Herpetological Conservation and Biology 15(1):69

Herpetological Conservation and Biology 15(1):69–78.

Submitted: 30 May 2019; Accepted: 30 January 2020; Published: 30 April 2020.

Dietary ObservatiOns Of fOur sOuthern african

agamiD LizarDs (agamiDae)

Wei Cheng Tan

1,2,3

, anThony herrel

, and John Measey

3,4

Unité Mixte de Recherche 7179 Centre National de la Recherche Scientique/ Muséum National d’Histoire Naturelle,

Département Adaptations du Vivant, 55 rue Buffon, 75005, Paris Cedex 5, France

Université de Poitiers - Laboratoire EBI Ecologie and Biologie des Interactions, Unités Mixtes de Recherche Le Centre

National de la Recherche Scientique 7267, Equipe Ecologie Evolution Symbiose, 6, rue Michel Brunet,

TSA 51106, 86073 Poitiers, France

Centre for Invasion Biology, Department of Botany and Zoology, Stellenbosch University, Stellenbosch, South Africa

Corresponding author, email: [email protected]

Abstract.—Analysis of stomach contents can provide insights into foraging mode, habitat use, and dietary

specialization of animals. In this paper, we make observations on the poorly known diet of four southern African

agamid species, Agama aculeata distanti (Eastern Ground Agama), Agama armata (Peter’s Ground Agama), Agama

atra (Southern Rock Agama), and Acanthocercus atricollis (Southern Tree Agama). We examined the diet of 67

individuals by identifying and weighing prey items after stomach ushing lizards in the eld. We found that these

agama species fed on a broad spectrum of arthropods (11 orders). A high relative importance of ants was present for

all agama species examined here, which suggests that ants are a major food source in the arid ecosystem. We found

that active prey such as ants, beetles, and highly mobile ying insects like wasps and ies to be major components

of the diet, indicating that these lizards are ambush predators. We also found that 43% of the stomachs contained

herbaceous material and 39% contained sand particles. Agama atra had the most diverse dietary niche, eating

fewer ants and more beetles, hemipterans, and dipterans than other species, whereas A. armata had a narrower

dietary niche consisting mainly of ants. Lastly, although low in sample size, we found that juveniles qualitatively

had a diet of functionally similar prey items, albeit with a narrower niche breadth, when compared to adults. We

discuss how diet corresponds with differences in foraging behavior and habitat specialization.

Key Words.—diet; index of relative importance; lizards; reptiles; southern Africa

intrODuctiOn

Diet plays an important role in the daily life of animals

as it provides a source of energy for growth, maintenance,

and reproduction (Huey and Pianka 1981; Zug et al.

2001). Many animals specialize in different prey items

and develop complex feeding behaviors based on their

anatomy or dietary requirements (Schwenk 2000).

Studying the prey eaten by an animal can provide insights

into their ecological roles and the relative importance of

each prey species in their diet (Losos and Greene 1988;

Ortega-Rubio et al. 1995; Znari and El Mouden 1997).

Studies of diet in an inter-specic context may further

provide information on niche overlap and thus on how

animals partition resources (Norval et al. 2012). The

optimal foraging model of MacArthur and Pianka (1966)

predicts that dietary specialization depends on the range

and abundance of available prey, as well as the energetic

gains and losses (handling and search time) associated.

For example, specializing on a narrow range of food

usually results in higher foraging efciency (Britt and

Bennet 2008), eliminating potential competition with

generalists. In a patchy environment with few food

resources, however, generalists tend to do better than

specialists. Generalists have a more diverse diet and a

reduced travel time between suitable patches (less search

time), which makes up for their lower foraging efciency

compared to specialists. This difference has been a

major interest to ecologists as the distribution pattern

of organisms is largely dependent on the degree of diet

specialization (MacArthur 1972).

Agamids are Old World lizards that have successfully

colonized a variety of habitats ranging from hot deserts

to tropical forests (Greer 1989). Some agamids even

appear to favor peri-urban (rural-urban transition zones)

landscapes more than natural or protected areas (Whiting

et al. 2009). Surprisingly, and despite their ubiquitous

nature, southern African agamids remain relatively

poorly studied in terms of their ecology (but see

Anibaldi 1998; Reaney and Whiting 2003; Van Berkel

and Clussela-Trullas 2018). Although most agamids

are generally terrestrial, some specialized saxicolous

and arboreal species exist. Agamids only occasionally

forage outside their home range (Whiting et al. 1999),

suggesting that the microhabitats used by different

species may constrain their diet.

Tan et al.—Feeding ecology of southern African Agamid lizards.

Agamas are widespread diurnal lizards that are

widely distributed in Africa. Eleven agamid species are

common and widespread throughout southern Africa

(Bates et al. 2014). Agama aculeata distanti (Eastern

Ground Agama; Fig. 1A) and A. armata (Peter’s

Ground Agama; Fig 1B) seem to prefer similar types of

macrohabitat: open canopy and semi-arid areas (Bates et

al. 2014). Although A. a. distanti is normally classied

as a ground dwelling species (Bates et al. 2014), some

A. a. distanti populations appear to be saxicolous,

occurring in rocky woodlands. They are normally

found basking on rocks, branches of bushes or

termitaria (Branch 1998). Similarly, A. atra (Southern

Rock Agama; Fig. 1C) is also a saxicolous species

found on rocky outcrops and mountain plateaus.

