BIOLOGY OF LARKS (AVES: ALAUDIDAE) IN THE CENTRAL NAMIB DESERT. Department of Zoology, Syracuse University, Syracuse, New York 13210

Transcription

1 BIOLOGY OF LARKS (AVES: ALAUDIDAE) IN THE CENTRAL NAMIB DESERT ERNEST J. WILLOUGHBY Department f Zlgy, Syracuse University, Syracuse, New Yrk 1321 ABSTRACT 1. The bilgy f six species f larks in the Namib Desert near Walvis Bay, Suth West Africa, was studied in 1964, 1965 and All species reprduced fllwing rainfall in summer and autumn mnths, with the appearance f green grass and abundant insects n which the birds fed. l. The primarily insectivrus species, Certhilauda alb/asciata, shws n seasnality in multing and testicular develpment, and reprduces whenever lcal rainfall results in favurable feeding and nesting circumstances, while the primarily granivrus Spizcrys starki and Erempterix ~erticalis have markedly seasnal testicular and multing cycles timed s that reprductin ccurs nly during the late summer and autumn mnths. Ammnwnes grayi is intermediate in fd habits and in timing f multing and reprductin, and C. albescens and Mira/ra nae~ia may als be intermediate in this respect. 4. S. starki and E. ~erticalis are the nly larks that nrmally drink, but flcks f bth species were bserved that were indepen~ent f drinking water. S. Larks avid heat stress during'daytime by keeping t the shade f stnes, tufts f vegetatin and muths f rdent burrws; and nests are nearly always placed n the shady side f a stne' r tuft f grass. 6. It is suggested that larks may be successful desert'ccupants because they have nt becme highly specialized and therefre can tlerate rapidly and severely changing envirnmental cnditins assciated with drught and irregular rainfall. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). INTRODUCTION The desert envirnment, with its shrtage f water, extremes f temperature, and paucity f fd and vegetative cver, presents special difficulties fr the survival f terrestrial animals. "The Namib Desert in Suth West Africa is ne f the driest, mst barren regins n earth; yet it is inhabited by a number f avian species. Prminent amng these birds are several species f larks, family Alaudidae, that live and reprduce successfully under these cnditins. The purpse f my study was t learn smething abut the ways in which these larks are adapted t their desert envirnment. I gave special attentin t the questins f hw and where these birds btain water, what they eat and hw they btain fd, what their breeding habits are, and hw reprductin and multing are related t changing envirnmental cnditins. I did my field wrk at the Namib Desert Research Statin at Gbabeb n the lwer Kuiseb River, abut 6 miles sutheast f Walvis Bay, during July and August, 1964, and frm July, 1965, t July, My apprach was extensive rather than intensive, since very little is yet knwn abut the avifauna and eclgy f the Namib Desert, r abut the bilgy f desert birds in general and larks in particular. Present address: Divisin f Science and Mathematics, McKendree Cllege, Lebann, Illinis 62254, U.S.A. Zlgica A/ricana 6 (1): (1971) III

2 134 ZOOLOGICA AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). FIGURE 1 Map f Suth West Africa, with the gegraphic limits f the Namib Desert indicated by stippling. The rectangle enclses the study area. (Frm Willughby & Cade 1967, with permissin f Namib Desert Research Assciatin.) THE NAMIB DESERT The Namib Desert extends sme 1 3 miles alng the west cast f suthern Africa, frm near Mssamedes in Angla t the muth f the Olifants River in the Cape Prvince. The gegraphy f this desert has been treated in detail by Wellingtn (1955) and Lgan (196). The latter reference in particular has a wealth f infrmatin pertaining t the study area, and the terminlgy used here fllws that given by Lgan where applicable. In Suth West Africa, this arid castal strip varies in width frm abut 6 t 1 miles (Fig. I), merging n the east with the semi-arid cntinental highlands. The study area includes all f Game Reserve Number 3 (Fig. 2), lying between the lwer Kuiseb River and Walvis Bay n the suth, the Swakp River n the nrth, and extending inland abut 8 miles. The area centres at abut the crdinates f 15 15' E. and 23 15' S. This area frms the gegraphic centre f the Namib and includes bth f the majr

3 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 135 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). FIaURB 2 Map f the study area indicating l-metre cntur lines, and utlines (but nt cnturs) f muntains and kppies. The heavy line is the brder f Game Reserve N.3 (als called Namib Desert Park); dark shading indicates areas f sand dunes; lighter shading indicates extent f currence and relative amunts f new grass resulting frm rainfall in sununer and autumn, , as assessed in May; dashed line enclses the area directly bserved and questin marks indicate areas that were nt bserved. gemrphic features f this desert. Frm the Kuiseb River suthward abut 4 miles t the Orange River, lies an expanse f high, barren dunes termed The Great! Sand Dunes by Lgan. Nrth f the Kuiseb, the dunes extend t the Swakp River nly al,ng the cast (Fig. 2). Otherwise, stretching nrth frm the Kuiseb is a vast, flat, gravel plain, The Namib Platfrm, which slpes gradually upward frm the cast fr abut 8 miles t the ft f the western escarpment f the cntinental highlands at abut 1 m elevatin. The Kuiseb River makes a remarkably distinct dividing line between The Great Sand Dunes and The Namib Platfrm. Althugh water flws at the surface f its bed nly when summer rainfall in the highlands is substantial, there is a perennial subterranean flw f water supprting a dense frest f tall trees alng its curse. This frest frms an extreme cntrast t the nearly vegetatinless expanses stretching nrth and suth frm the banks f the river, and prvides habitat fr many bird species frm the mre mesic highlands (Willughby and Cade 1967). Lwman (1966, p. 658) has published a clur phtgraph f this regin f the Namib taken frm uter space by U.S. Prject Gemini, which shws all the majr gemrphic features f this area, including the Kuiseb River with its frest.

4 136 ZOOLOGICA AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Climate The wrks f Lgan (196) and Schulze (1966) prvide descriptins f the climate f the Namib. Climatically, the Namib is divided int tw regins, the Outer r Castal Namib, and the Inner Namib. The Outer Namib is a fg desert, being influenced by the cld Benguela Current and within abut 2 miles f the cast, fg is f very frequent ccurrence, but rainfall is exceedingly scarce and averages nly abut 1 25 cm a year. Indeed, a large percentage f the ttal annual precipitatin at the cast cmes as cndensed fg water and dew (Walter 1936). Abut 4 miles frm the cast the cl, fggy climate f the Outer Namib gives way t a climate that reflects smewhat the climate f the cntinental highlands. Dew and"fg are unusual in the Inner Namib, and mre than a mnth may pass withut any cndensatin. Precipitatin in the Inner Namib, hwever, is appreciably greater than in the Outer Namib and cmes in the frm f lcalised shwers mainly in the summer and autumn mnths. Rainfall in the Inner Namib averages up t abut 152 mm annually, but is highly unpredictable fr any given lcality. Precipitatin. In the year this part f the Namib prbably had near nrmal rainfall, althugh just what the nrmal cnditin is has nt yet been determined because weather bservatins between the cast and the interir highlands have been discntinuus r lacking until 1963, when an fficial weather statin was established at the Namib Desert Research Statin, which lies right n the transitin between the Outer Namib and Inner Namib climatic regins. Lgan (196) gives a gd summary f weather data existing fr the study area t Figure 3 summarises rainfall data at lcatins frm Walvis Bay at the cast t Windhek in the interir highlands. Dnkerhuk is lcated just utside the nrtheastern crner f Game Reserve N.3, abut 2 miles east f the Tinkas water hles (Fig. 2), and represents rainfall cnditins at the eastern edge f the Namib Desert. The data fr mean mnthly rainfall were btained frm the Weather Bureau f the Republic f Suth Africa (1963), and s were mnthly rainfall figures fr 1965 and 1966 (Weather Bureau, Department f Transprt 1965, 1966). Figure 3 indicates that rainfall at Windhek was near nrmal fr bth years, but rainfall at Walvis Bay was practically nil ver the whle perid and was well belw the mean. Variability in ccurrence f rainfall in the Namib is well illustrated by the data frm Gbabeb (23 34' S., 15 3' E., 48 m), where slight but measurable rainfall came in January, February, March, April and September, 1965, and January, February, March, April and August, The data fr Dnkerhuk indicate the strng influence that the marked annual rainy perid f the interir has n rainfall in the Inner Namib. Except at Walvis Bay and Gbabeb, n rainfall figures are available fr the study area during the perid f this investigatin; but perids f ccurrence f rainfall within Game Reserve N.3 were easily determined by bservatin. During this study, rainfall in all instances was very lcalised, s that sme areas received several inches f rain (estimated) and thers less than 25 mm r nne at all. Figure 2 indicates general distributin f rainfall as assessed by the grwth f new grass during the summer and autumn f The shading n the map indicates the relative density and height f the grasses resulting frm the seasn's rainfall, as bserved in May, 1966, and s gives an indicatin f the relative amunts and distributin f the precipitatin. Perids when this rainfall.was bserved are indicated in Figures 5-8.

5 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 137 Ttll,...tlly Wel,,&,..,.. 1.'1, 1M,...: Wl... h... '" '... b (/) a: UJ I UJ :::E..J..J :::E z Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21)...J..J u... '" z '" a: MONTH FIGURE 3 Mnthly rainfall f lcalities in the central Namib and fr Windhek. lcated in the highlands f central S. W. Africa (Weather Bureau, Department f Transprt, Republic f S. Africa, 1963, 196', 1966). Vegetatin and Habitats One f the mst striking features f the Namib is its lack f vegetatin, and large expanses f the Inner Namib Platfrm may have hardly a trace f vegetatin after dry years. Perhaps the best way t treat the vegetatin here is as cmpnents f the majr avian habitats already defined by Willughby and Cade (I 967). Table 1 summarises the habitats and their characteristics. Surces f Drinking Water Fg is a pssible surce f drinking water fr birds in the Outer Namib, as it tends

6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). 138 Name~f habilat 1 Outer Gravel Flats Inner Gravel Flats Open Acacia Wdland Open Bush I.. Rcks Sand Dunes ZOOLOGICA AFRICANA VOL 6 TABLE 1 AVIAN HABITATS IN THE CENTRAL NAMIB General Descriptin Barren gravel plains f Outer Namib Platfrm Barren r sparsely vegetated gravel plains f Inner Namib Platfrm Scattered bushes 6 cm t 3 m tall anywhere in Namib Platfrm, but mstly in Inner Namib (where grund water that is nt t salty nears the surface) Large bulders, crevices and ledges with r withut vegetatin, anywhere in Namib Platfrm Shifting dunes, barren in Outer Namib, but with scattered tufts f perennial grasses in the Inner Namib Typical Vegetatin Nearly vegetatinless, r widely-spaced halphytic bushes 3 t 6 cm tall, Arthraerua leubnitziae and Zygphillum stapjii Sparse annual r perennial grasses, mainly Stipagrstis ciliata and S. btusa. Als S. uniplumus and S. hirtigluma. Sme areas with widely-spaced lw shrubs 15 t 6 cm tall, M nechma arenicla, Salsla spp., thers Primarily Acacia rejiciens, several ther species f shrubs In Inner Namib lw scrub dminated by Cmmiphra spp. and cactus-like euphrbias. In Outer Namib, lichens and small succulents (see Giess 1962; Walter 1936) In Inner Namib the grasses Stipagrstis sabulicla, S. lutescens, Eragrstis spinsa Characteristic Birds Layard's Chat Oenanthe tractrac Gray's Lark Ammmanes grayi, Stark's Lark Spizcrys starki Narrw strips f frest f Camelthrn Acacia giraffae Red-headed Finch very. pen aspect alng dry (large tree) and varius Amadina erythrcewater curses n Inner smaller trees and shrubs phala, Sciable Wea Gravel Flats ver Philetairus scius, Fiscal Shrike Lanius cl/aris Scaly-frnted Weaver Sprpipes squamifrns, Chat Flycatcher Bradrnis infuscatus, Schlegel's Chat Cercmela schlegelii Muntain Chat Oenan the mnticla, Familiar Chat Cercmela jamiliaris, Whitethrated Seed-eater Serinus albgularis Karr Lark Certhilauda albescens (Cntinued n next page)

