229 Dinosaurs and land plants

speci es of associated dinos aurs. Clearance is also a significant factor in locomotion across terrain upon which plants are growing or which arc strewn with fallen woody stems . T hus, in the cases of two massive sauro pods, th e be lly of Apatosaurus clea red th e gro und by slightly over a meter (G ilmore, 1936), while th at of B rachiosaurus cleared th e ground by nearl y three meter s (Ja nensc h, 1961). Perhap s th e habita t of th e giraffoid B rachiosaurus was littered with fallen tree trunks over which the anima l regul arly had to step, whic h could have effectively blocked the movement of an Apatosaurus.

Conve rgent selection in ecosystems

Biologists have long suspected that similar environmental conditions promo te the evo lution of similar plants and animals (Louw & See ly) 1982). On a gross biogeogra phic leve l, terrestrial ecosystems exhibi t marked similarities in structure and phys iognomy on different continents (W hitt aker , 1975; Pianka, 1978). If there is a limi ted number of efficien t solutions to a set of functional demands, natural select ion may force di ffering geno ty pes towar d sim ilar (convergent) phenotyp es (M ayr , 1963). All viable ph enotypes may not be realized with in one ecosystem; for example, large pseudo-kangaroos have not evo lved on northern continents. Nevertheless, many examples of convergence can be seen among organisms living under climatic conditions comparable to those that existed over broad areas of the planet during Mesozoic time.

T hus, clearings in humid West African forests are rapidly colonized by M usanga cecropioides, while similar habitats in South America are colonized by species of the genus Cecropia . These African and American species are remarkably convergent in their fast growth rates, soft stems , few terminal radiating branches and large palmate leaves. T he eco logical role of a pioneer apparently favors conve rgent solutions in rainforests of both regions. In the deserts and semi-arid grasslands of th e Old and N ew World, demands of water conservation, high temperatures and solar radiation have produced convergences between the eupho rbs of the former and the arborescent cacti of the latter.

T he dom inant mammalian land-dwelling terrestrial herbivores in West African forests are duikers and chevrotains, both of which possess the squat bodies, rounded backs and thick fur suited for life in dense secondary vegetation on or near the forest floor. In South America, capybaras and small deer are morphologically similar and fill simi lar ecological roles . Analagous convergences occur betwee n Old World

230 M. ] . Coe et at.

pangolins and New World arma dillos (Bour liere, 1973). Bot h the go lde n moles of southern Africa and the marsupial mole of Australia, in response to th e dem an ds of a fossaria l life in soft arid soil, have acquired tine silky fur, short and powerful digits in the forelimbs, and a horny prow on the ir snouts. In So uth America the fosso rial fairy armadillo bears a striking resemblance to these species.

Competitive interactions among dinosaurian herbivores led to the evolution of a wide variety of bod y shapes, their statures (reaches) bein g to some extent constrained, as arc those of mammals today, by the effect of rainfall on the sta tu re of plants fr om ar id grass lands, throug h low woodlands to th e tall fores ts of the humid tropics (Cae, C um ming & Phillipson, 1976 ; Cae, 1980b). As with modern mammals, d inosaurs were probably also separa ted eco logic ally by th e type of plant materials upon whic h they fed, their method of feeding, and seasonal patterns of dispersal and migration. Herbivorous dinosaurs may be sepa rated, at least at the superficial physiognom ic level , into quasi-fam iliar shapes (e.g. giraffe- like or elephant- like sauropods , rhino-like ccratopsians, camel­like hadrosaurs), wh ich sugg ests th at problems posed by Jurassic and Cretaceous vege tation to Mesozoic dinosaurs may have been rather similar to those posed by modern vege tation to modern mammals. T he number of basic mec hanical solutions ado pted by large verteb rates in the gather ing of plant food has probably not changed fundame ntally since Jurassic and Cretaceous time, and these solutions have appeared at different times in totally different phylogenetic lineages.

Jarman (1974) not ed that body size and muzzle shape in African antelopes are good p redictors of feed ing strategy . Sma ll forest- and scrub-dwelling duiker s seek ene rgy-rich food items such as fru its, antelopes of medium size are mixed (graze and browse) feeders, and th e largest ante lopes generally graze on coarse vegetation. Concentrate feeders in intermediate size ranges, such as the giraffe, kudu and gerenuk, have narrow snouts and select nutrient-rich shoots, pods and fruits on bushes and trees (H ofmann , 1973). Bulk-roug hage feeders, suc h as wildebeest, possess broad snouts. Concentrate feedi ng (shoot selec tor) strategies are suggested for many small ornithopods (e.g. Hy psilophodony

on the grounds of their narrow muzzles. Conve rsely, the broad rostrum of some large ornithopods (e.g. Edmontosaurus) resembles those of modern mammals that crop low swards of vege tation.

T he foregoing examples illu strate the possi ble importance of con­vergent selective pressures in evo lution. It is axiomatic that environ ­menta l constraints are critically important in determining which

organ isms survive in the strugg le for existe nce . On a broader sca le, mor e general convergences may becom e appa rent. Thus, interaction s with biological st resses may promote converge nt increases in the size of the central nervous system in mammals and birds (Russell, 1983 j Wyles, Kunkel & Wil son , 1983), and interactio ns with ph ysical st resses may promote con ver gen ces in she ll shape (Seilacher , 1984). Wi th the cu rrent emphas is on random processes in evo lution it is perhap s useful to consi der that non-random evo lutiona ry processes also occur; that is to

say , some evolutionary pathways are more frequently followed than others .

The browse line Specu lations on adapt ive int errelation ships between plants and herbi­vorous dinosaurs arc becoming mor e detailed and more firml y based on analogies with exis ting biologic systems . Some autho rs have argued that di nos aurian browse was indiscriminat e and gen erally not focu sed up on p lant reproducti ve structures . Only the smaller dinosaurs could have fed selectively on seeds or frui ts (H ughes, 1976). Other s have cited the existence of thorns, hairs and hard integ ume nts as defen sive adaptations of plant re prod uctive struct ures of Triassic and Jurassic age, before th e rise of the angiosperms . The diversity of these defen sive structures was taken as evidence of a degree of speci ficity in interact ions between plants and animals, including dinosaurs (W eisham pe l, 1984). Yet even sma ll dinosaurs must have been in ferio r gene rally to modern mammals and birds as dispersers of seed. Being ne ithe r as diver sified nor as intellig ent as modern hom oiotherms, sma ll dinosaurs would have been less well adapte d to di screte niches and less clever in seeking out att ractive propagul es (Regal, 1977).

Hughes (1976) and N orman & Weishampel ( 1985) st ress that specializatio ns for herbivory in dinosaurs becam e increasingly refined through Mesozoic time, and imply that dinosaurs fed abunda ntly on ang iosperm foliage. Furthe rmore , Bakker ( 1978) propose d th at th e feeding stra tegies of her bivor ous dinosaurs change d from browsin g p refer entia lly on tr ees to browsing preferentially on sh rubs and sma ller plants between the Late J urassic and Lat e C retaceous . Bushy angio­sperms were po stulated to have been able to withs tand browsing more successfully th an gym nosperm saplings . D inosaurian cropping thereby facilitated th e expansio n of primitive ang iosperms at the expense of gymnos perm forests . T his hyp othesis could be extended and applied to a much lon ger interva l of geo logic time.

232 M . J . Coe et at.

The Morrison Fo rmation, of Late Jurassic age, is exposed in the eastern Co rdillera and high plains of th e United Stat es and has produced one of the largest dinosaurian assemblages in the world (D odson et al. ,

1980). In most respects th is assemblage is similar to othe rs of Middle and Late Jurassic age around the globe, and it is taken here as typ ical of middle M esozoic terrestrial vertebrates dwe lling in semi-arid lowlands before the adve nt of the angiosperms.