Agama atra has a wide distribution throughout

southern Africa compared to the other agamids studied

here, perhaps suggesting a more generalist lifestyle

(Bates et al. 2014). Agama armata, however, tend

to be found in open deep sand savannah and calcrete

ats (Branch 1998). It has been suggested that widely

foraging lizards living in open habitats specialize on

feeding relatively sedentary and clumped prey such

as termites (Huey and Pianka 1981). Contrary to the

other species, Acanthocercus atricollis (Southern Tree

Agama; Fig. 1D) has an arboreal lifestyle, spending

much of its time on trees and logs. They consume

primarily mobile, diurnal insects such as ants, beetles,

and orthopterans, but also ingest occasional millipedes

and centipedes (Reaney and Whiting 2002).

The diet of only a few African agamid species has

been studied (Bruton 1977; Znari and El Mouden 1997;

Heideman 2002; Reaney and Whiting 2002; Ibrahim

and El-Naggar 2013). These lizards have been reported

to be ambush foragers (e.g. Whiting et al. 1999), feeding

almost entirely on insects, including predominance

of Hymenoptera, especially ants (Formicidae) and

Coleoptera. To date, little or no information is available

on the dietary niches of southern African agamids (Fig.

1; but see Bruton 1977; Reaney and Whiting 2002). In

this study, we examined the diet of four agamid species

from South Africa discussed above: A. a. distanti, A.

armata, A. atra and A. atricollis. We determined food

habits by examining their stomach contents. We sampled

across species in summer to gain insights into possible

differences in foraging strategies, niche overlap, and the

presence of possible dietary specialists. We captured

adult and juvenile lizards whenever possible to explore

the presence of ontogenetic shifts in diet.

materiaLs anD methODs

Study areas.—We conducted eldwork in three of

the nine biomes of South Africa: Fynbos, Savanna, and

Thicket (Fig. 2). We sampled A. atra primarily on Mui-

zenberg Mountain (34º05´S, 18º26´E) in March 2008.

Muizenberg is a 500 m tall mountain dominated by rich

endemic fynbos vegetation and is situated on the Cape

Peninsula in the Western Cape. We sampled the other

agamid species throughout the austral summer of 2017

(February-March). For every species, we captured in-

dividuals over the course of several weeks. We sam-

pled A. atricollis in Mtunzini (28º58´S, 31º45´E) and

Eshowe (28º52´S, 31º28´E) peri-urban areas, in Kwa-

figure 1. Four agamid species from southern Africa. (A) Agama aculeata distanti (Eastern Ground Agama), (B) A. atra (Southern

Rock Agama), (C) A. armata (Peter’s Ground Agama), and (D) Acanthocercus atricollis (Southern Tree Agama). (Photographed by

Wei Cheng Tan).

Herpetological Conservation and Biology

Zulu Natal, where these lizards live in close proximity

to humans. Acanthocercus atricollis has an arboreal

lifestyle and lives on big trees with signicant canopy

cover (Reaney and Whiting 2003). We collected A. ar-

mata in game reserves near Tshipise (22º31´S, 30º39´E)

in Limpopo. Tshipise contained at, open, sandy aca-

cia thornveld and occasionally dry mopane dominated

woodland. We caught Agama aculeata distanti at the

Welgevonden nature reserve, Waterberg District, also in

Limpopo (24º13´S, 27º54´E). The reserve is character-

ized by rocky woodlands and mountain bushveld at high

altitudes with minimal human activity. Note that this

population of A. a. distanti was found predominantly on

rocks.

We marked all lizards caught with a temporary non-

toxic marker to avoid recapturing and remeasuring

the same individual and returned them within 24 h to

their exact site of capture after we measured lizards

and ushed their stomachs. We considered individu-

als adults if they had a snout-vent length (SVL) above

100mm, while we considered lizards with a SVL be-

low 70 mm juveniles (Appendix Table 1). We assessed

lizards between these sizes individually according to

secondary sexual characteristics (e.g., coloration, scale

ornamentation, etc.) to determine if they were adults or

juveniles.

Stomach contents.—We collected stomach contents

of 67 individuals. We ushed stomachs within 4 h of

capture of a lizard according to the protocol of Herrel

et al. (2006). We held the lizard gently with one hand

while opening the mouth by tapping on the sides of the

jaw, which resulted in a jaw-opening threat response.

We then inserted a small plastic ring into the mouth to

keep the jaws open, allowing for a continuous ow of

water and food out of the digestive tract. We used a sy-

ringe with a round-tipped steel needle, which we insert-

ed gently into the stomach through the pharynx. Upon

feeling a slight resistance against the pyloric end of the

stomach, we pushed sufcient water into the stomach

to force food out without injuring the animal. We con-

tinued the sequence while slightly moving the syringe

up and down until a food bolus with fragmented matter

was regurgitated. We kept the diet samples in individual

vials with 70% ethanol and brought vials back to the

laboratory for examination.

