to sperm acceptance. Future work
exploring SST morphological changes through time is a challenging
but necessary step to understanding the process of sperm accep-
tance and release. (2) Bulbous tubules – whether these are typic al
structures for these species is unclear, but these tubules were seen
containing stored sperm and are clearly functional. Further work ex-
ploring intraspecific variation in tubule morphology will be needed
to explore this. If these structures do appear to be consistent fea-
tures within and across species, re- evaluation of their inclusion as a
separate tubule category may be needed.
5 | CONCLUSIONS
SSTs in birds are often described as ‘simple tubular invaginations’.
Whilst that may be true of the tubules typically seen in well studied
birds like chickens and turkeys, we find that across other Galliformes,
tubule structure is far more variable, diverse, and complex than
previous assumed. The variation in tubule and surrounding tissue
structure we obser ve here may have important implications for our
understanding of sperm storage tubule function in general, with
broader consequences for our understanding of post- copulatory
sperm selection. Fur ther research is needed to quantify this vari-
ation across Galliformes, explore variation across other species of
birds, and determine the functional significance of the structures
we observe here. An obvious first step will be to explore whether
storage tubule types are associated with sperm morphology, sperm
storage capacity and sperm competition intensity. It may also be
useful for future work to examine different sperm storage types
using scanning and transmission electron microscopy, including the
use of 3D imaging techniques, which will help confirm their ultra-
structure, internal morphology, and relationships with surrounding
tissues. Ultimately, these findings contribute to the growing body
of evidence across taxa that female sperm storage structures can
be complex, highly specialised and variable (Beese & Baur, 2006;

|
15 of 17
ASSERSOHN et al.
Berger et al., 2011; Holt & Fazeli, 2016; Hopkins et al., 2020; Lüpold
et al., 2013; Orr & Brennan, 2015; Ward, 2000).
AUTHOR CONTRIBUTIONS
Katherine Assersohn: Conceptualization (equal); data curation
(lead); formal analysis (lead); investigation (lead); methodology
(equal); project administration (equal); software (lead); validation
(lead); visualization (lead); writing – original draft (lead); writing
– review and editing (equal). J. Paul Richards: Conceptualization
(supporting); methodology (equal); writing – review and editing (sup-
porting). Nicola Hemmings: Conceptualization (equal); data curation
(supporting); funding acquisition (lead); investigation (supporting);
methodology (equal); project administration (equal); resources
(lead); supervision (lead); validation (suppor ting); writing – review
and editing (equal).
ACKNOWLEDGEMENTS
We thank Jamie Thompson, Emily Glendenning, Alex Ball, and
Rebecca Bastin for assistance during dissections and Ian Clark for
providing fresh cadavers for dissection.
FUNDING INFORMATION
KA was supported by the Natural Environment Research Council
ACCE Doctoral Training Part- nership (DTP) (grant number NE/
S00713X/1). NH was supported by a Royal Society Dorothy Hodgkin
Research Fellowship (DHF16020 0) and a University of Sheffield
Woman Academic Returners' Programme ( WARP) award which
funded JPR's position.
CONFLICT OF INTEREST STATEMENT
The authors declare there are no conflicts of interest.
DATA AVAIL ABILI TY STATEMENT
No quantitative data were generated.
ORCID
Katherine Assersohn https://orcid.org/0000-0002-1085-0266
Nicola Hemmings https://orcid.org/0000-0003-2418-3625
REFERENCES
Allen, T., & Grigg, G. (1957). Sperm transpor t in the fowl. Australian jour-
nal of agricultural research, 8, 788–789. https:// doi. org/ 10. 1071/
AR957 078 8
Assersohn, K., B rekke, P., & Hemmings , N. (2021). Physiological factors
influencing female fertility in birds. Royal Societ y Open Science, 8,
202274. https:// doi. org/ 10. 1098/ rsos. 202274
Bakst ,M.R.(1994).Fate offluorescent stainedsperm followinginsem-
ination – new light on oviducal sperm tr ansport and storage in the
Tur key. Biology of Reproduction, 50, 987–992. https:// doi. org/ 10.
