,5,*Negligible effect of tooth reduction on body mass in Mesozoic birds
ZHOU Ya-Chun1,2, Corwin SULLIVAN3,4, ZHANG Fu-Cheng,5,*收稿日期:2018-01-5网络出版日期:2019-01-20
| 基金资助: |
Corresponding authors: *zhangfucheng@ivpp.ac.cn,zhangfucheng@lyu.edu.cn
Received:2018-01-5Online:2019-01-20
作者简介 About authors
zhouyachun@ivpp.ac.cn
摘要
牙齿退化是中生代鸟类演化的一个重要过程,牙齿总重综合了牙齿大小和数量的信息,研究这一特征的演化趋势,有助于深入地分析中生代鸟类牙齿退化的原因和方式。然而,现生鸟类均不具齿,无法为研究中生代鸟类的牙齿重量提供参考。除鸟类外的现生脊椎动物中,鳄类与鸟类的亲缘关系最近,且牙齿形态、着生和替换方式与后者相似,因此可为估算中生代鸟类牙齿重量提供参考模型。对从8件现生暹罗鳄标本采得的31枚牙齿进行了形态和重量测量,基于缩放比例原理建立多组回归方程,依此方程对牙齿和齿列保存较完整的中生代鸟类标本进行了牙齿总重估计。结果表明多数中生代鸟类牙齿总重普遍占体重比例极小,据此推测其对飞行的影响可忽略不计,减轻体重的自然选择压力可能不是造成中生代鸟类牙齿退化的主要原因。中生代鸟类牙齿总重的多样性可能反映了其食性和取食行为的差异。
关键词：
Abstract
Tooth reduction and loss was an important evolutionary process in Mesozoic birds. Analysis of evolutionary trends in the total mass of the dentition, a function of tooth size and tooth number, has the potential to shed light on the evolutionary pattern of tooth reduction and loss, and on the causes of this pattern. Because modern birds lack teeth, however, they cannot provide the basis for a model that would allow estimation of tooth masses in their Mesozoic counterparts. We selected the teeth of crocodilians as analogues of those in Mesozoic birds because the former are the closest living relatives of the latter, and the two groups are similar in tooth morphology, tooth implantation, and tooth replacement pattern. To estimate tooth masses in Mesozoic birds, we formulated four regression equations relating tooth mass to various linear dimensions, which were measured in 31 intact isolated teeth from eight individual crocodiles (Crocodylus siamensis). The results for Mesozoic birds show that dental mass as a proportion of body mass was negligible, at least from the perspective of flight performance, suggesting that selection pressure favoring body mass reduction was probably not the primary driver of tooth reduction or loss. Variations in dental mass among Mesozoic birds may reflect the different foods they ate, and the different types of feeding behavior they displayed.
Keywords：
PDF (1278KB)元数据多维度评价相关文章导出EndNote|Ris|Bibtex收藏本文
本文引用格式
周亚纯, 舒柯文, 张福成. 中生代鸟类牙齿的退化及其可忽略的体重效应. 古脊椎动物学报[J], 2019, 57(1): 38-50 DOI:10.19615/j.cnki.1000-3118.180307
ZHOU Ya-Chun, Corwin SULLIVAN, ZHANG Fu-Cheng.
1 Introduction
Tooth loss in Mesozoic birds has traditionally been regarded as a phenomenon driven at least in part by the need to improve flight performance by reducing the mass of the body, and particularly of body parts situated far from the center of gravity (Dilger, 1957; Proctor and Lynch, 1998; Feduccia, 1999; Zhou et al., 2009; Louchart and Viriot, 2011). However, some recent studies (Zheng et al., 2011, 2014a) have linked tooth loss to dietary changes, and evo-devo results even demonstrated the tooth reduction seen in early birds was correlated with the evolutionary origin of beaks (Wang et al., 2017), implying that the mass reduction hypothesis requires critical reevaluation. Tooth loss occurred independently in several early avian lineages (Zhou and Zhang, 2006; Zhou et al., 2009; Davit-Béal et al., 2009), suggesting that reduction and eventual disappearance of the dentition might have conferred some broadly applicable selective advantage. Among toothed Mesozoic birds, the teeth vary widely in both number and size, indicating that tooth reduction was a gradual and not necessarily monotonic process. For example, Hesperornis had up to 47 teeth in the mouth (Marsh, 1880; Gregory, 1951), whereas Jeholornis palmapenis may have had only one or two maxillary teeth on either side of the skull (O’Connor et al, 2012). The crown height of the teeth is 1.2-1.3 mm in Archaeopteryx (London specimen, humerus length: 74.6 mm; Wellnhofer, 2009), but only about 0.1 mm in the admittedly smaller Longirostravis (humerus length: 24.0 mm; Hou et al., 2004). However, the pattern of variation in tooth size and number across individual Mesozoic bird species has never been examined in detail.Here we investigate tooth reduction using a more quantitative approach than has been brought to bear in previous studies (Dilger, 1957; Proctor and Lynch, 1998; Feduccia, 1999; Zhou et al., 2009).The parameter of interest is the total mass of the dentition, a function of both tooth number and average tooth size, as a proportion of body mass. This study aims to estimate the relative mass of the dentition in a range of Mesozoic birds, including representatives of key avian clades, in order to both assess the degree to which the dentition contributed to body mass across early birds in general and evaluate patterns of variation in relative dental mass within Aves. The results will shed light on the plausibility of the two underlying assumptions of the hypothesis that tooth loss in birds was driven primarily by selection for reduced body mass, namely that the teeth of the basalmost birds were massive enough to affect flight performance and that tooth mass was gradually reduced in proportional terms as flight capability improved. To the best of our knowledge, the analysis described here is the first to use empirically-derived estimates of the total mass of the dentition in a range of Mesozoic birds to explore the issue of tooth reduction.