We identied the stomach contents to the lowest pos-

sible taxonomic level using a binocular microscope. In

most cases, however, only the Order of the prey item

could be identied due to the fragmented nature of the

prey items. We decided to separate Formicidae prey

from the other Hymenopterans as they are reported to

be an important part of the diet of some agamids (Capel-

Williams and Pratten 1978; Znari and Mouden 1997;

Heideman 2002). We blot-dried the food items thor-

oughly with paper towels before measuring the mass

using an electronic microbalance (AE100-S, Mettler

Toledo GmBH, Zurich, Switzerland; ± 0.1 mg). Be-

figure 2. The ve collection sites used in this study fell into three of nine biomes of South Africa (Coastal Belt, Fynbos, and Savanna)

and three provinces (Kwa-Zulu Natal, Limpopo, and Western Cape).

Tan et al.—Feeding ecology of southern African Agamid lizards.

cause most prey items were crushed into fragments, we

decided to include all fragments, which we sorted into

different prey groups. We feel this approach provides a

reasonable estimate of prey volume as prey bodies were

rarely found intact.

The heads of the prey items were carefully enumerat-

ed to calculate the numeric abundance (N) of each prey

category. For each prey group, we also calculated the

index of relative importance (IRI) to quantify the sig-

nicance of a particular prey item in the diet (Pinkas et

al. 1971):

IRI = (%N + %V) × %Oc

where %N is the percentage of numeric abundance,

%Oc is frequency of occurrence of a certain prey group,

and %V is the proportion of mass of that prey group to

total prey mass. IRI is a compound index that provides

a balanced view of the diet of a lizard due to the

combination of unique properties affecting individual

measures (numbers, mass, and occurrence in the diet).

We also calculated a diet diversity index for each

species. Dietary breadth (B) was computed according

to Levins (1968):

B = 1 / ∑Pi ²

where P is the proportion of records of each species

in prey category, i. The index of Levins is a simple

computation that provides an indication of which species

has a more specialized diet if it had a relatively low

dietary niche breadth (B). In the course of the analysis,

we also discovered nematodes in the stomachs of some

individuals. We reported these separately as parasites.

In addition, we also did not include plant matter, sand

particles, or nematodes in the calculation of IRI or niche

breadth. We classied all prey groups according to their

evasiveness: evasive, sedentary, and intermediate based

on the denitions in Vanhooydonck et al. (2007), and

we used these categories to compare the prey consumed

between adults and juveniles.

Statistical analysis.—To meet normality and

homoscedasticity assumptions, we log10-transformed all

continuous data before further analyses. We conducted

multivariate analysis of variance (MANOVA) on the

IRI of the different prey groups to compare diet between

species. Following this we performed univariate

ANOVAs with LSD post-hoc tests on prey IRI to

determine their relative contribution among species. We

performed all analyses in IBM-SPSS v24 (SPSS Inc.,

Chicago, Illinois, USA). For all tests, α = 0.05.

resuLts

The IRI between species differed signicantly (Wilks'

lambda = 0.288; F

, 148.46 = 2.158, P < 0.001). There

were signicant differences within each prey group

(Appendix Table 2). Ants (Formicidae) were present

in the stomachs of all agamid species and represented

the majority of their diet (> 90% occurrence; Table 1).

Ants also appear to be the dominant prey by mass and

tabLe 1. The composition of diet of four agamid species from southern Africa (total n = 67): Agama aculeata distanti (Eastern Ground

Agama), A. armata (Peter’s Ground Agama), A. atra (Southern Rock Agama), and Acanthocercus atricollis (Southern Tree Agama).

Prey items are presented by Order and the corresponding evasive level (Fun.) according to Vanhooydonck et al (2011): e = evasive, i =

intermediate, s = sedentary, and na = not applicable. Ants are removed from other Hymenoptera as they are reported to be a principle

dietary item of agamids and have a different evasive level. Lepidoptera are larvae only. Diet is reported as occurrence (Oc), or the total

number of individual lizards that contain a particular food item, and its frequency (FOO) in percentage.