Weight is a very important parameter in assessing the eco logy of verteb rates. In the past it has been necessary to construct scale models of living din osaurs in order to assess their bod y weight s. N ow, using a method devised by Anderson , H all-Martin & Ru ssell ( 1985), these weights can be estimated using measurements of limb bones . T he weight s obtained, in combination with info rmation on the skeletal architecture of various Morrison dinosaurs together with estimates of the relative abundance of their remains in existing co llec tions) can be used to calculate thei r relative abundance in their original populations, the preferred vertical range of feed ing and the am ount of plant material required to sus ta in them (sec Appendix). The ca lculated total plant tissue consumed by all Morrison dinosaurian herbivores, apportioned in vertical IO-crn increments according to the feed ing range of the indi­vidual herbivore taxa, is shown graphically in Fi gure 9.2 (a) (thus sauro­pod s were the principal consumers of vege tation at ground level and beyond 3 m above ground level; stegosaurs were very important con sum ers near the l-rn level ). The histogram may be consider ed as a browse profi le. A comparable browse profi le was cons tructed for mammals now inhabiting the Amboseli Basin in southern Kenya (Figure 9.2 (c).

A comparison of the Morrison and Amboseli brow se profiles suggests that one lon g-term trend in plant-vertebrate interactions is a reduction in the height of feeding levels. The effect would be ind ependent of whether or not the vertebrates are din osaurs or mammals, or whether the plants are gymnosperms or angiosperms. The din osaurian assemblage from D inosaur Provincial Park, Alberta, is intermediate in age (75 million years BP) between th at of the M or r ison Plain an d th e Amboseli Basin . Like other Late Cretaceous assemblages , it is dominated by ornithischian dinosaurs with body we ight s intermediate betwee n those of most sauropods and most modern big-game mammals. Relevant data are listed in T abl e 9.8 (Appendix), der ived acco rding to the same procedur e as for the Morrison din osaurs (relative abundances have not been corrected for size effect s because the herbivores are more nearly

Dinosaurs and land planes 235

100 1 - 1- - - _ _ ~

C 75 Angio sperms

~ Gymnosperm s ~ 50 '"•::; ~ 25

Ferns o

100 Angiosperms

~ 75 Gymnosperms

e ------­.g~ 50 --------- .. <,

~ Spo re- bearing plants <, §: 25 <,

~-±-50 -;:----;-","""---:=-=-=;--;I.;----:O:----:J::-+--:'::-l-cl'"'40 130 120 110 100 90 80 70 60 Pal Jr

Cretaceous

Figure 9.3. Palaeobotanical diversity in various North American sites according to (a) megafloral data, and (b) palynofloral data. Di versities are give n according to the percentage of the total number of species that can be allocated to major plant groups. Jr, Jurassic ; Pal, Paleocene.

An Aptian (late Early Cretaceous) megaflora, again from southwestern Ca nada (Bell, 1956), is esse ntially similar to flor as of Late Jurassic age from this region. However, two species of angiosperms (representing 6 ~'ri of the species identified) were living near the basin of deposition. An Albian (latest Early Cretaceous ) pal ynoflora from Okl ah oma (H ed­lund & Norris, 1968) contain s 42 species of spo re-bearing plants, 14 gymnosperms, 19 angiosperms and five of uncertain affinity. Three western North American megaflora I assemblages of the same (Albian ) age contain an average of 5.7 species of ferns , no ginkgophytes, 3.3 cycadophytes, 7.3 conifers and 10.3 angiosperms (Bell, 1956 ; Berry, 1922 ; De!evoryas, 1971 ; W ard, 1900). Ferns and fern prair ies ma y have been regionally important plant formations, as is further suggested by cha rred plant material dominated by ferns in the Wealden of England (Harris, 1981). Gymnosperms remained diverse, but the angiosperm radiation was under way. Megafloral species compos ition ranged from

236 M . J. Coe el al .

o to 57 ~.~ angios perms, and was evidently influenced by enviro nme ntal conditions near the site of deposition.

In a palynofloral sample from th e cen tra l U n ited Sta tes (Linnenberger sha le) of Ceno ma n ian (early Late Cre taceous ) age, 46 % of th e species present are ferns, while only 13 0/0 of the species be long to thi s group in a mega floral sample from the same unit. Fern spores evidently dominated the regional spore rain, but 63 ~~ of the more locally derived me gafloral species are ang ios pe rrnous. Between 63 ~/~ and 85 ~';l of mega floral spec ies from the same geographical region (D akota Formation; Lesquereux, 1892), also of Ce no ma n ian age, represent flowerin g plants. Angiosperm pollen increases in relative diversity from about 1 5 ~/o of species near the base of the un it to about 33 % near the top (M ay & T raver se, 1973).

By the end of Cretaceous (Maastrichtian) time in the northern interior of the U nited Sta tes , th e ab un dance of fern spec ies had declined both in palynofloras (2 1%, H ell C ree k Format ion ; Norto n & H all , 1969) and megafloras (3 (Yo, Fox Hill s Fo rmation; Brown , 1939). Angiosperms were dominant both region ally (55%) and locally (76%). Indeed some modern ang iospe rm genera can be identified from the palynoflor aI record in strata of slightly grea ter (Cam pan ian) age (Ja rzen, 1980, 1983). Except for a few bog-loving ang iosperms and scouring rushes there is no record of any Late Cretaceous herbaceous plants other than ferns, and extensive fern savannas and prairies may have continued to exist well int o T ertiary tim e (Elsik, 1968a, b; Hickey, 1977).

A shift in vegetat ion lasti ng thro ug h much of the C retaceous peri od would probably have posed no serious threat to dinosau rian survival. Plants th at were tar gets of heavy predat ion wou ld have gradually deve loped eithermechanical orchem ical defenses analagous to the thorns of modern acacias, and anti- herbivory subs tances such as those of living poison ivy and Euph orbiaceae. T here is no reason to suspect that any of the major groups of plants were entirely safe from browsing dinosaurs. U nfortunately, the record of herbaceous angiosperms is very poor in strata of M esozoic age, and grasses do not appear until the Tertiary. Ferns const ituted a very important element in Jurassic and Cretaceous vegeta tion. The vast grass prairies and savannas of geo log ically recent origin may unfortunately inhibit a proper appreciation of the previous eco log ical importance of fern prairies .

An analysis of changes in the diversity of dinosaurian species cannot at present be carried out because of the paucity of available data at a low level of taxonomic resolution. D ata pertaining to the generic diversity of herbivorous dinosaurs in North America through Mesozoic time

they are not obvious, for th e skeleta l anatomy of the local sauropods is very poorly under stood. T he Late Cretaceous peak in th e di ver sit y of herbivorou s dinosaurs is du e to an increase in th e nu m ber of ornithis ­chian tax a, which appears to more than compensa te for waning sauris­chian di ver sit y. A glob al comp ilat ion of di nosaurian genera and fami lies is not cu rrently available, altho ug h some fauna l lists of representati ve Ea rly and Late Cretaceous faunas from various localiti es aro und the world have been published (Molnar, 1980). From th ese it wo uld appea r tha t th er e was no decline in th e wo rld -w ide di ver sity of saurischians throu gh Cre taceous tim e, an d that th e wo rld-w ide Late C re taceo us expans ion of or n ith ischia ns was mor e mod est th an th at suggeste d by th e N orth Am erican record .