Institutional abbreviation IVPP, Institute of Vertebrate Paleontology and Paleo-anthropology, Chinese Academy of Sciences, Beijing, China; STM, Shandong Tianyu Museum of Nature, Pingyi, Shandong, China.
2 Materials and methods
2.1 Fossil taxa
Mesozoic birds vary widely with respect to quality and mode of preservation, but for this study we focused on taxa represented by at least one specimen in which the teeth could be accurately counted and measured. A total of 15 specimens with well-preserved dentitions, representing 11 avian taxa, were measured for this analysis (Table 1). The number of teeth that could be measured in these specimens varied from only one in Jeholornis IVPP V 14978 (2 or 3 teeth on each side of mouth during life; Zhou and Zhang, 2003) to 14 in Longusunguis IVPP V 17964 (approximately 10 teeth on each side of the mouth during life; Wang et al., 2014). In most cases the measurements were taken directly from the fossils, but measurements for two specimens of Archaeopteryx were taken from published figures (Wellnhofer, 2009).Table 1
Table 1Fossil birds and dentition (both sides) measured in this study
| Taxa | Specimens | Premaxillary teeth | Maxillary teeth | Dentary teeth |
|---|---|---|---|---|
| Archaeopteryx | London specimen1) | 1 | 3 | 0 |
| Berlin specimen1) | 4 | 2 | 2 | |
| Jeholornis | IVPP V 13274 | 0 | 0 | 3 |
| IVPP V 14978 | 0 | 1 | 0 | |
| Sapeornis | IVPP V 13275 | 3 | 1 | 0 |
| IVPP V 13276 | 4 | 4 | 0 | |
| Longipteryx | IVPP V 12325 | 1 | 0 | 1 |
| Longirostravis | IVPP V 11309 | 4 | 0 | 5 |
| Pengornis | IVPP V 15336 | 1 | 2 | 1 |
| Cathayornis | IVPP V 9769 | 3 | 0 | 2 |
| Eoenantiornis | IVPP V 10533 | 4 | 0 | 0 |
| Longusunguis | IVPP V 17964 | 3 | 1 | 10 |
| Parabohaiornis | IVPP V 18691 | 3 | 4 | 6 |
| Bohaiornis | IVPP V 17963 | 4 | 2 | 6 |
| Yanornis | IVPP V 12558 | 2 | 3 | 4 |
| Tianyuornis | STM 7-53 | 0 | 1 | 1 |
新窗口打开|下载CSV
2.2 General approach to estimating dental mass in Mesozoic birds
Tooth masses cannot be measured directly in fossil vertebrates, even when isolated teeth are available, because the process of fossilization may alter the density of the hard tissues comprising the tooth (Rink and Hunter, 1997) and involve mineral infilling of pore spaces that would be occupied in life by fluids and soft tissues. Given this constraint, the ideal approach would perhaps be to use the volume of a fossil tooth, combined with data on tooth density from modern vertebrates, as a basis for estimating the tooth’s original mass in the living animal. This would be possible for complete, isolated teeth, but in the case of slab specimens would require that the dentition (including the tooth roots) be either excavated out of the jaws and surrounding matrix by a preparator or imaged three-dimensionally using a technique such as μ-CT scanning.Given time constraints and the technical difficulties involved in scanning large slab specimens with sufficiently high resolution, the more practical alternative is to use a regression approach to investigate the relationship between linear tooth dimensions and tooth mass in a living vertebrate whose teeth resemble those of Mesozoic birds. Linear measurements of the teeth of Mesozoic birds can then be exploited as a basis for estimating their masses. Regression analyses are widely used to estimate body mass in extinct animals (Bakker, 1972; Anderson, 1985; Christiansen and Fari?a, 2004; Campione and Evans, 2012; Campione et al., 2014), and a parallel methodology can be applied to the dentition alone. Because of their edentulous jaws, modern birds cannot serve as living analogues to their Mesozoic counterparts in this type of analysis. However, crocodilians are the closest living relatives of birds (Green et al., 2014), and broadly resemble Mesozoic toothed birds in the morphology, replacement pattern, and mode of implantation of the dentition (Martin et al., 1980). Accordingly, this study establishes statistical relationships between tooth linear measurements and tooth mass derived from isolated teeth of the extant Siamese crocodile (Crocodylus siamensis), and uses these relationships to estimate the total tooth masses of Mesozoic bird specimens from the linear dimensions of the preserved teeth.
