Dean Falk*, John C. Early hominid brain evolution: a new look Redmond, Jr and at old endocasts John Guyer Department of Anthropology, Early hominid brain morphology is reassessed from endocasts of University at Albany, SUNY, Australopithecus africanus and three species of Paranthropus, and new Albany, NY 12222, U.S.A. endocast reconstructions and cranial capacities are reported for four E-mail: [email protected]; key specimens from the Paranthropus clade. The brain morphology of [email protected]; Australopithecus africanus appears more human like than that of [email protected] Paranthropus in terms of overall frontal and temporal lobe shape. These new data do not support the proposal that increased encephali- Glenn C. Conroy zation is a shared feature between Paranthropus and early Homo.Our Departments of Anatomy and findings are consistent with the hypothesis that Australopithecus afri- Neurobiology/Anthropology, canus could have been ancestral to Homo, and have implications for Washington University School assessing the tempo and mode of early hominid neurological and of Medicine, St Louis, cognitive evolution. MO 63110, U.S.A. E-mail: 2000 Academic Press [email protected] Wolfgang Recheis Department of Radiology II, University of Innsbruck, Anichstr. 35, A-6020, Innsbruck, Austria. E-mail: [email protected] Gerhard W. Weber and Horst Seidler Institute of Human Biology, University of Vienna, Althanstrasse 14, A1091, Vienna, Austria. E-mail: [email protected]; [email protected] Received 6 May 1999 Revision received 2 August 1999 and accepted 5 November 1999 Keywords: Australopithecus, endocasts, frontal lobe, paleoneurology, Journal of Human Evolution (2000) 38, 695–717 Paranthropus, phylogeny, doi:10.1006/jhev.1999.0378 temporal lobe. Available online at http://www.idealibrary.com on Introduction casts (endocasts) that reproduce details of the external morphology of the brain from Much of what is known about hominid brain the internal surface of the braincase. evolution has been learned from endocranial Because these endocasts are usually from fragmentary pieces of fossilized skulls, *To whom correspondence should be addressed. their missing parts must be reconstructed. 0047–2484/00/050695+23$35.00/0 2000 Academic Press 696 . ET AL. Discoveries such as KNM-WT 17000 (P. for two of the specimens (Sts 5 and Stw aethiopicus) and KNM-WT 17400 (P. boisei) 505) by comparing several measurements (Leakey & Walker, 1988; Walker et al., obtained using calipers with measurements 1986; Brown et al., 1993) provide evidence taken on their corresponding virtual endo- of previously unknown parts of the Paran- casts that had been acquired with 3D-CT thropus brain. Prior to these discoveries, technology from the original skulls (Conroy Paranthropus endocasts were sometimes et al., 1998). The maximum length, height, reconstructed using endocasts of A. africa- and width obtained by measuring the virtual nus (e.g., Sts 5) as a model (see below). In endocast of Sts 5 on the computer screen this study we compare endocasts of Paran- were 0·98, 1·00, and 1·00 of the respective thropus (including P. robustus, P. aethiopicus, measurements obtained with calipers from P. boisei) with those of Australopithecus afri- the silicone endocast. The length of the canus and identify, quantify, and interpret fragmentary Stw 505 virtual endocast and previously unknown differences in the fron- the distance between its left frontal and tal and temporal lobe morphology between temporal poles were each 0·98 of the respec- these genera. In addition, we provide revised tive measurements obtained from the sili- estimates for the mean cranial capacity of cone endocast. A third measurement on the Paranthropus. virtual endocast of Stw 505 (between its highest point on the dorsal surface and its lowest point at the anterior end of the tem- Materials and methods poral lobe) measured 0·99 of the compar- Previously unknown parts of the cerebral able measurement of the silicone endocast. cortex in Paranthropus were observed and Thus, as detailed elsewhere (Weber et al., measured on both the endocast of 1998), endocasts prepared from museum KNM-WT 17000 (P. aethiopicus) and on a quality casts of skulls reproduce measure- silicone endocast prepared from a cast of ments obtained with 3D-CT technology KNM-WT 17400 (P. boisei). Corresponding from the braincases of the original observations and measurements for Austra- specimens with a high degree of accuracy. lopithecus africanus were obtained from sili- Eight measurements (described below) cone endocasts prepared from museum were obtained with calipers from basal views casts of Sts 5 (Mrs. Ples) and Stw 505 (Mr. of endocasts and projected onto the basal Ples), as well as from a copy of the natural plane. The procedure for orienting an endo- endocast of Sts 60. Other endocasts used for cast in basal view is to first determine the comparative purposes included KNM-ER maximum antero-posterior diameter of the 23000 (P. boisei), Sts 