Ther e is no sudden increase in the taxon omi c di versit y of dinosaurs afte r th e middle C retaceous or igin of th e angiospe rms tha t would indicate a radiation of herbivor ou s dinosaurs in respo nse to a new source of food . The high diver sity levels achieved by L ate Cretaceous dinosaurian herbivores cou ld be interpreted as a response to th e in­creasing importance of ang iosperms in pla nt com m un ities, but world­wide data imply that th e increase in d iversit y was pa rt of a trend whi ch began during Jurassic t ime . It m ight be countered, with equal validi ty , that herbivorou s dinosaurs wer e undergoing a long-su stained radi ation that would have occu rred reg ardless of the taxon omic affini t ies of th e plants upon wh ich they fed .

The first angiosp erms were early success iona l plants of moist hab itats (Retallack & D ilcher, 1981). M odern early successional species tend to

grow more rap idly and be less well defended chemically than late success iona l forms (Opler, 1978). If the same were true of ancestral angiosp erms, these feat ures wou ld have render ed th em attracti ve to large plant-eaters such as dinosaurs. Through feed ing and other activitie s, dinosaurian herbivor es ma y have opene d up wooded areas to early stages of succession (Far low, 1976), an d brou ght angiosperm propagules to th em in their guts. L on g passage tim es of digesta imposed by low metabolic rates and large size (Parra, 1978 ; D emment & Van Soest, 1985 ; Kar asov & Diam ond, 1985) may have rendered this form of propagul e di sp er sal hazardous. Nevertheless, it is not inconceivabl e, over the course of th e ten million yea rs required for angiosperms to spread around th e glob e fr om th eir area of origin (Retallack & Dil cher, 1981), that dinosaur-med iated di sp er sal was as important as th at of abi oti c agents such as wind and water.

Ther e wer e changes in the gro ss skeleta l morphology of dinosaurian

239Din osaurs and land plant s

C:-::::=~;':;:;;;;;~-;:;S 77 .5;:;;;A~ cm.~Edmontosaurus regalis W 3800 kg

.... S ty racosaurus albertensis SA 52 .5 em t

W 3700 kg

Cervus elaphus (E lk) SA 17.6 em" W 312 kg

C~ ~ ~ L eptoceratops gracilis SA 17.6 ern'

Mammuthus primigenius W 190 kg SA 90.0 em ! W 4000 kg

Figure 9 .5. Surface areas (SA) and shapes of dental batteries of Cre taceous din osaurs and subfoss il or Recent mammals of com par­able bod y weights (W). In each case the outl ine represents the occlusa l su rface of the dentition in one mandibular ramus. The comparisons suggest that , becau se the surface areas of the dental batteries and weights of Edmontosaurus (orn ithopod) , S tyracosaurus (ceratopsian), and Mammuthus (mammoth) , on one han d, and Leptoceratops (ceratopsian) and Cerc us (elk), on the othe r, are similar, th e metabolic rate s within the two groups of herbivores may have also been similar. Weights were calculated accord ing to eq uations 9.1 and 9.2 (see Appendix), and dental battery surface areas were measured directly from mandibles belonging to the same specimen as that on which the weight calculation was based .

herbivores between Late Jurassic and Late Cretaceous tim e. They too ma y well reflect general trends in the evolution of herbivorous verte­brates, rather than the special case of dinosaurian adaptation to angio­sperm consumption. Large, Late Jurassic herbivores (sauropods, stegosaurs) tended to have enormous bodies, small heads, and cropping but not chewing teeth. Small Late Jurassic ornithopods (hypsilopho­donts, camptosaurs) tended to have narrow beaks suggesting selectivity in cropping plant parts, and teeth that could crush plant t issues before they were swallowed.

By Late Cretaceous time the mean body size of herbivorous dinosau rs had declined substantially, and highly evolved grinding dentitions appeared within the Jarger, more muscular heads ofadvanced ornithopods

240 M . } . Coe er a/.

(hadrosau r -ids) and horned dinosaurs (ceratopsids) (F igu re 9 .5) . The greater taxonomic d iver sity of th ese two groups suggest s a higher degree of specificity in animal-plant rel ati onships than was the case with Jurassic d inosaurs. As in living herbivorous lizards) turtles, birds and mammals, most herbivorous dinosaurs probably employ ed a symbiotic gut m icroflora to ferment high-fiber plant mat eri als, and perhaps even

to det oxify poi sonous plan t seconda ry su bs ta nces. By chewing th eir food, ha drosaurs and ce rato ps ids m ay have incr eased digestive rates and the rat e of passage of food material through the gut. Hi gh er act ivity levels, higher ferm entation rate s and sm aller body m asses are suggestive

of a closer ap p roach to en dothermy than in the case of Late Jurassic dinosaurian herbivo res (Ost rom , 1980). Late Cre tac eous herbivores were intermediat e in size between th eir sma ll and large Late ] urassic counter pa rts, and were thus more uniform in weight. They di ffered from

small Late Jurassic orn ith isch ian he rbivores in thei r gre ater size and broad er beaks. Thus Late C retaceous orn ithopods evidently cou ld not be as di scriminating as th ei r sma lle r forebea rs in the q uality of the plant material th ey ing ested. T he y may have been select ive in the kind of plants they chose to feed on bu t not in what they croppe d from the m .

Sauropods may have ado p ted the strategy of stri pping vegetation and

retaining it in large quantities for slow fermentat ion . If so , this stra tegy wou ld appear to ha ve cont inued to be success fu l in some ar eas of the world during Late C retaceous time, notably in South Am eri ca and India (M olnar, 1980). Low metabo lic rates (see Appendix) enabled sau ro pods to grow to enormous d imensions, and yet avo id se rious su rface-volume

p roblem s in heat di ssipat ion that would hav e been en counter ed by endotherms of similar body weight. High d igest ive efficien cies cou ld have been attained in the absence of m ech ani cal food macerati on simply

by expos ing the d igesta to a m or e len gthy process of microbial fermen ­tation . Without moving the ce nt er of mass, a lon g neck suppo rted by a large skeletal fram e could carry the head to cro p plant material ove r a broad su rface area of terrain or hi gh in to trees. Inter est ingly, th e Late C retaceous sau ropod Nemegtosaurus possessed a long, d ist ally expanded muzzle rather like th ose of hadrosaurs (N owinski, 1971).

Many m od ern ferns are no t eaten by animals because of poisonous

secondary metabolites that accumulate in their fronds. These toxic compo unds are probably the result of a long hist or y of selec tive pressures, and m ay be m or e prevalent in ferns than in angiosperms

sim ply becau se ferns ar e a much m or e ancient group. Such poison s may well have originated in response to attacks by phytophagous ins ects , but

24 1 Dinosaurs and land planes

the y are now quite effective in de terring excessive feedi ng by warm­blooded herbivores (Janzen, 1970). However, simi lar levels of chemical defense ma y not have protected plants from cold-blooded herbivorous dinosaurs. T hese animals were proba bly ' animated compost heap s ', like the living gia nt to rto ises , which ingest relat ively sma ll quanti ties of food and retai n it in th eir bod ies for up to 20 days. Torto ises will to lerate highe r tanni n conce ntrations in the p lants they consume th an will herbivorou s hom oiotherms (Swain , 1976). In accordance with the dic­tates of the Red Queen Hypothesis, whereby an organism must continue to improv e in absolute adaptive fitn ess in an evo lving wo rld in order to maintain its relative evo lutiona ry fitn ess (Van Valen, 1973), the bio­chemical sophisticatio n of di nosa urian digestion must have kep t pace wit h the bioch em ical defenses of ang iospe rms. In any case, M esozoic fer ns provi ded lar ge quan tit ies of vegetation that could easily have been cropped by grazing rep tiles.