1095/ biolr eprod 50.5. 987
Bakst,M.R.(2011).Roleoftheoviductinmaintainingsustainedfertility
in hens. Journal of Animal Science, 89, 1323–1329. https:// doi. org/
10. 2 527/ jas. 2010- 3663
Bakst ,M.R.,&Bauchan,G.(2015).Apicalblebsonspermstoragetubule
epithelial cell microvilli: Their release and interaction with resident
sperm in the Turkey hen oviduct. Theriogenology, 83, 1438–1444.
https:// doi. org/ 10. 1016/j. theri ogeno logy. 2015. 01. 016
Bakst , M. R., & Bird, D. M. (1987). Localization of oviduc tal sperm-
storage tubules in the American kestrel (Falco sparverius). The Auk.,
104, 321–324.
Barnet t,M.,Telford,S. R., & Tibbles,B.J. (1995).Femalemediation of
sperm competition in the millipede AIIoporusuncinatus (Diplopoda:
Spirostreptidae). Behavioral Ecology and Sociobiology, 36, 413 –419.
Beese, K., & Baur, B. (2006). Expandable spermatheca influences sperm
storage in the simultaneously hermaphroditic snail Arianta arbus-
torum. Invertebrate Reproduction and Development, 49, 93–101.
htt ps:// doi. o rg/ 10. 10 80/ 07924 259. 2006. 9652198
Berger,D.,Bauerfeind,S.S.,Blanckenhorn,W.U.,&Schäfer,M.A.(2011).
High temp eratures reveal cryptic gen etic variation in a p olymorphic
female sperm storage organ: Evolvability of female reproductive
morphology. Evolution, 65, 2830–2842. ht tps :// doi. o rg/ 10. 1111/j .
1558- 5646. 2011. 01392. x
Birkhead, T. R. (1992). Sperm storage and the fertile period in the
Bengalese finch. The Auk, 109, 620–625.
Birkhead, T. R., & del Nevo, A . J. (1987). Egg for mation and the pre- laying
period of the common guillemot (Uria aalge). Journal of Zoology, 211,
83–88. https:// doi. org/ 10. 1111/j. 1469- 7998. 1987. tb074 54 . x
Birkhead,T.R.,&Møller,A.P.(1990).Numbersandsizeofspermstorage
tubules and the duration of sperm storage in birds: A comparative
study. Biological Journal of the Linnean Society, 45, 363–372. h t tp s://
doi . org/ 10. 1111/j. 1095- 8312. 1992. tb00 6 49. x
Bobr, L., Lorenz, F. W., & Ogasawara, F. X. (1964). Distribution of sper-
matozoa in the oviduct and fertilit y in domestic birds I. Residence
sites of spermatozoa in fowl oviducts. Journal of Reproduction and
Fertility, 8, 39–47.
Bobr, L., Ogasawara, F. X., & Lorenz, F. W. (1964). Distribution of sper-
matozoa in the oviduct and fertilit y in domestic birds II. Transport
of spermatozoa in the fowl oviduct. Journal of Reproduction and
Fertility, 8, 49–58 .
Brennan, P. L. R., Clark, C. J., & Prum, R. O. (2010). Explosive eversion
and func tional mor phology of the duck penis support s sexual con-
flict in waterfowl genitalia. Proceedings of the Royal Society B, 277,
1309–1314. https:// doi. org/ 10. 1098/ rspb. 2009. 2139
Brillard,J.P.,&Bakst,M.R.(1990).Quantificationofspermatozoainthe
sperm- storage tubules of Turkey hens and the relation to sperm
numbers in the perivitelline layer of eggs. Biology of Reproduction,
43, 271–275. https:// doi. org/ 10. 1095/ biolr eprod 43.2. 271
Briskie, J. (1996). Spatiotempor al patterns of spe rm storage and las t- male
sperm precedence in birds. Functional Ecology, 10, 375. ht t ps :// do i .