7 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 139 Name J habitat l General Descriptin Typical Vegetatin Characteristic Birds Kuiseb Riverine Frest Dense frest f tall trees and brushy undergrwth alng lwer Kuiseb and Swakp rivers 1 Willughby and Cade Acacia giraffae, A. albida, Ficus sycmrus, Tamara austr-ajricana, Euclea pseudebenus, Salvadra persica Tit-babbler Parisma subcaeruleum, Redeyed Bulbul Pycnntus nigricans, Redfaced Musebird Clius indicus Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). t cndense n slid bjects and ften gives rise t puddles n bare rck, and drenches vegetatin with drplets. Scattered water hles and man-made wells ccur alng the lwer Kuiseb River, and supply drinking water t many fthe birds that live in the Kuiseb Riverine Frest (Przesky 1963, p. 78; Willughby and Cade 1967). Hwever, these watering places are f minr imprtance t larks and ther species f the Sand Dunes Jlnd Gravel Flats. The few natural water hles in the Inner Namib are f much greater significance t this study. Tw such water hles that supply water t many thusands f birds and game mammals are Grss Tinkas and Klein Tinkas (Fig. 2). These cnsist f places in a dry stream bed where water seeps Ollt f the carse sand. Muntain Zebra Equus zebra and Gemsbk Oryx gazella paw ut depressins in the wet sand and drink the water that seeps in t fill the hles. Birds, including sme species f larks, visit these places by the thusands in ht dry weather. Occasinal salt-water seeps are t be fund in the Outer Gravel Flats, but the water des nt appear: t be utilised. The Suth West Africa Administratin has erected three game-watering wells in Game Reserve N.3 at Ganab, Htsas, and Zebra Pan (Fig. 2). Each cnsists f a windmill that pumps water frm a brehle int a watering trugh, thus supplying water t thusands f birds and mammals frm the sununding Inner Gravel Flats. Avifauna The avifauna f the Namib Desert is relatively prly knwn. Since the middle f the 19th Century rnithlgists have made ccasinal cllecting trips int r acrss the Namib, spending nly a few days at a time there, and treating the avifauna f the Namib incidentally t a mre general cverage f birds f suthwestern Africa (see fr example Anderssn 1872; Hesch and Niethammer 194). Willughby and Cade (1967) have listed the species f land birds fund in Game Reserve N.3 during my perid f study, and have indicated their habitat assciatins. Our list includes 17 species, f which 81 species belnging t 28 families ccurred in the Namib utside the Kuiseb Riverine Frest habitat. Mammals Ptential mammalian predatrs f birds that were bserved in the study area included the Chacma Babn Papi ursinus, Bat-eared Fx Otcyn megaltis, Black-backed Jackal

8 14 ZOOLOGICA AFRICANA VOL 6 Canis mesmelas, Striped Plecat Ictnyx striatus, Yellw Mngse Cynictis penicillata, and Suricate Suricata suricatta. Herbivrus mammals in the study area include the Muntain Zebra, Steenbk Raphicerus campestris, Klipspringer Oretragus retragus, Springbk Antidrcas marsupialis, Gemsbk, Cape Hare Lepus capensis,. Rck Rat Petrmus typicus, Cape Grund Squirrel Xerus inauris, Striped Muse Rhabdmys pumili, Namaqua Gerbil Desmdillus auricularis, Lesser Gerbil Gerbil/us paeba, and Brush-tailed Gerbil G. val/mus, plus a number f ther small mammals. Arthrpds The insect fauna in the Sand Dunes and Outer Gravel Flats is dminated by beetles f the family Tenebrinidae, which are the mst cnspicuus animals in these regins (Kch 1962). The insect fauna in the Inner Gravel Flats, in additin t the Tenebrinidae and ther beetle families, has cnspicuus additins f ants, termites (chiefty the harvester termites, Hdtermes spp.), lcusts and stick insects, and thers. Insects are especiauy cnspicuus and abundant fr a few weeks after lcal rainfall, when they mujtiply and swarm in the grass and ther vegetatin that grw after the rain. Other cnspicuus arthrpd grups are slifugids, scrpins, and spiders. METHODS Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). I made regular field trips int bth Inner Gravel Flats and Outer Gravel Flats habitats t make bservatins and t cllect specimens. I als made bservatins in smau areas f the dunes n the suth side f the Kuiseb River, but this wrk was limited by ftding f the Kuiseb during five separate ccasins in February, March and April, which prevented crssing the river t get int the dune,s. The bservatins reprted here represent abut 2 hr f field wrk. Specimens were cllected whenever time and circumstances permitted, with a single-sht 22/ 41 cmbinatin rifte and shtgun. I injected abut 5 ml f 1 per cent frmalin int the peritneal cavity f each specimen as sn as it was secured t preserve the stmach cntents and gnads fr future inspectin. In this way specimens culd be cllected ver a perid f up t three days in the field and prepared later upn returning t the Namib Desert Research Statin. Whenever pssible each specimen was preserved as a study skin and the whle stmach preserved in 7 per cent ethanl r 1 per cent frmalin. Study skins are depsited in the Fuertes Bird Cllectin, Fernw Hall, Crnell University, Ithaca, New Yrk. The preserved stmach cntents were later analysed fr the percentage cmpsitin by vlume f varius fd items. This was dne, in the case f larger samples, by pling the stmach cntents f all the specimens cllected during a given mnth, and manually separating the cmpnents int arthrpd remains, seeds, and ther vegetable, matter under a dissecting micrscpe. The cllected fd items were then drained f excess ftuid n filter paper, and their vlumes determined by sinking them in a vlume f 7 per cent ethanl in a small graduated test tube and recrding the change in vlume. In this stl,ldy the term cmplete mult is used t mean the replacing f bth bdy and ftight feathers. The term partial mult means the replacing f part r all f the bdy feathers alne.

9 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 141 LARKS OF THE NAMIB The larks, family Alaudidae, are typically small, brwn r dull-clured, grund-dwelling passerine birds f pen, arid r semi-arid plains, and are fund thrughut the desert areas f Eurasia and Africa; but the family is nt restricted t deserts r semi-deserts, and sme species ccur in alpine regins and n tidal flats and sand bars (Vaurie 1951). There are abut 75 species (Van Tyne and Berger 1959, p. 52), with abut 28 f them ccurring in the suthern third f Africa (Mackwrth-Praed and Grant 1962). The nmenclature and systematics f the larks have been in cnsiderable dispute, with mst argument cncerning generic relatinships (Bianchi 195; White 1957; Winterbttm 1962; Harrisn 1966; Maclean ]969), and abut all that is agreed upn is that the family is quite distinct and cherent. Table 2 lists all the species f larks recrded in the study area tgether with their habitat assciatins. The status f each species during this study is given belw. Large-billed Sabta Lark. This species is cnsidered cnspecific with M. sabta _by sme authrs. It is ne f the s-called bush-iarks-that is, it ccurs in pen bushy habitats and freely perches n tp f bushes when nt fraging n the grund. It ccurred sparsely but regularly alng the eastern edge f Game Reserve N.3, and in places farther t the west where extensive areas f lw, pen scrub ccur n rcky terrain, which is mstly abve an elevatin f 8 m. The species is fund thrughut Suth West Africa. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). TABLE 2 THE LARKS RECORDED IN THE CENTRAL NAMIB DESERT AND THEIR HABITAT ASSOCIATIONS l Species Habitats GFD GFI DB Large-billed Sabta Lark Mirafra naevia XXX Karr Lark Certhilauda albescem Lng-biUed Lark Certhilauda curvirstris Spike-heeled Lark Certhilauda albfasciata XXX XX Gray's Lark Ammmnes grayi X XXX Red-capped Lark Calandrella cinerea.. Stark's Lark Spizcrys starki.. XXX X Grey-backed Finch-lark Erempterix verticalis X XXX X Black-eared Finch-lark Erempterix australis.. R X S XXX 1 GFO = Outer Gravel Flats; GFI = Inner Gravel Flats; OB = Open Bush; S = Sand Dunes; R = Rcks; XXX = primary habitat; XX = secndary habitat; X = marginal habitat; - = rare r status uncertain (see text). X

10 142 ZOOLOGICA AFRICAN A VOL 6 FIOURE 4 Habitat f the Karr Lark in sand dunes suth f!he Kuiseb R. abut 1 miles sutheast f Gbabeb, 2 April, The tall II'1lSS is S,ipagrOSlis s/)u/icf. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Xarr lark. The subspecies f the Karr Lark fund in the study area, C. a. erythrchlamys. was frmerly knwn as the Red-mantled Lark C. erythrchlamys (Meinertzhagen 1951) because f its distinctive pale reddish-brwn. unmarked back, and because it is restricted in distributin t the dunes frm Liideritz Bay and Aus nrthwards t the Kuiseb River. Mre recently it has been cnsidered t be ne f the eight r nine subspecies f the Karr Lark that are scattered ver the drier areas f the suthwestern Cape Prvince and suthern Suth West Africa (Lawsn 1961), This lark appears t be the nly avian species that permanently inhabits the sand dunes. where it lives in assciatin with the tall, spiny dune grasses Stipagrstis sahulicia and S. /ulescens (Fig. 4). Lng-billed LArk. A few lng-billed larks were bserved in pen bush n rcky terrain, mainly arund the Tinkas water hles in the nrtheastern sectr f the study area. I secured nly ne specimen. Hesdt and Niethammer (194) fund the Lng-billed Lark in the fgfree Namib and "Pr-Namib" between the Brandberg and Chus Muntains, which lie nrth f the Swakp River. and in arid suthern Suth West Africa. Taxnmists recgnise abut seven subspecies that are scattered thrughut Suth Africa, Suth West Africa. and Angla (Mackwrth-Praed and Grant 1962). Spike-heeled Lark. The Spike-heeled Lark ccurs sparsely all alng the eastern edge f the study area, especially where the Inner Gravel Flats habitat has a scattered, wdy scrub less than 33 em tall. The species has a wide distributin in sut hern Africa, and as many as 16 subspecies have been recgnised, primarily n the basis f divergencies in plumage clu r (Winterbttm 1958). It is peculiar amng the larks f this study in having an unusually lng, straight h.ind claw, the functin f which, if any, is bscure. Gray'sLark. This lark is endemic t the Namib. and ranges frm Aus and near Liideritz Bay nrth int the Kakveld, at least as far as Orupembe. In the study area it ccupied the mre

11 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 143 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). r less barren gravel flats, aviding the sand dunes and sandy sils. Hesch and Niethammer (194, p. 223) cnsidered it a characteristic bird f the fg desert, but in Game Reserve N.3 during my study it ccurred in greatest abundance eastward frm abut 6 m elevatin t the eastern edge 'f the escarpment at abut 1 m elevatin. In ther wrds, it was primarily a bird f the Inner Gravel Flats. Westward it was encuntered with rapidly diminishing frequency, and appeared there presumably as transients that wandered int the Outer Gravel Flats frm the east. This was mst nticeable during strng east winds, when birds suddenly wuld appear near the cast where they therwise culd nt be fund. Stark's Lark. Stark's Lark was ne f the mst abundant f the avian species in the study area, and was the mst abundant f the larks. It is a small bird, with lengthened crwn feathers that shw as a crest when it erects them. Its habitat in the central Namib is Inner Gravel Flats where sparse grasses ccur, and it is typically assciated with gravelly r stny grund rather than sandy sil. Outside the study area Stark's Lark has a wide range in arid and semi-arid suthwestern Africa. Hesch and Niethammer (194) fund it in the Namib ut t the beginning f the fg zne (Outer Namib), but nt int it. Grey-backed Finch-lark. The Grey-backed Finch-lark was the secnd mst abundant lark in the study area, and in general habits and habitat it resembled the Stark's Lark, althugh it fraged n sandy grund mre readily than the latter. The plumage shws a strng sexual dimrphism in clur, a type f characteristic that is lacking in mst species f larks, but which is typical f the genus Erempterix. Outside the study area, this species is widespread in suthern and suthwestern Africa. Black-eared Finch-lark. I bserved the Black-eared Finch-lark n nly ne ccasin, 8 April, 1966, in grassy Inner Gravel Flats abut 11 miles suth f Ganab. Tw birds in male and female plumage appeared briefly n the grund at clse range, but I did nt clject them. Outside the study area this species ccurs in suthern Suth West Africa, Btswana, parts f the Cape Prvince, Orange Free State, and western Transvaal (Mackwrth-Praed and Grant 1962). Red-capped Lark. The Red-capped Lark was bserved nly three times during the study. A female was cllected by Dr. Tm J. Cade abut 33 miles sutheast f Walvis Bay n barren Gravel Flats habitat n 3 August, It was in a flck f abut six birds. Subsequently; I bserved ne at Zebra Pan n 18 September, 1965, where it drank at the watering trugh, and anther ne n Outer Gravel Flats n 13 January, 1966, near Ribank abut 2 miles sutheast f Walvis Bay, where it was fraging amng scattered saltbush Arthraerua leubnitziae. Hesch and Niethammer (194) fund this species n the cast between Cape Crss and Swakpmund in February, Outside the study area it is widely distributed in the drier regins f Africa and Asia. R~SULTS The fllwing accunts f the species vary in cmpleteness accrding t the abundance and accessibility f the species. The birds listed abve that are mitted frm further cnsideratin because f insufficient bservatins are the Lng-billed Lark, Black-eared Finch-lark, and Red-capped Lark. Table 3 summarises specimen data fr the Large-billed Sabta Lark, and

