Diagnostic study of the lakes Laborde (or Lake Cocoyer), Lachaux, and Douat to identify zones of protection.

Dr. Donald Huggins and Debra Baker

Kansas Biological Survey Report #181

Submitted 29 May 2015

to

Comité Interministériel d’Aménagement du Territoire

Kansas Biological Survey University of Kansas 2101 Constant Ave. Lawrence, Kansas 66047 USA

Diagnostic study of three lakes in southern Haiti 1 Contents Executive Summary ...... 4 Introduction and Background ...... 5 Weather and Climate ...... 8 Geology and Geomorphology ...... 8 Lakes and Lake Watersheds ...... 10 Study Objective ...... 16 Methods ...... 16 Land use and land cover (LULC) of watersheds ...... 17 Water level, Secchi depth, in situ water chemistry ...... 17 Nutrients and bacteria ...... 17 Macroinvertebrates ...... 18 Plankton ...... 18 Fish ...... 19 Macrophytes ...... 19 Results ...... 19 Bacteria ...... 19 Nutrients ...... 23 In situ water chemistry ...... 23 Macroinvertebrates ...... 24 Plankton ...... 24 Fish ...... 25 Macrophytes ...... 28 Landuse/landcover ...... 30 Slope ...... 34 Population ...... 38 Community meetings ...... 38 Springs and wells...... 42 Watershed findings and conclusions ...... 43 Soil erosion and land use ...... 43

Diagnostic study of three lakes in southern Haiti 2 Precipitation and runoff ...... 44 Lake quality and uses ...... 44 Identification of protection and restoration zones and areas...... 45 Protection area summary ...... 53 Institutional management options and actions ...... 57 Structural erosion and protection options and actions ...... 58 Non-structural erosion and protection options and actions ...... 59 Some future considerations ...... 59 Acknowledgements ...... 60 References ...... 61 Appendix A. Files and methods used by the Kansas Applied Remote Sensing Program to create Haiti watersheds and calculate slope. Written by Jerry Whistler...... 64 Appendix B. Study locations with approximate altitude...... 66 Appendix C. Photographs of fish collected during this study at Etang Lachaux and Laborde. ... 67 Appendix D. Fishes of Haiti with occurrence classes...... 69 Appendix E. Photographs of macrophytes collected during this study at Etang Lachaux and Laborde...... 71 Appendix F. 2011 landuse/landcover maps of each lake watershed...... 74

Diagnostic study of three lakes in southern Haiti 3 Executive Summary

Haiti’s Comité Interministériel d’Aménagement du Territoire (CIAT), an inter-departmental government agency, has been tasked with identifying and delineating zones within watersheds to protect against environmental degradation. As such, CIAT contracted the Kansas Biological Survey (KBS) to perform a study of three watersheds located northwest of Les Cayes in Haiti’s Southern Peninsula: the watersheds of Etang Lachaux, Etang Douat, and Etang Laborde. The goal of the study was to perform a bioassessment of the lakes and evaluation of landuse within the watersheds in order to make informed decisions about the best places and methods to protect or begin recovery activities.

Deforestation and erosion of Haiti’s land has been well documented. Less is known about Haiti’s aquatic systems and how landuse activities impact the ecological services provided by these ecosystems. Valuable ecological services include fish and bird habitat, biodiversity, hunting and fishing, and drinking water for humans and livestock. This study examines the physical (depth), chemical (in situ and nutrient parameters), and biological (zooplankton, macroinvertebrates, fish, phytoplankton, macrophytes) components of the lakes as well as the springs and wells located in the watersheds. Current and historical watershed landuse and landcover was examined through geographic information system (GIS), observation, and discussions with community members. The resulting information was used to delineate priority zones for protection or restoration, as well to make recommendations for protection and restoration activities. Specifically, in order of priority, the zones that should be protected or restored within each of the three watersheds are: 1. All areas within a 50-meter buffer or 5 hectare area (whichever is larger) around the most prevalent lake shoreline AND all areas within 50 m of stream banks AND all areas within 100 m of springs. 2. High slope (>30%) areas outside of the upper watersheds. 3. High slope (>30%) areas within the upper watersheds.

