<<

South African Archaeological Society

ANTIQUITY OF STONE-WALLED TIDAL FISH TRAPS ON THE COAST, SOUTH Author(s): PHILLIP HINE, JUDITH SEALY, DAVID HALKETT and TIMOTHY HART Source: The South African Archaeological Bulletin, Vol. 65, No. 191 (JUNE 2010), pp. 35-44 Published by: South African Archaeological Society Stable URL: http://www.jstor.org/stable/40985509 Accessed: 12-01-2016 09:10 UTC

REFERENCES Linked references are available on JSTOR for this article: http://www.jstor.org/stable/40985509?seq=1&cid=pdf-reference#references_tab_contents

You may need to log in to JSTOR to access the linked references.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at http://www.jstor.org/page/ info/about/policies/terms.jsp

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

South African Archaeological Society is collaborating with JSTOR to digitize, preserve and extend access to The South African Archaeological Bulletin. http://www.jstor.org

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions SouthAfrican Archaeological Bulletin 65 (191):35-44, 2010 35

ResearchArticle ANTIQUITY OF STONE-WALLEDTIDAL FISH TRAPS ON THE CAPE COAST, PHILLIP HINE,JUDITH SEALY, DAVID HALKETT& TIMOTHY HART DepartmentofArchaeology, University ofCape Town, Private BagX3, Rondebosch, 7701, South Africa E-mail:[email protected]; [email protected]; [email protected]; [email protected] (ReceivedMarch 2009. Revised March 2010)

ABSTRACT INTRODUCTION Stone-walledtidal fish Thispaper attempts to answera long-standingquestion in South traps(hereafter referred to as fish are a well knownfeature of the Africanarchaeology: the age ofstone-walled tidal fish traps generally traps) WesternCape coast, the Indian believedto date back to pre-colonial times. Since the stone walls cannot especiallyalong Ocean coastline(Fig. 1). These structuresconsist of bedirectly dated, we sought datable fish bone in nearbyarchaeological stone-walledenclosures which are sub- at sites. Four open shell middensat PaapkuilFontein, near Cape merged hightide, enabling fish to swim into them. The catch is then Agulhas,were excavatedand analysedand the contentsof two corralledwhen the water recedes, and canbe collected previouslyexcavated middens at StillBay were studied. Both areas are by nettingor spearing.Although some trapsremain in use renownedfor their numerous fish traps, but lack detailed archaeological today,they are generallybelieved to be of considerable studies.The middens yielded very little, if any fish bone, so areprobably antiquity.Goodwin (1946), on thebasis of his work at Oakhurst unrelatedto thetraps. There is, bycontrast, a greatdeal ofarchival RockShelter, first proposed that they might have been used by evidencefor the building and use of stonefish trapsby historical pre-colonialKhoe-San people. Avery(1975, 1976) proposed communities,with traps repeatedly built and dismantledin thelate datesof c. 3000-2000BR whilePoggenpoel (1996) suggested 19thand 20th centuries. Given the lack of any direct evidence in Later thatthey could have been in use as earlyas themid-Holocene. StoneAge sites,a pre-colonialage forthe practice of fishing with A widely-distributedposter published by the Directorateof stone-walledtidal fish traps can no longerbe entertained. Marineand CoastalManagement (a sectionof the Department ofEnvironmental Affairs and Tourism), and displayedin many Keywords:fish trap, weir, Later Stone Age, shell middens. localitiesalong the Cape coast,refers to 'ancient tidal fish traps7

I . 15 .Cape Barracouta ' 2. ló.Hotnotbaai ' 3.DangerPoint 17.Haaibaai ) 4.Haaiklip 18.Jongensfontein C 5.PearlyBeach 19.StillBay Harbour/Morrispoint/Noordkapperpunt' ó.DieDam 2O.Rietvlei c^iowJ

7.'fyverbaaisites 2 1.Dantes Point F«&nx* 8. SuiderstrandI & II 22.Riesiesbaan :*?_ ^ ; Aretenlar*edbelow 9. Agulhas 23.Meelefontein 10. I & II 24.Ystervarkpunt I, II, III I 1. 25.Between 12.Ryspunt 26.Skool se Bank 1 series 27.Platbank/NeliesGifTBorrelfontein 3. £ 14.Breede 28.Kannonstrand 19*a 24 26 3

I 27 ¿4« ,<• jT »' -18 "1 21 25 12* /r^ v<>~' •Sitesrecorded | byAvery (1975) pN-pr m N A 3* I T ^^^»^v ''+ Sitesrecorded by Goodwin (1946) 5* I |'^W A •• i i 0 20 km

FIG. 1. Distributionoffish traps along the south coast of the . Map adaptedfrom Kemp (2006).