12 144 ZOOLOGICA AFRICANA VOL 6 TABLE 3 SPECIMEN DATA FOR Mirafra naevia Specimen Bdy number Sex l Date weight Gnad measurements I Mult 282 M II Sept., 1 5 mm Late stage f cmplete 1965 mult 729 M 18 March, 26 7 g ca. 6 mm (testes dam- Early stage f cmplete 1966 aged by sht) mult 87 F 7 April, 24 1 g l'omm Heavy cmplete mult M 14 May, 25'3g 7 1 mm Cmplete mult just 1966 starting 1 M = male, F = female Ilength f left testis r diameter f largest fllicle Figures 5-8 summarise specimen data fr the Spike-heeled Lark, Gray's Lark, Stark's Lark, and Grey-backed Finch-lark. BODY WEIGHT Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Karr Lark. The mean weight f eight males was 29 4 g (range 26 3 t 33 1), and f 1 females, 26 8 g (range 24 8 t 28 9). Spike-heeled Lark. There was n apparent seasnal change in bdy weight. The data reflect the marked sexual dimrphism in bdy size, with males averaging abut 5 g heavier than females. Gray's Lark. There was n seasnal change f weight. Ther~ was definite sexual size dimrphism, reflected in the heavier bdy weights f males, which averaged abut 3 g mre than females. Stark's Lark. Bdy weight varied little with sex r seasn. Grey-backed Finch-lark. As in Stark's Lark, bdy weight did nt vary much with seasn r sex. BREEDING Large-billed Sabta Lark. I fund n nests. Anderssn (1872) described the nest as being f the dmed type with a side entrance. Reprductin very likely ccurred during summer mnths. /(Qrr Lark. I fund n nests, but the enlarged testes f males cllected in January and April, and a fully-develped egg ylk and viduct in a female cllected n 28 February indicate that egg-laying was ccurring during that perid. On IS Nvember, I bserved

13 ', WILLOUGHBY: BIOLOGY OF NAMIB LARKS I, 4 4 I POST- JUVENAL MOLT FlIG~T FEAT~ER5 1 I.....1::... I.. l..l. I..n:::.:::1 I BO~Y FEATHERS J.:...:-S. ::l::i::.i. I.:.:.I;j :..,".1... :::.1 QEGGS FLEDuLlN',S JUVENILES Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). 1 U) U) ~ ::: 8 ~ 141 u... ~ 7 ~ ~ 6...J u......j :!! 4 ~ ~ 3 (!) Z VI ~ ~ z ,...: ~ U) 24 ':Ii: «22 >::: (!) 2 In.. P "'8 OF ItAIN'ALL I II 8 ~ II 8 8 CERTHILAUDA ALBOFASCIAT A I "IM". '.'C.llt... FIGURE.5 SpeciJnen data fr the Spike-heeled Lark in Game Reserve N.3, Vertical bars fr mult indicate bserved inultm,; brken lines cnncctm, bservatins indicate perids when at least a part f the ppulatin was in mult. OMALE (MEAN WIlT OF 18 BIRDS FEMALE (MEAN WIlT. OF 13 8IRDS UNSEXED

14 146 ZOOLOGICA AFRICANA N' I T T T T I VOL " nedglings >( JUVENILES II Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). z ~ Z I Z 2 ILl...J IT II II "':Cf) ~ 211 ~ cc!) 17 III II > a:: I. (I) 8 8 "1111 ' RAINfALL I AMMOMANES ~ -.. II GRAY) 8 II FIGURE 6 Specimen data fr Gray's Lark in Game Reserve N.3, The length f the black lines fr multing indicate knwn duratin f multing in the ppulatin, and the width f the bar indicates the relative intensity f multing within the ppulatin (i.e. the narrw bar indicates that a few birds were multing, and the wide bar indicates that mst f the birds were multing). Brken lines indicate prbable multing, extraplated frm the data. s 8 'J' MALE.MEAN WT OF '22.7" 'EMALE,MEAN WT r» BIRDS' 2.2, I

15 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS N 4, 4 I T T 4 1 T 4 X I- 1 at ~ (/) ~ « ~ (/) ~ VEGETATION ::E => ~..J 4... > ARTHROPODS..J «... ll.. I- ffilillill SEEDS Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). POST-JUVENAL BODY MOLT ~.. ~.- FLIGHT (!) FEATHERS," tertial.") ("te,hals"') (!) ~ Z Z - BODY FEATHERS I- ".::.. ~ ~:..J... c ~ OEGGS a: " enestlings ::E CD e () e.fledglings (/) ~ 1 ~ (/) (/) ~ T 8 I- a: ~ I l- I- ll.. ~ ~ ::E 5..J.. ll....j 4..J ~ ;, X... ft 8 2 (.!) Z Z.. ~..J * 8 " JUVENILES MALE (MEAN WT OF Z 4 BIROS' ,.. 1.-: l (/) J I FEMALE ~ ::E e 1/ (MEAN WT Of' «3 BIROS' 11.1) >- a: (!) CD nll'odl Of' SPIZOCORYS STARKI UNSEXED FIGURE 7 Specimen data fr Stark's Lark in Game Reserve N.3, S)'Dlbls fr multing as in Fig. 6.

16 148 ZOOLOGICA AFRICANA VOL 6 Z... ac z en en '" 2 2 :) 1... '"...I > -l LL.... Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). C) C) Z Q Z... 'V...Ie '" a::: 2 CD en ~ fa en 1... a:::... '" LL..... '" '"...I 2 ~ Z... C) :J 4...I :i I Z ~ '"...I ~..,: ~ >- CD... 8 a IID a : 8 a ~, a g. 8 a,, ' I 8 I,, a EREMOPTERIX II VERTICALIS II I. I. I... ~ 1 'I.~I a FIOURE 8 Specimen data fr the Grey-backed Finch-lark in Game Reserve N.3, Symbls fr multing as in Fig. 6. a OEDGS.NISTLI"GS.fLIDGLI"GS " JUVE"ILIS -'E IMUII'WT. OF If _'lto I nmali II.AII.T." SO... lt'

17 1971 WILLOUGHBY : BIOLOGY OF NAMIB LARKS '" FIO URE 9 Nesl f Spike-heeted lark al the easlern brder f Game Reserve N_ 3 at Fann Onanis, 6 May, Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). birds chasing each ther vigrusly and assciating in pairs. and heard lud advertising sng, all f which indicate that breeding a<..livily was starting by the middle f Nvember. By the middle f June, there was n evidence f breeding. Prbably, then, the perid during which reprductin ccurred was frm late Oc:tber t early May, and s cincided with the rainy seasn f the Inner Namib. The nest f the subspecies C. a. ~ryrhrch/amys evidently has nt been described as such. The nests f ther ppulatins f the Karr Lark ften have a dme r bwer f stems ver them, and are usually placed under bushes (Vincent 1946, pp ; Maclean 1957; Mclachlan and Livers.idge 1957; Macwrth-Praed and Grant 1962). In the Sand Dunes. nests prbably are placed under the bush-like tufts f the dune grasses. The nest discvered by Heesch (1958) amng vegetatin n the sandy margin f the lwer Kuiseb River and described by him as the nest and clutch f Ammmanes gray; was mst likely that f C. a. eryrhrchlamys instead. Heesch described that nest as being rfed ver with dry grasses; and thus it is cnsistent with descriptins f the nests f the ther subspecies f the Karr Lark. Spike-heeled LArk. The fact that enlarged testes were bserved thrughut the year indicates that at all times f year the ppulatin had individuals that were prbably capable f reprductin under the right circumstances ; and it is reasnable t assume that this species breeds whenever envirnmental cnditins are prpitius, withut strict seasnality. A juvenile bird which was cllected in Oc:tber may have hatched the preceding August. Anther juvenile which was secured in April prbably had hatched in February r March. A nest with tw eggs was bserved n 6 May, These bservatins indicate that actual bree<ling did take place during bth dry and rainy perids. The ne nest that was bserved cntained tw eggs which were chalky grey finely speckled with brwn. It was an pen cup f ld, weathered grass leaves and plant fibres with a few twigs

18 ,,. ZOOLOGICA AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). FIGUU: 1 Nests f Gray's Lark. A, eight miles west f Zebra Pan, 26 April,1966, a scnlpcmeasuring IJ )( ', Scm. B, ne mile west f Zebra Pan, 14 May (the scale is I S ern lng), abut 25 em in diameter and em lng arund the rim, set in carse sandy sil against the su th-sutheast side f a tuft f grass (Stipagrsfis sp.) which was 3- Scm tall and 4 2 em wide at the base (Fig. 9). The nest had an inside diameter f 6 6 em, and a depth f 2 3 em. Gray's Lark. There was n clear-cut seasnal gnadal cycle in males, althugh the data suggest that there was a perid frm August t December when testes were mstly small and reprductively nnfunctinal. This perid cincided with the main cmplete mult. But there was n well-marked synchrny f gnadal recrudescence and regressin. and partly-

19 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS lsi " FlultE II Nesling habitat f Gray's Lark, Stark's Lark, and lhe Grey-backed Finch lark, 1 mile west f Zebra Pan, 14 May, The grass is primarily SlipagrO$(is b/u$q. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). enlarged testes OCCUlTed all year lng. This suggests that there may be sme individuals in the ppulatin leady t breed at all times f the year, as in the Spike heeled Lark, given the needed envirnmental circumstances. Actual breeding was detected during my study nly in March, April and May, which ci ncided with the summer autumn rainy seasn. The three nests that I fund aredesclibed as fllws: Nest number I was lcated eight miles west f Zebra Pan n 26 April, when it cntained tw nestlings estimated t be seven t nine days ld (Fig. loa). The "nest" was a bare scrape in the gravelly earth n the sutheast side f a wh.ite quartz stne abut 25 cm in diameter that shaded the nestlings during the afternn. Sparse green grass (Stipagrs/is ciliata) up t 3 cm tall, wh.ich had germinated after rainfall abut the middle f March, was grwing n the surrunding gravel flats. There was very UUle ther vegetatin. and prbably the eggs had been laid befre suitable nest materials in the frm f dry grass leaves and seed awns were available. Nest number 2 was lcated ne mile west f Zebra Pan n 14 May (Fig. lob), and it cntained a full clutch f tw eggs. The nest was a deep. thickly felted cup f very fine, dry grass fibres, stems, and feathery Stipagrstis seed awns n bare gravel with a small quartz stne at the nrth rim, which, hwever, prvided little shade fr the nest. The nest cup had an inside diameter f 5 7 cm, and a depth f 5 I em, and had been built in a little depressin with a surrunding aprn f small pebbles s that its rim was nearly flush with the level f the surrunding grund. The siuing bird sank deeply int the eup s that its back was level with the grund, and in such a psitin the bird was invisible frm just a few feet away because its drsal clratin matched that f the grund perfectly. The habitat in which tbis nest was situated was Inner Gravel Flats, with sparse green grass (S. b/usa) grwing abut 3 cm tall in patches with much bare gravel surface between (Fig. II). Nest number 3 was fund n 2 May, eight miles nrth nrtheast f Zebra Pan. It then

20 152 ZOOLOGICA AFRICANA VOL 6 TABLE 4 SUMMARY OF NESTING DATA OF Spizcrys starki Date Number Nest dimensins f eggs r (inside diameter nestlingsl X depth, cm) Prtective bject Cmpass directin frm prtective bject Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). 6 Apr. 3 e. 5 1 X Apr. 2 n. 4 5 X Apr. 2 e. 8 Apr. 2 e. 14 Apr. 2 e. 6 4 X? 15 Apr. 3 e. 5 8 X Apr. 3 e. 5 1 X 3 16 Apr. 3 e. 6 4 X? 16 Apr. 2 e. 5 4 X? 7 May I e., I n. 6 3 X May I e., 2 n. 5 5 X 3'7 15 May 2 e. 5 5 X May 2 n. 1 e. = eggs, n. = nestlings. Tuft f Stipagrstis ciliata 3 cm tall Tuft f S. ciliata 3 cm tall Small Zygphillum simplex Small Z. simplex Stne ca. 13 cm diameter Tuft f S. ciliata 3 tufts f S. ciliata Small Z. simplex Small Z. cyjindrijjium Tuft f S. ciliata 33 cm tall Small Z. simplex Tuft f S. ciliata 51 cm tall Stne 1 cm X 18 cm SSE S S S SSE S Sand E (Surrunded by branches) SE SSE SE S SSE cntained three eggs, ne f which was very small and turned ut t have n ylk. This nest was cnstructed just like nest number 2, and was set int gravelly sil at the suth side f a tuft f S. btusa that std 18 cm tall and affrded the nest sme shade during the afternn. It had a built-up aprn f small ( 6 t 1 5 cm diameter) pebbles arund it s that the nest rim was a little higher than the surrunding grund level. The inside diameter f the cup was 5 7 cm, and the depth was 3 8 t 4 5 cm, a little shallwer n the suth side away frm the tuft f grass. The habitat was Inner Gravel Flats with sparse but fairly evenlydistributed tufts f grass, S. btusa, 14 t 3 cm tall and spaced 8 t 8 cm apart.