2.3 Derivation of regression functions from measurements of Crocodylus siamensis teeth
We measured 31 isolated teeth (tooth height: 9.11-18.9 mm, Table S1) from eight individual immature Siamese crocodile skeletons (Crocodylus siamensis; total skeletal length ~1.3 m) housed at the IVPP, in order to examine the quantitative relationships between the linear dimensions of the teeth (determined by taking photos and measuring distances on the images) and their masses (determined using a Sartorius ME5 microbalance). Measurements taken included crown height (CH), tooth height (TH), fore-aft crown length (FACL), fore-aft base length (FABL) and fore-aft root length (FARL) (Fig. 1). The positions of the individual teeth within the dentitions of the eight individuals are unknown, because the teeth were ones that had dropped out of the jaws as the crocodile skeletons were shipped to the IVPP from a crocodile farm in Fujian Province, China. However, the teeth did not include any of the enlarged fangs that occur near the front of the mouth in crocodilians. The skeletons were individually packaged, making it possible to determine which teeth came from each individual in the sample. Because of hybridization practices in the crocodile farming industry, it is possible that the specimens are not pure C. siamensis but rather contain some genetic input from the salt-water crocodile (C. porosus) and/or other Crocodylus species. In the American alligator (Alligator mississippiensis), the tooth crowns tend to become more robust during ontogeny (Erickson et al., 2003), and similar changes may occur at least to some degree in Crocodylus. For this reason, the immature Crocodylus teeth used in this study probably represent better analogues for Mesozoic bird teeth, which are relatively slender, than would the teeth of older Crocodylus individuals.Fig. 1
新窗口打开|下载原图ZIP|生成PPTFig. 1Linear measurements applied to bird and crocodilian teeth in this study
A. isolated left premaxillary tooth of Archaeopteryx lithographica (London specimen; modified from Wellnhofer, 2009); B. isolated posterior tooth of Crocodylus siamensis (IVPP uncatalogued specimen). Dotted line: measurement line, solid line: indication line Abbreviations: CH. crown height; FABL. fore-aft base length; FACL. fore-aft crown length; FARL. fore-aft root length; TH. tooth height. Scale bars = 1 mm
Least-squares regression equations were fitted to the log-transformed measurement data in order to establish relationships between various linear dimensions and mass for the teeth of C. siamensis. We carried out bivariate regressions of mass as a dependent factor against all five linear measurements as independent factors, as well as ten multiple regressions that each incorporated two linear measurements as independent factors (Table 2). Carrying out ten multiple regressions enabled us to test every possible combination of two linear measurements.
A p-value was calculated as a metric for the statistical significance of each regression model (Table 2) and its corresponding equation, and R2 was calculated as a measure of the amount of variation in the C. siamensis tooth masses that could be explained by the model. For multiple regressions, a separate p-value was associated with each linear measurement, but a single R2 was calculated for the regression model as a whole. The regression models with the lowest p-values and highest R2 were selected as a basis for estimating tooth masses in Mesozoic birds. Specifically, regression models in which at least one linear measurement was associated with a p-value greater than or equal to 0.01 were immediately eliminated from consideration. The remaining models with acceptable p-values included all five bivariate models, in addition to the multiple regression model based on TH and FARL. Among these models, the TH and TH & FARL models exhibited the highest R2 values (0.822 and 0.887, respectively) and were accepted for inclusion in the Mesozoic bird analysis. The next highest R2 values were associated with the FABL and FARL models (0.636 and 0.649, respectively) and these models were also accepted given that FABL can be measured in many Mesozoic bird teeth whereas TH is impossible to measure in teeth with broken apices or concealed roots. Accordingly, a total of four regression models (Table 3) were utilized in the next stage of the analysis, whereas all others were rejected.
