Australia: The Land Where Time Began

A biography of the Australian continent 

Evolution of Avian Reproductive Behaviour - Preserved Avian Ovarian Follicles

According to the authors¹ it is indicated by data that crocodilians and birds, 2 groups of archosaurs, comprise an extant phylogenetic bracket for understanding the reproductive behaviour of dinosaurs, the behaviour being inferred from the nests and eggs that have been preserved, as well as gravid individuals (Zelenitski, 2006). There are many 'avian' traits that were already present in Parves, a clade that includes birds and closely related forms, and the early modern avian reproductive biology was already evolving (Zhou & Zhang, 2004). Non-avian maniraptorans already had daily oviposition and eggs that were asymmetrical and shells with a complex microstructure, and have been found to have protected their clutches (Norell et al., 1995; Sato et al., 2005; Varricchio et al., 1997). As with crocodilians, non-avian maniraptorans had 2 oviducts, unlike birds that have 1, eggs that were relatively smaller, and it has been suggested that they may not have turned their eggs as extant birds do (Zelenitski, 2006; Varricchio et al., 1997). In this paper the authors¹ report the first discovery of fossilised ovarian follicles, that were either mature or near mature, revealing a stage in dinosaur reproduction that was previously undocumented - reproductively active females that were near ovulation. These fossilised follicles were found in a specimen of Jeholornis and 2 enantiornithine birds from the Early Cretaceous in the lacustrine Jehol Biota of northeastern China, indicating the presence of 1 functional ovary in basal birds, though as a result of their lower metabolic rate relative to birds of the present, they retained morphologies that were primitive. As with crocodiles and parvian dinosaurs, they also indicate that sexual maturity was reached before skeletal maturity. According to the authors¹ follicular morphology differences between Jeholornis and enantiornithines have been interpreted as forming an evolutionary gradient leading from the reproductive condition that has been found in parvian dinosaurs to that of neornithine birds. Study of the 2 enantiornithines indicates this lineage may have evolved advanced reproductive traits in parallel with the neornithine lineage.

In the Tianyu Natural History Museum in Shandong, China (STM) there are 3 specimens of birds from the Mesozoic that have been found to contain mature ovarian follicles (a follicle is the primary oocyte and the containing sac (Gilbert, 1979)) preserved in their body cavity. STM2-51 slab and counterslab is the largest specimen and is referable to Jeholornis sp. is a basal bird that has a long bony tail, longer than that of Archaeopteryx. The specimen, that is in dorsoventral view, contains about 20 follicles that are ventral to the vertebral column and cranial to the pelvic girdle, and are overlapped by thoracic ribs. All the follicles are round and of subequal size, ranging in diameter from about 7.1-8.8 mm, though the authors¹ suggest there may have been more follicles than are observed in the specimens, possibly being obscured by overlap. The follicles are distributed in rows between the caudal end of the sternum and the pelvis in STM2-51, with the follicle numbers decreasing caudally, with proximally 3-4 and distally 1-2, though it has been suggested this pattern may be exaggerated by a crack laterally bordering the preserved follicles. The follicles are preserved permineralised by a dark microcrystalline mineral that is too thick to be eggshell. The follicles in the counterslab are preserved as pale pink impressions, though there are also some small crystals of the black mineral.

STM29-8 and STM10-45 are the 2 smaller specimens that are referable to the Enantiornithines, a large clade forming a sister group to the Ornithuromorpha, representing the first major avian radiation (O'Connor, 2009; O'Connor, Chiappa & Bell, 2011), is the clade to which extant birds belong, Neornithes. The proportions of the humerus/ulna and femur/tibiotarsus, and coracoid shape, all vary between the 2 specimens, indicating they are from different taxa, though it has not been found possible to identify them to species level as a result of poor preservation. Mature ovarian follicles have been preserved in the body cavity of both specimens that are visible in the dorsoventral, slab-counterslab, view. At least 12 follicles have been preserved in situ next to the synsacrum, along the left side of the vertebral column. There are also 2 follicles that have been displaced proximally that have been recognised near the left wing. All the follicles are almost circular, with a slight variation in size that ranges from a diameter of 5.8-8.8 mm. The follicles have been preserved as black carbonisation in the slab and as impressions in the counterslab. An uneven lattice structure of fibres has been preserved in the follicle surfaces that has been suggested to possibly be imprints of blood vessels from the perifollicular membrane that is highly vascularised.

Fewer follicles, at least 5, of subequal size ranging from 6.7-8.8 mm diameter, are preserved in Enantiornithes indet.STM10-45, that are located to the left of the vertebral column distal to the sternum and cranial to the pelvis, and the follicles are overlapped by a few rib elements that are disarticulated. There are 4 follicles that are associated together, that possibly obscure more follicles, that are well preserved, and caudally slightly displaced from the others is an additional follicle(s), making an estimated total of 5-7 follicles. There are small, black, circular spots of mineralisation on the surface of the follicles, and bacterial mediation of the preservation of the follicles in STM10-45 is indicated by the roundness of these structures.

