One hundred and fifty seven specimens of Ostrea puelchana were analyzed. These were collected live in the Las Grutas oyster bank (40º 48' S; 65º 05' W). Fifty three valves of ? Ostrea ? patagonica and 186 valves of ? O .? alvarezii were obtained from beds within the Puerto Madryn Formation, exposed at Puerto Pirámide (42º 35' 48'' S, 64º 15' 27'' W), province of Chubut.
Samples of the live specimens were taken randomly in the Las Grutas ground within the most densely packed zone of the oyster bank. The aim of such a choice was to facilitate collection and obtain the highest possible number of specimens. Sampling took place by divers during February 2003 at 3.5 meters depth.
For Ostrea puelchana , all borings were counted and, when possible, the specimens of Lithophaga patagonica were selected out and the following data were recorded: oyster specimen number, type of valve (left or right), sector of the valve containing the perforation, and specimen number (of L . patagonica ). The oyster valves were broken in pieces in order to ensure counting of all borings.
To record Lithophaga sp. borings in fossil oysters, the shells of these were superficially broken and the data recorded were: valve number, type of valve (left or right), and sector of the shell bored.
The location of borings on the shell was classified following a division of the shell surface into different sectors (figure 3). These divisions reflect distinct morphological features of the valves that may influence differential settling of Lithophaga sp. larvae on each one.
The left valves of Ostrea puelchana and ? O .? Alvarezii were divided into six different sectors and the right valves into five (figure 3). For the left valve, these sectors are: apex (3%), platform (5%), anterior margin (16%), ventral margin (30%), posterior margin (21%) and center (25%). In the right valve there is no platform, so the areas into which it was divided are: apex (3%), anterior margin (21%), ventral margin (30%), posterior margin (21%) and center (25%).
In ? Ostrea ? patagonica the areas and percentages for both valves were: apex (18%), anterior margin (11%), ventral margin (30%), posterior margin (11%), center (30%).
Statistic analysis and graphics were done using ?STATISTICA? Version 99. Tests were performed for differences between proportions. Therefore, in order to determine the independence of valve area and frequency of boring, contingency tables were performed (Chi-square Independence Tests). The expected perforation value for each area was obtained multiplying the total borings for each valve by the area occupied by each sector.