produced the extensive fish mortality observed in the study areas: (1) air and water 
temperatures maximas in late August were very high; (2) dissolved oxygen 
concentrations on August 23 experienced a rapid decline with a dissolved oxygen 
minima approaching short-lived hypoxia/anoxia, whereas persistent, long-lived 
hypoxic/anoxic conditions were observed in the Buttonwood Canal areas; and (3) large 
charges in the spring-neap lunar tidal cycle, the seasonal oceanic water level cycle and 
poor water circulation. Higher than average tides associated with higher than normal 
water levels, particularly in Snake Bight, may have permitted greater access to the 
shallowest inlets along the mainland. Subsequently, when the warmest waters were 
unable to provide sufficient oxygen (e.g., Garfield Bight), a shallow water "refuge- 
death trap" phenomenon occurred. It is believed that this is the first time that hypoxic 
conditions have been documented during a Park marine fish kill event. In 1990, three 
fish kills occurred at the same time in both the Snake Bight and Garfield Bight areas 
east of Flamingo. They were described as having the same general pattern as noted in 
1993 kills except that a larger number of fish may have died. No water quality 
measurements were made during the 1990 die-off. However, because the similarities 
involved, it is quite possible that the kills in earlier years were also caused by oxygen 
depletion. Peak periods of annual fish kills in the southeastern United States usually 
occur during mid-summer to early fall. These fish kills may result from stress related 
hypoxia (dissolved oxygen values <5.0 ppm) associated with high air/water 
temperatures, large rainfall events, extreme low water and or a combination of these 
factors. It is this time of the year dissolved oxygen concentrations in estuarine and 
coastal marine waters are experiencing their lowest levels of the year. Although 
findings have been significant with respect to the cause of the fish kills in the Flamingo 
area, they have also highlighted the complex and difficult nature of the Florida Bay 
ecosystem. It is not yet known what, precisely, from a myriad of sources contribute to 
hypoxia/anoxia in the Snake Bight area. It may represent a natural phenomenon with 
which the fishes are in a delicate balance. It should be noted that stresses associated 
with eutrophication could affect the oxygen budget as well as decomposition of organic 
material (e. g. seagrass die-off), and nutrient over-enrichment could result in hypoxic 
conditions. 
1990 0 
Textoris, S. D. (1990) Patch reefs in the Pleistocene of South Florida and their 
implications. The Compass . 67(2): 115. 
[ABSTRACT ONLY. DATE OF SAMPLING UNKNOWN OR NOT APPLICABLE.] The Pleistocene 
underlies much of the southern tip of the Florida mainland and extends southward 
through the Keys and beneath Florida Bay. Five units, designated Q1 through Q5 (oldest 
to youngest), are separated unconformably by freshwater limestones and supposed 
subaerial(?) crusts. The upper part of the sequence (Q4 and Q5) consists of the Miami 
Limestone for which three facies are recognized, coral, ‘bryozoan', and oolite, each the 
result of a different environment — shelf-edge, quiet protected back—reef, and shallow 
agitated marine. The lower part of the section, tentatively correlated with the Fort 
Thompson Formation, consists of quartzose sandstone (Q1 and Q2) followed by 
carbonate deposition (Q2 and Q3) as the source of elastics diminished and the water 
became shallower. The reef of the Key Largo Limestone (in part the coral facies of the 
Miami Limestone) began growing along the shelf edge in its present position in Q3 time 
contemporaneously with the development of patch reefs in the Bay area. At least three 
patch reefs are known and have been outlined by diamond-drill 11 coring and another is 
suspected in the backreef area. The presence of a configuration similar to present-day 
conditions supports the supposition that the Keys formed in part as a barrier reef 
during the late Pleistocene. Termination of the reef growth occurred when they either 
341 