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions 36 South AfricanArchaeological Bulletin 65 (191): 35-44,2010 and theiruse by Khoe-Sanpeople. There is, however,little theminimum period for sun- and wind-dryingof fish in the evidenceon which to base such claims. The primary aim of this southernCape in good weatherconditions was 4-5 days paperis to evaluatethe antiquity of fish traps in theWestern (Tothill1899 in litt.).The entireprocess of makingbokkoms, Cape. whichinvolves salting and thendrying, can takeup to two Previousresearch on fishtraps has beensporadic and has weeks(Anon. 2005). Any processing probably took place close focusedprimarily on mappingtheir distribution on theland- tothe catch site, in order to prevent spoilage. We would, there- scape,recording fish catches and assessingliving invertebrate fore,expect at least some fish bone to be presentin middens as a populationstherein (Goodwin 1946; Avery 1975, 1976; Gribble resultof meals consumedduring the processingperiod. If 2005;Kemp 2006). By contrast, our focus here is purelyon the processinginvolved removal of heads or otherparts, then antiquityof the traps. Since the stone walls themselves cannot evidenceshould be visiblein thearchaeological record. be dated,we excavatedseveral shell middens in close proximity Mulletbones are generallymore fragile and less likelyto to fishtraps. The aim was to investigatethe depositional survivein archaeological sites than those of larger bodied fish. historyof the middens, with particular emphasis on theidenti- Coastal shell middensin the winterrainfall area of South ficationand verticaldistribution of fishremains. Recognizing Africa,however, offer extremely good conditionsfor bone and datingfish species more likely to have been caught in fish preservation,and bonesof small fish species are preserved at trapsthan by othermeans should give an indicationof the a numberof sites (Poggenpoel1996; van Niekerk2004). antiquityof the traps. Consideringthe large quantitiesof mullettrapped today, Fishare well represented in coastal Holocene sequences of taphonomicprocesses ought not to have entirely removed this theWestern Cape, yet we knowrelatively little about their role speciesfrom fish-trapped archaeological assemblages. A recent in prehistoriceconomies (Poggenpoel 1996; Inskeep 2001; van study(Nagaoka 2005) demonstrated that the use of3 mmmesh Niekerk2004). Material remains relating to thetechnology of screensis adequateto ensure recovery of mullet remains from fishingare not well represented in southernAfrican archaeo- archaeologicaldeposits. All controlled archaeological excavations logicalassemblages. Stone 'sinkers' and fishgorges are known alongthe Western Cape coast,at leastduring the last 40 years, fromonly a fewsites and onlysome time periods (Louw 1960; haveused sieveswith 3 mmor smaller mesh sizes. This means Deacon 1970;Parkington 1977; Poggenpoel& Robertshaw thatarchaeological assemblages recovered during this time can 1981;Schweitzer & Wilson1982; Inskeep 1987; Orton & Halkett be usedto assess the importance ofmullet in the f aunal remains. 2007),and their use in the way the names imply is by no means certain.If coastalfish traps do, in fact,constitute a fishing PREVIOUS RESEARCH methodof considerable time-depth, there are wider implications The firstsystematic investigation into fishtraps was forour understanding of mid- to lateHolocene lifeways. The conductedby Goodwin (1946) in his paper "Prehistoric fishing labourrequired to buildand maintainthese traps, and the methodsin SouthAfrica". The stimulus for this work lay in his abilityto harvestlarge quantities of food would undoubtedly excavationsat OakhurstRock Shelter (Goodwin 1938), located havetied people to particular localities and influenced settlement about14 km inland from the coast. He noteda markedincrease patterns.From the mid-Holocene,hunter-gatherer societies inthe frequency of vertebrate fish remains in the mid- and late wereundergoing fundamental social and economic restructur- Holocene, compared with older layers. This suggested that the ing,viz. delayed return systems in theform of storage of plant inhabitantshad thetechnology to catchfish regularly and in foodsin the south eastern Cape (Deacon 1976; Hall 1990), possible quantityat thattime. Since no artefactswere foundthat processingand preservation ofshellfish along parts of the west seemed likelyto have been used for fishing,Goodwin coast (Jerardino1996), complex ritual behaviour (Hall & suggestedthe possible use of fishtraps, and he further Binneman1987; Hall 1990, 2000), and increased sedentism and proposedthat the best way to investigatethis hypothesis territoriality(Sealy 2006). Populations were growing and there would be througharchaeological excavation of nearbyshell appearsto have been pressureon foodresources. Were fish middens.Unfortunately, there is no detailedreport on the trapsfirst built and used at this time? Another possibility is that Oakhurstfish remains, so we do nothave speciesidentifica- theyare linked with the appearance of Khoe-San herders in the tionsand abundances.However, increased reliance on fish WesternCape after2000 BP. duringthe mid-Holoceneis consistentwith the pictureof Fishtraps are principally geared towards the exploitation hunter-gatherergroups widening their dietary breadth and of shoalingspecies. The speciesmost commonly caught are emphasizingsmall package food items (Hall 1990).Inskeep haardersor mulletbelonging to theMugilidae family (espe- (1987)reported a substantialincrease in thequantities of fish ciallyLiza richardsonii,the southernmullet and Mugil cephalus, recoveredfrom post-Wilton levels (i.e. those post-dating theflathead mullet), which favour inshore shallows and estu- 3300BP) at NelsonBay . ariesand are rarelycaught with hook and line (van der Eist GrahamAvery (1975, 1976) studiedfish traps between 1993).Recent research has shownthat Liza richardsoniimay Kleinmondand .As in Goodwin'searlier work, compriseup to100% of catches in fishtraps (Kemp 2006), and thefocus was on locationand mapping.Avery also provided of the up to 8000mullet have been reportedfrom a singletrapping importantinformation on theoperation and function event(Haddad 2003). Other species are also taken,most nota- traps,including statistics from local informants on thespecies wasthen known about bly Dichistiuscapensis (galjoen), Sparodondurbanensis (white andnumbers of fish caught. Using what of musselcracker),Sarpa salpa (strepie) and Pomatomussaltatrix past sea-levels,Avery suggested a likelyage forfish traps (elf)(Avery 1975; Kemp 2006). A widerange of species may be 3000-2000years, after sea levelhad stabilizedat approximately caughtin fish traps, and van Niekerk (2004) has suggested that itspresent position. thisdiversity should be reflectedin archaeologicalassem- Thirtyyears later, the SouthAfrican Heritage Resources ofUnderwater blages.We expect,however, that the use offish traps should Agency(SAHRA), as partof the National Survey and offish yieldlarge quantities oí Mugilidaespp. It is possiblethat not all Heritage,undertook extensive mapping surveying fishwere consumedat the trapsite. Some may have been trapsbetween False Bay and Mossel Bay. One aim was to locations preservedfor transport to otherlocations (as in themodern- producehigh-quality digital orthophotos snowing the day dryingof mullet, known locally as bokkoms).Historically, offish traps, and to verifythem by meansof ground surveys

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions SouthAfrican Archaeological Bulletin 65 (191):35-^4, 2010 37

P6O #P5 N ^' *P* OP10 A OP2 ^ ' P3 ' OP9 ' °OP8 • Vl °P1M ' ' P11 #P7

Vywerbaai ' Rasperpunt

i «- i Q^") HotagterklipX^^

Struisbaai^^^J ' -JNlSuiderstrand .^^^^^-^ '. ^ ( j j

RasperpuñT^-. ^ V~TT '___^__^A Cape Agulhas * of Southernmostpoint Africa g ^q 200m

FIG. 2. Locationsof Paapkuil Fontein shell middens (solid dots show excavated sites, open dots unexcavated) and nearby stone-walled tidal fish traps (dotted lines). Shadingindicates urban areas of Struisbaai, etc. (Gribble2005). Much of the work done by theSAHRA group theirpossible association. At Paapkuil Fontein, 11 LaterStone was describedby Kemp (2006). There is now extensivedocu- Ageshell middens have been located (Hart 2004), of which four mentationof all survivingfish traps along the southcoast, wereexcavated. These were chosen on thebasis of their prox- includinghigh-quality digital orthophotos. Kemp (2006) imityto the traps,their size, apparentlimited degree of focusedon theecology of fish traps, their possible impact on post-depositionaldisturbance, and (at threeof the four sites) fishpopulations and invertebratecommunities, and their thepresence in erodeddeposit of stone artefacts, bone, etc. as conservationas heritage resources. None of the above-mentioned well as shell.The remainingmiddens at PaapkuilFontein studiesprovided clarity on thearchaeological associations of consistedonly of thin scatters of shell, probably disturbed by thetraps, if such exist. road-buildingor other activities. AtStill Bay, two shell middens known as StillBay 1 (SB1) RESEARCH AREA and StillBay 2 (SB2)were investigated. Both are located above To explorethis issue, two localities along the south coast theexisting harbour and had beenexcavated in thecourse of a wereearmarked for archaeological investigation. The first was previousproject (Hart 1991), although there had been onlya ,part of the farm Paapkuil Fontein 281, near Cape preliminaryassessment of the contents. A fullanalysis of the This is also knownas fromthe materialwas carriedout for this Agulhas. bay' Vywerbaai, study. AfrikaansVywer meaning a fishtrap - evidencefor the promi- nenceof these structures in thebay (Figs1 & 2). The second EXCAVATION METHODS localitywas StillBay (Figs1 & 3), whichis famousfor its fish The PaapkuilFontein sites were excavatedin lm x lm traps(note numerous examples just north of Noordkapperpunt squareslaid acrossthe densest parts of themiddens. Where in Fig.3). Bothareas have shell middens in closeproximity to possible,sites were excavated according to occupational layers fishtraps, thereby providing an opportunityto investigate or changesin sedimentcolour, consistency or texture.In sites