T he geo log ic record thus shows that the dinosaurs continue d a lon g-term incr ease in di versit y coincident with the time of origin and initia l div ersi fication of angiosperms durin g middle and late C retaceous time. Land plant d iversity d id increase during this in terva l, and gym no­sperms steadily declined as the angiosperms divers ified . U nlike the gym nosperms, however, th e d inosaurian or de rs contin ued to d iversify un ti l the end of the C retaceous Period.

An giosp erms and the extinc tion of the dinosaurs

T he rise of th e ang iosperms has often been linked with environme nta l cha nges pos tu lated to have driven the dinosaurs gradually to extinc tion. Swain (1976) proposed th at angiosperms produce a variety of toxic chemicals th at were sta ted to be generally abse nt in more primi ti ve vasc ular plants, and tha t are less eas ily detected by reptiles than by mammals. The change in te rres tria l floras was thu s held to be at least partly resp on sible for de termin ing whi ch of th e rival classes of terrestrial vertebra tes would become dominant. In a similar manner , Krassilov (1981) linked the initial success and ultimate dem ise of the dinosaurs to the early M esozoic expans ion and late M esozoic decline of sclerophy llous shrub lands and fer n mar sh es. Dinosaurian success was aga in cons idered to have been a function of p lan t evolu tion . The contro versy surro un ding th e extinctio n of the d inosaurs has become part icularl y lively, with a di vision of opi nio n between supporte rs of more or less gra dua l st resses gene ra ted by te rres trially limited mechanisms, and sup po rters of

242 M . ] . Coe et al.

catastrophic st resses gene rated by extrate rrestr ial me chani sm s. It is not our intent her e to take par t in this debate , but sim ply to mention a few aspects of biotic changes that occurred during this interval as th ey could have related to d ino saur-p lant int eracti on s.

It has become ap pa rent that a flor al event was associa ted wit h the deposition of th e C retaceous-Tertia ry boundar y clay in some terrestrial environments (H att on , 1984 ; T schudy et al. , 1984 ; Nicho ls et al., 1986).

This event is mar ked by a grea t increase in th e relat ive ab undance of fern spores in sediments measuring on ly a few centime ters in thickness, and representing a very short in terva l of ba sal Tertiary time . T his ha s bee n interpreted as evidence of th e rapid recolonizati on of de vastated mid­continental reg ions by wind-d ispersed ferns followi ng th e extinction interval (Wolbach , L ewis & Anders, 1985). In itself, a vege tational simplification of this order would have produced a great reduction in th e biom ass of large vertebrate herbivores, which would probably have been acco mpanie d by extinctions .

In the longer term, the identifica tion of rem ains of th ermophilic plants in Early Tertiary stra ta suggests th at condit ions in the ph ysical environment were approximately the same during pre- an d post­extinc tion time (T schudy, 1970 ; Nicho ls et al., 1986). Neverthe less, basal Tertiary floras in two regions of th e N orthern H emispher e wer e cha racterized by an abundance of gymnospe rms of modern aspe ct (K rassilov, 1978; Sweet & Hills, 1984). It has be en suggested (Ja nzen & M artin , 1982) tha t, whe n th e mammalian megafauna becam e extinc t in Ce ntra l America 10 000 yea rs ago, the ranges of severa l tree spec ies wh ich depended on large mammals for di spersing th eir fru its wer e reduced (but see H owe, 1985). By analogy, if some Late C retaceous angiosperms also depended on dinosaurian herbivor es for di sp er sing th eir propagules, they too would have been placed at a di sadvantage in post-extinction floras. In thi s context , it is pe rhaps relevant to not e a rapid increase in the mean size of plan t propagul es in earliest T ertiary t ime (T iffney, 1984), which may have coinc ide d with a shift from foliage-eati ng dinosaurs (where, in the model proposed by Janzen (1984) th e fo liage surrounding th e seeds att racts lar ge herbivores in the same manner as frui ty tissu es su rrounding seeds att rac t small herbivores) to fruit-eat ing mammals and birds as d ispersal agents. T here ma y be evide nce here of th e eco logical integration of dinosaurs with plant life beyond th e stage of sim ply usin g plant vegetative organs fat food. The abrupt and com plete removal of large herbivores in itse lf would hav e

Dinosaurs an d land plants 243

resulted in altered forest com pos it ions, as evi dence d by the effects of the presence or ab sen ce of elephant in mod ern African woodlands (E ltr in gham , 1980 ).

It is doubtful that differ ences be tween Late C retaceou s and Early T ertiary vegetation were of sufficient magnitude to have been the primary cause of th e extincti on of th e dinosaurs. Ind eed , some of th e vegetationa l changes that ar e ob ser ved may have lar gely resul ted from th e absence of d inosaurs in terrestrial ecosystems. T he great herbivor ou s di nosaurs had evolved under rather cons tant and st rong selective pr es­sures that had prevailed during at least much of M esozoic time. Co uld these selective pressures have cont in ued broadly to direct th e T ertiar y radiation of large m ammalian herbivor es alon g pathways that resulted in large-scale st ru ctural convergences with dinosaurian prede cessors ?

Appendix: Derivation of browse profiles

According to An der son et al . ( 1985) dinosaurian weig hts (W , g) can be estimated

from th e circumfe rences of the humerus (Ch , mm) and femu r (Cp mm ) in quadrupeds, and the circumference of the femur in bipeds:

log" 117 ~ 2.73 log" (Gh + G,) - 1.11 (9.1)

(for which " ~ 0.99) 10 g1 0 W = 2.73 (log IOCf )-O.76. (9 .2)

The correla tion ind ex, or coefficient of determination ( r~) , is the p roportion of the total variation acco unte d for by the fin ed reg ression on the data given. T he

closer th e ind ivid ual data values lie to the linear regression , th e nearer is the r 2

value to 1. Note that no r value is given in eq uation 9 .2, because the equat ion was not derived through sta tist ical procedures. Weights listed for Morrison

dinosau rs (T able 9.1) were calcu lated for large spec ime ns of each taxon largely

on the basis of th ese rela tion ships. In order to obtain some appreciation for the relat ive abu ndance of different

din osaurs in the Morrison assem blage, minimum number s of individua ls (M N Is) were talli ed from publication s, from the field records of th e American Museum of N atural Hi story and from conversa tions with D an Ch urc, Bruce R. Er ickson,

James A. Jensen , Rob ert A. Long, James H . M adsen [r and Jo hn S. McIntosh , all of whom are author ities on Morrison dinosaurs. The percentage ab undances

of various genera are listed in Table 9.2, although it mu st be stressed that these numbers ar e working estimates that ma y be modified substan tially in future revisions. Sp ecimens of A llosaur us (M adsen, 1976) and Diplodocus (Brown, 1935) that occur in mass death assoc iations have not been included , nor have those from the Dry M esa Quarry in weste rn Co lora do that arc current ly bein g

245 Dinosaurs and land plants

Tab le 9 .2 . M inimum number of individuals (M N l) of Morrison dinosaurs

Observed Co rrec ted MNI ("0) M NI ('ifll)

Camarasaurus 20 .56 Nannosaurus 13.10 Apatosaurus 16.10 Dry osaurus 11.99 Diplodocus 14 .09 Othnielia 11.71 Stegosaurus 13.03 Allosaurus 10.00 Allosaurus 11.61 Coelurus 9.34 Camptosaur us 8.06 Camptosauru s 8.81 Dry osaurus 3.89 S tegosaurus 6.9 1 Barosau rus 1.98 Ornitholesres 6.12 Ceratosaurus 1.90 Camarasaurus 4.80 Othnielia 1.80 Diplodocus 4.42 Haplocanthcsaurus 1.73 Apatosaurus 3.33 Coelu rus 1.21 Stokesosaurus 2 .53 Brachiosaurus 0.99 Ceratosauru s 2.48 Ornitholestes 0 .60 M arshosaurus 1.85 Marshosaurus 0 .60 Elaphrosaur us 0 .97 Stokesosaurus 0.60 Hapiocanthosaurus 0 .79 N annosauru s 0.47 Barosaur us 0 .41 Elaphrosauru s 0.30 Nodosaur 0 .19 N odosau r 0.24 Brachiosauru s 0 .18 Sauropod, und escr ibed 0.2 4 Sauropod) undescribed 0.06

that the n umber of small juveniles becomes ab ou t equal to tha t of lar ge adu lts , a circumstance typical of condition s of attrit iona l mo rta lity (K lein & Cruz-U ribe, 1984).