Table 2
Table 2Regression results for Crocodylus siamensis tooth data
| Model | R2 | intercept | slope | p-value | ||
|---|---|---|---|---|---|---|
| M-CH | 0.423 | 1.081 | 1.483 | 0.000 | ||
| M-TH | 0.822 | -0.276 | 2.180 | 0.000 | ||
| M-FACL | 0.595 | 0.693 | 2.366 | 0.000 | ||
| M-FABL | 0.636 | 0.913 | 2.278 | 0.000 | ||
| M-FARL | 0.649 | 0.522 | 2.168 | 0.000 | ||
| M-(TH & FARL) | 0.887 | -0.346 | 1.606 (TH) | 0.940 (FARL) | 0.000 (TH) | 0.000 (FARL) |
| M-(CH & TH) | 0.855 | -0.463 | -0.749 (CH) | 2.836 (TH) | 0.017 (CH) | 0.000 (TH) |
| M-(TH & FABL) | 0.850 | -0.153 | 1.710 (TH) | 0.735 (FABL) | 0.000 (TH) | 0.030 (FABL) |
| M-(TH & FACL) | 0.838 | -0.249 | 1.821 (TH) | 0.603 (FACL) | 0.000 (TH) | 0.103 (FACL) |
| M-(FABL & FARL) | 0.703 | 0.571 | 1.187 (FABL) | 1.243 (FARL) | 0.031 (FABL) | 0.018 (FARL) |
| M-(CH & FARL) | 0.699 | 0.389 | 0.628 (CH) | 1.738 (FARL) | 0.039 (CH) | 0.000 (FARL) |
| M-(FACL & FARL) | 0.671 | 0.425 | 1.080 (FACL) | 1.409 (FARL) | 0.054 (FACL) | 0.006 (FARL) |
| M-(CH & FABL) | 0.636 | 0.904 | 0.50 (CH) | 2.227 (FABL) | 0.910 (CH) | 0.000 (FACL) |
| M-(FACL & FABL) | 0.636 | 0.883 | 0.229 (FACL) | 2.074 (FABL) | 0.854 (FACL) | 0.084 (FABL) |
| M-(CH & FACL) | 0.596 | 0.686 | 0.148 (CH) | 2.204 (FACL) | 0.757 (CH) | 0.002 (FACL) |
新窗口打开|下载CSV
2.4 Estimation of dental masses in Mesozoic birds
We used the regression equations selected during the previous stage of the analysis (Table 3) as the basis for calculating an average total tooth mass (ATTM, sample calculations for London Archaeopteryx given in Table 4) for each Mesozoic bird specimen included in our study. The three steps described below were repeated separately for each of the four regression equations, producing four alternative ATTM values per specimen.Table 3
Table 3Four least-squares regression models used to generate tooth mass estimates for Mesozoic birds
| Model | R2 | Log10Mass |
|---|---|---|
| M-(TH & FARL) | 0.887 | -0.346+1.606log10TH+0.940log10FARL |
| M-TH | 0.822 | -0.276+2.180log10TH |
| M-FARL | 0.649 | 0.522+2.168log10FARL |
| M-FABL | 0.636 | 0.913+2.278log10FABL |
新窗口打开|下载CSV
(1) We measured the linear dimension(s) (Table S2) on which the regression equation was based in as many preserved teeth as possible. The regression equation was then used to produce a single tooth mass (STM) estimate for each measured tooth.
(2) All STM values were averaged to calculate the specimen’s average single tooth mass (ASTM, Table S3).
(3) Minimum and maximum counts of the total number of teeth in the mouth (based on the observation of number of visible teeth and alveoli in this work, or interpreted in the light of previous work) were estimated for the specimen and then average to produce a final estimated tooth number. The ASTM value was multiplied by the estimated tooth number to obtain an average total tooth mass (ATTM, Table S3), which was retained as the optimal estimate of the total mass of the specimen’s dentition produced by the equation. Although it would have been preferable in principle to estimate the STM of every individual tooth in the mouth and then sum the STMs to produce a total mass estimate, this approach was not feasible because in most specimens some teeth were broken, unerupted or simply missing.
Table 4
Table 4Average total tooth mass (ATTM) estimates for the London Archaeopteryx (London specimen) based on different regression equations
| Taxa | Model | Teeth | STM (mg) | ASTM (mg) | Teeth (EN) | ATTM (mg) |
|---|---|---|---|---|---|---|
| Archaeopteryx London specimen | M-(TH & FARL) | I-PMT (1) | 2.082 | 2.082 | 50 | 106.182 |
| M-TH | I-PMT (1) | 4.667 | 4.667 | 50 | 238.527 | |
| M-FARL | I-PMT (1) | 2.803 | 2.083 | 50 | 142.953 | |
| M-FABL | I-PMT (1) | 1.845 | 3.091 | 50 | 157.679 | |
| MT (3) | 3.457 3.550 3.515 |
新窗口打开|下载CSV
To account for body size, each ATTM value was then divided by the estimated body mass (Table 5) of the specimen in question to produce a relative average total tooth mass (R-ATTM). Body mass was estimated using a regression equation (Log10BM=-0.762804+ 1.733282log10HumL) based on humeral length (Liu et al., 2012), where BM is body mass in grams and HumL is humeral length in millimeters.