A. a. distanti (n = 23) A. armata (n = 8) A. atra (n = 15) A. atricolis (n = 21)

Order Fun. Oc FOO (%) Oc FOO (%) Oc FOO (%) Oc FOO (%)

Formicidae i 23 100 8 100 14 93.33 20 95.24

Hymenoptera e 12 52.17 3 37.50 13 86.67 8 38.10

Coleoptera i 16 69.57 4 0.50 15 1 10 47.62

Hemiptera i 10 43.48 1 0.13 10 66.67 1 4.76

Diptera e 5 21.74 1 0.13 10 66.67 2 9.52

Lepidoptera s - - - - 3 20.00 1 4.76

Orthoptera e - - - - 2 13.33 1 4.76

Gastropoda s - - - - - - 1 4.76

Ephemoptera e - - - - - - 1 4.76

Isoptera s 1 4.35 - - - - - -

Diplopoda s - - - - 7 46.67 - -

Plant matter na 8 34.78 3 3.75 13 86.67 6 28.57

Sand particles na 16 69.57 3 3.75 7 46.67 - -

Herpetological Conservation and Biology

A. a. distanti (n = 23) A. armata (n = 8) A. atra (n = 15) A. atricollis (n = 21)

Formicidae No. 495 184 1160 406

Mass 1.444 0.105 0.808 0.433

IRI* 1.348 ± 0.457 1.503 ± 0.498 0.953 ± 0.560 1.428 ± 0.541

Other

Hymenoptera

No. 28 3 43 20

Mass 0.033 0.002 0.160 0.100

IRI 0.041 ± 0.058 0.011 ± 0.016 0.218 ± 0.300 0.036 ± 0.072

Coleoptera No. 45 9 140 12

Mass 0.191 0.052 0.533 0.019

IRI 0.161 ± 0.208 0.128 ± 0.197 0.362 ± 0.407 0.087 ± 0.230

Hemiptera No. 13 1 23 1

Mass 0.020 < 0.001 0.039 0.002

IRI* 0.019 ± 0.033 0.005 ± 0.014 0.036 ± 0.048 < 0.001

Diptera No. 9 2 16 2

Mass 0.004 < 0.001 0.005 < 0.001

IRI* 0.006 ± 0.017 0.002 ± 0.006 0.014 ± 0.024 < 0.001

Lepidoptera

(larvae)

No. - - 3 2

Mass - - 0.017 0.006

IRI - - 0.005 ± 0.012 < 0.001

Orthoptera No. - - 3 1

Mass - - 0.591 0.150

IRI - - 0.015 ± 0.054 0.002 ± 0.010

Gastropoda No. - - - 1

Mass - - - 0.002

IRI - - - < 0.001

Ephemoptera No. - - - 1

Mass - - - < 0.001

IRI - - - < 0.001

Isoptera No. 116 - - -

Mass 0.059 - - -

IRI 0.003 ± 0.015 - - -

Isopoda No. 1 - - -

Mass 0.001 - - -

IRI < 0.001 - - -

Diplopoda No. - - 9 -

Mass - - 0.089 -

IRI* - - 0.032 ± 0.075 -

Plant matter Mass 0.017 0.001 0.023 0.141

Sand particles Mass 0.152 0.005 0.058 -

Unidentied Mass 0.039 - 0.066 0.113

B (breadth)

1.909 1.166 1.427 1.203

tabLe 2. Diet comparison between Agama aculeata distanti (Eastern Ground Agama), A. armata (Peter’s Ground Agama), A. atra

(Southern Rock Agama), and Acanthocercus atricollis (Southern Tree Agama) from southern Africa, showing the number (frequency) and

total dry mass (g) of each prey taxon. The Index of Relative Importance, IRI, is displayed as mean ± standard deviation. Niche breadth,

B, is indicated at the bottom of the table. Asterisks (*) indicate prey taxon IRI that were signicantly different between agamid species

(P < 0.05): see Appendix 2.

Tan et al.—Feeding ecology of southern African Agamid lizards.

in IRI, followed by coleopterans, other hymenopterans,

hemipterans, and dipterans (Table 2). Agama atra

had a signicantly lower ant IRI (0.95) than the other

species, while A. armata had the highest ant IRI (1.5),

closely followed by A. atricollis (IRI: 1.43) and A. a.

distanti (IRI: 1.35), although these differences were

not signicant (Table 2). Moreover, Coleoptera seem

to be an important part of the diet of these species,

representing the second highest IRI among the other

prey items. Agama atra, which has a low ant IRI,

showed signicantly higher importance of evasive prey

in its diet (e.g., Hymenoptera and Diptera) compared

to other agamid species (Table 2). We observed large

evasive prey, such as Orthoptera, in the stomachs of two

A. atra individuals (Table 1). Additionally, A. atra was

the only species that was found preying on millipedes

(Diplopoda). Although rare, we also found mayies

(Ephemeroptera), snails (Gastropoda), woodlice

(Isopoda), and termites (Isoptera) in the stomachs of

these lizards.

In addition to arthropods, all species also ingested

plant matter (such as seeds, buds, and small twigs) and

ne particles of sand, observed in 43% and 39% of the

stomachs, respectively. Of the 67 stomachs examined,

36 (57%) also contained nematodes. They were present

in all species: 12 A. a. distanti, three A. armata, 13

A. atra, and 21 A. atricollis. The dietary breadth for

prey number was greatest in A. a. distanti (1.91) and

lowest in A. armata (1.17; Table 2). Juveniles showed

a lower dietary diversity than adults in all four species,

suggesting large quantities of prey items in few

categories (Table 3); however, functionally their prey

types appear to be very similar to that of adults (Fig. 3).