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions 38 South AfricanArchaeological Bulletin 65 (191): 35-44,2010

^^^^^^^H' StillBay

.- - -. SB°3 SB1 •ßJX ""

/ '

' ( '

^H Noordkapperpunt ? ■ ■

FIG. 3. Locationsof Still Bay shell middens (solid dots show excavated sites, open dots unexcavated) and nearbystone-walled tidal fish traps (dotted lines). Area enclosedby dashed line was carefullysearched for middens; only the three shown were visible in 2006. Shadingindicates the town of Still Bay.

wherestratigraphy could not be discerned,deposit was removed Quartzitewas the dominantraw materialin all foursites, in 10cm spits. accountingfor 72.5% of the stone artefacts recovered, followed Unlessotherwise stated, all depositwas passedthrough a byquartz (24.6%). At Paapkuil Fontein 5 and 11,quartz (espe- 3 mmmesh sieve. All material recovered from the sieves was ciallyquartz chips) was more common,suggesting some retained,clearly labelled and bagged forlater sorting and knappingon site.The numbersof stone artefacts were small analysisin thelaboratory. At site P5, a verylarge quantity of and therange of activities conducted may have been limited. materialwas recovered,and onlythe finds from the richest Radiocarbondates for the four sites indicate occupation span- square(Dll) have been analysed.These alone amountedto ningthe last 5000 years, but there are no detectabletemporal 82.9kg of material. patternsin theartefact assemblages. Retouched artefacts were A comprehensivedescription of the excavations and finds extremelyrare, with only four miscellaneous retouched pieces recoveredat PaapkuilFontein and StillBay is presentedelse- (MRPs).The lithics recovered at Still Bay comprised only seven where (Hine 2008). The most importantobservations are manuportsfrom SB1 and twomanuports and twoflakes from summarizedin Table1 and discussedbelow. SB2.

EXCAVATION RESULTS MARINESHELL Marineshell comprised the bulk of the material excavated LITHICS at PaapkuilFontein and StillBay. Nearly 240.9 kg of marine Stoneartefacts and manuportscomprised the majority of shellfrom Paapkuil Fontein was analysed,of which80.4 kg thecultural remains recovered from the Paapkuil Fontein sites. camefrom square Dll at siteP5. Considerablyless shellwas

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions South AfricanArchaeological Bulletin 65 (191): 35-44,2010 39

TABLE 1. Summarytable of important finds. P: PaapkuilFontein, SB: StillBay. The Still Bay sites are referred toas SB1 and SB2 in thispaper for consistency with Hart(1991), but it should be noted that SB2 is thesame midden as thatdesignated MPI byHenshilwood and Yates(n.d.), and samplessubmitted for radiocarbon datingwere labelled 'Morris Point 1 '. Numbersoffish bones and retouchedstone artefacts are in brackets. All radiocarbondates were on shell. Calibrations are based onthe Pretoria calibration curve for the (Talma & Vogel1993), updated in 2000. Thetwo dates for P5 arefor Shell Layer 1 and ShellLayer 2, respectively,the two major depositional units at thesite. The two dates for SB2 arefor shell taken from depths of 0.5 mand 1.5 m,respectively. Calibrated dates are givenat a one-sigmarange.

Site Deposit excavated Radiocarbondates Lab. number Lithics Fauna NISP Shellfish%MNI

P4 3.5 m2 4870±80BP GX-32533 Quartziteonly 54(1) T.sarmaticus 34% (1.5 m3) 3083(2969)2887BC n = 37 S. longicosta22.9% Oxystele(all spp.) 15.6% G oculus8.7% Burnupena(all spp.) 5%

P5 5 m2(1.7 m3) 2250±60BP GX-32529 Quartzite 54 Oxystele(all spp.) 68.6% Dll (0.4 m3) 221(278)370AD dominated O. tigrina47.2% n = 137 (2) T.sarmaticus 16.3% 2329 ± 70 BP GX-32531 C. oculus5.3% 120(207)278AD D. gigas 1.9% Burnupena(all spp.) 2.5%

P7 4 m2(0.8 m3) 1450 ± 60 BP GX-32530 Quartziteonly 21 T.sarmaticus 43.5% = 1043(1103)1191AD n 16 Oxystele(all spp.) 38.6% O. tigrina22.8% G oculus3.2% Burnupena(all spp.) 7%

PII 6 m2(0.53 m3) 1319 ±60 BP GX-32532 Quartzite 34 Oxystele(all spp.) 49.1% 1202(1259)1296AD dominated O. tigrina32.7% n = 290 (2) T.sarmaticus 31.7% G oculus1.5% Burnupena(all spp.) 9.2%

SB1 Quartzite 21(7) Oxystele(all spp.) 52.2% dominated Limpet (all spp.) 39.3% = n 7 Limpetspp. 25.9% S.cochlear8.4% P. perna5.3% Burnupena(all spp.) 0%

SB2 2455 ± 20 BP Pta-8465 Quartzite 202 S. longicosta27.4% 28(56)77AD dominated (94) Oxystele(all spp.) 40.5% n = 4 T. sarmaticus16.4% 2890 ± 60 BP Pta-8467 O. tigrina15% 552(466)388BC S. cochlear4% Burnupena(all spp.) 0.4% recoveredfrom the limited excavations at the two Still Bay sites. tionstrategies appear to have focused more intensively on the Approximately3.5 kg ofmarine shell was retrievedfrom SB1 shallowerinter-tidal. and 14.1kg of shell from SB2. Examinationof the shellfishassemblages at Paapkuil BONE Fonteinindicates that, in termsof food value, Turbo sarmaticus In thePaapkuil Fontein sites, bone was rareand usually (alikreukel)was themost important species at all sites.How- fragmented,making identification difficult. No fishbone at all ever,at Paapkuil Fontein 5, 7 and 11,all dating to within the last was foundat PaapkuilFontein 5, 7 and 11, whilePaapkuil 2000years, Oxystele spp. (periwinkles),especially O. tigrina, Fontein4 yieldeda singlesmall, incomplete fish vertebra, werealso abundant. The shellfish from SB1 were dominated by whichmay have been brought on to thesite by a non-human Oxystele,while the rangeof speciesat SB2 was similarto agent.The rarity of fish bone clearly indicates that fishing was PaapkuilFontein, with S. longicosta(27.4%), I sarmaticus notthe major attraction for prehistoric people who occupied (16.4%)and Oxystelecomprising the bulk of theassemblage. thisstretch of coastline, despite the importance of fish traps in Therewas little clearly patterned change through the sequence, thebay today. althoughthis site was >1.5 m deep. Fauna was almostas rarein theStill Bay assemblages.At In general,these sites show a patternof increased abun- SB1 sevenfish bones were recovered, none of which could be danceof smaller species such as Oxystele(particularly O. tigrina) identifiedto genusor specieslevel. At SB2 94 fishbones were and Burnupenaspp. and a declinein thenumber of limpets in found,representing a minimumof threeindividuals: two morerecent occupations. This matches the pattern described Rhabdosargusholubi (Cape stumpnose)and one Cymatoceps by Henshilwood(1995, 2008) for Later Stone Age middensat nasutus(black musselcracker). These fishwere small,about GarciaState Forest. In pre-2000BP sitesat GarciaState Forest, 120mm long (Poggenpoel pers. comm.). Stumpnose enter estu- thefocus was on shellfishfrom the lower littoral, such as Turbo ariesand lagoonsas juveniles,and remainthere until they spp.,Scutellastra argenvillei and S. tabularis,and Haliotisspp. reachsexual maturity. The smallsize of theseindividuals is After2000 BP, Oxystele spp. became more important and collec- consistentwith their having been caughtin themouth of the