21 1971 WILLOUGHBY : BIOLOGY OF NAMIB LARKS III Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). FIGURE 12 Stark's Lark fcedina yuna at nest, 8 miles nrtheast f Zebra Pan, 2S May, One nesfiina h.ad just tltered frm the nest, and the ther was still in the nest n the shady side f the stne. Slark's Lark. There was a definite seasnal leslicularcycle, with reprductin ccurring near the end f the rai ny perid frm late March 1 late May. Nesting ccurred in areas where green grasses had grwn after rains and had begun t mature and set seed (Fig. II). Thirteen nests were bserved. They were all simple pen cups made f fine, dry siems, leaves, panicles, and feathery seed awns f grasses, sunk in a slight depressin at the base f a small plant r beside a Slne (Fig. 12), s that the rim was mre r less even with the surrunding grund level. The nest was invariably placed n the suth t sutheast side f the prtecting bject s as t be shaded in the early and middle afternn. Table 4 summarises the nesting data. It appears that tw- and three-egg clutches arc abut equally prevalent, but statistics n actual clutch size are nt presented because I learned nthing abut mrtality in the nest, r whether the bserved clutches were cmpleted nes in every case. Nestlings leave the nest a few days befre they can fly, and ttter after the adults t t be fed (Fig. 12). As sn as they can fly, they jin large flcks f adults and ther immatures.

22 1>4 ZOOLOGI CA AFRICANA VOL 6 FIGVRf. \3 Nesl and eggs f the Grey-backed Finch-lark, 6 mi les sutheast f HQ(sas, 8 April, The gram is $ripagrsfis cilitlta. T AB LE 5 SU MMARY OF NESTI NG DATA FOR Erempterix vertica/is Dale Number Nest dimensins f eggs r (inside diameter nestlings' x deplh. em) PrOlectire bjeer Cmpass directin frm prtectilte bject Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). 8 Apr. 15 Apr. 16 Apr. 16 Apr. 16 Apr. 16 Apr. 3 Apr. 5 May 7 May 7 May 7 May 14 May 14 May 14 May 2 May 25 May 2 e. 2 e. 1 n. 2 e. 1 e. 2 e. 1 e. 2 e. 2 e. 2 e. 2 e. 2 e. 2 e. 2 e. 2 e. 2 e. 1 e. = eggs, n. = nestlings. 4 8 x x 2 5 5' 1 x x x x 3 g 5 6 x X j x X 3,) 5 1 X \ X X 2 5 s s X ) Tuft f Stipagrslis ciliata T uri f S. ciliala Tuft f S. ciliata 38 em tall Small Zygphillum simplex Cleme suffruticsa ) 5 em tall Tuft f S. ciliata 38 em tall Tuft f S. hfusa 17 8 em tall Tun f S. btusa 25 em tall Small Z. simplex 5 em tall Tuft f S. ciliata 41 em lall Tun f S. b/usa 23 em tall 4 tuns f Stipagrs/is sp. 25 t 3 em tall Tun f S. btusa 23 em tau Tun f S. b/usa 25 em tall 2 lufts f S. btusa 2 and 25 em tall Tuft f S. ciliata 41 em tall SE S SE ESE SSE SE S S S SSE SSE N, W, S S S SE S

23 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS ISS Grey-backed Finch-lark. The marked testicular cycle was crrelated with a breeding perid during the autumn mnths at the end f the rainy perid. Nesting ccurred side by side with Stark's Lark. The nest f the Grey-backed Finch-lark was practically indistinguishable frm that f Stark's Lark (Fig. 13), differing mainly by a tendency t have the materials built a bit mre lsely int the cup lining. The nest f the Grey-backed Finch-lark als ccasinally had a surrunding aprn f small pebbles which apparently had been taken frm the nest excavatin and placed arund the nest rim. Table 5 summarises the data fr the 16 nests that I bserved. Tw eggs made the usual clutch. In cntrast t Stark's Lark, nne f the nests was beside a stne. All nests were shaded frm midday thrugh the middle f the afternn by small plants beside them. MOULT AND PLUMAGES Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Large-billed Sabta Lark. Data in Table 3 indicate that a cmplete mult ccurred in autumn and winter mnths (March t September). These data d nt give any infrmatin n whether r nt there is a partial mult at sme time during the year. All the specimens were mature birds, judging frm the fact that their skulls were fully ssified. Prbably the tw males in early stages f cmplete mult (729 and 876) were at the end f a breeding cycle, judging by their enlarged testes. Karr Lark. Data are t meagre fr definitive statements n mult in this species. The juvenal plumage persists fr several weeks, the white edges t the feathers f the head and back gradually wearing ff befre the pst-juvenal mult begins. It is nt clear frm the few specimens f immature birds in mult whether there is a partial pst-juvenal mult, r whether it is cmplete, but I suspect the latter. Adult Karr Larks did nt shw any definite partial mult. Mst f the multi~g adult birds were underging cmplete mult after breeding. Macdnald cllected 18 specimens f C. a. erythrchlamys at Tsndab Vlei, nt far utside the present suthern brder f Game Reserve N.3, n 5 and 6 March, 195. One f these specimens was in juvenal plumage, and the adults were either in extremely wrn plumage with enlarged but apparently regressing gnads, r were just be~nning t mult (Macdnald 1953a, p. 333). My bservatins are cnsistent with these findings. Spike-heeled Lark. The fragmentary data indicate that multing may ccur thrughut the year, which is cnsistent with the evidence that reprductin may ccur at any time f year. It appears likely that an individual experiences a pst-nuptial bdy mult beginning when the gnads start t regress after a perid f sexual activity, fllwed immediately by, r merging int, a cmplete mult. Yung birds experience a pst-juvenal bdy mult starting perhaps within the first tw mnths after fledging. althugh it is impssible t determine precisely the time relatinships in multing frm the data that I cllected (Fig. 5). It is safe t say that

24 ls6 ZOOLOGICA AFRICANA VOL 6 there is little r n synchrny in either gnadal r mult cycles amng individuals f the particular ppulatin sampled. Gray's Lark. Individual birds in sme stage f mult were cllected all year lng. Mst specimens shwed synchrnised seasnal mult as diagrammed in Figure 6. A pst-juvenal bdy mult ccurred beginning a few weeks after fledging, and this merged int a cmplete mult during spring and early summer mnths at the same time that lder birds were underging a cmplete mult. Breeding birds wele either nt multing r were in very weak cmplete mult. The general picture f mult in Gray's Lark was ne f definite lng perids f strng multing by mst f the ppulatin, alternating with perids when multing ccurred at a lw level in a small part f the ppulatin. Stark's Lark. There was definite, synchrnised multing cycle in the ppulatin I studied, as diagrammed in Figure 7, In May, immediately fllwing breeding, adults went thrugh a heavy pst-breeding mult that invlved secndaries 7-1 (the tertials), lasting nly abut a mnth per individual and being s precipitus as t leave the birds with large bare patches n head, neck r breast cvered nly by the shrt pin-feathers cming in t replace the feathers that had drpped ~ut. At the same time, beginning within a mnth f fledging, juveniles went thrugh a heavy pst-juvenal bdy mult. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Grey-backed Finch-lark. Figure 8 summarises bservatins n moulting. There was a very rapid pst-breeding bdy mult that invlved secndaries 7-,1, and ccasinally the central pair f rectrices. Althugh I btained n immature specimens at the end f the 1~66 breeding seasn t shw it, presumably they began a pst-juvenal bdy mult abut the first part f July. The marked sexual dimrphism in plumage clur is f interest. The female is generally brwn with dark streaks, and has a black patch n the belly, whereas the mature male has a strngly cntrasting black and white head pattern, black breast, and grey-brwn back streaked with black. Very yung males were indistinguishable frm yung females and did nt attain any f the male plumage cluratin in the pst-juvenal mult. During their first cmplete mult, yung male Grey-backed Finch-larks tk n mre r less f the male plumage cluratin, but in few instances did they attain the full black-and-white pattern until their first pst-breeding mult. Sme f these yung males with streaked upper breast and nly the suggestins f the black and white face pattern shwing n the still streaked head were bserved at nests, and cllected specimens had fully enlarged testes, indicating that first-year males may breed despite their immature plumage. FOOD Large-biIJed Sabta Lark. The stmachs f three f the specimens were preserved. The ttal vlume f rganic matter was 5 mi, cnsisting f small seeds (mainly f Slipagrslis spp., Cleme spp., and thers), 6 per cent; insects (Hdtermes and a few ants), 4 per cent;

25 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21) WILLOUGHBY: BIOLOGY OF NAMIB LARKS 157 and a trace f green vegetatin (part f a grass stem). The stmach f the furth individual held nly small seeds. Karr Lark. The prprtinate vlumes f fd items frm 19 stmachs cllected in January, February, April and June are 32 per cent seeds, 68 per cent arthrpds, and traces f green vegetatin. Nearly all f the seeds were frm the grass Slipagrstis sabulicla. A larger prprtin f these seeds in stmachs in June (87 per cent, cmpared with abut 2 per cent in ther mnths) may have resulted frm increased availability f this fd, since this grass prduced seed in March and April. The arthrpds cnsumed were primarily large ants that swarm in the dunes arund the clumps f S. sabulicla. Besides ants, Karr Larks ate small lcusts, beetles, caterpillars, and pssibly Hdlermes. They prbably eat whatever insects are available. Spike-heeled Lark. Thrughut the year fd cnsisted chiefly f insects-ants, Hdtermes, small beetles, and ther arthrpds such as slifugids. Seeds cnsumed were mainly frm grasses (Stipagrstis spp.) and the herb Mnsnia umbel/ata. The average percentages f fd items fr the whle year are 16 per cent seeds, 84 per cent arthrpds, and traces f green vegetatin. Gray's Lark. Percentages f fd items fr mnthly samples are diagrammed in Figure 6. Small seeds frmed the bulk f the diet' in the driest seasns, but after rainfall that resulted in plant and insect grwth, insects were mre prevalent in the diet. The means f the percentages fr all mnths are 61 per cent seeds, 36 per cent arthrpds, and 3 per cent green vegetatin. When the ttal vlumes f fd items in the year's cllectin are cmbined, the percentages wrk ut t 56 per cent seeds, 43 per cent arthrpds, and I per cent green vegetatin. The discrepancy between the means f the percentages f mnthly samples and the percentages f the cmbined year's sample cmes frm the fact that the sample vlumes varied greatly frm mnth t mnth, s that a sample f small vlume cnsisting mstly f seed, fr example, cntributed mre strngly t the seed categry when taken as a mnthly sample than when added as part f the ttal vlume f cllected material. Hwever ne calculates the percentage cmpsitin f the diet, it is clear that seeds predminate. The seeds cnsumed by Gray's Larks were primarily grass seed (Stipagrstis spp.) with substantial amunts f seeds f the herbs Cleme spp. and Mnsnia umbel/ala, plus smaller amunts f ther species. The green vegetatin cnsisted mainly f grass, which was ingested bth directly, and as the stmach cntents f herbivrus insects such as lcusts. The arthrpds eaten were mainly insects-hdtermes, ants, varius small beetles, small stick insects and lcusts, flies, and mths, rughly in descending rder f frequency. Other arthrpds included in the diet were small spiders and slifugids. Tw nestlings ne day ld were crammed with small lcust nymphs, several spiders, and a beetle larva. Tw nestlings estimated t be seven t nine days ld cntained a large (2 cm lng) lcust nymph and a large fly, respectively. Nne f the nestlings cntained seeds, and the data suggest that the yung are fed exclusively n insects while in the nest.