Formicidae appear to be the most important prey type

in the diet of juveniles, and there was a higher ant IRI

in juveniles (mean = 1.47) than in adults (mean = 1.14;

Table 3).

DiscussiOn

Our results show that the agamids in our study are

insectivorous like most other agamids (see Huey

and Pianka 1981). Ants (Formicidae) were the most

common and important prey items in the diet of these

animals. Other major prey components included other

hymenopterans such as bees and wasps, coleopterans,

A. aculeata distanti A. armata A. atra A. atricollis

A (n = 9) J (n = 14) A (n = 2) J (n = 6) A (n = 10) J (n = 5) A (n = 14) J (n = 7)

Formicidae 1.262 ±

0.563

1.402 ±

0.390

1.109 ±

0.410

1.634 ±

0.480

0.705 ±

0.521

1.451 ±

0.150

1.393 ±

0.624

1.484 ±

0.403

Other

Hymenoptera

0.039 ±

0.057

0.043 ±

0.061

0.035 ±

0.009

0.003 ±

0.008

0.30 ± 0.34 0.059 ±

0.05

0.037 ±

0.087

0.036 ±

0.046

Coleoptera 0.171 ±

0.170

0.155 ±

0.236

0.409 ±

0.179

0.034 ±

0.075

0.484 ±

0.451

0.119 ±

0.110

0.118 ±

0.293

0.038 ±

0.026

Hemiptera 0.012 ±

0.014

0.023 ±

0.041

- 0.007 ±

0.017

0.038 ±

0.057

0.030 ±

0.026

- < 0.001

Diptera 0.010 ±

0.025

0.003 ±

0.009

- 0.003 ±

0.006

0.016 ±

0.030

0.010 ±

0.007

- < 0.001

Lepidoptera

(larvae)

- - - - 0.007 ±

0.015

<0.001 - < 0.001

Orthoptera - - - - 0.022 ±

0.066

- - 0.006 ±

0.016

Gastropoda - - - - - - - 0.001 ±

0.003

Ephemoptera - - - - - - - < 0.001

Isoptera 0.008 ±

0.024

- - - - - - -

Isopoda - < 0.001 - - - - - -

Diplopoda - - - - 0.030 ±

0.090

0.037 ±

0.039

- -

B (breadth)

2.281 1.352 2.072 1.072 1.691 1.204 1.282 1.173

tabLe 3. Dietary composition and breadths of adult (A) and juvenile (J) Agama aculeata distanti (Eastern Ground Agama), A. armata

(Peter’s Ground Agama), A. atra (Southern Rock Agama), and Acanthocercus atricollis (Southern Tree Agama) from southern Africa

based on Index of Relative Importance, IRI, and Levin’s niche index (B). The IRI is displayed as mean ± standard deviation. Note, small

sample sizes prevented statistical comparisons.

Herpetological Conservation and Biology

and dipterans. These observations agree with previous

studies on the ecology of African agamids (Capel-

Williams and Pratten 1978; Znari and Mouden 1997;

Heideman 2002). Ants are small and hard prey of

intermediate evasiveness (see Vanhooydonck et al.

2007). Hence, they likely provide little energetic value

relative to the handling time needed to process them,

as agamids crush their prey before swallowing (Herrel

et al. 1996; Meyers and Herrel 2005; Schaerlaeken et

al. 2008). Yet, ants are a common component of arid

ecosystems and present an important feature of the diet

of many lizards simply due to their abundance and ease

of capture (Pianka 1986; Branch 1998).

Using quantitative measurements, Agama atra,

A. planiceps and A. atricollis have been previously

classied as ambush foragers (Whiting et al. 1999;

Reaney and Whiting 2002). Not surprisingly, there

was a high relative importance of ants and beetles

(Coleopterans) in the diet of agamids in our study. This

suggests that these lizards are indeed ambush predators

(Pianka 1986) although our study further showed that

A. atra also fed on prey items classied as evasive

prey that typically jump or y to escape. These prey

types are thought to be captured by ambush predators

possibly with very short, quick dashes from a standstill

(Vanhooydonck et al. 2007), suggesting that this species

may be a more specialized ambush predator.

Diet composition can vary greatly according to the

prey availability in specic localities or microhabitats

(Measey et al. 2011). For example, diet of agamids can

change across an urban to rural gradient (Balakrishna

et al. 2016) and this may account for some of the

differences we found between A. atricollis and the other

species. Our results showed a greater proportion of ants

and few ying insects in the diet of A. armata, which is

more of a ground dwelling species.

Our results indicate that the diet differs signicantly

among the four agamid species, but with a very similar

functional grouping of prey types dominated by prey of

intermediate activity with some evasive invertebrates.