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions 40 South AfricanArchaeological Bulletin 65 (191): 35-44,2010

nearbyGoukou River. They would not have been swimming in sion thattidal traps had a destructiveeffect on fishstocks. theopen sea. Section10 of thesummary of theS.C.R. (Select Commission Regarding)recommended that any future Act to do withthe ARCHIVAL EVIDENCE: HISTORICAL DOCUMENTS fishingindustry should prevent the destruction of fish stocks RELATING TO STONE-WALLED TIDAL FISH TRAPS throughthe practice of making 'kraals' or 'enclosures of stone' In a surveyof early traveller accounts at the Cape, no refer- (Anon.1892 in litt). encewas foundto fishingalong the coast with stone-walled Asa result,in August of 1893, the Fish Protection Act of 1890 tidalfish traps. There are, however, references to fishingwith was amended.Section 2 stipulatedthat: "it shall not be lawful hookand line(Kolbe 1738 in Thompson1913), spearing with forany personor personsto constructor makeuse of any sharpenedwooden sticks (Tavenier 1660 in Raven-Hart1971; 'kraal'or enclosures below high- water mark, for the purpose of Langhans1694 in Raven-Hart1971; Burchell 1824; see also snaringor catching fish of any description" (Anon. 1893). This Raven-Hart1967), basket traps (Barrow 1806; Stow 1905), and regulationwas reiteratedin Proclamations353 of 1894, 393 of theuse ofnets (Kolbe 1738 in Thompson 1913). Van Riebeeck' s 1895,and 81 of1897. Another Parliamentary commission was diarydescribed how, in 1657, the Dutch bought a largequantity establishedin April1904, this time to investigatethe state of ofsteenbras (enough to feed the garrison for 3-4 days) from the fisheriesin the Caledon district,which included the farm local'Kaapmans' who had spearedthem in a shallowlake at knownas PaapkuilFontein. Mr H. van Breda,then owner of FalseBay. Some of the fish had been saltedto preservethem PaapkuilFontein, commented that he allowedfishermen to (Thorn1954). The only mentionthat we could find of camp on his propertyduring the haarder(mullet) season stone-builtfish traps used by indigenous people come from the (H. van Breda1904 in litt.).Unfortunately, the exact location interiorof southernAfrica. In his journal,General Janssens wherefishing took place and the methodsused were not describedthe fishing methods of the 'Bosjemans' of the Orange mentioned.The informationdoes, however,confirm that Riverarea: ". . .ifthey expect a swellingof the stream, while the historically,the area was a favouredplace for fishing, in partic- wateris still low, they make upon the strand a largecistern, as it ularfor mullet. were,enclosed by a wallof stones, which serves as a reservoir, In the earlyyears of the 20thcentury, local authorities whereif fortune be favourable,a quantity of fish are deposited repeatedlyappealed to have the ban on fishtraps lifted, on the atthe subsiding of the waters" (Lichtenstein 1815: 55). Schapera groundsthat they provided livelihoods and sustenancefor (1930:138) described people in theOkavango building stone bywoners(tenant farmers) and otherpoor people living along wallsand reed fences to funnel fresh- water fish into reed traps. the Riversdalecoastline (e.g. A. Badenhorst1924 in litt.).In Coastalstone-walled fish traps have, since the late 19th November1905, the DivisionalCouncil of Riversdaleintro- century,been allowedto operateonly under strict licensing ducednew regulations stating that owners of land abutting the conditions.As a result,records in local archivescontain a sea, or theirauthorized representatives, would be allowedto wealthof information about the location of traps, size of catches use fishtraps during the months of August to January to catch and species diversity.There are threeimportant periods: mulletduring spring tides at new moon.Fish less thaneight 1892-1905,1910-1913 and 1924-1933. incheslong could not be harvested(Anon. 1905 in litt.). During The 1890swere a tumultuousperiod for the Cape fishing December1905 Mr Morris Fox, acting Fishery Commissioner of industry.In 1890,the Fish Protection Act was passedto protect theRiversdale District, inspected the fish traps at StillBay and fishstocks (mainly those of commercial value) by regulating foundthat walls had notbeen maintained and thetraps were catches.Nevertheless, by 1892it was clearthat fishing at the fullof sand,making them ineffective (Fox 1905in litt.).It is Cape was in declineand a Parliamentarycommission was set importantto note that the lack of maintenance since the use of up toinvestigate the cause. Commissioners interviewed stake- fishtraps had beenbanned in 1893(only 12 yearspreviously) holdersin the industry, including professional fishermen, boat renderedthem unusable. owners,harbour administrators and ownersof fishing compa- The reprievewas short-lived:the use of fishtraps was nies.The first mention of fish traps found in the archives dates bannedonce again by Proclamation 456 of 1908. The reluctance to 1892. to liftthe restrictions was based primarilyon thedifficulty of JohanStephan of Stephan Bros., owner of a largefishing managingthe use oftraps, which were in many cases in remote companyat theCape, reported on theuse offish traps on the locations.There were a limitednumber of mounted policemen Western coastbetween and Saldanha.He to monitorfish trapping, and it was fearedthat inadequate Cape" HoetjiesBay testifiedthat thereis a practiceamong the farmers who reside policingwould lead to abuse and unauthorized proliferation of nearthe reefs of rocks on thecoast, of making 'kraals' or enclo- traps(Janisch 1910 in litt.). There was also uncertainty over who suresof stone for entrapping fish. . ." (Stephan1892: 17 inlitt.) was responsiblefor individual traps, since coastal farms were and wenton to notethat these were excessively destructive. oftenowned jointlyby a numberof farmerswho visited JohnLouis McLachlan of Stumpnose Bay echoed this: "certain periodicallythroughout the year to fish. A majorproblem was partiesin thevicinity destroy vast quantities of young fish by thatafter these visits, fishing parties departed without breach- buildingsea wallsamong the rocks sufficiently high to allow ingthe walls to enable fish to escape from the untended traps. theflood tide to cover the same, thereby entrapping fish which In a letterdated 11 November1910, H. and P. Lowrens, cannotescape at low water" McLachlan 1892: 19 in litt.). Morris Attorneysat Law,petitioned the ProvincialGovernment on Fox(1892: 25 inlitt.), who lived near the mouth of the Goukou behalfof farmers of the Riversdale and MosselBay Districts to Riverat StillBay, noted that people built "fibre walls of stone" grantpermission "to againtake up vywersto catchfish along whichretained fish as thetide receded, and he proposedthat the sea coast on theirrespective properties" (Lowrens & thisshould require a license. Lowrens1910 in litt.). The farmersprovided a mapindicating Today,recognizable fish traps are known from only a few the locationswhere they wished to build fishtraps (Fig. 4). isolatedlocalities along the west coast, far fewer than the testi- Membersof the Provincial Council pointed out that traps were moniesof MessrsStephan and McLachlanwould have us beneficialto farmersvisiting the coast for holidays with their believe,suggesting that many late 19thcentury traps are no families.Unlike professional fishermen, these farmers were not longervisible. It was clearfrom the testimonies to the Commis- equippedwith the means (nets, boats etc.) to acquirefish for