Table 5
Table 5Average total teeth mass (ATTM) values calculated for Mesozoic birds using four alternative regression models
| Taxa | Specimens | ATTM (mg) | Body Mass (g) M-HumL | |||
|---|---|---|---|---|---|---|
| M-(TH & FARL) | M-TH | M-FARL | M-FABL | |||
| Archaeopteryx | London specimen | 106.182 | 238.527 | 142.953 | 157.679 | 307.000 |
| Berlin specimen | 143.521 | 227.000 | ||||
| Jeholornis | IVPP V 13274 | 6.462 | 597.400 | |||
| IVPP V 14978 | 10.826 | 285.400 | ||||
| Sapeornis | IVPP V 13275 | 80.496 | 70.148 | 680.600 | ||
| IVPP V 13276 | 111.254 | 141.115 | 712.900 | |||
| Eoenantiornis | IVPP V 10533 | 6.028 | 14.344 | 16.280 | 30.602 | 60.900 |
| Cathayornis | IVPP V 9769 | 10.208 | 52.285 | |||
| Pengornis | IVPP V 15336 | 25.200 | 64.568 | 49.392 | 74.270 | 235.200 |
| Longipteryx | IVPP V 12325 | 47.976 | 67.956 | 109.392 | 60.132 | 126.700 |
| Longirostravis | IVPP V 11309 | 0.406 | 2.016 | 0.980 | 1.052 | 42.600 |
| Longusunguis | IVPP V 17964 | 37.336 | 58.872 | 50.016 | 64.680 | 104.627 |
| Bohaiornis | IVPP V 17963 | 59.883 | 179.886 | 54.705 | 144.550 | 153.636 |
| Parabohaiornis | IVPP V 18691 | 39.004 | 134.279 | 37.716 | 78.413 | 135.081 |
| Yanornis | IVPP V 12558 | 98.753 | 186.388 | 284.920 | 284.838 | 336.000 |
| Tianyuornis | STM 7-53 | 7.775 | 22.971 | 14.664 | 17.602 | 43.545 |
新窗口打开|下载CSV
3 Results
Average total tooth mass (ATTM) estimates for particular specimens varied widely across the four regression equations used in this study (Table 5), and this variation was unsystematic in the sense that no equation consistently produced larger estimates than any other. The inconsistency reflected variability in tooth shape among the Mesozoic birds included in the study, as each equation used different shape information as a basis for estimating tooth mass. The mass of a tall and slender tooth, for example, would likely have been overestimated by the TH equation and underestimated by the FABL and FARL equations, whereas the reverse would have been true for a tooth with short and robust proportions.Table 6
Table 6Relative average total tooth mass (R-ATTM) values (‰) calculated for Mesozoic birds using different regression models
| Taxa | Specimens | R-ATTM | |||
|---|---|---|---|---|---|
| M-(TH & FARL) | M-TH | M-FARL | M-FABL | ||
| Archaeopteryx | London specimen | 0.346 | 0.777 | 0.466 | 0.514 |
| Berlin specimen | 0.632 | ||||
| Jeholornis | IVPP V 13274 | 0.011 | |||
| IVPP V 14978 | 0.038 | ||||
| Sapeornis | IVPP V 13275 | 0.118 | 0.103 | ||
| IVPP V 13276 | 0.156 | 0.198 | |||
| Eoenantiornis | IVPP V 10533 | 0.27 | 0.236 | 0.267 | 0.502 |
| Cathayornis | IVPP V 9769 | 0.195 | |||
| Pengornis | IVPP V 15336 | 0.107 | 0.275 | 0.210 | 0.316 |
| Longipteryx | IVPP V 12325 | 0.379 | 0.536 | 0.863 | 0.475 |
| Longirostravis | IVPP V 11309 | 0.010 | 0.047 | 0.023 | 0.025 |
| Longusunguis | IVPP V 17964 | 0.261 | 0.563 | 0.478 | 0.618 |
| Bohaiornis | IVPP V 17963 | 0.390 | 1.171 | 0.356 | 0.941 |
| Parabohaiornis | IVPP V 18691 | 0.289 | 0.994 | 0.405 | 0.580 |
| Yanornis | IVPP V 12558 | 0.294 | 0.555 | 0.848 | 0.848 |
| Tianyuornis | STM 7-53 | 0.179 | 0.528 | 0.337 | 0.404 |
新窗口打开|下载CSV
Variation in dental mass estimates across taxa was also considerable. FABL-based estimates of ATTM, which are available for every specimen in the study because FABL was the easiest of the linear dimensions to measure, range from just over 1 mg for the small enantiornithine Longirostravis V 11309 (estimated body mass 43 g) to 285 mg for the ornithuromorph Yanornis V 12558 (estimated body mass 336 g). Other regression equations also assigned the smallest ATTM values to Longirostravis. Only the FABL and FARL equations assigned the largest ATTM values to Yanornis, with the TH & FARL equation assigning a larger value to the London Archaeopteryx and the TH equation assigning larger values to both the London Archaeopteryx and Yanornis V 12558. Nevertheless, all four equations ranked Yanornis among the taxa with the most massive dentitions.
Unsurprisingly, taxonomic variation in relative average total tooth mass (R-ATTM) (Table 6) is less extreme. Again using estimates derived from the FABL equation as an example, the smallest R-ATTM (0.011‰) belongs to the large long-bony-tailed bird Jeholornis V 13274, while the largest (0.941‰) belongs to Bohaiornis V 17963. The 86-fold difference in R-ATTM between Bohaiornis and Jeholornis is of course much less than the 270-fold difference in ATTM between Yanornis and Longirostravis, reflecting the normalizing effect of dividing by body mass. It is also notable that the R-ATTM values produced by all four regression equations were rather small, with the estimated R-ATTM exceeding 1.000‰ in only one case (1.171‰ for Bohaiornis, using the TH-based equation).