Agama aculeata distanti appeared to be the most

generalist species followed by A. atra, based on the niche

breadth index of Levin. One problem, however, with

this index is that it does not take into account the mass

of the prey items, which could play an important role in

the energy gained from ingested prey items. When we

added mass into our diet analyses (IRI calculation), A.

atra stood out as having a more diverse and generalist

diet, with ants being relatively less important compared

to the other species studied here. Agama armata, the

smallest southern African agamid (Branch 1998), preys

primarily on ants and appears more specialized than

the other species, although this may simply reect the

paucity of other prey in their environment. Agama atra

and A. atricollis both consumed larger prey items like

grasshoppers and caterpillars. Perhaps only the small to

medium-sized prey items (e.g., ants and small beetles)

can be consumed by smaller predators simply due to the

limitations on gape or bite force (Capel-Williams and

Pratten 1978). Another explanation may be that the

A. atra population from Western Cape encounters the

greatest diversity of potential prey items compared to

the subtropics where A. armata lives. To better interpret

these differences, future studies should investigate the

availability and abundance of prey, in addition to diet, to

understand whether agamids actively select certain prey

types, avoid others, or are opportunistic predators.

Although not included in the dietary analyses, we

observed that nearly half of the stomachs examined

contained plant matter. Many insectivorous lizards

are known to consume herbaceous material. Capel-

Williams and Pratten (1978) found that plant material is

an important food or water source during the dry season

due to scarcity of animal food or water (Joger 1979).

This feeding strategy has also been observed in Agama

impalearis (Bribon’s Agama; Znari and Mouden 1997)

and Stellagama stellio (Starred Agama; Ibrahim and El-

Naggar 2013). The high frequency of vegetation found

in this study indicates that it could be an important

food source for these agamids as well. We also found

sand particles in the stomachs. These agamids (except

A. atricollis) live in sandveld or rocky habitats, which

likely explains the indirect ingestion of mineral particles

with prey captured on sandy substrates (Capel-Williams

and Pratten 1978; Ibrahim and El-Naggar 2013).

The difference in diet, but not prey functional types,

between juveniles and adults could correspond to the

different microhabitats occupied by juveniles and

adults. For instance, ants are a major component of the

diet of juveniles whereas orthopterans are an important

component of the diet in adult A. impalearis (Capel-

Williams and Pratten 1978). Juveniles may be more

figure 3. Diet as a composition of functional prey types of

Agama aculeata distanti (Eastern Ground Agama), A. armata

(Peter’s Ground Agama), A. atra (Southern Rock Agama) and

Acanthocercus atricollis (Southern Tree Agama). Functional

prey types are sedentary (black), intermediate (yellow) and active

(orange) prey (see Table 1 for classication) for both adults and

juveniles (juv.).

Tan et al.—Feeding ecology of southern African Agamid lizards.

associated with open ground between shrubs and under

rocks where ants are found, and we have observed that

adults and juveniles are rarely in close proximity. This

apparent specialization in juveniles, however, may also

result from morphological and mechanical constraints

that would prevent young (and thus smaller sized) lizards

to feed only on larger prey items (Capel-Williams and

Pratten 1978). Our low sample sizes prevented us from

performing statistical comparisons to test for age effects

and prevented us making any rm conclusions.

Acknowledgments.—This study was carried out under

the permit for scientic collection from CapeNature

(056-AAA041-00168-0056), SANParks, and Ezemvelo

KZN Wildlife (OP 550/2017). This Project was

nancially supported by the European Commission

through the program Erasmus Mundus Masters Course

- International Master in Applied Ecology (EMMC-

IMAE; FPA 532524-1-FR-2012-ERA MUNDUS-

EMMC). WCT and JM would like to thank the DSI-NRF

Centre of Excellence for Invasion Biology and National

Research Foundation of South Africa incentive fund

for rated researchers. This project was also supported

by the National Research Foundation (NRF) of South

Africa (Key International Science Capacity Fund

Program) to AH. For their invaluable assistance during

eldwork in South Africa, we are indebted to members

of the MeaseyLab, Gareth Coleman, Amy Panikowski,

John Wilkinson, François Meyer, Jason Savage, Krystal

Tolley, and Frederik Igelström for access to various

sites, local knowledge of species, data collection, and

generous hospitality. Additionally, we are extremely

grateful for all the help by the staff and volunteers in the

Welgevonden Game Reserve.

Literature citeD

Anibaldi, C., L. Luiselli, and F. Angelici. 1998.

Notes on the ecology of a suburban population of

rainbow lizards in coastal Kenya. African Journal of

Ecology 36:199–206.

Balakrishna, S., A. Batabyal, and M. Thaker. 2016.

Dining in the city: dietary shifts in Indian rock

agamas across an urban-rural landscape. Journal of

Herpetology 50:423–428.

Bates, M.F., W.R. Branch, A.M. Bauer, M. Burger, J.

Marais, G.J. Alesander, and M.S. De Villiers. (Eds.).

2014. Atlas and Red List of the Reptiles of South

Africa, Lesotho and Swaziland. Suricata 1. South

African National Biodiversity Institute, Pretoria,

South Africa.

Branch, W.R. 1998. Field Gide to Snakes and Other

Reptiles of Southern Africa. Ralph Curtis Books,

Sanibel Island, Florida, USA.