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions South AfricanArchaeological Bulletin 65 (191): 35-44,2010 41

• MeJkhout.•■""•• ^ _^__^ ^^^ ' W * FonlpinGcjcgcn '. -- .^^ ^^^ ' ''

'"• / Rietvallei ' 1 ' V. "'■•■/ •■■ gelegen ' yS ' ^^"^Sv^X ' •' Un KaffirKuils Riviei 1 ' ' / V J -*■*«*■ ' ~yS Rici™ei '^ 9'X' 'S- '. ' {' - a/v^jn"V / > c t '• I _-■ - ~~~~A ' '• 4AMouthü;thc' y douritzRiver S ' '• '. ) ' / X 'f- ' «"ontcin •. 1 ' -, '-/ : " - ' D ,/ A / /~>^'i V>v i /Grootljzcrvark BuijelshockR,iffi.Uh«vlf I ' '"• - -^ yS U ^-^~J /B / Fontcin ' Langefontein' / s>^ / '•-'.'■.¡^ ^^ - ^^ ' Bosch / (Ka$rkuilsRiver Mouth s^ ^^'^-~-^ / "'• '' I ">" / ' / y^ C ~" ^buII Point / '^ ^^^^J jT^~^

' >^^ MorrisPoint Point ^^^ '^^ ^V-^íjzeiA-ark ^^^

FIG. 4. Map accompanyingRivendale farmers' 1910 petition to the Provincial Government. Areas marked A-B and C-D indicatewhere they wished to build fish traps;note how these are adjacent to demarcated farms. The Kafferkuils River is nowcalled the Goukou River. Cape Archives PAN 6 A120/B/13.

theirown consumption.These appeals were successful, and mentionedin his recommendations tothe Provincial Secretary. revisedregulations were published in the Fisheries Ordinance, He notedthe Natal Government's decision to abolishall fish no.12 of 1911, Proclamation 223. These permitted limited use of trapsin its waters, except for a fewin the mouths of the Tugela, fishtraps, on conditionthat, first, any personwishing to Umzimkuluand TongaatRivers and observedthat there were constructa fish trap obtained a writtenpermit from an officer stillhundreds of fish traps in PortugueseEast Africa that were authorizedby the Provincial Administrator. Second, applicants graduallybeing demolished (Gilchrist 1924 in litt.). had tosubmit sketch plans showing the locations and dimen- In 1925,Section 17 of the Fisheries Ordinance No. 30of 1920 sionsof the proposed traps. Third, successful applicants were was amendedto allow constructionand use of fishtraps, requiredto demolishtheir traps if subsequently instructed to providedthat these retained sufficient water at low tide to keep do so bythe Administrator (Anon. 1911 in litt.).Furthermore, fish alive until the next high tide, when they would be ableto fishcaught in traps could not be soldcommercially. Seventeen escape(Anon. 1925). Because no permitswere needed, it was farmersapplied for permits, of whom only three were success- difficultto identifythe owner of any particular trap, and the ful:Mr D.P. du Toitof the farm Prins Kraal, Mr T. Wilsonof regulationof trappingbecame much more difficult. There is Skipskopand Mr J.W. Myburgh of the farm Vogelgezang. The littleinformation from this period, although a fewindividuals grantingof these permits caused considerable tension, espe- continuedto apply to the ProvincialGovernment to build ciallybetween farmers and localfishermen who thoughtthat traps,despite this not being necessary under the new regula- theyshould hold the permits.There was concernover fish tions.Some applications came from as faraway as Johannes- trapsbeing situated at thebest locations, thereby restricting burgand Natal (Webber1930 in litt; Stansfeld1933 in litt). theplaces traditionally used by fishermen,particularly when Thereare references to theuse offish traps along the Still Bay fishingfor mullet. coastline,from the DuivenhoksRiver to the GouritzRiver In 1913,the Fishery Officer inspected the coastline from mouth(Fig. 4) (Badenhorst1924 in litt). PortSt Johns to Cape Townto report specifically on thedistri- In Februaryand March1931, Dr. C. van Bondeinspected butionof fish traps. None werereported from Port St Johns, the fishtraps at Cape Agulhas,Struis Bay, Arniston and PortAlfred, Port Elizabeth, Jeffreys Bay, Plettenberg Bay, Skipskop(van Bonde 1931in litt).He was concernedabout KnysnaLagoon, George or Mossel Bay. Fish traps were noted theirproliferation along the coastline, leading to onlyin the River sdale District. The Fishery Officer wrote: "the the destructionof immaturehaarders, but nothingcould be methodof trapping fish, by constructing the fish kraal or viper, done because the trapscomplied with the 1925regulations. seemsto have been a regularpractice engaged in by both Thiswas thelast mention of fish traps found in thearchival re- farmersof the district and fishermen alike, and the whole coast, cordbut even today the battle between the fish trap advocates whereverthere is a rockyreef, shows signs of dismantled walls and theauthorities continues. ofthese kraals which were used someyears ago" (Cripps1913 Fromthe evidencesummarized above it is clear that, in litt.).At 'Riet Vlei' fishtraps were foundto be in use duringthe late 19thand early20th centuries, fish traps were (Weisbecker1913). This was reportedto the police in Albertinia usedprimarily by farmers and theirbywoners. Itis strikingthat and a mountedtrooper was requestedto demolishthem. All fishtraps occur most often where farms had been established threeof the traps authorized under Proclamation 223 of 1911 immediatelyadjacent to the coast, especially in the Bredasdorp werefound to be in contraventionofthe legislation. The traps regionbetween Cape Agulhasand Skipskop,and along the weredemolished and theowners informed that none would StillBay coastbetween Noordkapperpunt and the Gouritz be allowedin thefuture. RiverMouth. These two areas contain the highest densities of Thereare few references to fishtraps in thearchives from fishtraps found anywhere along South African coast. Other 1913until the mid-1920s, when the police began to implement regionssuch as theNamaqualand coastline lack tidal fish traps theregulations more vigorously. In a letterto theProvincial entirely,despite the presence of many suitable localities along Secretary,the Magistrate of Riversdale stressed the importance rockyshores with gently sloping topography (Hart & Halkett, offish traps for the subsistence of poor people in the area, espe- personalobservation). This is poorfarming country and was ciallyat StillBay (Badenhorst1924 in litt).J.D. Gilchrist, the notdensely settled in historicaltimes, although there is abun- FishingAdministrator, reiterated some of the concerns already dantevidence of pre-colonial occupation.