26 ISS ZOOLOGICA AFRICANA VOL 6 Stark's Lark. The average percentages f fd items are 75 per cent seed, 2 per cent arthrpds, and 5 per cent green vegetatin (Fig. 7). When the whle year's stmach cntents were cmbined and vlumes wrked ut, the percentages were 77 per cent seed, 19 per cent arthrpds, and 4 per cent green vegetatin. The reasns fr the discrepancy are the same as thse fr Gray's Lark discussed abve. It is clear that seeds frm the bulk f the diet. Small seeds f all kinds were cnsumed, but were primarily f grasses (Stipagrstis spp.), Cleme spp., and M. umbel/ata. The green vegetatin cnsumed was mstly grass. The birds btained this by pulling at the culm until it slipped ut f the sheath, and then nibbling at the tender, juicy basal jint and discarding the rest f the r.tem. I saw a bird take bites ut f the yellw flwers f Tribulus sp., a prstrate, creeping perennial herb that blmed after rains. Arthrpds in the diet were primarily ants and Hdtermes, with additins f small beetles, bugs, flies, and spiders and slifugids. It shuld be nted that a large amunt f green grass had been cnsumed by birds cllected the first week f Octber fllwing rains n September 3, which had caused grass seed t germinate, and that cllectins in February and March als had cnsiderable amunts f green vegetf.tin which had grwn up fllwing rainfall in thse mnths. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Grey-backed Finch-lark. Fd cnsisted verwhelmingly f small seeds. In a series f specimens cllected in January just after a rainfall, harvester termites predminated; but thi& was an unusual situatin and very likely was wing t the sudden cnspicuus emergence f winged sexual individuals, at which time all kinds f birds were feeding n the termites. The means f the mnthly percentages f fd items are 91 per cent seed, 8 per cent arthrpds and J per cent green vegetatin. Because sme f the mnthly stmach cllectins cntained nthing but seeds, they were nt preserved, s n figures based n ttal vlume f stmach cntents fr the year are given. The seeds mst cmmn in the stmachs were f grass Stipagrstis spp., and f the herbs Cleme spp. and Mnsnia umbellata. There were lesser amunts f small, unidentified seeds. The green vegetatin was chiefly grass shts and spruting grass seeds. Arthrpds included ants, Hdtermes, and small beetles; and unmeasurable traces f insects, such as muthparts and chitin fragments, were usually present in the stmachs. DRINKING Large-billed Sabta Lark. I did nt see this species drink, althugh it ccurred in the rcky terrain arund the Tinkas water hles and wuld thus seem t have the pprtunity t d s. lt prbably gets mst f the water it requires frm the insects it eats. Karr Lark. Surface water des nt exist in mst f the habitat f the Karr Lark, and dew r ther cndensatin is rare in dunes f the Inner Namib where Karr Larks are mst abundant. I did nt see this species drink at any f the watering places in the bed f the Kuiseb River adjacent t parts f its habitat, such as the fresh-water pls at Ribank

27 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 159 sme 2 miles sutheast f Walvis Bay. The Karr Lark prbably btains all its water frm its fd. Spike-heeled Lark. This species was never bserved t drink, and since its diet cnsists mainly f insects, which prvide a relatively gd surce f water, it prbably never has t drink. Gray's Lark. Gray's Lark des nt nrmally drink. Althugh n ne ccasin at the Zebra Pan water installatin several birds walked alng the edge f the water briefly, nly ne f them stuck its beak int the water, which it did nce. This was the nly time in many hurs f bservatin that I saw this species at water. Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Stark's Lark. Stark's Lark was ne f the tw larks that gathered at watering places t drink. During dry times, August thrugh the first half f January, Stark's Larks ften carne in flcks t the Grss Tinkas water hle, but rarely t the Ganab and Htsas watering places. I als bserved this species drinking at a stck-watering installatin at the eastern brder f Game Reserve N.3 n the farm Onansis, and at a water hle in a wide part f the Kuiseb Canyn where the main Walvis Bay Rad crsses the river. Althugh several hundred individuals wuld smetimes drink in a day at Grss Tinkas, which was the watering place mst regularly frequented by Stark's Lark during August t January, it became clear early in the study that these birds represented a relatively small prprtin f the ppulatin in the surrunding gravel fiats. On the basis f rad cunts and ppulatin estimates, I cncluded that smething n the rder f 1 per cent r less f the ppulatin within a five-mile radius f the water hle drank there per day. Later in the summer and autumn mnths after rains had brught up the grasses and when insects were abundant, it was unusual t see Stark's Lark at water anywhere, althugh it literally swarmed in grassy areas n the surrunding flats. The distributin f Stark's Lark shwed n definite relatinship t watering places in the Namib. It is thus evident that sme individuals f this species did nt nrmally drink every day, if at all, even in dry cnditins, and that when there was green grass and abundant insects. they were independent f surface water. Grey-backed Finch-lark. The Grey-backed Finch-lark drank mst frequently f all the larks f the Namib. At all times f the year they culd be fund at watering places that were near their ppulatin cncentratins. The watering place mst frequently visited was Grss Tinkas. Htsas, Ganab and Zebra Pan were visited during the breeding seasn when many breeding birds had cllected in the fresh grass arund these lcatins. Nevertheless, the distributin f this lark did nt appear t be bund t the lcatins f drinking water, as I bserved large flcks in the Inner Gravel Flats which appeared t be independent f drinking water. The distributin f this lark was mre strngly crrelated with the presence f its preferred grassy habitat than with surface water.

28 16 ZOOLOGlCA AFRICANA VOL 6 BEHAVIOUR Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Large-billed Sbta Lark. I had little pprtunity t bserve behaviur in this species. Karr Lark. The Karr Lark spends its life n the Sand Dunes, fraging between and arund the clumps f dune grasses and seeking refuge within the bases f the grass clumps. It takes fd by several methds. One methd is simply t pick up fd, such as ants, frm the surface f the sand. Anther methd is t uncver buried fd with the beak by rapid sideways jerking f the head s that the tip f the beak scatters the sand aside. Besides using these methds f feeding, Karr Larks frequently hp up severe! centimeters, r as high as a meter, t seize insects and seeds directly frm the verhanging stems f plants, but in ding s they d nt perch n the stems. One f the mst marked aspects f the behaviur f the Karr Lark, particularly with regard t water cnservatin, is its inactivity during the warmest time f day. The Karr Larks frage and display actively in the mrning until arund 1. r 11. n a ht summer day, when the grund surface temperature appraches 5 C. Then all birds disappear int the bases f the large shady grass clumps and d nt stir again until abut 16.3, when the grund cls ff smewhat. Then they suddenly reappear and frage and display actively until dusk. Sme representative temperatures taken during a time when the birds were inactive will shw hw they avid undue heat stress by remaining still in the shade: On 2 April, at 13. at a lcatin in the S~nd Dunes abut 1 miles sutheast f Gbabeb, the surface temperature f the sand n level grund was just 5 C. The air temperature in the shade 1 cm abve the grund was 33, and the air temperature in the shade f a clump f Stipagrstis sabulicla at the spt where a Karr Lark had been squatting a few secnds befre was als 33. Sand surface temperatures in the sun ften exceeded 5 C, It is quite evident frm these figures that the bird cruching in the shade at a temperature f 33 was aviding cntact with the ht grund, and radiatin frm the grund and sun, and was in this way minimising its water lss t evaprative cling. I did nt have a chance t study the advertising sng and displays f this species in detail, but the fllwing ntes are included because they add t general bilgical infrmatin n the Karr Lark. The advertising sng, given bth when flying and when perching, was fairly steretyped and nt very lud. I transcribed it int syllables as fllws: tip-tip-tip-tip-tipzree'-trrrr. The first five ntes are incisive, staccat ntes, the sixth a buzzy, high-pitched nte, and last a lwer-pitched trill. The entire sng has a duratin f tw t three secnds. This sng was given by birds standing n munds f sand, perching n grass stems several feet abve the grund, and when flying slwly arund with slw, deep wing beats 3 t 2 m abve grund. Besides the advertising sng, a rapid series f mellw bubbling ntes was given, usually by birds hvering abut 3 m up with feet dangling. When ne bird was vigrusly chasing anther, as if a territrial dispute were in prgress, a buzzy nte was repeatedly uttered. During the breeding seasn I usually saw Karr Larks in pairs, r in what appeared t be family grups f tw adults and ne r tw juveniles. Later, when breeding had finished, I fund them mst ften in small parties f fur t eight individuals that included bth adult and immature birds. Thus there is n evidence that they frm large flcks as d sme ther species f larks.

29 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21) WILLOUGHBY; BIOLOGY OF NAMIB LARKS 161 Spike-heeled Lark. The Spike-heeled Lark ccurs in Inner Gravel Flats with sparse grass, and I fund it mst frequently where there were widely-spaced, lw, wdy perennial shrubs up t abut 6 cm tall. The birds ccasinally perched n these shrubs, thus aviding cntact with the ht grund during midday; but mre ften they std in the shade beneath. I seldm encuntered single Spike-heeled Larks, but usually' fund them in pairs and small parties f up t six individuals. The latter grups were made f birds f bth sexes and all ages, even including lder fledglings still being fed by adults. It thus appears that the Spike-heeled Lark is sciable except when nesting, and tends t assciate in small grups rather than large flcks when nt tied t a specific nest site. Nestlings leave the nest and run after the adults befre they can fly. Spike-heeled larks have a marked tendency t stay arund rdent burrws, especially the large burrw cmplexes f the grund squirrel Xerus inauris, and the birds spend cnsiderable time fraging in and arund the burrw entrances. This may be partly because wind-blwn seeds and surface-running insects tend t drp int the muths f burrws and t cllect there. At the same time the birds can take advantage f the shade and cler grund temperatures available in the burrw entrances when fraging, which they d at all times f day, and thus reduce the need t use evaprative cling. Spike-heeled Larks readily g dwn rdent burrws when wunded, and several times I was bliged t dig birds ut f these refuges. Macdnald (l953b) nticed this same escape reactin in the nrth-western Cape Prvince. It appears that the burrws f rdents serve the Spike-heeled Lark well as cl feeding spts and ready predatr-escape tunnels in an therwise nearly featureless habitat. Besides fraging arund the muths f rdent burrws, Spike-heeled Larks als frage n the pen grund. They take fd items directly ff the surface f the grund and als uncver buried items by digging with rapid sideways jerks f the beak, as described fr the Karr Lark. I have n definite bservatins n advertising sng and sexual behaviur. The usual nte heard was a mellw, lw-pitched chatter r trill. Otherwise this species appears t be cmparatively silent. Gray's Lark. Gray's Lark inhabits gravel flats exclusively, with r withut sparse grass; but it was encuntered mst frequently where there was sparse grass standing 15 t 5 cm tall with much bare gravel surface between tufts. All year lng this species assciated in small parties f frm three r fur up t as many as 3 birds, including bth sexes and ld and yung birds. The mean size f 121 grups ttalling 396 individuals bserved in January, February and March was 3 3 birds, with a range frm I t 1. Individuals did nt appear t be strngly assciated in these small grups, and when fraging they spread ut ver wide areas cvering several acres and smetimes wuld jin ther similar grups that passed by. But wherever Gray's Lark ccurred, there was a marked tendency fr several birds t assciate tgether and t travel as a grup. These grups wandered ver large expanses and nwhere shwed definite signs f staying in a circumscribed range.