19 (A. africanus) and endocast in left lateral view (using the right the Sterkfontein Number 2 natural endo- hemisphere, if left is not present) that con- cast (A. africanus). Comparative endocast nects the frontal and occipital poles as measurements were taken (by JG) from ten described and illustrated by Connolly gorillas (G. gorilla), nine chimpanzees (P. (1950:124–125). The endocast is then troglodytes), nine bonobos (P. paniscus), and turned upside down and secured so that the ten modern humans. Associated cranial maximum anterior-posterior diameter is in capacities were obtained with mustard seed the horizontal or basal plane and the mid- for the gorilla and chimpanzee sample and sagittal plane is vertical. In cases of partial from the literature for the human, bonobo, endocasts (e.g., Sts 60, Stw 505; KNM-WT and early hominid sample. 17400), basal orientations were estimated As a preliminary step, GWW and DF by aligning them next to correctly oriented validated the size of the silicone endocasts full endocasts from the same genus (e.g., Sts 697 5; KNM-WT 17000). The fossil hominid 2 measurements were from undistorted and unreconstructed portions of endocasts. In 1 order to reduce potential observer error or bias, measurements were taken together by three observers (DF, JG, and JCR) on two rof different occasions and the results averaged. 7 rob JG took the measurements with sliding cali- 6 pers, while JR and DF confirmed his selec- 3 cob tion of landmarks and readings from the mbat 5 calipers, and made sure the calipers mat mat remained oriented so that measurements bat 4 bat 8 were projected onto the basal plane. mcp In order to quantify remeasurement error, the three workers together repeated all of the measurements on the fossil hominids one year after the first measurements were obtained and then compared their results with the earlier ones. For each of the eight measurements, remeasurement error was calculated as the mean of the absolute dif- ferences (determined for each specimen) bpc between the first and second sets of Figure 1. Measurements obtained from basal views and projected onto the horizontal (basal) plane from endo- measurements. Remeasurement error was casts of australopithecines, apes, and humans (see text then expressed as a percentage of the aver- for details and Table 1 for data). Landmarks: bat, most age length for each measurement. The anterior point on temporal lobe from basal view; mat, most lateral point on endocast at the level of bat in results were 1% for measurements 1, 2, 6, basal plane; mbat; middle of the line connecting the and 8; and ranged from 2–8% for the other two bats; rof, the most rostral point on the orbital four measurements. The highest remeasure- surfaces of the frontal lobes; mcp, middle clinoid process; cob, caudal boundary of olfactory bulbs (cri- ment errors expressed as percentages of briform plate) in midline; rob, rostral boundary of mean lengths were for the shortest lengths. olfactory bulbs in midline; bpc, most posterior point on Similarly, JG remeasured 12 endocasts cerebella in basal view. (three each from humans, bonobos, chim- panzees, and gorillas) one year after taking that rof should not be confused with the the initial measurements from these speci- frontal pole); (4) mcp–mbat, the shortest mens. The results ranged from 1–8%, with distance between the middle clinoid process the greatest relative remeasurement error (or anterior border of the sella turcica) and associated with the shortest lengths. mbat; (5) mcp–rof(tan), the shortest dis- The measurements included (Figure 1): tance between the middle clinoid process (1) bat–bat, the distance between the most and rof(tan); (6) cob–rof(tan), the short- anterior points of the temporal lobes in basal est distance between the caudal boundary of view; (2) mat–mat, maximum width of the the olfactory bulbs (cribriform plate) and frontal lobes at the level of bat; (3) mbat– rof(tan); (7) rob–rof(tan), the shortest dis- rof(tan), the shortest distance between the tance between the rostral boundary of the middle of the line connecting the two bats olfactory bulbs and rof(tan); (8) rof(tan)– and the tangent to the most rostral point on bpc(tan), the shortest distance between the orbital surfaces of the frontal lobes (note rof(tan) and the tangent to the most 698 . ET AL. posterior point on the cerebella in basal view observations and statistically significant (bpc). results were synthesized and the key features Measurements 3, 6, 7, and 8 were used to summarized for endocasts from apes, early calculate three additional lengths for each hominids, and humans (Table 4). specimen: [3–6] the length between mbat Additionally, because previously un- and cob;[6–7], the length of the olfactory known parts of Paranthropus endocasts are bulb (cribriform plate); and [8–3], the now available, new endocast reconstructions length of the basal aspect of the endocast were made for Paranthropus specimens caudal to mbat.