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions 42 SouthAfrican Archaeological Bulletin 65 (191):35-44, 2010

DISCUSSION be a promisingavenue for future research. The documentsand evidence fromfishermen who stilluse Could itbe thatin pre-colonialtimes, fish were caught in the trapstoday (Avery1975, 1976; Kemp 2006; see also docu- tidaltraps and processed at sites other than middens, which we mentaryfilm Johnny Appels - TheLast Strandloper),emphasize havenot yet identified? This is possible,but unlikely. If signifi- thatstone-walled tidal fish traps are dynamic structures. In the cantquantities of fish were caught, they would have to have late19th and early20th centuries they were repeatedly built, beenprocessed and preservedimmediately, probably by salt- alteredand demolishedas localcommunities complied with, ing,smoking and/or sun- and wind-drying. As outlined above, challengedor triedto circumventchanging legislation. Fish thistakes at leastseveral days, and mayneed as muchas two trapswere highly politicized facilities, and as the summary weeks.The processwould have required close supervision to aboveshows, their ownership and use wereheavily contested. keepaway scavengers and flies,necessitating camping nearby. Thiscontestation occurred because the traps were so effective.This Withan abundanceof fish at hand,people would surely have effectivenessunderscores the significance of the rarity of fish eaten some of it duringthis time.We simplydo not see bone in the PaapkuilFontein and StillBay sitesdescribed evidenceof such behaviour. above. Stone-walledtidal fish traps occur all overthe world and We also examinedother middens, although we did not the anthropologicaland archaeologicalliterature describes excavatethem. Among these were a seriesof very large examplesin (Dortch 1997; Randolph 2004; Angeles middensat Waenhuiskrans(Arniston), adjacent to fishtraps 2005),the (Bannerman & Jones1999; Williams thatare stillin use (as documentedin the 2006film 'Johnny & McErlean2002; O'Sullivan2003), the Netherlands(Low Appels- TheLast Strandloper'). Careful search of the surfaces of Kooijmans1987), Denmark (Pedersen 1995), parts of Africa thesemiddens revealed bird and mammalbone, but not fish (Breenet al. 2001), North America (Treganza 1945; Keegan 1986; bone.The sameapplies to themiddens at Noordkapperpunt Mosset al 1990;Lutins 1992; Tveskov & Erlandson2003, Foster (Fig.3). 2005),al-Bahrain (Serjeant 1968) and (Munita et al 2004). Couldthe records of the 19th-20th centuries merely reflect In several instances,stone-built fish traps previously themost recent end of a longhistory of the building and use of thoughtto be ofpre-colonial origin have been shownto be of fishtraps extending back into the pre-colonial past? We argue more recentdate, forexample, through the use of aerial thatthe archaeological evidence fails to support this idea. With photographyin Australia(Randolph 2004). At Lake Cahuilla, the exceptionof Stofbergsfonteinmidden (which is on the California,a number of 'ancient stone fish traps' occupying a LangebaanLagoon, not on thecoast) (Robertshaw 1978/1979), seriesof rocky terraces 90 feet below the present high water line noneof the assemblages of archaeological fish bone reported in are in facthouse depressionsrather than fishtraps. Local theliterature include large numbers of Mugilidae spp. (haarders CahuillaIndian stories of how these 'fish traps' were operated or mullet),as one would expectfrom the use of traps.We probablyoriginated among whites, and theNative American also notethat areas with high densities of fishtraps do not communityfound it amusing to pass them on (Treganza1945). necessarilyhave high densities of archaeological sites. This is This illustratestwo points,first, that features can easilybe particularlystriking at StillBay, with dozens of tidal fish traps, misinterpreted,and second,that such misinterpretations can butwhere repeated surveys of the area adjacent to the shore by be absorbedin the storiesand folkloreof local indigenous ourselvesand otherarchaeologists have located relatively few peoplesand can be difficulttodebunk. Caution should be exer- middens.Sites we identifiedduring a visitto thearea in 2006 cisedwhen assuming the antiquity of features such as these are shownin Figure3. An earliersurvey of the same area by withoutstrong evidence. In thecase ofthe Cape coastalfish ChrisHenshilwood and RoydenYates reported additional traps,one ofthe problems has been that they were regarded as middensthat were not visiblein 2006,but these authors 'static'features, as artefacts'captured' in time.As a result, commented"Given the extent of therocky inter-tidal in the suggestionsthat they might be ancient,as madeby Goodwin area,the absence of highernumbers of largemiddens ... is and others,gradually became received wisdom. surprising. . . Some . . . are,in effect,short occupation sites or 'lunchspots' that represent a singlemeal or overnightstop/' CONCLUSIONS Theyalso notedthat "All the sites... are relativelyclose to Despitestrenuous efforts, we couldfind no evidenceof any viswywersbut there is scantevidence for the consumption of associationbetween stone-walled tidal fish traps and excavated fish"(Henshilwood & Yatesn.d.). By contrast, some areas have pre-colonialsites. At Paapkuil Fontein, fish was absentfrom all middenswith fish bone, but apparently no fishtraps. butone site,which yielded a singlevertebra. At the Still Bay It is worthmentioning one furthersite here:a kitchen sites,fish remains were rare and the size of the specimens iden- dump associated with a local fisherman'shouse in the tifiedsuggests that they were caughtin the estuaryof the Hotagterkliparea (now ArgonautaPark) in Struisbaai(see GoukouRiver. In starkcontrast, there is clear documentary and Fig.2). Thismidden probably dates to the early part of the 20th distributionalevidence linking fish traps with historical settle- century,and containedlarge quantities of fish bone (Halkett ment,especially between Cape Agulhasand Arnistonand at 1996).Preliminary analysis showed a rangeof species includ- StillBay. Active fishing in precolonialtimes is notin question: inghaarder, black and whitemusselcracker, elf, silverfish, red numerousLater Stone Age sites contain quantities of fish bone. and white stumpnose,kabeljou, galjoen, dassie, sand Manyof these sites are nowhere near any known traps, and we steenbras,white steenbras and shark.With the exceptionof knowprecolonial fishermen had otherways of catching fish; dassie,silverfish and red stumpnose, the species present are all some coastal midden sites have yielded the bones of commonlycaught in fish traps. We cannot be certainhow these deep-waterfish species that could only have been obtained by fishwere caught, and morethan one methodmay have been linefishing (Deacon 1970, Inskeep 1987; Poggenpoel 1996). This used,including line-fishing, netting, etc. The siteis, however, studyset out to searchfor evidence of a pre-colonialage for veryclose to fish traps, and datesto a timewhen we knowthat stone-walledtidal fish traps, but has foundnone whatsoever. thesewere in use. The sheerquantity of fish bone makesthis There is, therefore,no reason to believe that the use of middentotally unlike pre-colonial sites in the area. Further stone-walledtidal fish traps on theCape coaststretches back excavationof historical middens in areas with fish traps would morethan a fewcenturies. This scenario provides opportunities