30 162 ZOOLOGICA AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). In parts f the study area where Spike-heeled Larks ccurred, I ccasinally saw small grups f bth species clse tgether arund the same rdent burrws. Hwever, Gray's Lark tended t be scarce in areas f lw scrub where the Spike-heeled Lark was mst prevalent. Like the Spike-heeled Lark, Gray's Lark has the habit f fraging in and arund the muths f rdent burrws, and n dubt avids daytime heat t sme extent this way. This species als dives dwn burrws when wunded. Hwever, these habits are nt s marked as in the Spike-heeled Lark, and Gray's Lark spends cnsiderable time fraging in the pen n the bare gravel surface, even at midday when temperatures are arund 38 C. Between buts f fraging during the heat f the afternn, hwever, Gray's Larks stand n stnes r twigs a few centimetres abve the gravel surface with the flded wings held away frm the sides f the bdy s as t expse the thinly feathered surfaces f the sides and under the wings, and face int the prevalent westerly sea breeze, thus dubtlessly facilitating the lss f bdy heat t the cler air by cnvectin. Gray's Lark takes fd principally frm the surface f the grund. but it als uncvers buried items with sideways jerks f the beak as d the ther lark species. It feeds n grass the way Stark's Lark des, discarding the culm after nibbling the tender, juicy basal nde. Nests f Gray's Lark were exceedingly difficult t find. The incubating r brding adult n the nest was practically invisible because f its cncealing cluratin. I saw birds get ff their nests and walk away fraging while I was appraching in a vehicle r n ft and still abut 2 r 3 metres away, making it hard t find the nest when its lcatin was nt already marked. Nestlings leave the nest well befre being capable f flight and fllw the adults t be fed. When able t fly, but still being fed by adults, they jin with the usual small mixed scial grups. Curtship and sexual behaviur was difficult t study because f the lack f sexually dimrphic plumage. One rather unusual trait f Gray's Lark is that the flight sng-what ne wuld rdinarily cnsider the full sng r advertising sng-is perfrmed primarily in the dark. Frm early February thrugh May, but mst frequently in March and April, there wuld be a burst f cailing and singing fr abut half an hur after sunset, and again fr abut tw hurs befre dawn and ending befre sunrise. Singing and calling f this srt were als bserved in the daytime, but this vcalisatin was rarely as vigrus then as at dusk and befre dawn. The sng was a cmbinatin f high-pitched, very sharp tinkling ntes with a lud, up-slurring whistled nte, and when given in flight was accmpanied by a lud, reedy whirring f the wings. Tw variants f the sng are diagrammed in Figure 14. Althugh the number and rder f tinkling ntes varied sme, the slurring whistle was invariable. When this sng was given in flight, the wing whirring preceded the vcal phrase and verlapped the initial tinkling ntes, as diagrammed in Figure 14. When I bserved the flight sng perfrmed during the day, the bird fllwed an undulating flight path abut seven metres abve the grund (Fig. 14). In this display the nly time the wings were beating was when they prduced the whirring sund. In rdinary level flight the wing sund was much less audible, and it thus seems that the wing sund is prduced as a definite part f the display. Other vcalisatins include a high-pitched tseet, given by birds disturbed by the presence

31 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 163 x X x 9 ',.cgd. (A) WING WHIRRING FLIGHT PATH VOCALIZATION FIGURE 14 Diagrams f advertising sng (A) and sng flight (B) f Gray's Lark. An x dentes a sharp, tinkling nte, and a thin line dentes a whistled nte. (8) Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). f man, a mellw tew, given in an indefinite series, and a series f three piping ntes which were uttered by birds assciating in grups. Curtship and pairing appear t take place within the small grups, since I bserved adults feeding ther adults (presumably males feeding females) and munting f ne bird by anther several times within grups f fur r five birds in April and May. At the same time territriality was nt markedly expressed, and the impressin these bservatins gave me was that birds paired while still mving abut with a grup, left the grup t establish a nest, and then rejined a grup as sn as the fledglings culd fly. Stark's Lark. When fraging, the Stark's Lark creeps alng hlding its bdy clse t the grund and pecking at the surface f the grund r ccasinally using the sideways jerking f the beak t uncver buried fd items. When a bird finds a seed r ther fd item it usually raises its head and squats dwn while nibbling and swallwing the bject. It thus gives the impressin f being smewhat lethargic. The bserver may walk within three r fur metres f several birds withut being aware f them until they suddenly rise and fly a shrt distance making sft chirping sunds. Stark's Lark shwed a tendency t fly twards vehicles and hver abve r land beside them, as if inspecting them ut f curisity. I nearly always encuntered Stark's Lark in flcks f fur r five t several hundred individuals. These flcks were cntinually wandering arund frm ne grassy area t anther s that I did nt always encunter them at the same places frm week t week; and after rainfall they had a marked tendency t cllect where green grass was grwing, where they smetimes cngregated by the thusands in an area f a few square miles. Their habitat preference was Inner Gravel Flats with grass standing 25 t abut 51 cm tall and separated by plenty f bare grund (Fig. II).

32 164 ZOOLOGlCA AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). I first heard singing n 6 December, and thereafter sng became mre frequent, reaching a high level frm February thrugh April and then declining. The sng, a lng series f sft churring chirps and whistles, is given n the grund r in the air. I transcribed an example f this sng int the fllwing syllables t try t recrd the character f it: "chur-chirrcheer-chr-cheer-reel'." In perfrming the flight sng display, the bird typically rises steadily frm the grund nearly vertically, singing cntinuusly until reaching a height f 5 t 2 m, where it hvers against the wind, r circles slwly if there is n wind s that it remains abve the same spt n the grund, and sings cntinuusly fpr several minutes; and then it clses its wings and dives vertically t the grund where it may ften indulge in a brief aerial chase f anther Stark's Lark that has been n the grund. Smetimes birds sing while flying distances f several metres a few feet abve grund with feet dangling and wings fluttering slwly. Birds singing n the grund ften accmpany the sng with crest raising. A male displaying t a female n the grund sings cntinuusly and faces tward the female with bdy hrizntal, crest raised, and head bbbing slwly up and dwn. During the mnth r s prir t actual breeding, thusands f Stark's Larks gathered in areas where green grass was grwing and displayed and sang; but there was a great deal f milling arund and mving frm place t place during this time, and actual nesting did nt begin until the grasses had begun t mature and set seed. Then suddenly nests appeared and the singing and milling arund that had been prevalent nly a week befre was much reduced. Meanwhile in adjacent areas where the grasses had nt yet begun t mature, Stark's Lark cntinued t fly abut and sing cnspicuusly. The densest ppulatins f this species that I fund in these grassy areas I estimated at tw t five birds per acre. Nesting, nce begun, prceeded rapidly, with nn-flying yung leaving the nests in a mnth and then jining pst-breeding flcks f adults and juveniles shrtly after they culd fly. During the httest times f day, Stark's Larks usually avided the heat by cruching in the shade f stnes, verhanging rcks, r, mst frequently, f vegetatin. Occasinally birds wuld perch atp lw shrubs r rcks and stand facing int a breeze with the flded wings held away frm the sides s as t expse the thinly-feathered sides f the thrax and axillae. In tw such instances, I measured a grund surface temperature abve 5 C (ff the scale f my thermmeter) and air temperatures at a height f 61 cm f 36 and 33 C., respectively. Grey-backed Finch-lark. In general behaviur the Grey-backed Finch-lark resembles Stark's Lark in many ways. It frages in the same manner as Stark's Lark, rests in the shade f small plants, and ccurs in large flcks like Stark's Lark. It differs mainly in nt perching n bushes r rcks, but staying cnstantly n the grund when nt flying. It is much harder t apprach than is Stark's Lark, and althugh it was actually mre numerus than the latter in sme areas, it was much mre difficult t cllect. Like Stark's Lark, the Grey-backed Finch-lark wandered in flcks frm place t place prir t breeding, accumulating in great numbers where green grass was grwing, and nesting when these grasses reached heights f 2 t 5 cm and beg~n t mature. In fact, the Grey-

33 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 16S backed Finch-lark wandered much mre freely than Stark's Lark befre and during the breeding seasn in late summer and autumn, and it was at this time that flcks turned up in the Inner Sand Dunes and far ut tward the cast n the Outer Gravel Flats. The flight sng display is less cnspicuus than that f Stark's Lark because the flight is nt usually s high - nly 15 t 3 m high - and the sng is weaker and less sustained. In flight the displaying bird circles abut slwly in undulating flight rather than hvering ver ne spt; and the sng ntes are sharp and tinkling rather than churring and chirping as in Stark's Lark. This sng flight typically ends in a vertical dive t the grund. I estimated ppulatin densities ffrm tw t five birds per acre in the best breeding areas. DISCUSSION REPRODUCTION AND MOULT Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). It is pssible t draw sme cnclusins abut hw the reprductive and mult cycles f larks in the Namib Desert are adaptive t the envirnment. The five lark species that are mst abundant in the central Namib shw tw general patterns f respnse t the envirnment with respect t reprductin and multing. On the ne hand Gray's Lark, the Spike-heeled Lark, and the Karr Lark are sedentary species that remain cnstantly in the Namib, and d nt wander far. These species include a large prprtin f insects and ther arthrpds in their diets - mre than 4 per cent in Gray's Lark, and mre than 6 per cent in the ther tw species. In them, with the pssible exceptin f the Karr Lark fr which data are lacking, gnadal activity and multing are nt strictly synchrnised thrughut their respective ppulatins. Thus Spike-heeled and Gray's Larks with enlarged testes r underging sme stage f mult were cllected at all times f year. On the ther hand Stark's Lark and the Grey-backed Finch-lark are very mbile, mving frm regin t regin in large flcks and gathering temprarily in places with favurable cnditins fr feeding and breeding. These birds cnsume primarily seeds - mre than 75 per cent in Stark's Lark and mre than 9 per cent in the Grey-backed Finch-lark. They shw a high degree f synchrny f gnadal develpment and multing within their respective ppulatins. Reprductin in all these species ccurred during r at the end f the summer-autumn rainy perid f the Inner Namib, and thus was timed t crrespnd with perids when insect fd was at a peak f abundance, wing t the new grass that came up fllwing the rains. Hwever, in the primarily insectivrus Spike-heeled Lark there was evidence that reprductin tk place at ther times f year as well; and in Gray's Lark, the fact that there was n clear-cut synchrny f multing and testicular develpment in the whle ppulatin suggests that reprductin prbably can ~ake place at ther times f year, given favurable envirnmental circumstances such as abundant insect fd and green grass. The data fr the Grey-backed Finch-lark and Stark's Lark give n suggestin at all that they bred, r were capable f breeding, at any time except during the summer and autumn rainy seasn. There-

34 166 ZOOLOGIC A AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). fre, I suggest that the fllwing relatinships exist between these species and their envirnments: The Spike-heeled Lark, specialising in a diet f small insects, is restricted t areas in the Namib with a relatively rich and perennial arthrpd fauna. Sme segment f the ppulatin is capable f reprducing at any time that lcal rainfall prduces r maintains favurable cnditins fr raising yung. Such rainfall may ccur at any time in the Namib, but is mst likely t ccur in summer and autumn at the latitude f Game Reserve N.3. Gray's Lark eats a large prprtin f seeds, and is widely distributed ver the Namib Platfrm, but must have abundant insects t9 feed its yung. Sme segment f the ppulatin is ready t breed at any time lcal rainfall prduces the right envirnmental cnditins, but since this was mst prevalent in summer and autumn during the time f this study, the majrity f birds shwed apprximately the same multing and reprductive timing. The Karr Lark prbably has the same multing and reprductive relatinship t the envirnment as Gray's Lark, althugh data t make a gd case either way are lacking. The Grey-backed Finch-lark and Stark's Lark specialise in a diet f seeds, but must have insects t raise their yung. At the same time they are highly mbile and travel lng distances in large flcks in search (s t speak) f favurable feeding and nesting cnditins, and mve int the Namib frm the semi-arid highlands t the east when these favurable cnditins ccur. Therefre, they are strngly influenced by the marked annual dry and rainy cycle f the central interir f Suth West Africa. They nest when grass grwth is sufficient t prvide nesting material and shelter frm the sun, when insect fd is abundant fr the nestlings, and when seed becmes abundant fr fledglings and adults. MOULTING. AND PLUMAGE Thus far I have treated multing and reprductin tgether, because in mst birds the time f multing is clsely assciated with the.sexual cycle. Because multing - the replacement f feathers - takes energy, its timing is sensitive t natural selectin (Amadn 1966), and it is apprpriate t ask hw the multing and plumages f the larks in the Namib are adaptive t their envirnment. The multing patterns f larks are nt well knwn, but Vaurie (1951) distinguished three main patterns f mult amng Asiatic representatives f the Alaudidae. One is a single cmplete annual mult ccurring as pst-breeding r pst-juvenal mult, the first winter bird being indistinguishable frm the adult. This pattern ccurs in mst f the larks that Vaurie studied. A secnd pattern invlves a single annual mult, but with nly a partial pst-juvenal mult in which the bdy feathers are changed. The third pattern cnsists f a cmplete pst-breeding and pst-juvenal mult tgether with a mre r less restricted prebreeding bdy mult. Sme f the larks in the Namib present still ther multing schemes. The Large-billed Sabta Lark pssibly has ne cmplete mult a year, withut a partial mult, but data are