Britt, E., and A.F. Bennet. 2008. The energetic

advantages of slug specialization in garter snakes

(Genus Thamnophis). Physiological and Biochemical

Zoology 81:247–254.

Bruton, M.N. 1977. Feeding, social behaviour and

temperature preferences in Agama atra Daudin

(Reptilia, Agamidae). African Zoology 12:183–199.

Capel-Williams, G., and D. Pratten. 1978. The diet of

adult and juvenile Agama bibroni (Reptilia: Lacertae)

and a study of the jaw mechanisms in the two age

groups. Journal of Zoology 185:309–318.

Greer, A.E. 1989. The Biology and Evolution of

Australian Lizards. Surrey Beatty & Sons, Chipping

Norton, New South Wales, Australia.

Heideman, N.J.L. 2002. A comparison of the breeding

and nonbreeding season diets of Agama aculeata and

Agama planiceps (Reptilia: Agamidae) in Windhoek,

Namibia. Journal of Herpetology 36:515–520.

Herrel, A., J. Cleuren, and F. De Vree. 1996. Kinematics

of feeding in the lizard Agama stellio. Journal of

Experimental Biology 199:1727–1742.

Herrel, A., R. Joachim, B. Vanhooydonck, and

D.J. Irschick. 2006. Ecological consequences

of ontogenetic changes in head shape and bite

performance in the Jamaican lizard Anolis lineatopus.

Biological Journal of the Linnean Society 89:443–

454.

Huey, R.B., and E.R. Pianka. 1981. Ecological

consequences of foraging mode. Ecology 62:991–

999.

Ibrahim, A.A., and M.H. El-Naggar. 2013. Diet and

reproductive biology of the Starred Agama, Laudakia

stellio (Linnaeus, 1758) (Squamata: Agamidae), in

the northern Sinai, Egypt. Zoology in the Middle

East 59:136–143.

Joger, U. 1979. Zur Ökologie und Verbreitung wenig

bekannter Agamen Westafrikas (Reptilia: Sauria:

Agamidae). Salamandra 15:31–52.

Levins, R. 1968. Evolution in Changing Environments:

Some Theoretical Explorations (No. 2). Princeton

University Press, Princeton, New Jersey, USA.

Losos, J.B., and H.W. Greene. 1988. Ecological

and evolutionary implications of diet in monitor

lizards. Biological Journal of the Linnean

Society 35:379–407.

MacArthur, R.H. 1984. Geographical Ecology: Patterns

in the Distribution of Species. Princeton University

Press, Princeton, New Jersey, USA.

Macarthur, R.H., and E.R. Pianka. 1966. On optimal

use of a patchy environment. American Naturalist

100:603–609.

Measey, G.J., A.D. Rebelo, A. Herrel, B. Vanhooydonck,

and K.A. Tolley 2011. Diet, morphology and

performance in two chameleon morphs: do harder

bites equate with harder prey? Journal of Zoology

285:247–255.

Herpetological Conservation and Biology

Meyers, J.J., and A. Herrel. 2005. Prey capture kinematics

of ant eating lizards. Journal of Experimental Biology

208:113–127.

Norval, G., S.C. Huang, J.J. Mao, S.R. Goldberg, and K.

Slater. 2012. Additional notes on the diet of Japalura

swinhonis (Agamidae) from southwestern Taiwan,

with comments about its dietary overlap with the

sympatric Anolis sagrei (Polychrotidae). Basic and

Applied Herpetology 26:87–97.

Ortega-Rubio, A., A. González-Romero, and R. Barbault.

1995. Food analysis and resource partitioning in a

lizard guild of the Sonoran Desert, Mexico. Journal

of Arid Environments 29:367–382.

Pianka, E.R. 1986. Ecology and Natural History of

Desert Lizards: Analyses of the Ecological Niche and

Community Structure. Princeton University Press,

Princeton, New Jersey, USA.

Pinkas, L., M.S. Olpihant, and I.L.K. Iverson. 1971.

Food habits of Albacore, Bluen Tuna and Bonito

in Californian waters. California Department of Fish

and Game: Fish Bulletin 152:1–105.

Reaney, L.T., and M.J. Whiting. 2002. Life on a limb:

ecology of the Tree Agama (Acanthocercus a.

atricollis) in southern Africa. Journal of Zoology

257:439–448.

Reaney, L.T., and M.J. Whiting. 2003. Picking a tree:

habitat use by the Tree Agama, Acanthocercus

atricollis atricollis, in South Africa. African Zoology

38:273–278.

Schaerlaeken, V., A. Herrel, P. Aerts, and C.F. Ross.

2008. The functional signicance of the lower

temporal bar in Sphenodon punctatus. Journal of

Experimental Biology 211:3908–3914.

Schwenk, K. 2000. Feeding in lepidosaurs. Pp. 175–

291 In Feeding: Form, Function and Evolution in

Tetrapod Vertebrates. Schwenk, K. (Eds.). Academic

Press, San Diego, California, USA.

Van Berkel, J.B., and S. Clusella-Trullas. 2018.