The circular structures on the slab and counterslab, that have been interpreted as follicles, are clearly located in the body cavity of these fossil birds as they are preserved overlapping with bone. The suggestion that these structures could be seeds preserved in the stomach has been discounted because of their morphology and position (E.M.Friis, pers.com. to authors¹⁾, as well as their comparison with other seed types that have been observed in the Jehol fossils. A specimen that has been found that had been preserved with seeds in its stomach, Jeholornis prima, at the Institute of Palaeontology and Palaeoanthropology in Beijing, V13274 (Zhou & Zhang, 2002), there are more seeds, they are larger, with a diameter of about 10 mm, are of a different shape, being tapered at one end not circular, and have an ornamented surface. Unlike the follicles in Jeholornis STM2-51 that are located dorsally along the vertebral column and just cranial to the pelvis, which is consistent with the ovaries of living archosaurs. Some of the Jehol Birds, such as Sapeornis, Hongshanornis (Zheng et al., 2011). The crop is found cranial to the sternum in these specimens and the seeds are different from the follicles in that they are of uneven size and morphology - they are less round. The authors¹ say these are clearly not gastroliths, which are often found in many of the Jehol ornithuromorph birds (Zheng et al., 2011). It has been found that these stones are always preserved almost as in life being 3D, in large numbers, of various shapes, sizes and mineral composition, and with polished surfaces, and proportionally much smaller than structures of soft tissue that have been preserved in the case of these follicles. The preserved structures have a circular shape, which is consistent with the 2D preservation of a spherical structure, such as the mature follicle of the ovary, single edged cells. The authors¹ suggest that the follicles had not entered the oviducts as there is no evidence of preserved eggshell, and their clumping together, with no follicles caudally that remain in situ.

Crocodilians and birds, that are extant archosaurs, have reproductive habits that differ greatly, in terms of clutch size, nesting behaviour, degree of parental care, developmental strategy of the young, pose the question of when and why the avian reproductive traits evolved within the Dinosauria. It is indicated by the known evidence that the reproductive system of non-avian theropods retained some archosaurian traits that were primitive, such as 2 functional oviducts, eggs that were hyper ellipsoidal, and offspring that were precocious, they also had some derived bird-like characteristics, such as the deposition of 1 egg per oviduct per day, eggs that were relatively to body size larger, complex microstructures of eggshells, asymmetrical eggs and parental care (Zelenitski, 2006, Sato et al., 2005; Clark, Norell, Chiappe, Varricchio et al., 1997). The authors¹ say that within the Aves class itself information on the soft tissue reproductive organ anatomy, and the sequence and timing of changes in reproductive strategy are not well understood. They say these new birds provide the first glimpse of the reproduction of basal birds, also recording a fairly wide bracket within the phylogenetic tree of the birds from the Mesozoic, thereby revealing different stages of evolution.

The reproductive system of living archosaurs, as in all vertebrates, is usually divided into 2 separate parts, the ovary and the oviduct (Girling, 2002; Gill, 2007). Among the amniotes the birds are unique in having 2 ovaries and oviducts in the embryo but in the adult only the left ovary and oviduct are functional. The kiwi is unique among birds in a different way, the left and right ovaries developing and being functional, though the left oviduct is the only one that develops, and there are other exceptions, particularly among birds of prey (Gill, 2007). The female reproductive oocytes are all present in the ovary, the number diminishing over the lifespan of the individual, though as the bird approaches reproductive maturity the ovary enlarges. Some of the oocytes that are normally small, mature into follicles, the deposition of yolk (vitellogenesis) causing them to enlarge in preparation for ovulation. The follicles do not mature all at once in crocodiles, as in birds, forming a hierarchy reflecting the order in which the follicles will ovulate. Vitellogenesis takes an extended period of time in crocodilians because of their low metabolic rate when compared to birds, months instead of 4-16 days in birds (Gill, 2007; Lance, 1989), the result being that the follicular hierarchy is much less distinct in crocodilians and the follicles are of subequal size (Lance, 1989). There is a follicular hierarchy that is strongly apparent in living birds in which deposition of yolk is rapid as a result of their high metabolic rate (Gilbert, 1979). Crocodilians lack medullary bone, a source of calcium that is mobilised rapidly for eggshell production in living birds where it forms before oviposition (Schweitzer et al., 2005; Schweitzer et al., 2007), crocodilians accessing calcium from structural bone (18). Female crocodilians ovulate a full clutch of up to 60 eggs, the average clutch size ranging from 12-48 among the living crocodilians (Thorbjarnarson, 1996), storing the eggs for several weeks in the distal end of the oviduct, and when the embryos are at about the 18-somite stage (Richardson et al., 2002) the eggs are laid en masse. The eggs are incubated passively, though the adult female guards the nest and young, and the young are not typically fed by the adult, a highly precocial developmental strategy. Eggs are oviposited daily by living neornithine birds, apart from paleognaths, which lay 1 egg every 2-6 days.