35 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 167 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). meagre. The Karr Lark als appears t have a single cmplete mult a year-a pstbreeding and pst-juvenal mult. The Spike-heeled Lark seems t have a pst-breeding and pst-juvenal bdy mult that merges int a cmplete mult. Gray's Lark definitely shws a rapid pst-breeding and pst-juvenal bdy mult that is fllwed by a prlnged cmplete mult. Bth the Grey-backed Finch-lark and Stark's Lark have intense pst-breeding and pst-juvenal bdy mults fllwed after an interval by a prlnged cmplete mult. It is evident that the rapid pst-breeding and pst-juvenal bdy mults that ccur in Gray's Lark, Stark's Lark, and the Grey-backed Finch-lark are timed t ccur when fd in the frm f insects and seeds is still abun<lant at the end f the breeding perid. The subsequent prlnged cmplete mult during the dry perid is sustained mainly by the seeds prduced in the previus rainy perid. It is hard t assess the adaptive significance f having tw changes f bdy plumage, especially when there is n seasnal clur variatin in any f these birds. It may be that tw bdy mults are nt adaptive in these species, but that this multing pattern has been inherited frm ancestral species in which it did at ne time have stne adaptive functin, and the descendants have been unable t eliminate r suppress the pstbreeding bdy mult withut assciated mal-adaptive side-efi'ects. In this case they have dne the next best thing in advancing and accelerating ne bdy mult s that it cmes within a time f plenty. Or n the ther hand, the extra bdy mult may be adaptive; and a pssible way that it culd be s is in replacing excessively wrn and faded bdy plumage that results frm the cnstant expsure t wind, sand, and intense sunshine in the desert. It is a fact that even with the tw changes f bdy plumage a year, feathers in these larks becme quite ragged and faded, and I am inclined t favur this secnd alternative. ENVIRONMENTAL STIMULI FOR BREEDING The breeding f the larks in the Namib clearly cincided with the seasnal rainfall that brught an abundance f succulent fds. This is true generally f birds in arid and semi-arid regins f Africa (Mreau 195) and Australia (Immelmann 1963). A pertinent questin is what specific envirnmental factrs stimulate the birds t cmmence breeding. The data prvide n answer, but they d suggest sme pssibilities. It has been shwn that in nrth temperate regins, lengthening daily phtperids in the spring stimulate gnadal recrudescence, nrthward migratin, and breeding in numerus species; and the literature n phtperidism in birds is extensive (fr a recent review, see Lfts and Murtn 1968). Hwever, in equatrial regins where phtperid varies little during the year, and in arid and semi-arid regins where rainfall and favurable breeding cnditins ccur irregularly, phtperid has little influence in timing events in the nrmal reprductive cycles f birds. Instead, respnses t varius envirnmental stimuli cnnected with the circumstances favurable t breeding as well as endgenus rhythms have been pstulated as timing mechanisms fr breeding (Marshall 196; Immelmann 1963). In the case f the larks f the central Namib, phtperid is prbably nt imprtant as a timing factr, and sme ther exgenus factrs r endgenus factrs r bth are respnsible fr the timing f the breeding cy'cle.

36 168 ZOOLOGICA AFRICANA VOL 6 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). In the Spike-heeled Lark, which shws what appears t be a tendency t breed any time that envirnmental circumstances are favurable, the stimulus might be falling rain itself. Imrnelmann (1963) bserved birds in the arid interir f Australia in which the falling rain appeared t trigger latent breeding tendencies s that the birds were well alng with breeding even befre the increased fd supply and nesting material resulting frm the subsequent plant grwth had appeared. Marshall and Disney, in experiments with Que/e que/e in East Africa, fund that watering the cages with sprinklers apparently had the effect f accelerating the reprductive cycle (Marshall and Disney 1957; Marshall 1961, p. 323). I think it likely that green grass and an abundance f insect fd are als imprtant stimuli fr breeding in the Spike-heeled Lark. These same factrs may be at wrk in the timing f reprductin in Gray's Lark. The ne nesting that had ccurred befre the grasses resulting frm rainfall had grwn enugh t prvide nest-building material (Fig. 1) may have been stimulated partly by the rain itself. The fact that rainfall n 3 September which caused sme grwth f grass did nt result in any detected breeding activity in this species suggests that an endgenus physilgical prcess may have been at wrk in the bulk f the ppulatin which made mst individuals unrespnsive t the envirnmental stimuli at that early date when the majrity f birds were underging the cmplete mult. Stark's Lark and the Grey-backed Finch-lark may be influenced in timing f breeding by bth endgenus factrs and by exgenus factrs. Birds f bth species captured in August and Nvember, 1965, kept t the same mult schedules as wild birds thrugh June, 1966, and several f the captive Stark's Larks sang vigrusly in their cages at the time wild birds were breeding, althugh all f them were kept in individual cages inside a building. In the wild, hwever, actual nesting did nt ccur until the grasses reached maturity and began t flwer, at which time insects were at a peak f abundance, and nesting material in the frm f dry grass stems and feathery seed awns became available. Experiments are needed t clarify the rle f such envirnmental factrs as stimuli fr breeding; and Stark's Lark and the Grey-backed Finch-lark might make gd subjects wing t the ease with which they can be maintained in captivity (Willughby 1968). BEHAVIOURAL ADAPTATIONS TO DESERT CONDITIONS One f the first things a bird can d t amelirate the effects f its envirnment is t find shelter. In the Namib, where water is at a premium, it is advantageus fr a bird t avid excessive heat transfer frm the envirnment t its bdy s that the expenditure f water by evapratin t eliminate the excess bdy heat is kept t a minimum. The birds d this by keeping in the shade f bushes, grass tufts, stnes, and muths f rdent burrws during the httest times. In additin t using these tactics, Gray's Lark and Stark's Lark avid cntact with the ht grund by standing abve it n stnes r bits f wdy vegetatin, and althugh they are still expsed t the intense sunshine, the relatively cl sea breeze that regularly blws acrss the Namib and seldm has a temperature abve 38 C dubtlessly

37 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 169 carries away a gd part f their excess bdy heat by cnvectin as they stand facing int the air flw with wings held away frm the sides f the bdy. Trst (1968) has bserved that hrned larks, Eremphila alpestris, in the Mjave Desert f Califrnia rarely alight n ht grund but perch in the wind n tp f creste bushes, r stay in the shade beneath the bushes. The larks build their nests n the surface f the grund, usually next t sme prtective gbject such as a stne r tuft f grass s that the nest and its cntents receive shade during the httest time f day, frm abut nn t 16. hr. ~;;J In the Spike-heeled, Gray's, and Stark's Larks, and prbably als in the Grey-backed Finch-lark, nestlings leave the nest befre they can fly, and run after the adults t be fed. This habit may have the func~n f dispersing the yung away frm the nest as sn as pssible t reduce their vulnerability t predatin by such animals as snakes, mngses and jackals. SEXUAL SIZE DIMORPHISM Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). The significance f the sexual dimrphism in bdy size bserved in sme f the larks in the Namib is an intriguing questin. Mst species f birds d shw sme difference in size between the sexes, and either male r female may be larger (Amadn 1959). Amng the larks in the central Namib, the male is the larger sex when there is clear sexual size dimrphism. The mst interesting aspect f this dimrphism is that it is expressed t. different degrees amng the different species, and seems t be crrelated with fd habits. Thus the Spikeheeled Lark, the mst insectivrus ne, shws a marked dimrphism, with males averaging 24 per cent heavier than females; and Gray's Lark, feeding n nearly equal prprtins f seeds and insects, shws a less marked size dimrphism, with males averaging 12 per cent heavier; and Stark's Lark and the Grey-backed Finch-lark, specialising in a diet f small seeds, shw n size dimrphism. These weight differences are reflected genera1ly in linea: measurements, but there is n striking difference between the sexes in prprtinate size f structures such as beak and feet. I have mitted the Karr Lark frm cnsideratin here because f the lack f data, but the small sample I have indicates a mderate degree f size dimrphism, as in Gray's Lark. The questin f the eclgical significance f sexual size dimrphism has been discussed recently fr varius grups f birds by several investigatrs (Rand 1952; Amadn 1959; Cade 196; Selander 1966; Strer 1966); and the bilgical functins f sexual dimrphism in size are n dubt nearly as diverse as the species and families in which it ccurs. In the larks f the Namib, the fact that this dimrphism seems t be crrelated with fd habits gertainly suggests the hypthesis that it is cnnected in sme way with differential utilisatin f feeding niches. The hypthesis is that the sexual dimrphism in the insectivrus larks reduces intra-specific cmpetitin fr fd because the larger males tend t take larger insects, which results in an effective widening f the size range f prey items that can be explited by the

38 17 ZOOLOGICA AFRICANA VOL 6 whle ppulatin. The nes that are granivrus, n the ther hand, and eat unifrmly small seeds such as thse f the grasses f the genus Stipagrstis, are nt affected by sexual differences in bdy size in regards the size range f fd items made available t them. Unfrtunately, the bilgy f the larks f arid regins is t little knwn yet t shw whether r nt this relatinship between fd habits and size dimrphism hlds generally. I d have anther bit f evidence frm the Spike-heeled Lark f the Namib that tends t supprt the hypthesis that the tw sexes explit different feeding sub-niches. When J prcessed the specimens, I made brief ntes n the kinds f insects that predminated in the stmachs. In 16 stmachs frm males, 13 (81 %) included remains f beetles (mstly small tenebrinids abut five t ten mm lng). Ants, harvester termites, r small rthpterans predminated in the three ther stmachs. In the stmachs f 12 females, hwever, nly three (25%) cntained beetle remains, and the rest held ants r harvester termites r bth. This suggests that the males take a greater prprtin f small, hard beetles, and females take a greater prprtin f ants and sft-bdied termites. I saw n difference between males and females f the less dimrphic Gray's Lark with respect t the kinds and sizes f insects taken by them. WATER REQUIREMENTS OF NAMIB LARKS Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). The fact that nne f the species f larks that live in the central Namib depends n drinking water indicates that the birds must have efficient physilgical mechanisms fr cnserving bdy water, which permit them t subsist n the prefrmed and metablic water frm their fd. This is mst remarkable in thse species that subsist primarily n seeds, Stark's Lark and the Grey-backed Finch-lark, because air-dried seeds cntain nly abut ten per cent water. A study f the water metablism f these tw species in the labratry (Willughby 1968) has shwn that they have such lw evaprative, urinary and faecal water lsses that under mderate cnditins f humidity and ambient temperature they can maintain water balance when cnsuming nthing but air-dried millet seed. In the wild, an ccasinal insect, bit f succulent vegetatin, r an infrequent visit t a watering place wuld prvide enugh water t cmpensate fr increased lsses during ht, dry weather. Thse species, such as Gray's Lark, that include a large prprtin f insects in their diet, are in a mre favurable psitin t maintain water balance in the desert, wing t the increased amunt f free water in the fd. The ability f the larks f the Namib t live withut drinking, whatever the physilgical and behaviural mechanisms permitting it may be, is ne f the prime factrs fr the success with which these species ccupy an extremely dry and barren habitat. SUCCESS OF ALAUDIDAE IN DESERTS There appears t be n ther passerine family that equals the Alaudidae in successful desert ccupancy, as regards the numbers f species filling a variety f desert niches. It is tempting t speculate upn the basis fr the adaptive success f the larks in the desert envirnment.

39 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 171 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). As a grup the larks have specialised in living n pen grund in dry situatins. Within the brad limits f this specialisatin, hwever, they have remained unspecialised (except fr individual species) with respect t fd and feeding, multing, reprductin, and behaviur. It is difficult t discern any features unique amng the Alaudidae ther than the mrphlgical feature that gives the family its cherence, namely that the back f the tarsus is runded and scutellated instead f keeled and encased in an undivided sheath (Stresemann 1934, p. 85). These traits f the Alaudidae suggest that this family has prbably experienced a lng histry f ccupancy f arid and semi-arid envirnments, and has experienced adaptive radiatin t explit varius feeding, habitat, and climatic niches in these regins. Winterbttm (1962) cncluded that despite cnflicting taxnmic evidence fr the antiquity f the Alaudidae, the fact that the family is rather cherent tends t supprt the ntin f its antiquity; but the indistinctness f many f the suppsed genera f larks and the high ratis f species per genus and subspecies per species suggest recent radiatin. It may be that the basis fr the success f the larks in arid regins is their ability t tlerate rapidly and severely changing envirnmental cnditins assciated with seasnal drught and irregular rainfall by virtue f their lack f rigid specialisatin in fd, reprductive habits, multing and plumage, seasnal migratin and scial behaviur. A recent study by Trst (1968) f physilgical respnses f several Nrth American ppulatins f the Hrned Lark, Eremphila aipestris, t stressful envirnmental cnditins indicates that this ne species shws cnsiderable physilgical variability between different ppulatins, and is physilgically tlerant f wide ranges in envirnmental temperature, barmetric pressure, humidity, and availability f water. This tends t indicate that the Alaudidae have remained physilgically adaptable t changing cnditins als. It appears, therefre, that the retentin f evlutinary plasticity may be adaptive t life in an pen arid envirnment. ACKNOWLEDGMENTS I am greatly indebted t Dr. Tm J. Cade fr his guidance and supprt during this study. A grant frm the U.S. Public Health Service (Envirnmental Health) ES 8 t Dr. Cade financed the wrk. I must give special thanks t Dr. V. FitzSimns and Dr. C. Kch fr making the facilities f the Namib Desert Research Statin available t me. The Suth West Africa Administratin kindly granted permissin t carry ut the study within Game Reserve N.3. I want t thank Messrs. Rdney Baxter and Bernard Viljen f the Suth West Africa Department f Nature Cnservatin fr their helpful advice and infrmatin abut watering places and wildlife in the reserve. Identificatin f plant specimens was furnished by the Btanical Research Institute, Department f Agricultural Technical Services, Pretria. I thank L. Greenwald, O. P. M. Przesky, and D. Gde fr their help in cllecting specimens. D. Gde als helped in the preparatin f Figure 2. Dr. Cade, Dr. W. R. Dawsn, and Dr. C. Trst helpfully criticised drafts f the manuscript. My wife, Paula, helped with the preparatin f the manuscript.

