Experimentally determined ranges of thermal tolerance and requirements for complefion of the life history of some 60 seaweed species from the North Atlantic Ocean were compared with annual temperature regimes at their geographic boundaries. In all but a few species, thermal responses accounted for the location of boundaries. Distribution was restricted by: (a) lethal effects of high or low temperatures preventing survival of the hardiest life history stage (often microthalli), (b) temperature requirements for completion of the life history operating on any one process (i.e. [sexual] reproduction, formation of macrothatli or blades), (c) temperature requirements for the increase of population size (through growth or the formation of asexual propagules). Optimum growth/reproduction temperatures or lethal limits of the non-hardiest stage (often macrothalli) were irrelevant in explaining distribution. In some species, ecotypic differentiation in thermal responses over the distribution range influenced the location of geographic boundaries, but in many other species no such ecotypic differences were evident. Specific daylength requirements affected the location of boundaries only when interacting with temperature. The following types of thermal responses could be recognised, resulting in characteristic distribution patterns: (A) Species endemic to the (warm) temperate eastern Atlantic had narrow survival ranges (between ca 5 and ca 25 ~ preventing occurrence in NE America. In species with isomorphic life histories without very specific temperature requirements for reproduction, northern and southern boundaries in Eur/Africa are set by lethal limits. Species with heteromorphic life histories often required high and/or low temperatures to induce reproduction in one or both life history phases which further restricted distribution. (B) Species endemic to the tropical western Atlantic also had narrow survival ranges (between ca 10 and ca 35~ Northern boundaries are set by low, lethal winter temperatures. Thermal properties would potentially allow occurrence in the [sub)tropical eastern Atlantic, but the ocean must have formed a barrier to dispersal. No experimental evidence is so far available for tropical species with an amphi-Aflantic distribution. {C) Tropical to temperate species endemic to the western Atlantic had broad survival ranges (< 0 to ca 35 ~ Northern boundaries are set by low summer temperatures preventing (growth and} reproduction. Thermal properties would permit occurrence in the (sub)tropical eastern Atlantic, but along potential "stepping stones" for dispersal in the northern Atlantic (Greenland, Iceland, NW Europe) summer temperatures would be too low for growth. (D) In most amphi-Atlantlc (tropical-) temperate species, northern boundaries are set by low summer temperatures preventing reproduction or the increase of population size. On European shores, species generally extended into regions with slightly lower summer temperatures than in America, probably because milder winters allow survival of a lar...

Erosion rates and sources of sediment ingested were quantified for the 2 most abundant parrotfish species on a leeward fringing reef of Bonaire, Netherlands Antilles: Scarus vetula and Spansoma viride. Direct estimates of erosion by different size classes were obtained from daily feeding rates and grazing scar frequency, scar volume and substrate density. Foraging preference and distribution of fish on the reef were used to examine patterns of bioerosion at 2 spatial scales: reef zones and individual substrates used for grazing. Sediment mass ingested by fish provided an independent check on erosion rates, and was partitioned according to source. S. vetula, employing a scraping feeding mode, removed less material from grazed substrates than similar sized S. vjride, which forages by excavating the substrate. Eros~on rates increased strongly with fish size in both species. The (indigestible) carbonate derived from epilithic algae accounted for all sediment ingested by juvenile fish. In adult fish, the proportion of freshly eroded carbonate substrate ingested increased with fish size. The distribution of adults of these large scarids over different reef zones determines the rate of bioerosion on a large spatial scale. The highest bioerosional rates occur on the shallow reef (ca 7 kg m-2 y r ' ) , and they decrease with depth. Parrotfish foraging preferences, and the effects of food type and skeletal density of substrates on the size of the grazing scars, cause large differences in bioerosional rates on a small spatial scale. The highest rates of bioerosion occur on substrates infested with boring algae and of low skeletal density, while high-density substrates and substrates covered with crustose corallines undergo lower rates. Living coral is rarely eaten by scarids, and largely escapes erosion by grazing.