Evidence obtained from the fossil of a non-avian maniraptoran dinosaur shows that it had 2 active oviducts, as it had the paired association of eggs in its body cavity (Sato et al., 2005). According to the authors¹ it is not clear why living birds lost the right ovary and oviduct, the most widely accepted hypothesis being that it was to reduce weight for flying in the reproductive season, so the female need carry only a single egg inside rather than 2 she would have had with paired ovaries and oviducts, though it has been suggested it may have also been to reduce problems associated with the requirement for calcium at the time of ovulation (Taylor & How, 1970). The number of functional oviducts cannot be commented on as the follicles have been interpreted as being in the ovaries, though both enantiornithines, that are in dorsal view, have clearly preserved the association of follicles on the left side of the body. In Jeholornis STM2-51 the position of the ovary is not clear, though dorsal to the follicles the thoracic vertebrae are preserved in left lateral view in which the follicles appear to be isolated to the left side, as occurs in the enantiornithines, suggesting that the left ovary was the only one active. It is believed the loss of function in the right ovary occurred very close to the transition from non-avian maniraptoran to avian as the dinosaurs that are most closely related had 2 functional oviducts, which supports the hypothesis that the loss is related to flight. The authors¹ suggest it is notable that the right ovary is larger than the left in some crocodiles which suggests the loss of the right ovary would save weight for flight (Lance, 1989).

It is suggested that these female individuals were ready, or almost ready, to ovulate, as indicated by the large, even size of the follicles that were preserved in these specimens, because if they had been in the early stages of vitellogenesis the follicles would not have been of such large and even size. In STM29-8 there is a slightly higher degree of variation among the preserved follicles which suggests that vitellogenesis was not complete. An alternative suggestion has been that vitellogenesis may have occurred more rapidly and over a shorter time interval producing a follicular hierarchy that was more distinct, similar to the that found in living birds. It has been suggested that the latter hypothesis is consistent with the lower number of follicles relative to the number in Jeholornis and the phylogenetic placement of enantiornithines that is more derived. The entire ovaries have not been preserved, so it has been inferred that oocytes that were immature, and follicles that were pre-vitelline and follicles that were atretic in the ovary are either not visible, small or obscured by overlap, or were not preserved, or potentially a combination of both. The authors¹ propose that preservation of the mature follicles was facilitated by a perivitelline layer, and other protective layers that were not present in immature oocytes (Gilbert, 1979). It is suggested that ovulation had not yet begun as there were no eggs in the oviduct(s); or that the animal had died in the time interval between the laying of 1 egg and ovulating another, which in living birds is typically less the 1 hour, or egg(s) in the oviduct(s) could possibly have been expelled from the body after the female had died, as has been reported to have occurred with a female pterosaur from the Jurassic of China (Lü et al., 2011).

The size of the follicles is very consistent in Jeholornis, STM2-51 the long-tailed bird, which suggests a style of reproduction that is more crocodilian, in which a large number of follicles attain maturity almost simultaneously, resulting in a minimal follicular hierarchy (Astheimer, Manolis & Grau, 1989), suggesting vitellogenesis occurred more slowly, similar to what occurs in crocodilians, which is consistent with the lower metabolic rate relative to living taxa, a rate which has been predicted for basal birds (Erickson et al., 2009). It has been suggested that in basal birds eggs would have been too large to store in the body, as the preserved follicle size suggests egg size was large relative to the body size, despite morphological restrictions such as pubes that contact distally; oviposition would have occurred between ovulation, as occurred in other maniraptorans and living birds, but unlike what occurs in crocodiles (Sato et al., 2005).