4 Discussion
The large and unsystematic variation in the mass estimates produced by the four regression equations suggests that the Mesozoic birds in our analysis vary widely in their tooth proportions, relative to both extant crocodilians and each other. Nevertheless, the results of the regression equations contain some clear and evidently meaningful patterns.4.1 ATTM values for fossil birds
The wide differences in ATTM estimates for the fossil birds in the analysis are partly driven by variation in body size, with larger taxa generally tending to have larger dentitions. However, our sample included significant exceptions such as Jeholornis, a relatively large bird with a small dentition, and ATTM values are often informative regarding the nature of the dentition despite the influence of size. Yanornis, for which large ATTM estimates were recovered by all four regression equations, was estimated to have about 68 teeth in the mouth (Zhou and Zhang, 2001; Zheng et al., 2014a), and our measurements indicated that these teeth are fairly large (CH: 0.649-1.312 mm). By contrast, the smallest ATTM estimates were consistently recovered for Longirostravis (CH: 0.070-0.145 mm), in which we estimate 14 teeth to have been present in the mouth although only 10 are actually preserved (Hou et al., 2004).4.2 R-ATTM, a measure of proportional tooth mass
By representing tooth mass as a proportion of body mass, R-ATTM reveals meaningful patterns of tooth mass distribution among Mesozoic birds of varying sizes more clearly than does ATTM. The fact that all taxa were found to have R-ATTM values of less than 1.2‰, regardless of the regression model used, is important because of the improbability that an increment or decrement of ~1‰ of a bird’s body mass would have any detectable impact on its flight capability.Abundant observations suggest living birds can function as capable fliers even when carrying objects considerably heavier than 1‰ of body mass, as exemplified by the findings of Yosef (1993) with respect to the carnivorous passerine Lanius ludovicianus (loggerhead shrike). When experimentally offered mice, shrike with an average body mass of 48 g were able to transport mice averaging 18 g in mass an average distance of 65 m using the beak, and were able to transport mice averaging 40 g an average of 35 m using the pedal claws. Prey carried using the feet ranged up to 129%, and prey carried in the beak up to 64%, of the average shrike body mass reported in the study. Although the masses carried by the shrikes clearly did reduce their capacity for sustained flight, their ability to transport prey items of large proportional mass over considerable distances nevertheless suggests that a burden amounting to ~1‰ of body mass would have a negligible impact on flight performance. It is particularly notable that the shrikes were able to fly for tens of meters while carrying large prey items in their beaks, as this undermines otherwise plausible arguments that selection strongly favored dental reduction because the position of the teeth far anterior to the center of gravity caused them to exert a large, destabilizing torque during flight (Zhou et al., 2009).
Studies of birds of prey with body masses considerably greater than that of the loggerhead shrike also indicate that sizeable loads can be carried with little impact on flight performance. Videler et al. (1988a, b) applied artificial loads exceeding 30% of body mass to the hindlimbs of a female and a male kestrel (Falco tinnunculus), whose respective body masses were ~190 g and ~160 g. Over a flight distance of 125 m, these loads caused a reduction in cruising speed of less than 7% for the male kestrel and less than 10% for the female kestrel, compared to unloaded runs. Pennycuick et al. (1989) found that two female Harris’ hawks (Parabuteo unicinctus) with body masses of about 920 g were both able to fly a 50 m course with loads amounting to about 68% of body mass applied to their torsos, although when loaded to this degree they were unable to climb. Climbing flight was possible with loads approaching 50% of body mass, though climbing speed decreased sharply as load increased. These examples not only reinforce the observation that extant birds remain capable fliers while carrying loads that exceed ~1‰ by orders of magnitude, but show that this result applies to birds that span a fairly wide range of body sizes.
Apart from the ability of birds to fly while carrying external burdens, species such as the bobolink (Dolichonyx oryzivorus) (Zheng, 1995) and Alaskan bar-tailed godwit (Limosa lapponica) (Pennycuick, 2008) put on additional fat that may increase body mass by more than 50% before embarking on long-distance migrations. The added mass inevitably affects the birds’ flight performance during the early part of the migration by increasing their minimum power speed (Pennycuick, 2008), but they can nevertheless cover extremely large distances (10300 km; Pennycuick, 2008) while their mass is significantly elevated. Although we are unaware of any compelling evidence that any Mesozoic birds engaged in aerial migrations, the ability of some modern species to do so while carrying vast reserves of body fat suggests that the mass of the dentition would have had only a trivial effect on performance and the energetic cost of transport even if Mesozoic birds sometimes flew for considerable distances.