org/ 10. 2307/ 2390286
Briskie, J. V., & Birkhead, T. R. (1993). A review of the methods used to
study the anatomy of avian sperm storage. Ornis Scandinavica, 24,
323. https:// doi. org / 10. 23 07/ 3676795
Briski e, J. V.,& M ontgomeri e, R. (1993). Patter ns of sperm sto rage in
relation to sperm competition in passerine birds. The Condor, 95,
44 2–45 4. https:// doi. org/ 10. 2307/ 1369366
Chiba, A.,& Nakamur a,M.(20 01).Microscopic structure ofthesperm
storage tubules in the polygynandrous alpine accentor, Prunella col-
laris (Aves). Acta Zoologica, 82, 299–306. htt ps:// doi. org/ 10. 1046/j .
1463- 6395. 2001. 00091. x
Cramer, E. R. A., Yilma, Z. B., & Lifjeld, J. T. (2023). Selection on sperm
size in response to promiscuity and variation in female sperm stor-
age organs. Journal of Evolutionary Biology, 36, 131–143. ht t ps :// do i .
org / 10. 1111/ jeb. 14120
Denissenko, P., Kantsler, V., Smith, D. J., & K irkman- Brown, J. (2012).
Human spermatozoa migration in microchannels reveals boundary-
following navigation. Proceedings National Academy of Sciences
United States of America, 109, 8007–8010. https:// doi. org/ 10. 1073/
pnas. 12029 34109
Eamer, L., Vol lmer, M., Nosr ati, R., S an Gabri el, M. C. , Zeidan, K ., Zini,
A., & Sinton, D. (2016). Turning the corner in fertility: High DNA

16 of 17
|
ASSERSOHN et al.
integrity of boundary- following sperm. Lab on a Chip, 16, 2418–
2422. https:// doi. org/ 10. 1039/ C6LC0 0490C
Eberhard, W. G. (1996). Female control: Sexual selection by cryptic female
choice. Princeton University Press.
Freedman ,S., Akuffo,V.,& Bakst,M. (2001). Evidence fortheinner va-
tion of spe rm storage tubules in the ovidu ct of the Turkey (Meleagris
gallopavo). Reproduction, 121, 809–814. https:// doi. org/ 10. 1530/
rep.0. 1210809
Gasparini, C., Daymond, E., & Evans, J. P. (2018). Extreme fertilization
bias towards freshly inseminated sperm in a species exhibiting
prolonged female sperm storage. Royal Society Open Science, 5,
172195. https:// doi. org/ 10. 1098/ rsos. 172195
Harmon, L. (2023). Treeplyr: “Dplyr” functionality for matched tree and
data objects. R package version 0.1.11, https:// github. com/ uyedaj/
treeplyr
Hatch ,S. A.(1983 ).M ech anismandeco log icalsign ifi ca nce ofsper mstor-
age in the northern fulmar with reference to its occurrence in other
birds. The Auk, 100, 593–600 .
Hemmings, N., & Birkhead, T. (2017). Differential sperm storage by fe-
male zebra finches Taeniopygia guttata. Proceedings of the Biological
Sciences, 284, 2 0171032 .
Hemmings, N., & Birkhead, T. R. (2015). Polyspermy in birds: Sperm
numbers and embryo survival. Proceedings of the Royal Society B:
Biological Sciences, 282, 20151682. https:// doi. org/ 10. 1098/ rspb.
2015. 1682
Hemmings, N., & Birkhead, T. R. (2020). Extraordinary sperm to egg ra-
tios in seabirds. The Auk, 137, ukaa052. https:// doi. org/ 10. 1093/
auk/ ukaa052
Hemmings, N., Birkhead, T. R., Brillard, J. P., Froment, P., & Briere, S.
(2015). Timing associated with oviductal sperm s torage and release
after artificial insemination in domestic hens. Theriogenology, 83,
1174–1178. https:// doi. org/ 10. 1016/j. theri ogeno logy. 2014. 12. 022
Higginson, D. M., Miller, K. B., Segraves, K. A., & Pitnick, S. (2012).