The number of mature follicles is a good proxy for the size of the clutch Gilbert, 1979; Lance, 1989), as in females that are healthy and undisturbed all mature follicles will ovulate. In crocodilians clutches are much larger than in birds, but compared to the adult body size the size of the follicle, as well as the complete egg, is much smaller. The mature follicle of Alligator mississippiensis mature follicles are 45 mm wide - the femur/follicle ratio changes with age, averaging 0.017 (15,24). According to the authors there is only limited comparative data on the size of follicles, but follicles reach 40 mm in Gallus (Gilbert, 1979), as the size had been increased by domestication, being comparable to the size in A. mississippiensis, in spite of the very large body size difference (follicle diameter/femur length ratio 0.51). The absolute size of the preserved follicles, that are inferred to have been slightly exaggerated by compression, is similar among the 3 specimens, though there is a considerable difference in the adult body sizes - the body mass of Jeholornis STM2-51 has been estimated to be 676 g; enantiornithine STM29-8 is estimated to be 125 g and enantiornithine estimated to be 105 g (Liu et al., 2012) was used to estimate body mass. The result is that there is a distinct ratio of egg/body for each specimen. Jeholornis sp. STM2-51 has preserved the highest number of follicles that are the smallest, proportionally, compared to the overall size of the body - follicle diameter/femur length ratio of 0.087 - which is consistent with the basal position within Aves of this taxon (only more derived than Archaeopteryx). STM8-29, the larger of the 2 enantiornithines preserves a high number of smaller follicles, with a follicle/femur ration of 0.171, while STM10-45, the smaller enantiornithine preserves only a few follicles that are almost the same size as those of Jeholornis, in spite of the disparity of their overall body size - follicle/femur ratio 0.217. The same trade-off between the size of eggs and the size of the clutch are the same as those that have been observed in living birds (Godfray et al., 1991). In Jeholornis STM2-51 and Enantiornithines indet. STM29-8 the large clutch size is consistent with the precocial reproductive strategy that has been inferred for basal birds (Varriccio et al., 2008). In Enantiornithines indet. STM10-45 the smaller clutch size that is preserved, though possibly artefact of preservation, suggests there may have been a diverse range of reproductive strategies among the enantiornithines.

The timing of sexual maturity is another key difference between living archosaurs, early onset of reproductive maturity in means crocodiles reach reproductive maturity before they reach skeletal maturity. Growth is rapid in living birds, skeletal maturity typically being reached within 1 year, though they don't typically reach sexual maturity until between 2-8 years, and domestic chickens 6 months (Gilbert, 1979; Farlow et al., 2005). The reproductive pattern involving the reaching sexual maturity (Erickson et al., 2007) before skeletal maturity that occurs in crocodiles has been shown by histological analyses to have also occurred in paravian dinosaurs. There is a clear lack of fusion in the compound bones, such as the carpometacarpus, of enantiornithine STM10-45. The compound bones of Enantiornithines are known to fuse late in ontogeny (O'Connor, 2009), and it is believed to be a subadult, indicating that among the enantiornithines at least this lineage sexual maturity preceded skeletal maturity, as occurs in crocodilians and paravian dinosaurs. The authors attempted histological analysis on the 3 specimens but they were unable to sample bone in STM10-45. The medullary cavity was not preserved clearly in STM2-51 and STM29-8 as the bones were badly crushed which prevented medullary bone identification. Enantiornithine STM29-8 shows a greater degree of fusion in the compound bones, though the exact degree of fusion could not be determined because of poor preservation, and it was found to be skeletally mature, requiring 1 year to reach adult size.

The recognition of sexually dimorphic traits is made possible the preservation of reproductive organs, such as ovarian follicles, which allowed the determination of the gender of these specimens. In some enantiornithines elongate paired tail feathers strongly resemble similar feathers in Confuciusornis, that have been interpreted and sexually dimorphic traits that were present only in males. The feathers do not co-vary with size, as was expected, resulting in this interpretation being controversial (Chiappe et al., 2008). STM29-8 is the only specimen in which feathers have been preserved, and as the feathers around the pygostyle are clearly visible showing that there are no paired elongate rectrices in this mature (adult) female.

The preserved ovarian follicles in these birds from the Early Cretaceous, that are the first discovery of this kind, provide new clues to the reproduction of these birds. The consistent placement of the follicles on the left side of the body in all 3 specimens, based on the bracket placed by the fossil record, suggests the loss of the right functional ovary occurred at the dinosaur-avian transition. The size of clutches of basal birds, inferred to be plesiomorphically large, which is consistent with the precocial development strategy that is inferred from embryos and juveniles, and information from dinosaurs that are closely related. There is a trend that is observable within the Aves towards smaller clutch size, with eggs that are relatively larger, extending from birds with long, bony tails that are closely related to Archaeopteryx into Neornithines, and reproductive specialisation strategies that are clade-specific that are not limited by the crown group. In birds from the Mesozoic the ovary was large with a follicular hierarchy that was minimal, and with skeletal maturity preceding sexual maturity, as occurs in crocodilians, both conditions being consistent with the lower metabolic rate inferred for birds and paravian dinosaurs compared to neornithine birds.

Sources & Further reading

1. Zheng, Xiaoting, Jingmai O/'Connor, Fritz Huchzermeyer, Xiaoli Wang, Yan Wang, Min Wang, and Zhonghe Zhou. "Preservation of Ovarian Follicles Reveals Early Evolution of Avian Reproductive Behaviour." Nature 495, no. 7442 (03/28/print 2013): 507-11.