Furthermore, the mass of the gizzard stones in some living birds is approximately 1% of body mass (Wings and Sander, 2007), making them about an order of magnitude heavier in proportional terms than the dentition of any Mesozoic bird considered in our study. If gizzard stones are in some sense a functional replacement for teeth, the change from dental to gastric trituration has increased rather than reduced the mass of the body and its contents, although it has admittedly also concentrated the body’s mass closer to the center of gravity. The Mesozoic birds Sapeornis (Zhou and Zhang, 2003; Zheng et al., 2011), Bohaiornis (Li et al., 2014) and Yanornis (Zhou et al., 2004) in fact possess both gizzard stones and teeth, and it is striking that the non-ornithothoracine Sapeornis in particular could apparently fly adequately with both types of triturating apparatus in place despite lacking the ossified sternum and elongate coracoid (Zheng et al., 2014b) seen in more aerially capable avians. While some other non-ornithothoracine birds do possess an ossified sternum and elongate coracoid, a number of additional derived, apparently flight-related features are restricted to ornithothoracines, including a triosseal canal and a well-developed sternal keel (Brusatte et al., 2015). Mesozoic ornithuromorphs and to some degree enantiornithines may have broadly resembled typical extant birds in their mode of flight, but non-ornithothoracine birds including Sapeornis were clearly less well adapted to aerial locomotion. It is possible, and indeed likely, that non-ornithothoracine birds were somewhat more easily impaired by loads than their modern counterparts. Nevertheless, the occurrence of both gastroliths and teeth in Sapeornis, a non-ornithothoracine bird that was probably a particularly weak flier, suggests that the mass of the triturating apparatus was not a major target of selection pressure even in birds of this grade. More broadly, the impressive load-carrying capabilities of extant birds strongly imply that a dentition making up no more than ~1‰ of body mass would not have significantly encumbered even non-ornithothoracines, despite their lower level of aerial adaptation.
5 Conclusions
Our results indicate that the total mass of the dentition in Mesozoic birds was uniformly small, amounting to less than 1.2‰ of total body mass in most taxa examined in our study. Even allowing for the fact that the teeth would have been in a cranial position, giving their weight a large moment arm about the center of gravity, their effect on flight performance was probably negligible given that some modern birds carry proportionally much larger burdens in flight (including in the beak) and undergo dramatic fluctuations in body mass in connection with migration flights (Videler et al., 1988a, b; Yosef, 1993; Pennycuick, 2008).Nevertheless, the reduction and eventual disappearance of the dentition in avians requires an evolutionary explanation. It is possible that replacement of the teeth by a rhamphotheca resulted in greater general feeding efficiency (Dilger, 1957; Louchart and Viriot, 2011), or at least was a pathway to exploitation of specific feeding niches (Zheng et al., 2011, 2014; O’Connor and Zhou, 2015). However, a modified form of the mass reduction hypothesis may still be viable, given that the presence of teeth requires the tooth-bearing bones to be deep enough to enclose tooth roots. Despite the small mass of the dentition itself, it is possible that flight-related lightening of the head, not the teeth per se, favored reduction of the premaxilla, maxilla and dentary to the point where retention of teeth was no longer feasible.
Acknowledgments
This work was supported by grants from the National Natural Sciences Foundation of China (41688103), National Science Fund for Distinguished Young Scholars (41125008), the National Basic Research Progam of China (2012CB821900), the Chinese Academy of Sciences (KZCX2-EW-105) and Linyi University to Zhang Fu-Cheng, and by a Natural Sciences and Engineering Research Council of Canada Discovery Grant (RGPIN-2017-06246) and start-up funding awarded by the University of Alberta to C.S.参考文献 原文顺序
文献年度倒序
文中引用次数倒序
被引期刊影响因子
DOIURL [本文引用: 1]
DOIURL [本文引用: 1]
[本文引用: 1]
DOIPMID [本文引用: 1]
Background: Body size is intimately related to the physiology and ecology of an organism. Therefore, accurate and consistent body mass estimates are essential for inferring numerous aspects of paleobiology in extinct taxa, and investigating large-scale evolutionary and ecological patterns in the history of life. Scaling relationships between skeletal measurements and body mass in birds and mammals are commonly used to predict body mass in extinct members of these crown clades, but the applicability of these models for predicting mass in more distantly related stem taxa, such as non-avian dinosaurs and non-mammalian synapsids, has been criticized on biomechanical grounds. Here we test the major criticisms of scaling methods for estimating body mass using an extensive dataset of mammalian and non-avian reptilian species derived from individual skeletons with live weights.;Results: Significant differences in the limb scaling of mammals and reptiles are noted in comparisons of limb proportions and limb length to body mass. Remarkably, however, the relationship between proximal (stylopodial) limb bone circumference and body mass is highly conserved in extant terrestrial mammals and reptiles, in spite of their disparate limb postures, gaits, and phylogenetic histories. As a result, we are able to conclusively reject the main criticisms of scaling methods that question the applicability of a universal scaling equation for estimating body mass in distantly related taxa.;Conclusions: The conserved nature of the relationship between stylopodial circumference and body mass suggests that the minimum diaphyseal circumference of the major weight-bearing bones is only weakly influenced by the varied forces exerted on the limbs (that is, compression or torsion) and most strongly related to the mass of the animal. Our results, therefore, provide a much-needed, robust, phylogenetically corrected framework for accurate and consistent estimation of body mass in extinct terrestrial quadrupeds, which is important for a wide range of paleobiological studies (including growth rates, metabolism, and energetics) and meta-analyses of body size evolution.