Female reproductive tract for m drives the evolution of complex
sperm morphology. Proceedings. National Academy of Sciences.
United States of A merica, 109, 4538–4543. https:// doi. org/ 10 . 1073/
pna s. 11114 74109
Hiyama, G., Matsuz aki, M., Mizushima, S., Dohra, H., Ikegami, K.,
Yoshimura, T., Shiba, K., Inaba, K., & Sasanami, T. (2014). Sperm ac-
tivation by heat shock protein 70 suppor ts the migration of sperm
released from sperm storage tubules in Japanese quail (Coturnix
japonica). Reproduction, 147, 167–178. https:// doi. org/ 10. 1530/
RE P- 1 3- 0 43 9
Holm, L ., Ekwall, H ., Wishart, G. J., Ridderstråle, Y., & Ridderstrale, Y.
(2000). Localiz ation of calcium and zinc in the sperm storage tu-
bules of chicken, quail and Turkey using X- ray microanalysis. Journal
of Reproduction and Fertility, 118, 331–336. https:// doi. org/ 10.
153 0/ rep rod/ 118.2 . 331
Holt, W. V.(2011). Mechanismsofspermstorage in the female repro-
ductive tract: An interspecies comparison. Reproduction in Domestic
Animals, 46, 68–74. htt ps:// doi . org/ 10. 1111/j. 1439- 0531. 2011.
01862. x
Holt, W. V., & Fazeli, A . (2016). Sperm storage in the female reproductive
tract. Annual Review of Animal Biosciences, 4, 291–310. ht t ps :// do i .
org/ 10. 1146/ annur ev- anima l- 02181 5- 111350
Hopkins, B. R., Sepil, I., & Wigby, S. (2020). Structural variation in
Drosophila melanogaster spermathecal ducts and its association
with sperm competition dynamic s. Royal Societ y Open Science, 7,
20013 0. https:// doi. org/ 10. 1098/ rsos. 200130
Huang, A ., Isobe, N ., Obitsu, T., & Yoshimura, Y. (2016). Expression of li-
pases and lipid receptors in sperm storage tubules and possible role
of fatt y acids in sperm survival in the hen oviduct. Theriogenology,
85, 1334–1342. https:// doi. org/ 10. 1016/j. theri ogeno logy. 2015. 12.
020
Huang, A ., Isobe, N ., & Yoshimura, Y. (2017). Changes in localization
and density of CD63- positive exosome- like subst ances in the hen
oviduct with artificial insemination and their effec t on sperm via-
bility. Theriogenology, 101, 135–143. https:// doi. org/ 10. 1016/j. theri
ogeno logy. 2017. 06. 028
Ito, T., Yoshizaki, N., Tokumoto, T., Ono, H., Yoshimura, T., Tsukada,
A., Kansaku, N., & S asanami, T. (2011). Progesterone is as
sperm- releasing factor from the sperm- storage tubules in birds.
Endocrinology, 152, 3952–3962. https:// doi. org/ 10. 1210/ en.
2011- 0237
Khilla re, G. S., Sas try, K. V. H., Agr awal, R., Pr akash, A ., Mohan, J., &
Singh,R.P.(2018).M ole cul ara ndb ioc hem icalev ent sinspe rms tor-
age tubules of oviduct in birds. World's Poultry Science Journal, 74,
35–44. https:// doi. org/ 10. 1017/ S0 043 93391 700085X
Kustra,M.C.,& Alonzo,S.H.(2023).Thecoevolutionarydynamics of
cryptic female choice. Evolution Let ters, 7, 191–202. https:// doi. org/
10. 1093/ evlett/ qrad 025
Levine, B . A., Schuett, G. W., & Booth, W. (2021). Exceptional long- term
sperm storage by a female vertebrate. PLoS One, 16, e0252049.
https:// doi. org/ 10. 1371/ journ al. pone. 0252049
Lüpold,S.,Manier,M.K.,Berben,K.S.,Smith,K.J.,Daley,B.D.,Buckley,
S. H., Be lote, J. M., & Pit nick, S. (201 2). How multiv ariate ejacu-
late traits determine competitive fertilization success in Drosophila
melanogaster. Current Biology, 22, 1667–1672. https:// doi. org/ 10.
