Abstract Middle-late Miocene (ca. 13-10 Ma) Ponce Limestone exposures in southern Puerto Rico provide an opportunity to evaluate development of a tropical carbonate ramp system during a time of known regional upwelling in the Caribbean. Three sequences (DS1, DS2, and DS3) developed in response to relative sealevel fluctuations. Each sequence is characterized by basal heterozoan-larger benthic foraminifera (LBF) facies that grades upward to a photozoan facies composed of corals tolerant of cool and turbid water at the top. DS1 transgressive deposits include Kuphus (?incrassatus), Amphistegina-Archaias packstone interbedded with Amphistegina packstone, and Archaias angulatus and gastropod-rich packstone. Maximum flooding is indicated by a Globigerinid planktonic foraminiferal facies. Upper DS1 strata consist of Montastraea imperatoris, Goniopora imperatoris , and several species of Porites coral rud-floatstone and framestone, which were deposited during highstand and sea-level fall. DS1 is capped by a surface of subaerial exposure (SB1). DS2 transgressive deposits consist of Amphistegina -coralline red algae packstone-grainstone that grade upward to coralgal- Amphistegina packstone deposited during highstand and sea-level fall. DS2 is capped by a surface of subaerial exposure (SB2). A rapid sea-level rise for DS3 is interpreted due to the apparent lack of transgressive deposits. Preserved strata consist of prograding coralgal clinoforms developed during highstand. SB1 (~13-12 Ma) and SB2 (~11-10 Ma) may correlate with unconformities in other Caribbean areas, which could indicate regional tectonic or eustatic control on sequence development. The dominance of heterozoans and larger benthic foraminifera tolerant of mesotrophic and temperate water conditions and the presence only of those photozoan corals tolerant to turbidity and cooler water are consistent with a system affected by upwelling. The presence of photozoan corals only in the highstand and regressive portions of sequences suggests highest upwelling intensity and/or transport of upwelled water and nutrients to shallowest water ramp environments during transgressions. Our results have direct implications for other similar age-equivalent systems developed in the Caribbean, including those forming important reservoirs, and other tropical systems in the rock record affected by adverse photic zone conditions.

Abstract From October 5–8,1985, a tropical wave centered about 25 km northeast of Ponce, Puerto Rico, produced as much as 560 mm of rainfall in 24 hours and as much as 70 mm in one hour. This extraordinarily heavy rainfall triggered a rock-block slide that destroyed much of the Mameyes residential area, on the northwest outskirts of Ponce. The Mameyes landslide failed in three distinct phases between 3:00 and 4:00 a.m. (local time) on October 7, 1985. The first two phases of sliding involved translational sliding of two 12-m-thick slabs of calcareous-sandstone bedrock along bedding-plane surfaces parallel to the slope surface. The third phase of sliding involved the toppling failure of a block that disaggregated and formed a rock fall on the western part of the slide. Subsidiary flow failures onto the toe and from the downstream face of the toe were triggered by the heavy rainfall and the rupture of a water pipe that emptied as much as 4 million liters of water onto the slide. The Mameyes landslide is approximately triangular in plan view; the maximum width and length are both about 250 m. The total area of the landslide is ~35,000 m2, and the slab of bedrock and soil that failed is about 12 m thick and comprises ~300,000 m3. The landslide moved downslope 30 m, parallel to the 17°–24° bedrock dip. At most, an estimated 120 homes were destroyed by the landslide, and at least 129 people were killed, though only 39 bodies were recovered. This death toll is the largest from a single landslide in North American history. Calculations of the prelandslide slope stability at the site indicate that when the water table was below the potential shear surface, the factor of safety against failure was about 1.26. A water table about 6 m above the shear surface, which would saturate half the thickness of the slide block, reduces the factor of safety to 1.0. A water table at this level is a reasonable result of the conditions preceding failure. The landslide area has been stabilized by engineering measures and was developed as a memorial park to the landslide victims.

Of the many types of landslides common to Puerto Rico, debris flows—the mobilization and flow of rock and soil down steep slopes—are the most abundant in many areas. On October 5–8, 1985, a tropical storm produced extreme rainfalls and consequent widespread debris flows along the south-central coast of Puerto Rico. Locally, 24-hr rainfall exceeded 560 mm, 4-day rainfall exceeded 750 mm, and intensities reached 70 mm/hr. Most of the flows occurred in an area where Tertiary limestone and detrital sediments form 20° to 40° slopes covered by less than 2 m of colluvium or residuum. Many areas received more rainfall and had steeper slopes but contained far fewer flows. The distribution of flows probably was caused by a small cell of very intense rainfall or by localized engineering properties of the colluvium that rendered it more susceptible to failure in the area of concentration. Most of the flows resulted from storm-induced buildup of pore pressure at the colluvium/bedrock contact. The debris generally failed in disk-shaped slabs as much as 15 m across and 0.5 to 1.5 m thick. All the colluvium in the source areas mobilized and exposed bedrock surfaces parallel to the original ground surface. The flowing debris scoured surficial soil down to bedrock and destabilized preexisting gully walls, which triggered additional thin debris slides into the channel. Material contributed by channel scouring and side-slope debris sliding commonly made up 90 to 95 percent of the debris deposit. Debris flows are recurrent phenomena in southern Puerto Rico. Several flows incised channels into older debris-flow deposits, and in some areas, as many as three successively older debris deposits were exposed. Preliminary dating of debris-flow deposits suggests that recurrence intervals for flows at a given site could range from several decades to several hundred years.