Food selection by the Canbbean stoplight parrotfish Spansoma vlrlde was lnvestlgated on a fnnging coral reef of Bonalre, Netherlands Antllles For different reef zones, the diet composition for each life phase was determined by descnption of randomly selected bites, and compared to the avallabllity of food resources, as determined w~t h the aid of chaln-llnk transects S viride employs an excavating grazing mode and feeds almost exclus~vely on algae associated with dead coral substrates Preferred food types are large and sparse turfs growlng on carbonate substrates inhabited by endohthic algae Crustose corallines, with or without algal turfs are not preferred Feeding forays were longer on the preferred food types Foraging preferences are related to nutntlonal quahty of the food types and their yleld, i e the amounts of blomass, proteln and energy that can be Ingested per bite, as calculated from the size of grazing scars and the biochemical composition of the algae In spite of selective foragIng, a large proportion of b~t e s is taken on infenor food types Endolithic algae constitute an Important food resource for scraping herbivores, such as S vlnde These algae have relatively hlgh energetic value, and allow a high yield as a result of weakening the carbonate matrix by thelr boring filaments The yield of algal resources also depends on the skeletal density of the limestone substrates On deeper reef parts ( > 3 5 m depth), low-density substrates predomnate, resulting in higher y~e l d s of algae per bite than are attalned from high-denslty substrates that predominate on shallower reef parts The increased availablllty of hlgh-yleld food and substrate types coinc~des w~t h the occurrence of harermc terrltorlal behavlour m S v i r~d e males on the deeper reef parts Terntones are defended agalnst conspeclflcs and have a n Important function as spawning sites It 1s argued that the access to supenor food resources on the deeper reef makes terntonal defence feaslble for S viride

On the fringing reef of Bonaire, Netherlands Antilles, a comparative study was made of habitat use, diet selection, foraging behaviour and food acquisition of the parrotfish species Scarus vetula and Sparisoma vinde. The species are sympatric and hve in the same reef habitats (depth zones). Both species show similar foraging selectivity, but exploit algal resources differently. Preferred food items are turf algae on substrates infested with endolithic algae, whereas crustose corallines are avoided. Foraging preferences are related to yield, i.e. the amount of AFDW (ash-free dry weight), protein and energy that can be harvested per bite. Foraging behaviour differs between the species. S. vetda takes more bites in long forays, has higher bite rates (no. of bites S-'), and makes fewer and smaller scars on grazing substrates than S. viride. Furthermore, S. vetula prefers flat substrate surfaces while adult S. viride graze by preference on concave surfaces. Species-specific differences in preference and utilization of grazing substrates are related to feeding mode. S. vetula employs a scraping feeding mode by which mainly epilithic algae are ingested. In contrast, S. viride is an excavating grazer that ingests large amounts of endolithic and crustose algae. Intake and assimilation of algal AFDW, protein and energy were quantified through a combination of laboratory feeding trials and field observations. S. vetula has lower food intake (mg AFDW bite-') than S. vin'de (0.8 X I O -~ X fish wet wt, FWW, and 2.3 X I O -~ X FWW respectively), resulting from smaller (shallower) bites. Assimilation efficiencies of total AFDW, protein and energy by S. vetula were higher than in S. viride grazing on the same dead coral substrates. In spite of different feeding modes and different fractions of the primary production harvested, daily amounts of assimilated nutrients and energy are similar for both species, resulting from higher feeding rates (no. of bites h-') and higher assimilation efficiency in S. vetula.

The sensitivity to UV‐B radiation (UVBR: 280–315 nm) was tested for littoral (Palmaria palmata[L.] O. Kuntze, Chondrus crispus Stackhouse) and sublittoral (Phyllophora pseudoceranoides S. G. Gmelin, Rhodymenia pseudopalmata[Lamouroux] Silva, Phycodrys rubens[L.] Batt, Polyneura hilliae[Greville] Kylin) red macrophytes from Brittany, France. Algal fragments were subjected to daily repeated exposures of artificial UVBR that were realistic for springtime solar UVBR at the water surface in Brittany. Growth, DNA damage, photoinhibition, and UV‐absorbing compounds were monitored during 2 weeks of PAR + UV‐A radiation (UVAR) + UVBR, whereas PAR + UVAR and PAR treatments were used as controls. The littoral species showed a higher UV tolerance than the sublittoral species. After 2 weeks, growth of P. palmata and C. crispus was not significantly affected by UVBR, and DNA damage, measured as the number of cyclobutane‐pyrimidine dimers per 106 nucleotides, was negligible. Photoinhibition, determined as the decline in optimal quantum yield, was low and decreased during the course of the experiment, coinciding with the production of UV‐absorbing compounds in these species. In contrast, no UV‐absorbing compounds were induced in the sublittoral species. Growth rates of P. pseudoceranoides and R. pseudopalmata were reduced by 40% compared with the PAR treatment. Additionally, constant levels of DNA damage and pronounced photoinhibition were observed after the UVBR treatments. Growth was completely halted for Phycodrys rubens and Polyneura hilliae, whereas DNA damage accumulated in the course of the experiment. Because Phycodrys rubens and Polyneura hilliae showed the same degree of photoinhibition as the other sublittoral species, it appears that the accumulation of DNA damage may have been responsible for the complete inhibition of growth. The results suggest an important role of DNA repair pathways in determining the UV sensitivity in red macrophytes.