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions South AfricanArchaeological Bulletin 65 (191): 35-44,2010 43 forfuture work: how was thismethod of fishing introduced to http://www.parks.ca.gov/default.asp?page_id=23153 M.H. 1892. of the Select Committee.S.C.R. on theCape? Was it importedby immigrantsettlers (or slaves)? Fox, Reports Fishing is therean offish between Industry. House ofAssembly. Fourth Session of Why especiallyhigh density traps Parliament.C.2-'98. D. Hermanusand Mossel Whatwere the historical Appendix Bay? processes Fox,M.H. 1905. Letterto C.J.Roux. Archives.AGR 1/512G.18. this Cape thatled to theirdevelopment and proliferationalong Gilchrist,J.D. 1924. Letterto the ProvincialSecretary. Cape Archives. particularstretch of coast?These and manyother questions PAN 6. A120/B/13.Volume 1. remainto be answered. Goodwin, A.J.H.1938. Archaeologyof the Oakhurst Shelter,George. Transactionsof the Royal Society of South Africa 25: 229-324. ACKNOWLEDGEMENTS Goodwin, A.J.H. 1946. Prehistoricfishing methods in South Africa. 20: 134-141. Wewould like to thank Mr Christiaan Mostert for allowing Antiquity Gribble,J. 2005. The ocean baskets:pre-colonial fish traps on the Cape us to do researchon hisproperty at PaapkuilFontein. Charlie south coast. The Stick22: '-A. Ntandazo Digging Arthur,Shadreck Chirikure, Mjikiliso,Mpakamo Haddad, D. 2003.Fishing the Khoisan way. GoodTaste September/Octo- Sasa,Liesbet Schietecatte and BrianStewart helped carry out ber,pp. 102-106. thefieldwork. Cedric Poggenpoel identified the fish remains. Halkett,D.J.H. 1996. Phase 2 archaeological investigationof Erf 853 We thankJohan Binneman, Mariagrazia Galimberti, Simon Struisbaai,south west Cape Province.Unpublished Report.Archae- Contracts of Town. Hall,Lucy Kemp, Alex Mackay, Tim Maggs, John Parkington, ology Office,University Cape AraWelz and Yatesfor discussion.Gerrit Mars Hall, S. 1990.Hunter-fisher-gatherers of the Fish RiverBasin: a contri- Royden helpful butionto theHolocene ofthe Eastern and te Vinkof Still us informationand prehistory Cape. Unpublished Jan Bay kindlygave DPhil thesis.Stellenbosch: University of Stellenbosch. documentsabout the fish traps at Still Bay. Antonieta Jerardino Hall, S. 2000.Burial and sequence in the LaterStone Age ofthe Eastern providedrelevant literature, some of which was translatedinto Cape Province,South Africa.South African Archaeological Bulletin 55: Englishby Mariagrazia Galimberti. The assistance of Lance van 137-146. Sittertisalso gratefully acknowledged. Funding for this research Hall, S. & Binneman,J. 1987. Later Stone Age burial variabilityin the camefrom the National Research Foundation of South Africa Cape: a social interpretation.South African Archaeological Bulletin 42: 140-152. and theWenner Gren Foundation to PH). (bursary Hart,T. 1991.An archaeologicalassessment of shell middens,harbour development area, Stillbaai. Unpublished report. Archaeology REFERENCES ContractsOffice, University of . G.A. 2005. Fish - a of our health and Angeles, traps significantpart Hart,T.J.G. 2004. Heritageimpact assessment of Portion15 of Paapkuil Medical Australia 541-543 wellbeing. Journalof 182(10): (consulted Fontein 281, Cape Agulhas, South Africa. Unpublished report. Month year): http://www.mja.com.au/public/issuesl82_10_160505/Archaeology Contracts Office, University of Cape Town. anelO811 fm.html) Henshilwood, C.S. 1995. Holocene archaeology of the coastal Garcia Anon. 1892. of the Select Committee.S.C.R on the Reports Fishing State Forest,southern Cape, South Africa.Unpublished PhD thesis. of Good House of FourthSession of Industry.Cape Hope Assembly. Cambridge:Cambridge University. Parliament.C.9-'98. Notes and Proceedings. Henshilwood, C.S. 2008. Holocene prehistoryof the Southern Cape, Anon. 1893.Fish ProtectionAct. Government Gazette. 24/11/1893. South Africa.Excavations at Biombos Cave and the Blombosfontein Anon. 1905. the of mulletsor haarders Regulationsregarding catching Nature Reserve.British Archaeological Reports International Series. in kraalsor Archives.AGR G.18. enclosures, (vywers).Cape 1/512 CambridgeMonographs in AfricanArchaeology 75. Oxford: Archaeo- Anon. 1911.Proclamation 223 of 1911. GovernmentGazette 26/09/1911. press. Anon. 1925. Amended Ordinance of 1925. Archives. Cape PAN 6. Henshilwood, C.S. & Yates,R.J. n.d. Reportprepared forthe StillBay A120/B/13. Town Council, StillBay, Southern Cape. Unpublished manuscript. Anon. 2005. About bokkoms (consulted February2009): http://www. Hiñe, P.J.2008. Stone-walled tidal fish traps: an archaeological and bokkom.com/aboutbokkoms.htm archivalinvestigation. Unpublished MPhil thesis.Cape Town: Uni- Avery,G. 1975. Discussion on the age and use of tidal fish-traps versityof Cape Town. South (visvywers). AfricanArchaeological Bulletin 30: 105-113. Inskeep, R.R. 1987. , Cape Province,South Africa: The Avery,G. 1976.A systematicinvestigation of open stationshell middens Holocenelevels. Oxford: British Archaeological Reports. the along southwesternCape coast. Unpublished MA thesis.Cape Inskeep,R. 2001. Some notes on fishand fishingin Africa.In: Milliken, Town: Universityof Cape Town. S. & Cook, J.(eds) A VeryRemote Period Indeed: Papers on thePalaeolithic Badenhorst, A. 1924. Letter to the Fishery Administrator.Cape Presentedto DerekRoe: 63-73. Oxford:Oxbow Books. Archives.PAN 6. A120/B/13.Volume 1. Janisch, N. 1910. Letter to Mr. Vincent. Cape Archives. PAN 6. Bannerman,N. & Jones,C. 1999. Fish trap types: a component of A120/B/13.Volume 1. maritime cultural landscape. The InternationalJournal of Nautical Jerardino,A.M.S. 1996. Changing social landscapes of the Western Archaeology28: 70-84. Cape of southernAfrica over the last 4500 years.Unpublished PhD Barrow,J. 1806. Travels into the Interior of Southern Africa. Vol. 1. London: thesis.Cape Town: Universityof Cape Town. Cadell & Davies. - JohnnyAppels TheLast Strandloper. 2006. Documentary film directed by Breen,C, Forsythe,W, Lane, P., McErlean,T, McConkey,R., Omar, Michael Raimondo forthe South AfricanBroadcasting Corporation. A.L., Quin, R. & Williams,B. 2001. Ulster and the ? South Africa. Recent maritimearchaeological research on the East Africancoast. Keegan, WE 1986. The ecology of Lucayan Arawak fishingpractices. Antiquity75: 797-798. AmericanAntiquity 51: 816-825. Burchell,W.J. 1824. Travels into the Interior of Southern Africa. Vol. 2. Lon- Kemp, L.V 2006. Ancientstone wall fishtraps on the south coast of don: Longman,Hurst, Rees, Orme & Brown. South Africa:documentation, current use, ecological effectsand Cripps.1913. Report of the Fishery Officer on certainfishing stations on management implications.Unpublished MSc thesis. Cape Town: the coast of the Cape Province.Cape Archives.PAN 6. A120/B/13. Universitvof Cape Town. Deacon, H.J. 1970. Two shell midden occurrencesin the Tsitsikama Lichtenstein,H. 1815.(reprinted 1930) Travelsin SouthAfrica in theYears National Park, Cape Province: a contributionto the study of the 1803,1804, 1805 and 1806. Vol. 2. Cape Town: Van RiebeeckSociety. ecologyof the Strandlopers.Koedoe 13: 37-49. Louw, J.T.1960. Prehistoryof the Matjes RiverRock Shelter.National Deacon, H.J.1976. Where Hunters Gathered: a StudyofHolocene Stone Age Museum, Bloemfontein,Memoir No. 1. Peoplein theEastern Cape. Cape Town: South AfricanArchaeological Low Kooijmans,L.P. 1987. Neolithicsettlement and subsistencein the SocietyMonograph 1. wetlands of the Rhine/MeuseDelta of the Netherlands.In: Coles, Dortch,C.E. 1997. New perceptionsof the chronologyand develop- J.M. & Lawson, A.J.(eds) EuropeanWetlands in Prehistory:144-180. mentof Aboriginal fishing in south-westAustralia. World Archaeology Oxford:Oxford University Press. 29: 15-35. Lowrens,H. & Lowrens,P. 1910. Letterto Mr.Vincent. Cape Archives. Foster,J.W. 2005. Ahjumawi fish traps (consulted February 2009): PAN 6. A120/B/13.Volume 1.