In Table 9.3 •corrected • M NIs and weights are m ult ip lied and normalized to

100 % (or to 100 g) to ca lcu late the relati ve biomass contributed by vario us taxa. The relat ions hip between longevity and body weight has not been we ll stud ied in living rep tiles) but a review of zoo data (for longevit ies see Bowl er , 1977 ; weights were compiled fro m many sources) suggests th at , with th e exception of tu rtles, th e relationsh ip is approximately the same as for mammals (Sacher ) 1975). If th is is true) the values for the ratio p roduct ivit y (P) :b iomass (B ) may

also be similar . However, there is a considerable amount of sca tte r in the relations h ip . Fo r example, low values of P: B in lon g-lived fishes (6 % to 7 % ) contrast strongly with th ose in sho rt- lived fishes, which may exceed 60 % (M cNeill & Lawt on ) 1970). T he annua l tu rnover of turtle biomas s on Aldabra

is estimated to be just under I % for animals with a me an body weight of 22 kg (H am ilton & Coe, 1982). A population of herbivoro us mammals with th e same body weigh ts would be expected to tu rnover as mu ch as 80 % of the ir biom ass eac h year (Coe, 1980a). T he mamma lian rela tionship (Farlow, 197 6 ; p. 846; see a lso Ban se & Mosher) 1980) bet ween the P :B ra tio and body weigh t (W , g) is

00

N

""

Table 9.5 . Data fo r Amboseli mammals (b iom ass = 4848 kg krrr" )

Weight Biomass Intake Animals N umber (kg) (kg km") (kcal km") per km 2

Elephant 2 18 3000 75 1.4 20288.9 0.2 5 Hippo 75 1000 86.3 3 132.5 0 .09 Rhino 40 8 16 37.3 1433 .5 0. 05 G iraffe 115 772 101.8 3960.3 0 .13 Buffalo 457 450 236. 1 1062 4.4 0 .52 Eland 37 363 15.5 740.0 0.04 Ze bra 2200 200 505 .6 28366.8 2.53 O ryx 88 167 16.0 942.3 0 .10 W ildebe est 2473 165 468 .8 27 706.2 2.84 Kon goni 87 136 13.6 844.3 0.10 Gram's gazelle 1230 50 70.8 5 768 .6 1.42 Impala 700 45 36. 4 3050.6 0.81 Thomp son 's gaze lle 788 20 17.9 1872 .7 0.90 Ostri ch 142 114 18.4 1203.0 0 .16 Cattle 9 184 227 2394.9 129804.2 10.55 Sh oat 2550 16 47.0 5 215.1 2.94 D onkey 20 3 130 30 .5 1924.2 0 .23

Total 4848.3 246877.6 23 .66

Jurassic and mod ern vert ebrate faunas both lived in sem i-a rid envi ronme nts . Weights and populat ion numbers of Amboseli herbivores ar e by W estern (1975 ) and Behrensm eyer et al. (1979), and estimates of hippo populat ion s were prov ided by A. K . Behrensm eyer (personal communication). T he relat ive int akes of different componen ts of the Amboseli herbivore biom ass we re calcu lated accord ing to Farlow's (1976) eq uation for endothe nns :

log" Intake (keel/ day) ~ 0.23 +0.72 log" IV (9.7) (for whi ch T' ~ 0.95 ).

Fo r a total Amboseli biomass of 4848 kg km"? (Coe et al., 1976), abo ut 247 000 kcal of plant food are on the average rem oved fr om each sq ua re kilometer of the Basin per da y (T ab le 9 .5).

T he dinosaurian herbivores of th e Morrison Pla in were predominantly sau ro pods. Saurop ods ma y be conside red as possessing reptil ian metabol ic rates on the basis of brain-body propor tions (Hopson, 1980 ), surface-volume proportions and heat loss (Spotila, 1980), th e amo unt of time necessar y for feeding (Weaver, 1983), and th e presen ce of th ermally induced (seasonal) ring s in their bones (Reid , 1981 ; Ricqles, 1983). M od em ana logs of mid-Mesozoic gymnos pe rms often have relativ ely slow rates of grow th (R ussell, Beland & McIntosh , 1980), but th ese may have been offset by higher conce nt rations of

atmospheric carbon dioxide (Bern er. La saga & G arrels , 1983 ; Rogers, T homas & Bingham, 1983). Sh ould differences in tr ophic relation ships, suc h as a relati ve

increa se in primary producti vity and a decrease in resource ut iliza tion by secondary consu mers, cance l each othe r out between Amboseli and M orrison

environments, th e M orrison Plain would have sustaine d a biomass of larg e herbivores amo unt ing to 93000 kg km" (T able 9.6 ). The factor of nearly 20 separating the two biomass conce nt rations is du e both to relatively lower reptilian met abolic rates and to th e great differen ce in mean weights between th e herbivores of the two ages (200 kg vs 7000 kg). In terms of number s of individuals, Amboseli herbivore population s are twice as den se. However , in th e Amboseli Basin , elephan t population den sit ies amo unt to on ly one animal per 4 km", while on th e M orrison Plain ove r 50 animals, each weigh ing tw ice as much as a large elephant, would occu r in the same area.

That suc h de nse populations of giant herbivores cou ld ever exist might hardly seem cred ib le. T he on ly potential mod ern an alog for the huge, cold- b loode d herbivorou s dinosaurs are th e giant tor to ises th at inhabit the Aldabra Atoll (240 km northwest of M adagascar ), and th e G alapagos Islands (800 km off the coas t of Equador ). Population studies (Bourn & Coe, 1978) have reveal ed tha t about 150000 tortoises occur on Aldabra , 60 ( :/~ of which are concentrated

in an area of 3360 ha. T hese figur es tr an slate into biom asses of 58000 kg km "? in th e area of concen trat ion, and 53000 kg km " ! for the whole of th eir range. In Afri can wildlife ecosystems dominated by large herbivorous mammals, the highe st biomasses recorded are; 17500 kg krn' " on th e Rwindi Plain, Zalre , 19000 kg km "! in th e Man yara Na tional Park, T an zani a ; and 20000 kg km '" in

Table 9.7. V ertical fee ding range of herbiv ores

1. A1orrison dinosaurs N annosaurus Othnielia Dryosaurus Ca mptosaurus Stegosa urus stenops S tegosaurus ungulatus Haplocanthosaurus Diplodocus Apatosaur us B arosaurus Ca marasaurus Brachiosaurus

II. Amboseli mammals Impala Thomp son's gazelle Catt le Shoat Donkey Wildebeest Kongoni Grant 's gazelle Hippo Oryx Rhino Buffalo Eland Zebra Ostrich Giraffe Elephant

Range (m)

0 ­0.4 0 ­2.0 0 ­2.0 0- 2.8

0. 3 - 1.1 0.4-1.3

0 - 4 .5 0 - 4 .5 0 - 4.5 0 - 4.5 0- 6.5

0.4 -12.0

0- 1.0 0 - 1.0 0 - 1.0 0- 1.0 0- 1.0 0- 1.2 0- 1.2 0- 1.2 0 - 1.5 0 - 1.5 0- 2.0 0- 2.0 0 - 2 .0 0- 2.0 0 - 2 .0 0- 5.5 0-10.0

Tab le 9.8. Data fo r Dinosaur Pa rk megafauna! herbiv ores

A B C 0 E

Euop/ocephalus 14.69 2000 11.02 8.47 0- 0.50 Panoplosaurus 4.28 2250 3.6 1 3.06 0 - 1.00 Ceratopsids 25.12 3500 32 .97 40.01 0- 2.00 Hadrosaurids 55 .91 2500 52.41 48.37 0- 4.00

A, Percentage of herbivore individuals. B. Weight of individu al herbivore (kg). C. Percentage biomass of herbivore groups in assemblage. D . Percentage of total kilocalories ingested by megafaunal herbivores per day (individual intake calculated from equat ion 9.6). E. Vertica l feeding range (m).