1016/j. cub. 2012. 06. 0 59
Lüpold, S., Pitnick , S., Berben, K. S., Blengini, C. S., Belote, J. M., &
Manier,M.K.(2013).Femalemediationofcompetitivefer tilization
success in Drosophila melanogaster. Proceedings. National Academy
of Science s. United States of Amer ica, 110, 10693–10698. ht t ps :// do i .
org / 10. 1073/ pna s. 130 09 54110
Magdanz,V.,Koch,B., Sanchez,S.,& Schmidt,O. G.(2015).Spermdy-
namics in tubular confinement. Small, 11, 781–785. https:// doi. org/
10. 1002/ smll . 2014 0 1881
Matsuzaki, M., Mizushima, S., Hiyama, G., Hirohashi, N., & Shiba, K.
(2015). Lactic acid is a sperm motilit y inactivation factor in the
sperm storage tubules. Scientific Reports, 5, 1–12. https:// doi. org/
10. 1038/ srep1 7643
Mendon ca, T., Cadby, A. J. , & Hemmings , N. (2019). Sperm gate keep-
ing: 3D imaging reveals a constricte d entrance to zebra finch sp erm
storage tubules. Biophysical Journal, 117, 2180–2187. ht t ps: // doi .
org / 10. 1016/j. bpj. 2019. 10. 038
Miller,G.T.,&Pitnick,S.(2002).Sperm-femalecoevolutioninDrosophila.
Science, 298, 1230–1233. https:// doi. org/ 10. 1126/ scien ce.
1076968
Muniz-Da-Silva,D.F., Passos, J., Siegel,D. S., & Almeida-Santos, S. M.
(2020). Caudal oviduct coiling in a viperid snake, Crotalus durissus.
Acta Zoologica, 101, 69–77. http s:// doi. org / 10. 1111/ a zo. 12 271
Ogasawara, F. X., Lorenz, F. W., & Bobr, L. W. (1966). Distribution of
spermatozoa in the oviduct and fertility in domestic birds III. Intra-
uterine insemination of semen from low- fecundity cocks. Journal of
Reproductive Fertility, 11, 33–41.
Orr, T. J., & Brennan, P. L. R. (2015). Sperm storage: Distinguishing
selective processes and evaluating criteria. Trends in Ecology &
Evolution, 30, 261–272. https:// doi. org/ 10. 1016/j. tree. 2015. 03.
006
Orr,T.J.,&Zuk,M.(2012).Spermstorage.Current Biology, 22, R8–R10.
https:// doi. org/ 10. 1016/j. cub. 2011. 11. 003
Orr, T.J. , & Zuk, M. (2013). D oes delaye d fertil ization f acilitat e sperm
competition in bats? Behavioral Ecology and Sociobiology, 67, 1903–
1913. http s:// doi. or g/ 10. 10 07/ s0026 5- 013- 1598- 2
Payne, L . F. (1914). Preliminar y report of vitality and activity of sperm cells
and artificial insemination of the chicken. Oklahoma Agricultural and
MechanicalCollege.
Pitnick , S., Hosken, D. J., & Birkhead, T. R. (2009). Sperm biology: An evo-
lutionary perspective (pp. 69–149). Academic Press.
R Core Team. (2023). R: A language and environment for statistical comput-
ing. R Foundation for Statistical Computing. https:// www. R- proje
ct. org/

|
17 of 17
ASSERSOHN et al.
Sasanami, T. (2017). Avian reproduction: From behaviour to molecules.
Springer. htt ps:// doi. o rg/ 10. 10 07/ 978- 981- 10- 3975 - 1
Sasanami,T.,Izumi,S.,Sakurai,N.,Hirata,T.,Mizushima,S., Matsuzaki,
M.,Hiyama,G.,Yorinaga,E.,Yoshimura,T.,Ukena,K.,&Tsutsui,K.