Behavioral thermoregulation is highly repeatable

and unaffected by digestive status in Agama atra.

Integrative Zoology 13:482–493.

Vanhooydonck, B., A. Herrel, and R. Van Damme.

2007. Interactions between habitat use, behavior, and

the trophic niche of lacertid lizards. Pp. 427–449 In

Lizard Ecology: The Evolutionary Consequences of

Foraging Mode. Reilly, S.M., L.D. McBrayer, and

D.B. Miles (Eds.). Cambridge University Press, New

York, New York, USA.

Whiting, M.J., P. le FN Mouton, J.H. Van Wyk, and

W.E. Cooper. 1999. Movement-and attack-based

indices of foraging mode and ambush foraging in

some gekkonid and agamine lizards from southern

Africa. Amphibia-Reptilia 20:391–399.

Whiting, M.J., K. Chetty, W. Twine, and P. Carazo. 2009.

Impact of human disturbance and beliefs on the Tree

Agama Acanthocercus atricollis atricollis in a South

African communal settlement. Oryx 43:586–590.

Znari, M., and E. El Mouden. 1997. Seasonal changes

in the diet of adult and juvenile Agama impalearis

(Lacertilia: Agamidae) in the central Jbilet Mountains,

Morocco. Journal of Arid Environments 37:403–412.

Zug, G.R., L.J. Vitt, and J.P. Caldwell. 2001.

Herpetology. An Introductory Biology of Amphibians

and Reptiles. 2

Edition. Academic Press, San Diego,

California, USA.

appenDix tabLe 1. Mean and range values of snout vent length (SVL in mm) and mass (in g) for adults and juveniles of

Agama aculeata distanti (Eastern Ground Agama), A. armata (Peter’s Ground Agama), A. atra (Southern Rock Agama)

and Acanthocercus atricollis (Southern Tree Agama).

A. a. distanti A. armata A. atra A. atricollis

adult juvenile adult juvenile adult juvenile adult juvenile

SVL mean 78.8 40.5 71.6 32.6 83.1 55.4 119.2 61.9

SVL range 71.8–

85.9

30.2–48.2 70.8–72.3 28.0–

39.6

71.2–95.0 35.4–69.4 100.82–

139.45

40.54–

93.63

Mass mean 18.8 2.5 15.5 1.7 20.2 8.9 65.8 16.4

Mass range 16.0–

24.0

1.0–

5.0

14.0–17.0 1.0–

3.0

12.1–28.1 4.8–

11.3

40.0–

112.0

2.5–

30.0

Tan et al.—Feeding ecology of southern African Agamid lizards.

appenDix tabLe 2. Results of the ANOVAs performed on the prey Index of Relative Importance (IRI) comparing Agama

aculeata distanti (Eastern Ground Agama), A. armata (Peter’s Ground Agama), A. atra (Southern Rock Agama) and

Acanthocercus atricollis (Southern Tree Agama). Asterisks (*) indicate prey taxa that were signicantly different between

species of agamids. The degrees of freedom for all prey taxon is 3 and 61.

IRI F P

Formicidae 0.200 0.022*

Other Hymenoptera 0.566 0.345

Coleoptera 0.106 0.832

Hemiptera 2.537 < 0.001*

Diptera 3.374 < 0.001*

Diplopoda 2.134 < 0.001*

Lepidoptera 0.433 0.114

Orthoptera 0.103 0.345

Gastropoda 0.050 0.532

Ephemoptera 0.110 0.532

Isoptera 0.597 0.619

Isopoda 0.597 0.619

Wei cheng (nichOLas) tan is a Masters graduate from the Université de Poitiers, France, and

Universidade de Coimbra, Portugal. He received his B.Sc. with a major in zoology at the Australian

National University, Canberra, Australia. Wei Cheng is interested in the evolution of phenotypic

diversity and the mechanisms that drive biodiversity. He is currently a Ph.D. Candidate at the

Alexander Koenig Research Museum, Bonn, Germany, where he is conducting further research in

herpetology. Wei Cheng is passionate about conservation of reptiles and enjoys studying them in

the eld. (Photographed by Jason Savage).

anthOny herreL is a Research Director at the Centre National de la Recherche Scientique

working at the Natural History Museum in Paris, France. His research interests encompass the

evolution of form and function. He is specically interested in both the drivers of phenotypic

diversity and the constraints that hamper phenotypic diversication. Anthony works at one of the

oldest and largest natural history museums in Europe where he is lucky enough to be able to study

the anatomy and function of a diversity of vertebrate species. He has a keen interest in herpetology

and loves doing eld work to study organisms in their natural environment. (Photographed by

Colin Donihue).

JOhn measey is a Chief Researcher in Invasion Biology at the Department of Science and

Innovation-National Research Foundation, Centre of Excellence for Invasion Biology at

Stellenbosch University, South Africa. His research interests encompass all things herpetological,

with a special focus on invasions. John lives and works in the unique and diverse fynbos biome

of South Africa where many species are impacted by invasive species. (Photographed by Thalassa

Matthews).