252 M. J. Coe et at.

A brief digression is necessary to exp lain how the upper limit of sau ro pod brow se was defined . Bakker (1978 ) proposed) as had severa l previous aut hors,

that sauropods could thrust th eir necks h igh into t rees by standing on th eir hind legs and bracing themselves in a t ripodal posture with their tail s on th e gro und .

Alexander (1985 ) a lso suggested th at sau ropods cou ld rotate their bod ies abo ut their hindlimbs, holding the forelimbs off the ground . Hi s arguments are furth er supported by th e extreme lightening of th e sauro pod neck (although taphonomic ev ide nce suggests th at if the neck was su ppo rt ed b y a powerful dor sal tendon , the cervical vert ebrae none the less became disassociated more rapidly than did other region s of th e vertebral co lumn). Elephants can also stand rather eas ily on thei r hind legs. However, th ey do not usually feed in th is posit ion, and sauro pods probably did not often feed from a tripodal stance either . The upper limit of sauropod bro wse is set here by th e maximum' com forta b le ' level to wh ich it is beli eved th e head cou ld be brought by th e neck , withou t th e animal risin g on its bac k legs.

W e are particul arl y gra te fu l to William Chalone r and Peter C rane for th eir cons truc tive counse l in th e cou rse of prep ar ing this manuscript.

References

Alexander , R . M eN . (1985). M ech an ics of posture and gait of some large

dinosaurs. Zoological Jo urnal of the Linnean Society , 83 , 1- 25. Anderson , J. F. , Hall -Martin , A. & Ru ssell, D . A. (198 5). Long-bon e

circum ference and weight in mammals, birds and dinosaurs. Jou rnal of Zoology, 207, 53-61.

Bakk er , R. T . (1978). Dinosaur feeding behaviou r and th e origin of flowering plants. Na ture, 274 , 661- 3.

Ban se, K . & M osher , S . (1980). Adult bod y mass an d annua l

production /biomass relation ships of field populati on s. Ecological AJonographs, 50, 355-79 .

Behren sm eyer , A. K. , Western, D . & Boaz, D . E. D . (19 79). New perspectives in verteb rate pale oecology fro m a recent bone assemblage. Paleobiology ,S, 12- 21.

Beland, P . & Russell, D . A . (19 78). Pal eoecology of Dinosaur Provin cial Park (C retaceous), Alberta , inte rpreted from th e distribution of a rt icu lated

vertebrate rem ains. Canadian J ournal of Earth S ciences, IS , 1012- 24 . Bell, W . A. (1956). Lower Cretaceous floras of western Ca nada. M emoirs

Geological Survey of Canada , 285, 1-331.

Berner , R. A., La saga, A. C . & Garrels, R . M . (198 3). The carbona te-s ilicate geochemical cycle and it s effect on atmos p heric carbon d iox ide ove r th e

pa st 100 million years. American J ournal of Science, 283. 64 1-83 . Ber ry, E. W . (19 22). The flora of the Cheyenne Sands to ne of Kan sas. U.S.

Geological S urvey Prof essional Papers. 1291, 199-225 .

Bird, R. T. (1985 ). Bones f or Barnum Brown, cd . V. T . Schreiber &

J . O . Farlow. Fort Worth : Texas Christian Universit y Press. Bond, G. ( 1955). A note on dinosaur remains from the Forest Sandsontc

(U pper Karoo). Occasional Papers, Na tional Mu seum of Rhodesia, 2, 795-800.

Bourliere, F . (197 3). The comparative ecology of rain forest mammals in Afri ca and tropical America . In Tropical Forest Ecosy stems in Africa and South America, ed . B. J. Meggars , E . S . Ayensu & W. D . Duckworth, pp . 279-92 . Wa shington, DC: The Smithsonian In stitution .

Bourn, D . & Coe, M . J. (1978). The size, st ruc tu re and distribution of th e giant tortoise population of Aldabra. Philosophical Transactions of the Roy al Society , series B , 282 , 139-75.

Bowler, J . K . ( 1977). Longevity of reptiles and amphibian s in North

American collections. Society for the Study of Amphibians and Reptiles, Mi scellaneous Publications in Herpetology, 6.

Brown, R. (1907). Gastro lith s. Science, 25, 392. Brown, B. (1935 ). Sinclair Dinosaur Expedition, 1934. Natural Hi story, 42,

J-15. Brown, B. (1941 ). The last dinosaurs . Natural History, 48 , 290-5. Brown, R. W . (1939 ). Fo ssil plants from the Colgate Member of the Fox

Hills Sandstone and adja cent strata. U.S. Geological Survey Professional Papers, 1891, 239-75 .

Chaloner, W . G . (196 8). The paleoecology of fossil spores. In Evolution and Environm ent . A Symposium presented on the Occasion of the 100th

Anniversar y of the Foundation of the Peabody Museum of Natura l Hi story, pp . 125-38. New Haven : Yale University Press.

Chandler, M . E. J . (1966 ). Fruiting organs from the Morrison Formation of Utah, U .S .A. Bulletin of the Br itish Mu seum (N atural History ) Geology, 12, 137-71.

Coe, M . J . (l 980a). African mammals in savanna habitats . In International Symposium on Habitats and their Influen ce on Wildlife, pp. 8}-109. Pretoria: Endangered Wildlife Trust.

Coe , M . J. (l 980b). The role of modern ecological studies in the reconstruction of palaeoenvironments in sub-Saharan Africa. In Fossils in the Making , ed . A. K . Behrensmayer & A. K . Hill, pp . 55-71. Chicago : University of Chicago Press.

Coe, M . J., Cumming, D . H . & Phillipson, J. (1976). Biomass and production of large Afri can herbivores in relation to rainfall and primary production. Decologia, 22, 341-54.

Collins, S . L. & Uno, G. E . (198 5). Seed predation, seed dispersal, and disturbance in grasslands : a comment. American Naturalist, 125,866-72.

Delevoryas, T . (1971). Biotic provinces and the Jurassic-Cretaceous floral transition. Proceedings of the N orth American Paleontological Convention, ed . E. L. Yochelson, v01. L, pp- 1660-74. Lawrence, Kansas : Allen Press .

Demment, M. W . & Van Soest , P . J . (1985 ). A nutritional explanat ion for body-size patterns of ruminant and nonruminant herbivores. American Naturalist. 125, 641- 72.

Dodso n, P., Behrensrneyer, A. K., Bakker, R. T. & Mcintosh, j . S. (1980). Taphonomy and paleoecology of the dinosaur beds of the Jurassic Morrison Formation. Paleobiology , 6, 208-32 .

Dubost, G. (1979 ). The size of Afr ican forest artiodacty la as determined by

the vegetation structure . African Journal of Ecology, 17, 1-17. Elsik , W . C. (1968a). Pa lyno logy of a Pa leocene Rockda le lignite, Milam

Coun ty, Texas. I. Morphology and taxonomy. Pollen et Spor es, 10,

263-314. Elsik, W. C . (1968b ). Pa lynology of a Paleocene Ro ckdale lignite , Milam

County, Texas. II . Morphology and taxonomy. Pollen et Spores, 10, 59'H>64 .