40 J72 ZOOLOGICA AFRICANA VOL 6 This paper is based n part f a dissertatin submitted in partial fulfilment f the requirements f the degree f Dctr f Philsphy in Zlgy in the Graduate Schl f Syracuse University, June, SUMMARY Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). The central Namib Desert is very barren, with extremely lw precipitatin. A study f six indigenus species f larks, family Alaudidae, gives a basic understanding f the ways by which these birds are adapted t their envirnment. Table 6 summarises the majr findings n the bilgy f the larks f the central Namib. Reprductin in all species cincided with the end f the summer-autumn rainy perid f the eastern part f the Namib, and came at a time when insects, maturing grass seed, and nest-building materials were mst abundant. Althugh tw seed-eating species, Sp;zcrys stark; and Erempter;x vert;cal;s, smetimes came t watering places by the hundreds t drink, these birds represented nly a small prprtin f the ppulatin in the regin, and nne f the lark species shwed dependence n surces f drinking water. Ammmanes grayi, S. stark; and E. vertical;s have a rapid pst-breeding bdy mult fllwed later by a prlnged cmplete mult. The pst-breeding bdy mult f these species ccurred when insect fd and grass seed were still abundant fllwing breeding. It is suggested that the change f bdy plumage twice a year might help t cunteract the attritin and fading f the plumage wing t cnstant expsure t wind, sand, and sun in the Namib. The larks in the Namib minimise heat influx frm the envirnment and thus minimise their need t evaprate bdy water fr cling by squatting in the shade f stnes, rcks, tufts f grass and bushes during the heat f midday. Certhilauda alb/asciata and A. gray; als frage in the muths f rdent burrws, and thus avid the ht surface f the surrunding grund. A. gray; and S. stark; frequently perch abve the surface f the grund n stnes r shrubs and face int the relatively cl sea breeze with the flded wings held away frm the sides f the bdy, thus aviding cntact with the ht grund and facilitating lss f bdy heat by cnvectin. The fact that the species that are mst insectivrus shw the greatest difference in size between the sexes (males are larger) while granivrus species shw n sexual size dimrphism supprts the hypthesis that the size dimrphism functins t reduce intra-specific cmpetitin fr fd by widening the range f size r type f prey items that the species can explit. It is suggested that the basis fr the adaptive success f the Alaudidae in desert envirnments is the absence f rigid specialisatin in fd, reprductive habits, multing and plumage, behaviur and physilgy, making the larks better able t cpe with rapidly and severely changing envirnmental cnditins assciated with drught and irregular rainfall.

41 1971 WILLOUGHBY: BIOLOGY OF NAMIB LARKS 173 TABLE 6 SUMMARY OF BIOLOGY AND WATER RELATIONS OF LARKS IN THE CENTRAL NAMIB DESERT, 1965 AND 1966 Species Occurrence l (habitats and mvements) Fd Timing / reprductin Mu/ting Drinking habits Primary surces / water Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). Large-billed Sabta Lark Mira/ra naevia Karr Lark Certhilauda a/bescens Spike-heeled Lark Certhilauda a/b/asciata Gray's Lark Ammmanes grayi Stark's Lark Spizcrys stark; Grey-backed Finch-lark Erempterix vertica/is Uncmmn, in pen bush n eastern edge f Namib Cmmn, sedentary, n dunes with tall perennial grasses Cmmn, sedentary, n grassy Inner Gravel Flats with very sparse scrub Cmmn, restricted t Namib, primarily n grassy Inner Gravel Flats Cmmn, at times very abundant, wandering, n grassy Inner Gravel Flats Cmmn, at times very abundant, wandering, n grassy Inner Gravel Flats Small seeds and insects abut equal prprtins 32% grass seed,68% insects 16% seeds, 84% insects 56% small seeds, 43% insects, 1% green vegetatin 77% small seeds, 19% insects, 4% green vegetatin 91% small seeds, 8% insects, 1% green vegetatin Prbably during r at end f summer-autumn rainy perid Octber t May, cinciding with rainy perid Breeding ccurred at all times f year Cmplete pstbreeding mult; n data n partial mult Cmplete pstbreeding mult ; n definite data n partial mult Mult ccurred all year; individuals have pstbreeding bdy mult that merges with cmplete mult Occurred all year, mst individuals in lse synchrny, pstbreeding bdy mult fllwed by cmplete mult Des nt drink Des nt drink Des nt drink Late summer and autumn, cinciding with rainy perid, but withut a definite seasnal gnadal cycle Late summer Synchrnised May drink in Des nt drink Metablic water and insects Metablic water and insects Metablic water and insects Metablic water and insects Mainly metablic water; and autumn, seasnal mult numbers durat end f rainy with rapid pst- ring dry insects and perid, with a breeding bdy weather when green marked sea- mult fllwed water is availsnal gnadal by a mre pr- able, but nt seasnally vegetatin cycle lnged cmplete dependent n mult drinking water Late summer Synchrnised May drink in Mainly metaseasnal- mult numbers buc water, and autumn, at end f with rapid pst- whenever perhaps ccabreeding bdy water is avail- sinally rainy perid, with marked mult fllwed able, but nt drinking seasnal by a mre dependent n water gnadal cycle prlnged cm- drinking plete mult water 1 "uncmmn" = species bserved n fewer than % f visits t suitable habitat. "cmmn" = species bserved n mre than % f visits t suitable habitat.

42 174 ZOOLOGICA AFRICANA VOL 6 REFERENCES Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). AMADON, D The significance f sexual differences in size amng birds. Prc. Am. phil. Sc. 13: AMADON, D Avian plumages and mults. Cndr 68: ANDERSSON, C Ntes n the Birds f Damara Land and the Adjacent Cuntries f Suth-west Africa. (Arranged and edited by J. H. Gurney.) Jhn van Vrst, Lndn. 394 p. BIANCHI, B Dbersicht der Frmen des Genus Ammnanes Cab. der Fam. Alaudidae. J. Om, Lpz. 53: CADE, T Eclgy f the peregrine and gyrfalcn ppulatins in Alaska. Univ. Calif. PubIs. Zl. 63: GIESS, w Sme ntes n the vegetatin f the Namib Desert with a list f plants cllected in the area visited by the Carp-Transvaal Museum Expeditin during May, Cimbebasia 2 (August): (Als Sci. Pap. Namib Des. Res. Stn., N.3.) HARRISON, C. 1. O The validity f sme genera f larks (Alaudidae). Ibis 18: HOESCH, w Nest und Gelege der WUstenlerche Ammmanes grayi. J. Om, Lpz. 99: HOESCH, w. and G. NIETHAMMER Die Vgelwelt Deutsch-SUdwestafrikas. J. Orn, Lpz. 88 Supplement. 44 p. IMMELMANN, K Tierische Jahresperidik in klgischer Sicht. Zl. Jb. Syst. 91 : KOCH, C The Tenebrinidae f suthern Africa XXXI. Cmprehensive ntes n the tenebrinid fauna f the Namib Desert. Ann. Transv. Mus. 24: LAWSON, w The races f the karr lark Certhilauda albescens (Lafresnaye). Ostrich 32: LOFTS, B. and R. K. MURTON Phtperidic and physilgical adaptatins regulating avian breeding cycles and their eclgical significance. J. Zl., Lnd. 155: LOGAN, R. F The Central Namib Desert Suth West Africa. Publicatin 758, Natinal Academy f Sciences-Natinal Research Cuncil, Washingtn D.C. (Freign Field Research Prgram, Office f Naval Research, Reprt N.9). 162 p. LOWMAN, P. D. Jr The earth frm rbit. Natn. gegr. Mag. 13: MACDONALD, 1. D. 1953a. Taxnmy f the karr and red-back larks f western Suth Africa. Bull. Br. Mus. nat.' Hist. (Zlgy) I: MACDONAW, 1. D. 1953b. Sme aspects f variatin in the spike-heel lark (Certhilauda albfasciata). Prc. zl. Sc., Lnd. 122: MACLEAN, G. L Nests f pallid karr lark in Namib Desert. Ostrich 28: 124. MACLEAN, G. L Suth African lark genera. Cimbebasiil, Ser. A, I: MACKWORTH-PRAED, c. w. and c. H. B. GRANT Birds f the Suthern Third f Africa, vl. I. Lngmans, Green and C., Lndn. 688 p. MARSHALL, A Annual peridicity in the migratin and reprductin f birds. Cld Spring Harb. Symp. quant. BiI. 25, Bilgical Clcks: MARSHALL, A Breeding seasns and migratin, p in Marshall, A. J. (ed.) Bilgy and Cmparative Physilgy f Birds, vl. 2. Academic Press, New Yrk.

43 1971 WILLauGHBY: BIOLOGY OF NAMIB LARKS 175 Reprduced by Sabinet Gateway under licence granted by the Publisher (dated 21). MARSHALL, A. J. and H. J. DE S. DISNEY Experimental inductin f the breeding seasn in a xerphilus bird. Nature, Lnd. 18: MCLACHLAN, G. R. and R. LIVERSIDGE Rberts Birds f Suth Africa, Revised Editin. Central News Agency, Suth Africa. 54 p. MElNERTZHAGEN, R Review f the Alaudidae. Prc. zl. Sc. Lnd. 121: MOREAU, R. E The breeding seasns f African birds.-1. Land birds. Ibis 92: PROZESKY,. P. M Ornithlgical results f the Transvaal Museum Namib Expeditin, May, 1959, and the subsequent trip t Sandwich Harbur during January, 196. Ostrich 34: RAND, A. L Secndary sexual characters and eclgical cmpetitin. Fieldiana, Zl. 34: SCHULZE, B. R Climate f Suth Africa, Part 8: General Survey. Publicatins f the Weather Bureau, Republic f Suth Africa, W.B. 28. The Gvernment Printer and Weather Bureau, Pretria. 33 p. SELANDER, R. K Sexual dimrphism and differential niche utilizatin in birds. Cndr 68: STORER, R. w Sexual dimrphism and fd habits in three Nrth American accipiters. Auk 83: STRESEMANN, E Handhuch der Zlgie, vl. 7, part 2, Saurpsida: Aves. Walter de Gruyter, Berlin and Leipzig. 899 p. TROST, C. H Adaptatins f hrned larks t stressful envirnments. Ph.D. dissertatin, University f Califrnia, Ls Angeles. VAN TYNE, 1. and A. J. BERGER Fundamentals f Ornithlgy. Jhn Wiley & Sns, New Yrk. 624 p. VAURIE, c A study f asiatic larks. Bull. Am. Mus. nat. Hist. 97: VINCENT, A. w On the breeding habits f sme African birds. Ibis, 88: WALTER, H Die klgischen Verhliltnisse in der Namib-Nebelwiiste (Siidwestafrika) unter Auswertung der Aufzeichnungen des Dr. G. Bss (Swakpmund). Jb. wiss. Bt. 84: WEATHER BUREAU, DEPARTMENT OF TRANSPORT, REPUBLIC OF SOUTH AFRICA Climate f Suth Africa: Part 7: Average Mnthly Rainfa// Suth West Africa, W.B. 27. Pretria. 28 p. WEATHER BUREAU, DEPARTMENT OF TRANSPORT, REPUBLIC OF SOUTH AFRICA Mnthly Weather Reprt, fr the mnths January thrugh December (separate issue fr each mnth). Pretria. WEATHER BUREAU, DEPARTMENT OF TRANSPORT, REPUBLIC OF SOUTH AFRICA Mnthly Weather Reprt, fr the mnths January thrugh August (separate issue fr each mnth). Pretria. WELLINGTON, J. H Suthern Africa, a Gegraphical Study. Cambridge University Press. 2 vl. WHITE, C. M. N Ntes n African larks-part IV. BuJ/. Br. Orn. Club 77: WILLOUGHBY, E. J Water ecnmy f the Stark's lark and grey-backed finch-lark frm the Namib Desert f Suth West Africa. Cmp Bichem. Physil. 27: