INTRODUCTION
Estimating parameters such as population size and minimum viable population size is becoming a key component of population biology. With the pressing issues of habitat loss and fragmentation, conservation biologists are focusing more on the importance of spatial structure, whether it refers to genes, individuals, populations or communities. In the last few decades, many studies have been conducted that aim to quantify populations of different species ranging from fast-moving organisms, such as mammals and birds, to more sedentary species such as those present on rocky shores.

The intertidal zone is the shoreward fringe of the seabed between the highest and lowest extent of the tides (Levinton, 1996). The uppermost part of the intertidal zone is very calm, but on open coasts' waves may splash many feed above normal levels. There are many variations among rocky shores, depending on the slope, texture, size, and shape. They may consist of sudden cliffs with immense walls to gentle slopes that seem to merge with the sea. Rugged pilars and pinnacles, eroded by constant wave action form disfigured shapes.

There are predominantly five zones, characterized by the organisms that live in each zone and that zone's distance from shore. The zones are as follows: white, gray, black, yellow, and pink; in order from closest to shore, to furthermost extent. The white zone may be distinguished as the zone that is literally the shore itself and may contain different sorts of plant life and/or land animals. The gray zone is the widest of the zones above the intertidal. Its upper portion may remain dry for days on end; its lower margin is close to the intertidal and may receive almost daily immersion. The black zone is wetted completely during spring tides and some high tides, but is mostly dry. The yellow zone is truly the intertidal zone. It is inhabited by many exclusively marine rather than amphibious organisms. Immediately below the lower yellow zone on vertical rock faces there is a pink encrusting layer of algae known as the coralline lip (Levinton, 1996). Close examination of the crust may reveal deposits of calcium carbonate, and tube-building organisms.

The site chosen for this study is a rocky intertidal habitat formed by a fossilized mangrove forest located on the northeastern coast of Key Biscayne, in Crandon Park. Rocky intertidal communities are very dynamic and rich in species composition, making them optimal study sites for many biological processes such as predation and competition, as well as population dynamic studies, such as species recruitment, distribution and abundance.

In this study we attempt to evaluate the applicability of two techniques for estimating population size, the Lincoln Index and a random sampling approach. The Lincoln Index is a capture/recapture technique and it is the simplest variant of the many capture/recapture methods available. The random sampling approach is an actual census of the organism of interest using random quadrats in which size and abundance are determine by the area of the habitat available for the organism being considered.

The objective of the study is to look at the population dynamics of the bleeding tooth nerite, Nerita peloronta, a gastropod of the rocky intertidal communities of South Florida and the Caribbean, by starting from estimating its density on Crandon Park. This species was chosen for two main reaosns: it is easy to identify, due to a red spot in the inner edge (therefore bleeding tooth) and it is not extremely abundant on the reef. It is interesting to look at population dynamics to see the way a particular species interacts with other species. Observing competition and predation may provide some insight to how the ecosystem/community operates. Population dynamics may also be key to conservation purposes.

METHODS
Site Selection
The site selection was a rocky intertidal area on the northern part of Key Biscayne, FL. The rocky intertidal borders Bear Cut, which is a tidal cut between Key Biscayne and Virginia Key, connecting the Atlantic Ocean with Biscayne Bay. The intertidal area is a carbonate platform produced by the fossilized roots of black mangroves. It is cut by small sinkholes and eroded into a rough topology. The area experiences a semi-diurnal tide with an average range of 0.8 m.

Physical Characterization
The area of this site was measured using surveyors tape and compass. Length and compass heading of the edges of the site were plotted and approximate area was calculated from a scaled plot. The scale used to map the site and calculate the total area was 1 cm to 500 cm. The profile of the intertidal zone was measured by chain transect over the same distance. The ratio of chain distance and linear distance provide an estimate of the rugosity and complexity of the habitat.

Biological Characterization
A species diversity check-list was compiled by observation throughout the whole area of the rocky intertidal community (Table 1). Species abundance and distribution of rocky intertidal fauna and flora were measured using a 1m² best quadrat sampling method along a 25 m transect running from inshore to offshore. Species abundance and diversity were recorded. In addition, species diversity of the subtidal portion of the reef was measured using the point intercept technique (16 points/m²). Species distribution was obtained by correlating species abundance and diversity to substrate complexity and distance from shore.

Population Size Estimate
Two methods were used to estimate population size, the Lincoln Index and the random quadrat sampling. The Lincoln Index is the simplest form of mark/recapture approach. On a first visit, bleeding tooth shells were exhaustively searched and marked with red nail polish. These shells are referred to as V1. On a second visit, two markers were used: black and purple. Black was used to remark the shells that were recaptured (V2) and purple was used to mark the newly found shells (V3). The Lincoln Index estimates the population size by using the following argument: the ratio of the shells marked on the first visit (V1) to the total number of shells (N) is the same as the ratio of the number of remarked shells (V2) to the total captured on the second visit (V2 + V3) (Roughgarden, 1997). That is:

where:

N = the total # of individuals
V1 = number of shells captured and tagged on the first visit
V2 = number of shells captured and tagged on the second visit
V3 = number of shells captured and tagged on the third visit

The population estimate was calculated on the computer program MATLAB. Because this is an estimate, there is also a standard error to be considered. The standard error means that 95% of the time the population estimate ranges between a value that is N + two times the value of the standard error. The standard error was calculated using the equation given by the Lincoln Model and the computer program MATLAB. The equation is:

It is important to note that this method is reliable only when as many individuals as possible are tagged during the first visit. Fifty percent are recaptured on the second visit. This is to minimize the variance in the data.

The random quadrat sampling consisted of haphazardly placing a 1 m² on the rocky intertidal and counting the number of bleeding tooth shells present in the quadrat. This method is based on the assumption that nerites are randomly distributed on the rocky platform. A total of 100 m² were sampled. The number of N. peloronta was estimated by multiplying the average number of shells in a 1 m² by the total area of the study site.

RESULTS
Using a scaled plot of the study site, the total area of the reef was estimated to be 2316.2 m². Furthermore, the slope of the reef was very gentle (Figure 2). For this reason, this rocky intertidal community is characterized by a very large yellow zone with very narrow white, gray and black zones. The southern half of the reef had higher complexity than the northern half. That is, it had a greater number of depressions, offering a more diverse habitat for marine organisms. The northern part, on the other hand, was quite flat, offering little refuge and shade to organisms.

Species Abundance
Species abundance was measured by direct count of organisms using a 1 m² quadrat along a transect runnning from inshore to offshore for 25 meters. The most abundant organism was the snail Batillaria minima, reaching numbers ars high as 1200/m² (Figure 3). The abundance of this species decreased in quadrats closer to the offshore edge of the reef. Whiel the number of B. minima decreased with distance from mean high water (MHW), the overall diversity of the mollusk community increased, showing the typical trend of a rocky intertidal habitat (Figures 4 and 5). Figure 5 shows the distribution of mollusks in the rocky intertidal community.

Point Intercept
Point intercept technique was used to observe frequency and abundance of the subtidal community of the study site (Figure 6). Beside high algal cover, organisms such as echinoderms (sea urchins), bivalves, chitons and sponges were recorded. Halimeda sp., a calcareous green alga, was the most frequently observed algae species (Table 2). Finally, mollusk density was correlated to the complexity of the substrate (Figure 7). Mollusk density and diversity were greater on moderately complex than on highly complex substrate.

Population Estimate, Capture/recapture
Field measurements:
On the first visit, 120 N. peloronta were marked with red nail polish (V1). On the second visit 63 were recaptured (V2) and 62 new individuals were found (V3). The population (N) is estimated to be 238 individuals +29 (see methods for a description of Lincoln Model). Therefore, the population estimate ranges between 209 - 267 individuals (95% confidence interval)

Random Sampling
The number of N. peloronta was estimated by multiplying the populatiohn density (mean numnber of shells per quadrat) by the total area of the site. The number of individuals per quadrat was calculated to be 0.6/m² + 1.5 (mean + sd, N=100). The total area was estimated to be 2316.2 m². Therefore, according to random sampling, population size of beeding tooth nerites is 1,274 (Table 3).

DISCUSSION
Site Characterization
The rocky intertidal community of Crandon Park is a carbonate platofrm cut by depression in eroded areas. The vertical structure is caused by erosion rather than steep inclinations. Therefore, the site is characterized as a flat, gentle rolling, wide yellow zone. The most abundant species on the rocky intertidal platform was the gastropod Batillaria minima. However, the abundance of this species decreased in quadrats closer to the offshore edge of the reef. This may have occurred because Batillaria minima cannot withstand wave action since it lacks the capability of having a strong grip on the rocks. Species abundance and diversity increases within quadrats closer to the water. This is explained by several factors such as less exposure to dryness and high temperatures, higher complexity of the substrate closer to the water and greater food availability. Species living on rocky shores have to withstand very stressful condidions. The further from the water they are located, the more difficult life becomes. Also, the wave action on the edges of the reef erodes the reef away, creating a more structured substrate that, in turns, offers a better habitat for organisms to live. Finally, with the presence of water, the availability of food increases such as turf, encrusting algae, and other small organisms. This means that as the complexity of the substrate increases so does the complexity of the biotic community. Figure 5 shows the relative density and diversity of mollusks along the 25m transect. Even though the rocky platform is mainly represented by a yellow zone, the mollusks are distrubted along a gradient within it, and follow a specific pattern from the upper to the lower edge of the yellow band. Species present from the 22^nd to the 25^th meter were not recorded in preceding quadrats.

The point intercept technique was used to observe frequency and abundance of organisms of the subtidal community. Figure 6 shows the frequency distribution of fauna and flora of this portion of the reef, which is always submerged. The algae of the genus Halimeda were the most abundant among all other species. However, new organisms typical of the subtidal zone were recorded such as oysters, loggerhead sponges, coralline algae, and echinoderms. Finally, mollusk density was correlated to substrate complexity. This was greater on moderately complex substrates rather than more complex substrates and it is in contrast to what is normally observed in this type of community. Unsualy, the more complex the substrate is, the more space available, and the more abundant the number of species is found. One possible explanation for this result may depend on the large number of B. minima counted along the transect. One way to avoid this situation would be recording individual species separately rather than all together.

Population estimate
Two values of population estimates were obtained using the Lincoln Model and the random sampling approach, 209-267 and 1274-3600, respectively. The most accurate population estimate is probably somewhere in between these two sets of values. In fact, the Lincoln Index underestimates the population size of Bleeding Tooth Nerites, while the random sampling method overestimates it. The first model relies on an exhaustive search of the shells for both visit one and visit two. Bleeding tooth is a mollusk found on the upper boundary of the intertidal community. For this reaosn it is exposed to air and high temperatures for longer periods of time than other organisms. During low tide it seeks shade, making it very hard to be seen. On the recpaturing day the tide was low and the temperature very high, and we were able to recapture only 50% of the shells tagged during the first visit. We returned to the site at high tide and we observed a high number of individuals that were never tagged. This suggests that we were unable to thoroughly search the reef the first day, and that a significant number of individuals were missed during the count.

On the other hand, random sampling overestimates the population size. the numbers we obtainewd with this method ranged from 1274 to 3600, therefore showing a great variability. This model is based on the assumption that Bleeding tooth is randomly distributed throughout the reef. However, it is not. As mentioned above, they are mainly distributed on th uppermost zones of the rocky platform and where the substrate offers enough shade and refuge. Therefore they are not distributed along a uniform gradient but in patches, depending on the substrate morphology.

Recommendations
For future studies of this sort it is recommended to visit the site at high tide rather than low tide in order to maximize capture rate of the Bleeding Tooth. The random sampling should be avoided all together because it overestimates the population size. Instead, random habitat stratified sampling should be used. This method takes into acocunt the distribution of the organism and habitat structure, allowing to invest greater sampling effort on zones where the organisms are sure to be found.

References

Brehaut, Roger N. (1982). Ecology of Rocky Shores, 58pp. London: Edward Arnold.

Carson, Rachel (1971). The Rocky Coast, 118pp. New York: The McCall Publishing Company.

Earll, R. and D. G. Erwin (1983). Sublittoral ecology: The ecology of the shallow sublittoral benthos, 277pp. New York: United States by Oxford University Press.

Kaplan, Eugene H. and Roger T. Peterson (1988). Southeastern and Caribbean Seashores, 425pp. New York: Houghton Mifflin Company.

Levinton, Jeffrey (1996). Marine Biology 271-288pp.

Newell, R.C. (1979). Biology of Intertidal Animals, 781pp. Marine Ecological Surveys LTD, Faversham, Kent. U.K.

Raffaelli, David and Stephen Hawkins (1996). Intertidal Ecology, 356pp. London: Chapman & Hall.

Roughgarden, J., (1998). Primer of Ecological Theory, 456pp. New Jersey: Prince Hall, Inc.

Stephenson, T.A. and A. Stephenson (1972). Life Between Tidemarks on Rocky Shores, 423pp. San Francisco: W.H. Freeman and Company.