Can lizard richness be driven by termite diversity? Insights from the Brazilian Cerrado G.C. Costa, G.R. Colli, and R. Constantino Abstract: We test predictions of the Morton and James hypothesis, which states that high termite diversity promotes high lizard diversity. We explore consumption of termites by lizards in the Brazilian Cerrado, a system that shares many simi- larites with arid Australia whose fauna formed the basis for the original hypothesis. We found that Cerrado lizards prey heavily on termites. Several species had >40% of their diet consisting of termites, some species reached up to 80%. How- ever, lizards prey on termites independently of their diversity in the environment and do not show niche segregation in re- lation to termite resource. Hence, our results in the Cerrado do not support the hypothesis that termite diversity can promote lizard diversity. The diets of Cerrado lizards have a high proportion of termites; however, the diets of desert liz- ards from the Australian and the Kalahari deserts have a much higher proportion of termites when compared with those from the Cerrado and the Amazon. Differences in termite consumption by lizards across ecosystems do not seem to be re- lated to local termite diversity. We hypothesize that overall prey availability can explain this pattern. Several arthropod groups are abundant in the Cerrado and the Amazon. In deserts, other prey types may be less abundant; therefore, termites may be the best available resource. Re ´sume ´: Nous testons les pre ´dictions de l’hypothe `se de Morton et de James selon laquelle une forte diversite ´ des termites favorise une diversite ´e ´leve ´e des le ´zards. Nous e ´tudions la consommation des termites par les le ´zards dans le Cerrado bre ´- silien, un syste `me qui posse `de plusieurs ressemblances avec la re ´gion aride d’Australie dont la faune a servi de base a ` l’e ´laboration de l’hypothe `se initiale. Les le ´zards du Cerrado consomment beaucoup de termites. Plusieurs espe `ces ont un re ´gime alimentaire qui contient >40 % de termites et certaines espe `ces atteignent 80 %. Cependant, la consommation des termites par les le ´zards est inde ´pendante de leur diversite ´ dans le milieu et les le ´zards ne font aucune se ´gre ´gation de niche en relation avec les ressources en termites. Nos re ´sultats dans le Cerrado n’appuient donc pas l’hypothe `se qui veut que la diversite ´ des termites puisse favoriser la diversite ´ des le ´zards. Les re ´gimes alimentaires des le ´zards du Cerrado contiennent une forte proportion de termites; cependant, les re ´gimes alimentaires des le ´zards des de ´serts d’Australie et du Kalahari en ont des proportions encore beaucoup plus e ´leve ´es par comparaison a ` la situation au Cerrado et en Amazonie. Les diffe ´ren- ces de consommation des termites par les le ´zards d’un e ´cosyste `me a ` un autre ne semblent pas relie ´es a ` la diversite ´ locale des termites. Nous posons l’hypothe `se selon laquelle c’est la disponibilite ´ globale des proies qui peut expliquer ce patron. Plusieurs groupes d’arthropodes sont abondants dans le Cerrado et l’Amazonie. Dans les de ´serts, les autres types de proies peuvent e ˆtre moins abondants et les termites peuvent ainsi constituer la meilleure ressource disponible. [Traduit par la Re ´daction] Introduction Worldwide, termites constitute one of the most common categories of prey taken by lizards (Vitt et al. 2003a). Ter- mites are social insects, which feed mainly on cellulosic ma- terial. They are extremely abundant in tropical forests and savannas and are predictable food source for many verte- brates (Wood and Sands 1978; Redford and Dorea 1984). In desert assemblages from Africa, Australia, and North Amer- ica, termites are widely exploited by many lizard species. In some species in the Kalahari desert and Australian arid zones, termites can constitute >90% of the diet volume (Pianka 1986). In these lizard assemblages, termites are probably the most important food resource. Termites appear most important among autarchoglossan lizards, a clade con- taining the majority of squamate species and in which most species rely on chemical discrimination of prey (Vitt et al. 2003a; Vitt and Pianka 2005). One of the world’s most diverse lizard faunas occurs in Australia, where local richness can reach 50 species (Pianka 1969, 1986). To explain this high diversity, Morton and James (1988) formulated a complex hypothesis where infer- tile soils and low and unpredictable rainfall lead to domi- nance by sclerophyllous vegetation with nutritionally poor aboveground biomass and litter, which favors detritivores such as termites. The high termite abundance and richness provide food for a variety of lizards, resulting in lizard radi- ation (Morton and James 1988). Later studies argued against the Morton and James (1988) hypothesis and proposed other explanations for high lizard diversity in arid Australia (Pianka 1989; James and Shine 2000; Colli et al. 2006). Received 13 April 2007. Accepted 14 September 2007. Published on the NRC Research Press Web site at cjz.nrc.ca on 25 January 2008. G.C. Costa. 1,2 Po ´s–graduac ¸a ˜o em Ecologia, Universidade de Brası ´lia, Co ´digo de Enderec ¸amento Postal 70910-900, Brası ´lia, DF Brasil. G.R. Colli and R. Constantino. Departamento de Zoologia, Universidade de Brası ´lia, Co ´digo de Enderec ¸amento Postal 70910-900, Brası ´lia, DF Brasil. 1 Corresponding author (e-mail: costagc@ou.edu). 2 Present address: Sam Noble Oklahoma Museum of Natural History, 2401 Chatauqua Avenue, Norman, OK 73072, USA. 1 Can. J. Zool. 86: 1–9 (2008) doi:10.1139/Z07-107 # 2008 NRC Canada