DOIURL [本文引用: 1]
Body mass is strongly related to both physiological and ecological properties of living organisms. As a result, generating robust, broadly applicable models for estimating body mass in the fossil record provides the opportunity to reconstruct palaeobiology and investigate evolutionary ecology on a large temporal scale. A recent study provided strong evidence that the minimum circumference of stylopodial elements (humerus and femur) is conservatively associated with body mass in living quadrupeds. Unfortunately, this model is not directly applicable to extinct bipeds, such as non-avian dinosaurs. This study presents a new equation that mathematically corrects the quadruped equation for use in bipeds. It is derived from the systemic difference in the circumference-to-area scaling relationship of two circles (hypothetical quadruped) and one circle (hypothetical biped), which represent the cross-section of the main weight-bearing limb bones. When applied to a newly constructed data set of femoral circumferences and body masses in living birds, the new equation reveals errors that are significantly lower than other published equations, but significantly higher than the error inherent in the avian data set. Such errors, however, are expected given the unique overall femoral circumference-body mass scaling relationship found in birds. Body mass estimates for a sample of bipedal dinosaurs using the new model are consistent with recent estimates based on volumetric life reconstructions, but, in contrast, this equation is simpler to use, with the concomitant potential to provide a wider set of body mass estimates for extinct bipeds. Although it is evident that no one estimation model is flawless, the combined use of the corrected quadrupedal equations and the previously published quadrupedal equation offer a consistent approach with which to estimate body masses in both quadrupeds and bipeds. These models have implications for conducting large-scale macroevolutionary analyses of body size throughout the evolutionary history of terrestrial vertebrates, and, in particular, across major changes in body plan, such as the evolution of bipedality in archosaurs and quadrupedality in dinosaurs.
DOIURL [本文引用: 1]
DOIURL [本文引用: 1]
DOIURL [本文引用: 3]
DOIURL [本文引用: 1]
[本文引用: 2]
DOIURL [本文引用: 1]
DOIURL [本文引用: 1]
DOIURL [本文引用: 2]
DOIURL [本文引用: 1]
A new specimen of Bohaiornis guoi from the Jiufotang Formation, comprising a nearly complete skeleton, sheds light on enantiornithine morphological variation and ecological specialization. The new specimen was collected from near Lamadong Village in Liaoning Province, which is the same area where the sub-adult holotype specimen was reported. It provides new information on the cranial and pectoral girdle anatomy of the species, e.g., broad nasal, strikingly robust acromion, medially curved acrocoracoid process. In contrast to the holotype, the newly referred specimen has small rounded stones in the thoracic region that in other extinct taxa has been interpreted as direct evidence of diet. Direct evidence of diet is so far unknown in other Enantiornithes. Specifically the lack of "stomach stones" or gastroliths in enantiornithines despite their excellent fossil record has been proposed to be related to their insectivorous diet as well as to their arboreal ecology. We hypothesize that cranial morphology as well as the number and shape of the preserved stones in Bohaiornis may be most consistent with a raptorial ecology previously unknown for Enantiornithes and considered rare for Avialae. While rostrum shape has a strong relationship to feeding ecology in living birds, in basal avialan birds most diversity is in dental morphology, number, and distribution of the teeth.
[本文引用: 1]
DOIURL [本文引用: 2]
All living birds are toothless, constituting by far the most diverse toothless vertebrate clade, and are striking examples of evolutionary success following tooth loss. In recent years, an unprecedented number of Mesozoic birds have been described, illustrating the evolution of dentition reductions. Simultaneously, major advances in experimental embryology have yielded new results concerning avian edentulism. Reviewing these lines of evidence, we propose hypotheses for its causes, with a prominent role for the horny beak during development. A horny beak and a muscular gizzard functionally 'replaced' dentition for food acquisition and processing, respectively. Together with edentulism itself, these features and others contributed to the later success of birds, as a result of their high performance or additional functionality working in concert in these complex organisms.
[本文引用: 1]
[本文引用: 1]
[本文引用: 1]
DOIURL [本文引用: 1]
The Early Cretaceous long bony-tailed bird Jeholornis prima displays characters both more basal than Archaeopteryx and more derived, exemplifying the mosaic distribution of advanced avian features that characterises early avian evolution and obfuscates attempts to understand early bird relationships. The current diversity of Jeholornithiformes is controversial, since multiple possibly synonymous genera were named simultaneously. Here, we provide the first definitive evidence of a second species belonging to this clade, and erect the new taxon J. palmapenis sp. nov. This new specimen reveals the tail integument of Jeholornithiformes, the morphology of which appears to have no aerodynamic benefit suggesting this clade evolved plumage patterns that were primarily for display.
DOIURL [本文引用: 1]
[本文引用: 4]
[本文引用: 2]
[本文引用: 1]
DOIURL [本文引用: 2]
DOIURL [本文引用: 2]
[本文引用: 1]
DOIURL [本文引用: 1]
[本文引用: 4]
[本文引用: 1]
DOIURL [本文引用: 2]
[本文引用: 1]
DOIURL [本文引用: 3]
DOIURL [本文引用: 3]
DOIURL [本文引用: 1]
[本文引用: 1]
DOIURL [本文引用: 2]
DOIURL [本文引用: 1]
DOIURL [本文引用: 1]
[本文引用: 4]