(2015). A unique mechanism of successful fertilization in a domes-
tic bird. Scientific Reports, 5, 7700. https:// doi. org/ 10. 1038/ srep0
7700
Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH image to
ImageJ: 2 5 years of image analysis. Nature Methods, 9, 671–675.
https:// doi. org/ 10. 1038/ nmeth. 2089
Shankar,G.,Gagan,T.A.,Kumari,T.R.S.,&Marathe,G.K.(2022).Sperm
storage by females across the animal phyla: A survey on the oc-
currence and biomolecules involved in sperm storage. Journal
of Experimental Zoolog y Part B – Molecular and Developmental
Evolution., 340, 283–297. https:// doi. org/ 10. 1002/ jez.b. 23189
Snook, R. R., & Hosken, D. J. (2004). Sperm death and dumping in
Drosophila. Nature, 428, 939–941. htt ps:// doi. o rg/ 10. 10 38/ nat ur
e02455
Steele,M.,&Wishart,G.(1996).Demonstrationthattheremovalofsialic
acid from t he surface of chicken spermatozoa impedes their trans-
vaginal migration. Theriogenology, 46, 1037–1044 . https:// doi. org/
10. 1016/ S0093 - 691X(96) 00268 - 3
Stein, R.W.,Brown,J.W.,& Mooers, A.Ø.(2015). Amoleculargenetic
time scale demonstrates Cretaceous origins and multiple diversi-
fication rate shifts within the order Galliformes (Aves). Molecular
Phylogene tics and Evolution, 92, 155–164. https:// doi. org/ 10. 1016/j.
ympev. 2015. 06. 005
Stoffel,M.A.,Nakagawa,S.,&Schielzeth,H .(2017).rptR:Repeatability
estimation and variance decomposition by generalized linear
mixed- effects models. Methods in Ecolog y and Evolution, 8, 1639–
1644. https:// doi . org/ 10. 1111/ 2041- 210X . 12797
Van Drimmelen, G. C. (1946). “Spermnests” in the oviduct of the do-
mestic hen. Journal of the South African Veterinary Association, 17,
42–52 .
Verma, O., & Cherms, F. (1965). Appearance of sperm and their persistenc y
in storage tubules of Turkey hens after a single insemination. Poultry
Science, 44, 609–613. https:// doi. org/ 10. 3382/ ps. 0440609
Verma, O. P., & Cherms, F. L. (1964). Obser vations on the oviduct of tur-
keys. Avian Diseases, 8, 19. https:// doi. org / 10. 2307/ 1587814
Wanless,S.,&Harris,M.P.(1986).Timespentatthecolonybymaleand
female guillemots Uria aalge and razorbills Alca torda. Bird Study, 33,
168–176. https:// doi. org/ 10. 1080/ 00063 65860 9476916
Ward, P. I. (2000). Cryptic female choice in the yellow dung fly
Scathophaga stercoraria (L.). Evolution, 54, 1680–1686. ht t ps : // doi .
org/ 10. 1111/j. 0014- 3820. 2000. tb007 12. x
Yoshimura, Y., Ok amoto, T., & Tamura, T. (1997). Localis ation of MHC
class II, lymphocytes and immunoglobulins in the oviduct of laying
and moulting hens. British Poultry Science, 38, 590–596. ht t ps : // doi .
org/ 10. 1080/ 00071 66970 8418041
Zaferani,M., Palermo,G. D.,&Abbaspourrad,A.(2019).Stricturesof a
microchannel impose fierce competition to select for highly mo-
tile sperm. Science Advances, 5, ea av2111. https:// doi. org/ 10. 1126/
sci adv. aav2111
SUPPORTING INFORMATION
Additional supporting information can be found online in the
Suppor ting Information section at the end of this article.
How to cite this article: Assersohn, K., Richards, J. P., &
Hemmings, N. (2024). The surprising complexity and diversity
of sperm storage structures across Galliformes. Ecology and
Evolution, 14, e11585. https://doi.org/10.1002/ece3.11585