Eltringham, S . K . (1980). A quantitat ive assessment of range usage by large African mammals with particular reference to the effects of elephants on trees , African Journal of Ecology, 18, 53-71.

Fa rlow, J . O. (1976). A consideration of the trophic d ynamics of a late Cretaceous large-dinosau r community (O ld man Formation). Ecology, 57, 841-57 .

Gi lmore, C. W . (1936). Osteo logy of Apatosaurus, with specia l referenc e to specimens in the Ca rnegie Museum . M emoirs of the Carnegie Museum, 11, 171- 298.

Graham, A, (1968 ). The Lake Rudolph crocodi le (C rocody lus niloticu s Laurenti) population . M .Sc. th esis, University of Nairobi .

Hamilton, J . & Coe, M. J . (1982). Feeding, digestion and assimilation of a popu lation of giant tortoises (Geochelone gigantea Schweigger ) on Aldabra

Atoll. Journal of Arid Emnronmentsc S, 127-44. Ha rri s, T . M . (1981 ). Burnt ferns from the English Wea lden . Proceedings of

the Geologists' A ssociation, 92, 47-58 .

He d lun d, R. W . & Norris, G . (1968). Spores and pollen grains from F redericksburg ian (Aptian) strata, Marsha ll County, Oklahoma. Pollen er Spores, 10, 129-59 .

H ickey, L. J . (1977 ). Stratigraphy and pa leobotany of the Golden Valley Formation (Early Tertiary) of western North Dakota. M emoirs of the

Geological Society of America, 150, 1- 183. H ill, C. R. (1976). Coprolites of Pt ilophyllum cuticles from the Midd le Jurassic

of north Yorksh ire. Bulletin of the British A1useum (N atural History ) Geology, 27,289-94 .

Hofmann, R. R . (1973 ). The ruminant stomach. East African M onographs in Biology, no . 2. Nairobi : Ea st African Publishing House.

Hopson , J. A. (1980) . Relative brain size in dinosaurs : implications for

dinosaurian endothermy. American A ssociation for the Advancement of Science S elected Symposia, 28 , 287-310.

Hotton , C. (198 4). PalynofloraI changes across th e C retaceo us-Terti ary

bounda ry in east cent ral M ontana, U. S.A. Abstracts, Sixth International Paly nological Confe rence, Calgary 1984 , p. 66 .

Howe, H . F. (1985). G omphot he re fru its: a cr itiq ue. American Naturalist, 125, 853-M,

Hughes, N. F . (1976 ). Pala eobiology of A ngiosperm Origins. Ca m bridge: Cambridge U niversi ty Press.

[an ensch , W. ( 1929) . Magen st eine bei Sauro pod en der T endaguru-Sch ichten . Palaeontographica, Su ppl. 7, Erste Reihe, T eil 2. 135-44.

[a nensch, W. (1961). D ie Skel ettrekon struktion von B rachiosaurus brancai, Palaeontographica, S uppl. 7, Erste Reihe, T eil 3, 237-40.

Jan zen , D . H . (1970). Herbivores and th e number of t ree species in trop ical fores ts. American Naturalis t, 104, 50 1- 28.

Janzen, D . H . (1984). D ispersal of sma ll seeds by big herbivores : foliage is the fruit. American Na turalist, 123, 338-53.

Janzen , D . H . & Martin, P. S . (1982). Neotropi cal anac h ron isms : th e fru its th e gomphotheres ate. S cience, 215,19-27.

Jarman , P . J . (197 4). The soc ial organisati on of an telope in rela tion to their eco logy. Behaviour, 48, 215-66.

Jarzen, D . M . (1980 ). The occurre nce of Gunnera pollen in th e fossil record. Bictropica, 12, 117- 23.

jarzen, D . M . (1983) . The fossil pollen record of th e Pandanaceae. Gardens' Bulletirl of Singapore , 36, 163-75.

Karasov, W . H . & D iam ond, J . M . (1985). Dig est ive ada pta tions for fue ling the cos t of endo thermy. Sc ience, 228, 202-4.

Klein , R. G . & Cruz -U ribe, K . (198 4). Th e Analy sis of Animal Bones fro m A rcheological Sires. Chicago: University of Chicago Pr ess.

Kr assilov, V. A. (1978). Late C retaceous gymnosperms from Sakhalin ,

U .S.S .R. , an d the terminal C retaceo us event . Palaeontology , 21, 893--905,

Krassilov , V. A. (198 1). Changes of M esozoic vegeta tio n and the extinct ion of th e d inosaurs. Palaeogeography , Palaeoclimatology , Palaeoecology , 34, 207-24.

Lesquereux, L (1892 ). The flor a of the Dakota Group. U. S . Geological S urvey M onographs, 17.

Lou w, G. N . & See ley, M . K . (1982). Ecology of Desert Organisms. Harlow :

Longman. McNeill, S . & Lawton , J . W . (1970). Animal product ion and respirati on in

anima l populations. Nature, 255, 472-4. Madsen , J. H . (1976). Allosaurus fr agilis : a revised osteology. Uta h Geological

and M ineral Survey, Bu lletin, 109 . Majer , S . (1923 ). Spuren von dinosauriem de r Oberkreide im liegenden des

Kosder Eocanen Kohl enflot zes. Foldta ni Kozlony , 51-2, 66-75, 113-114.

May, F . E . & T raverse, A. (1973 ). Palynology of the Dakota Sandstone

256 M.]. Coe el al.

(midd le Cretaceous) near Bryce Ca nyo n Na tional Park, southern U tah. Geoscience and M an, 7, 57-64.

M ayr , E . (1963). Animal S pecies and thei r Evolution. Ca mb ridge, M assachusetts : Har vard U nivers ity Pr ess.

M olnar, R . (1980). Au stralian late M esozoic terres trial tetrap od s : some

implications . M emoires Societe Geologique de France, 139, 131-43. Nichols, D. J., jarzen , D . M ., Orth, C. J. & O liver, P . Q . (1986 ).

Palynological and irid ium anomalies at the Cretaceous-Tertia ry bounda ry,

south-cen tral Saskatchewan . S cience, 231, 714-17. No rm an, D . B. & Weishampcl, D . B. (1985 ). Ornithopod feeding

mechani sms: their bearing on the evolut ion of herbi vor y. Americarl

Natura list , 126, 151-64.

N orto n, N . J. & Hall , J. W . (1969) . Palynology of the Upper Cre taceo us and Lower Tertiary in the type localit y of the H ell Creek Formation , M ontana, U .S. A . Palaeontograph ica B , 125, 1--64.

Nowinski, A. (1971). Nemegtosaurus mongoliensis n . gen. , n . sp. (Sauropoda)

from the uppermost Cretaceous of M ongolia . Pala eontologie Polonica , 25 , 57-81.

Opler, A . (1978). Interaction of plant life hi story com ponents as related to

arboreal herbivory. In Th e Ecology of Arboreal Folivores, ed . G . G . Montgomery, pp. 23-3 1. Wa shington, D C : Smithsonian In stitution

Press . Ostrom, J . H. (1980). The evide nce for endo the rmy in d inosaur s. American

A ssociation f or the A dv ancement of S cience, S elected Symposia, 28, 15-54. Parker, L. R . (1980). Paleobot an y (cd. L. J . Hi ckey). Geotimes, Februar y

1980 , p . 40 .