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions 44 South AfricanArchaeological Bulletin 65 (191): 35-44,2010

Lutins,A. 1992.Prehistoric fishweirs in easternNorth America. Unpub- Sealy,J. 2006. Diet, mobility,and settlementpattern among Holocene lished MA thesis.Binghamton: State Universityof New York(con- hunter-gatherersin southernmostAfrica. Current Anthropology 47: sultedAugust 2005): http://www.lutins.ore/thesis/ 569-595. McLachlan, J.L.1892.Reports of the Select Committee.S.C.R. on the Serjeant,R.B. 1968. Fisher-folkand fish-trapsin al-Bahrain.Bulletin of FishingIndustry. Cape of Good Hope House of Assembly.Fourth theSchool of Oriental and AfricanStudies 31: 486-514. Session of Parliament.C.2-'98. Appendix D. Stansfeld,H.W 1933. Letterto the Magistrateof Bizana, Pondoland to Moss, M.L., Erlandson,J.M. & Stuckenrath,R. 1990.Wood stakeweirs constructfish kraal. Cape Archives.PAN 6 A12/B/13.Volume 2. and salmonfishing on theNorthwest coast: evidence fromsoutheast Stephan, J. 1892. Reportsof the Select Committee.S.C.R. on Fishing Alaska. CanadianJournal of Archaeology 14: 1-15. Industry.Cape of Good Hope House ofAssembly. Fourth Session of Munita,D., Alverez,R. & Ocampo, C. 2004. Corralesde Piedra, Pesca Parliament.C.2-'98. Appendix D. pasiva en la costa interiorde Chiloé. Boletínde la SociedadChilena de Stow, G.W. 1905. The Native Races of South Africa.London: Swan Arqueologia37: 61-74. Sonnenschein. Nagaoka, L. 2005. Differentialrecovery of PacificIsland fishremains. Talma,A.S. & Vogel,J.C. 1993. A simplifiedapproach to calibrating14C journalof Archaeological Science 32: 941-955. dates. Radiocarbon35: 317-322. Orton,J. & Halkett,D. 2007. Excavationsat Noetzie Midden: an open Thorn,H.B. (ed.) 1954. Journalof Jan van Riebeeck.Vol. IL Cape Town: siteon the Cape south coast. TheDigging Stick 24(3): 5-7. Balkema. O'Sullivan,A. 2003.Place, memoryand identityamong estuarinefish- Thompson, W.W 1913. The Sea Fisheriesof theCape Colony,from Van ingcommunities: interpreting the archaeology of early medieval fish Riebeeck's Days tothe Eve of the Union: With a Chapteron Troutand Other weirs.World Archaeology 35: 449-568. FreshwaterFishes. Cape Town: Maskew Miller. Parkington,J.E. 1977. Follow the San. Unpublished PhD thesis. Tothill,F. 1899. Reportsof the Select Committee.S.C.R. on the Fishing Cambridge:Cambridge University. Industry.Cape of Good Hope House ofAssembly. Second session of Pedersen,L. 1995. 7000 years of fishing:stationary structures in the Parliament.G.52-1899. Mesolithicand afterwards.In: Fischer,A. (ed.) Man and Sea in the Treganza,A.E. 1945. The Ancient Stone Fish Traps' of the Coachella Mesolithic,Coastal Settlement above and BelowPresent Sea Level:75-86. Valley,southern California. American Antiquity 10: 285-294. Oxford:Oxbow Monograph 53. Tveskov, M.A. & Erlandson, J.M. 2003. The Haynes Inlet weirs: Poggenpoel,CA. 1996.The exploitationof fish during the Holocene in estuarinefishing and archaeologicalsite visibilityon the southern thesouth-western Cape, SouthAfrica. Unpublished MA thesis.Cape Cascadia coast.Journal of Archaeological Science 30: 1023-1035. Town: Universityof Cape Town. van Bonde, C. 1931. Letterto the ProvincialSecretary. Cape Archives. Poggenpoel, C.A. & Robertshaw, P.T. 1981. The excavation of PAN 6. A120/B/13.Volume 2. SmitswinkelbaaiCave, . SouthAfrican Archaeological van Breda, H. 1904. Reports of the Select Committee.S.C.R. on the Bulletin36: 29-35. Caledon Fisheries.Cape of Good Hope House of Assembly.First Randolph, P. 2004. Lake Richmond'fish traps'? WorldArchaeology 36: Session of Parliament.A.3-1904. 502-506. van der Eist,R. 1993.A Guideto the Common Sea Fishesof Southern Africa. Raven-Hart,R. 1967. Beforevan Riebeeck:Callers at SouthAfrica from 3rd edn. Cape Town: Struik. 1488-1652.Cape Town: Struik. van Niekerk,K.L. 2004. The archaeology of fishing:establishing a Raven-Hart,R. 1971.Cape o/Good Hope, 1652^-1702: TheFirstFifty Years of comparativecollection and subsequent analysisof marinefish bone DutchColonisation as Seenby Callers. Vol. 2. Cape Town: Balkema. fromHolocene levels at Biombos Cave, South Africa.Unpublished Robertshaw,P.T. 1978/1979. Archaeological investigations at Langebaan MPhil thesis.Bergen: University of Bergen. Lagoon, Cape Province.Palaeoecology of Africa and the Surrounding Webber,A. 1930. Letterto the ProvincialSecretary to constructfish Islands10/11: 139-148. kraal.Cape Archives.PAN 6. A120/B/13.Volume 2. Schapera,!. 1930.The Khoisan Peoples of South Africa. London: Routledge. Weisbecker,A. 1913.Reply to A.L. van Wyk'sapplication to constructa Schweitzer,F.R. & Wilson,M.L. 1982. 1: a Late Quater- 'Vijver' at RietVley. Cape Archives.PAN 54. K13/23. naryliving site in the southernCape Province,South Africa.Annals Williams,B. & McErlean, T. 2002. Maritimearchaeology in Northern ofthe South African Museum 88: 1-203. Ireland.Antiquity 76: 505-511.

This content downloaded from 137.158.114.126 on Tue, 12 Jan 2016 09:10:37 UTC All use subject to JSTOR Terms and Conditions