Parra, R. (1978) . Compariso n of foregut and hindgut fermentation in herbivores. In The Ecology of Arboreal Potioores, ed . G . G. M ontgomer y,

pp . 205-9. Washington , D C: Smithsonian In stitu tion Pr ess. Pianka, E . R . (1978). Evolutionary Ecology . N ew York : Harper & Row. Pocock , S. A. J . (1962). M icroftoral ana lysis an d age de termina tio n of strata at

the Jurassic- Creta ceou s boundary in the western Canada plains. Palaeontograph ica B , Ll I , 1- 95.

Potter , F . W. (1976). In vest igat ion s of angios perms from the Eoce ne of

southeas tern No rt h Am er ica : pollen assemb lages from Mill er Pit , Henry County, T ennessee. Paiaeontogra phica B, 157,44-96.

Raath, M. (1974). Fo ssil vertebrate stud ies in Rhode sia : further evidence of gastro liths in prosau ropod dinosaurs. A moldia, 7,1-7.

Regal, P . J . (1977). Ecology and evo lut ion of flowering plant dominance. Science, 196, 622-9.

Reid , R. E . H . (1981). Lam ellar-zonal bone with zones and annuli in the pelvis of a sauropod dinosaur. Na ture, 292, 49-51.

Retallack, G . J . & Dil cher, D . L. (198 1). A coastal hypothesis for the

disper sal and rise to dominan ce of th e flower ing plants. In Paleobotany , Paleoecology , and Evolution, vol. II , ed . K . J. Nikl as, pp. 27-77. New

York : Praeger Publishers. Ricql es, A. de (1983). Cyclical growth in th e long limb bones of a sau ropod

dinosaur. A cta Palaeontologica Polonica, 28 , 225-32. Roger s, H . H ., Thomas, J. F , & Bingham , G . E. (198 3). Response of

agronomi c and forest spec ies to elevated atmos pheric car bon dioxide. Science, 220, 428-9.

Roth , J. L. & Di lcher, D . L. (1978 ). Some conside ra tions in leaf size and leaf margin ana lysis of fossil leaves. Courier Forschungs-Lnstitut , Se kenberg, 30, 165-71.

Russell, D . A. (1983 ). Exponential evolution : implications for int elligent extraterrestria l life. A dvances in Space Research, 3, 95-103.

Russell , D . A. ( 1984a), The gradual decline of th e din osaurs - fact or fallacy? Nature, 307 , 360-1.

Russell , D , A. (l 984b). A check list of th e famili es and gen era of North American din osaurs. Syllogeus, 53.

Russell , D . A., Belan d , P. & McIntosh , J. S . (1980). Paleoecology of the dinos aurs of Tendaguru (T anzan ia) . M bnoires de la S ociete Geologique de France, 139, 169-75 .

Sacher, G . A. (1975). Maturati on and longevit y in relati on to cran ial capacity in hominid evolution. In Primate Functional Morphology and Evolution, ed . R. H . Tutt le, pp. 417-41. The Hague : M outon Publ ishers .

Seilach er , A. (1984). Co ns truc tiona l morphology of bivalves : evolutionary pathways in primary ver sus secondar y soft- botto m dwellers. Palaeontology, 27,207- 37 .

Spotil a, J. R . (1980 ). Co ns traints of bod y size and env ironm ent on th e temperature regulation of dinosaurs. American A ssociatiml for the Advancement of Science S elected Symposia, 28, 23}-52.

Stocker, G . C . & Ir vine , A. K . (1983 ). Seed di spersal by Cassowaries (Casuarius casuarius) in North Queen sland's rainforests. B iotropica, 15, 170-6.

Stokes, W. L. (1964). Fossilized stomach cont ents of a sauropod dinosaur . Science, 143, 576-7 ,

Swain, T . (1976). Angiosperm-reptile co-evolution . Linnean Society Symposia, 3, 107-22 .

Sweet , A . R. & Hills, L. V. (1984). A palyn ological and sed imen tolog ical

anal ysis of th e Cretaceous-Tertiary boundary, Red De er Valley , Alberta, Canada . Abstracts, S ixth International Palynological Conference, Calgary 1984, p . 160.

Thoma s, R. D . K . & Olson , E . C. (eds.) (1980) . A cold look at warm-blooded

din osaurs. A merican A ssociation fo r the A dvancement of Science S elected Symposia, 28.

258 M.]. Coe et al.

Tiffney, 8. H. (1984 ). Seed size, dispe rsa l syn dromes, and the rise of th e angiospe rms : evidence and hypothesis. Annals of the M issouri Botanical Garden, 71, 55 1-76.

T royer, K . (1984). Structure and function of th e digestive tract of a herbivorous lizard Iguana iguana . Physiological Zoology, 57, 1-8.

Tschudy, R. H. (1970). Palynology of th e C retaceous-Tertia ry boundary in the northern Roc ky M ountain and M ississipp i Em bay ment regions. Geological So ciety of America Special Papers, 127, 65-111 ,

Tschu dy, R. H ., Pill mure, C . L. , Orth , C. J., Gilmore, J. S . & Knight , J. D . (1984) . Disru ption of th e ter res trial plant ecosystem at the Cretaceous-Tertiary boundary, wes te rn interior. Science, 225 , 1030-2 .

Van Valeri, L. (1973). A new evo lutiona ry law. Evolutionary Theory , 1, 1- 30. Viera, 1.. I. & Torres, J. A. (1979 ). El We aldico de la zona de Enciso (S ierra

de los Cameros) y su fauna de gra ndes reptiles. M unibe, Sociedad de

Cienciat A ranzadi, San Sebastian , 31. 141- 57. Ward, L. F. (1900 ). Status of M esozoic floras of the United States. U .S.

Geological Survey, Twentieth Annual Report, 2, 2 11-7 48.

Weaver , J. C. (1983 ). The improbable endo therm : th e ene rge tics of th e sau ropod dinosaur Brachiosaurus . Paleobiology , 9, 173-82.

Wei shampe l, D. B. (1984). Int eract ion s between M esozoic plants and

vertebrates: fruc tifica tions and seed p reda tion. Ne ues Jahrbuch fu r Geologie und Paliiontologie, A bhandlungen, 167, 224-50 .

Wes tern. D . (1975 ). Wate r avai labi lity and its influence on the st ructu re and dynamics of a savannah large ma mma l comm unity . East African W ildlif e

J ourna l, 13, 265-86. Whi ttaker, R. H . (1975 ). Communities and Ecosystems, 2nd edn . N ew York :

MacMillan . Wo lbac h, W . S .• L ewis, R . S . & Ander s, E. (198 5). Cretaceous extinct ions :

evidence for wildfire s and search for meteo ri t ic material. S cience, 230 , 167- 70 .

Wyles, J. S ., Kunkel , J. G. & Wil son , A. C. (1983). Birds, behavior and anatomical evolution. Proceedings of the U.S. Academy of Sciences, 80 ,

4394-7.

The origins of angiosperms and their

biological consequences --------- . --------­

EDITED BY

Else M ar ie F riis Section of Palaeobotany , Swedish M useum oj Na tural Hi story, Stockholm

Will iam G. Cha loner FRS Department of Biology , Roy al Hal/away and Redf ord N ew College, London Unioersity

Peter R. Crane Department of Geology , Field M useum of Natural H istory , Chicago

TIo.­ ..,M 41~.

t'~,,,,,,,_. "f Comb'Hi,. 'Ol"'~' 0...J •• 11

all ""'~~" at Iro...l.. ~o., "o~"J ~.'

li e"" f ill ,. IJ H TI>r t'.,,,,,,,, h", p".'d " • .1~I,,~<J (Q~ II"",,~,I)'

"." IH 4

CA M BR IDGE UN IVE RSIT Y PR E S S

Ca mbridge

New Y ork Port Chester M elbourne Sydney