Diagnostic study of three lakes in southern Haiti 4 Introduction and Background In May 2013, upon request of the Ministry of the Environment of the South Department of Haiti, the Central Plains Center for BioAssessment (CPCB) performed a preliminary baseline study of Etang Lachaux, a 54 hectare shallow lake located between Les Cayes and Camp Perrin. The goal was to evaluate if the lake could sustain a harvestable fish population. Etang Lachaux is one of a series of shallow lakes that stretches east to west in this area of the southern peninsula (Figure 1). In January 2014 in conjunction with returning to the area to report our Etang Lachaux findings, we did a walking assessment of Etang Laborde, a 36 hectare lake located 5 km the southeast. The Haitian organization Comité Interministériel d’Aménagement du Territoire (CIAT), and inter-departmental government agency whose mission includes watershed management and protection, then funded us to perform a full-bioassessment in March 2015 on the largest three lakes in the region, Etang Lachaux, Etang Douat, and Etang Laborde (Fig. 2 and 3). The Organization for the Rehabilitation of the Environment (ORE), a non-profit group Camp Perrin, coordinated field logistics, including providing drivers and a translator and meeting space.

Figure 1. Location of the study watersheds (blue) in the western half of Haiti’s southern Peninsula.

Diagnostic study of three lakes in southern Haiti 5

Figure 2. Topographic map showing the location of the study lakes near Camp Perrin Haiti in the South Department. The new National Route 7 highway is not on this map. Topographic lines are 20 m intervals. Watersheds are delineated in light blue.

Diagnostic study of three lakes in southern Haiti 6

Figure 3. Aerial photo map showing the location of the study lakes near Camp Perrin Haiti in the South Department. The new National Route 7 highway is not on this map. Watersheds are delineated in light blue.

Diagnostic study of three lakes in southern Haiti 7 Weather and Climate Haiti has a hot and humid tropical climate and is semiarid where the mountains in east cut off the trade winds. Haiti is characterized by diurnal temperature variations that are greater than the annual variations and are modified by elevation. The southern region has two rainy seasons, lasting from April to June and from August to October, whereas other regions experience rainfall from May to November. Average annual rainfall at Camp Perrin was 87.3 inches (221.7 cm) for the 20-year period beginning in 1993 and ending in 2012 (http://www.oreworld.org/rainfall.htm, accessed April 2015). Annual variations of precipitation can cause droughts, widespread crop failures, and famine. The southern peninsula is more vulnerable to hurricanes (tropical cyclones) than other parts of Haiti. The average annual rainfall is 140 to 200 centimeters, but it is unevenly distributed. Heavier rainfall occurs in the southern peninsula and in the northern plains and mountains. Rainfall decreases from east to west across the northern peninsula. The eastern central region receives a moderate amount of precipitation, while the western coast from the northern peninsula to Port-au-Prince, the capital, is relatively dry.

Geology and Geomorphology The dominant topography of the central and western portions of the island of Hispaniola is reflected in Haiti’s name that is derived from the indigenous Arawak place-name Ayti or “mountainous land” and is also the local name for cone karst. Madden and Minson (2010) state that the mountain chains of Haiti are often composed of cone karst with many sinkholes, springs, caves and caverns and disappearing and losing streams with few continuous stream or river systems (Figure 4).

Cone karst topography, which is common in the tropics, is typically characterized by closed depressions (e.g. sinks, sinkholes) located at the base of many steep-sided, cone-shaped hills that often create narrow, steeply-walled valleys. The following figure from the literature illustrates typically karst features, many of which can be seen throughout the southern peninsula of Haiti and in our study areas.

These sinks and sinkholes can be cylindrical, conical, bowl- or dish-shaped and range from a few meters in diameter to hundreds of meters. The study lakes appear to be part of a series of ancient depressions that have become sealed and fill with water during wet periods. This interpretation is, in part, supported by McFadden’s study of the geography setting of nearby Macaya and La Visite National Parks (MacFadden 1982). He note that within and on the

Diagnostic study of three lakes in southern Haiti 8

Figure 4. Karst topography landscape cross section from http://www.slideshare.net/valentic/planet-earth-groundwaterpowerpointpresentation

limestone substrate an extreme karst topography has developed with numerous solution features such as sinks, sinkholes and vertical-walled pipes (openings to the surface from caves and caverns). There are two different mechanisms for the formation of sinks and sinkholes: solution and collapse. Caves tend to migrate to the surface through repeated ceiling collapses that systematically raise the roof (collapse) and floor (deposition) of the cave until the ceiling collapse breaks to the surface of the landscape. The other cause of sinkholes and sinks is related to the corrosive solution of limestone by rainwater in areas where cracks in the limestone promotes water movement into the subsurface limestone formations. This normally process typically forms the bowl-shaped type of sinks. All of our study lakes as well as three other lakes in this region tend to be shallow, bowl-shaped waterbodies with large surface areas when completely full. That these lakes are likely to be water-filled sinks is further suggested by the fact that the karst solution process often produces large amounts of clay that can be deposited in these bowls which then act to seal the bottom of the sink leading to the creation of little lakes of rain water. These sinkhole lakes are a rare aquatic feature in karst areas where most precipitation rapidly infiltrates into the caves, underground rivers and other subterranean karst features.

In addition, a complex system of underground caves and caverns has developed throughout this region and there are many examples of dry valley collapses that are also called hanging valleys. These hanging valleys can be observed throughout the region where these lakes occur. The hanging valleys often lack a distinct surface water connection to the lower main valley or valleys because surface flows into these valleys drain to the underground karst drainage systems. As is

Diagnostic study of three lakes in southern Haiti 9 characteristic of karst regions, much of the precipitation in this region enters the hydrologic cycle by vertical movement into the subsurface through joints and solution features.

Lakes and Lake Watersheds The oldest reference to Etang Lachaux that we found was from a 1934 expedition by P.J. Darlington Jr. to collect ground insects. He said “On October 26th and again on the 27th I collected along the shores of Etang Lachaux, a fine, small lake an hour's walk over a ridge east of Camp Perrin. This was perhaps the best single locality I found below 1,000 ft. for ground collecting.” This paints a picture of a vegetated landscape surrounding the lake at the time of Darlington’s expedition. Now all that remains are few large trees, shrubby plants, and crops such as corn planted up to the receding lake edge and on the hillsides overlooking the lake. As with much of Haiti, deforestation has led to bare hillsides prone to erosion during heavy rains. To our knowledge there are no published accounts describing the historic or even more current (1940s to 1990s) landscape characteristics and land use within the watersheds draining Etang Douat or Laborde.

While the land use and land cover of these watersheds is only vaguely documented, it seems safe to assume that they were historically forested but have suffered from continuous deforestation and intensive agricultural grazing and cultivation. As of the date of this report these watersheds continue to suffer from massive, basin-wide erosion with little or no organized or programmatic efforts to contain or reduce soil erosion and instill land management practices.

These karst watersheds are typical of the karst landscape of the southern peninsula and as such share similar topographic, geological and soil characteristics. Soils are predominately Udepts – new, shallow soils associated with forests. In many areas of these watersheds, these thin, exposed forest soils have quickly eroded from the high-slope terrain that characterizes all the watersheds in this mountainous region. It should be noted that karst environments such as these watersheds have been recognized as important landscape worthy of protection (Day 2011). Karst landscapes play a critical part in water conservation and the maintenance of biodiversity and much of the world’s karst areas lay within developing countries like Haiti. Day (2011) lists a number of reasons for the protection and preservation of karst landscapes. They include: 1) important areas for scientific study across a variety of disciplines; 2) religious and spiritual areas; 3) areas of specialized agriculture and industry; 4) critical areas to the understanding of regional hydrology; 5) recreation and tourism areas with important economic and aesthetic value; 6) habitats for endangered species of flora and fauna; 7) areas possessing rare minerals and/or unique landscape features; and 8) important historic and prehistoric areas with cultural importance. Only about one percent of the karst regions in Haiti is offered some level of protection from human impacts and land use changes (Kueny and Day 1998).

Diagnostic study of three lakes in southern Haiti 10 Recognizing the importance of karst landscapes, and their protection in Haiti, adds incentive to the protection of these study lakes thus assuring the simultaneous protection of these karst watersheds.

Much of what we know about these watersheds comes from two primary published sources – a 2007 thesis by Jean-Louis Enold and a 2012 land use and land cover baseline report produced by The Earth Institute, Columbia University in partnership with Catholic Relief Services and the Organization for the Rehabilitation of the Environment (ORE), Haiti. Enold’s thesis concentrated on landownership, erosion, agriculture uses and management practices within the Laborde watershed. Both of these publications proved of value and are cited on a number of occasions in reference to general observations in some or all of our study watersheds and lakes. However, we have used our own estimates of a number of lake characteristics and watershed features since our methods and techniques provide more pertinent accuracy and precision than that of Enold’s and the Earth Institute reports.

During the last 60 years (± 10 years) water level of all of these lakes has been reported to fluctuate from local flood to near dry conditions (personal communications, Lachaux, Douat and Laborde community meetings, March 2015). The current study and the past study of Lachaux were both conducted during low water level and rainfall conditions as noted by local authorities and the historic rainfall records of ORE (Table 1). The average rainfall in 2013 and 2014 was 136 and 167 cm, about 32% less than the previous 20 year average with 2013 having the lowest rainfall on record for Camp Perrin. Thus it should be noted that study results may not represent typical lake conditions due to the drought-like conditions that preceded lake sampling events.

Table 1. Watershed and lake size (square meters) at and before the time of this study. Watershed area was determined from boundaries created in GIS from ASTER GDEM source data (see Appendix A for details). 2011 lake area was determined from Earth Institute (2012) landuse/landcover calculations. 2013 – 2014 lake area was determined by authors by recording lake perimeters while walking with handheld GPS unit. Lake Etang Lachaux Etang Douat Etang Laborde 2011 lake area m2 585,874 119,246 470,504 2013 May lake area m2 539,361 -- -- 2014 January lake area m2 -- -- 389,346 2015 March lake area m2 676,252 0 363,744 Watershed area m2 3,684,745 2,710,137 7,128,851

Low-water and no water lake conditions as experienced in our studies also impacts land use and land management practices that more directly affect the lake ecosystem and water quality.

Diagnostic study of three lakes in southern Haiti 11 While general land use within each watershed is thought to be fairly similar from year to year, the area immediately surrounding the lakes dramatically changes during low and no water conditions. Farmers nearest the lake and with access to the lake bottom lands exposed by decreasing water levels immediately put the exposed lake bottom into production. The increase in agriculture cultivation around and in the lake basin is also accompanied by great livestock use of the lake for both forage and drinking water. Both the increased livestock access and increased cultivation in and near the lake bottom decrease biological and water quality.

The use of the lake basin and bottom areas (during drought conditions) in each lake are very similar as is the general land use and practices noted in each watershed (Table 2). The actual land uses within each watershed are very similar when the highly fluctuating water category (i.e. lake surface area) is removed with the biggest differences in watershed land use occurring in the “other” category and agricultural cultivation (i.e. Cultures Agricoles). The Lachaux watershed has about 25 % less cultivation in it than the Laborde watershed which might affect water quality except that portions of each watershed drain to and through large alluvial fans (e.g. alluvial plains) which are functioning to reduce sediment delivery and associated containments directly to each lake. This suspected phenomenon is discussed in the recommendations section of the report. In summary, it appears that watershed-wide (i.e. basin-wide) land use and land cover are extremely similar, thus impacts from land uses should be similar except for affects related to near lake land slope, farming and buffer conditions.

Table 2. Percent land use and land cover as determined from 2011 data provided by Earth Institute, calculated after removing the water category from the total percentage. Total watershed area estimates do include the water land cover category. See Earth Institute report (2012) for land use definitions. Land use category Etang Lachaux Etang Douat Etang Laborde Agroforestry (Systèmes Agroforestiers) 26.0% 28.2% 24.3% Grassland (Savanes) 3.1% 2.9% 2.9% Forest (Fôrets) 0.2% 0.2% 0.1% Agriculture (Cultures Agricoles) 17.0% 21.1% 22.8% Pasture (Pâturages) 29.6% 31.0% 29.1% Urban (Urbain) 0.4% 0.3% 0.5% Other (Autres) 23.8% 16.2% 20.3% Total watershed (Bassin versant) (m2) 3,684,745 2,710,137 7,128,851

Diagnostic study of three lakes in southern Haiti 12

Figure 5. Etang Lachaux showing sampling sites in white, and depths (cm) of two sites in yellow. Watershed boundary is represented by a blue line.

Diagnostic study of three lakes in southern Haiti 13

Figure 6. Etang Douat showing locations of a residual pool of water (D_hole), cave opening, and the hibiscus bramble pictured in Figure 6. Watershed boundary is represented by a blue line. Aerial photo shows water but basin did not contain water during the month of the study.

Diagnostic study of three lakes in southern Haiti 14

Figure 7. Etang Laborde showing sampling sites in white and depths (cm) of three sites in yellow. #1 is hand dug well #1, #2 is hand dug well #2, NGO well is a pump well. Watershed boundary is represented by a blue line.

Diagnostic study of three lakes in southern Haiti 15 Study Objective Our objective was to perform a baseline ecological assessment of Etang Laborde, Etang Douat, and Etang Lachaux in the South Department of Haiti (Fig. 5 - 7) and offer suggestions for zones of protection. Water quality measurements included in situ measurements and bacteria and nutrients from samples returned to the laboratory. We performed a complete biological assessment (bioassessment) including using quantitative methods to collect representative samples of the macroinvertebrates, zooplankton, and phytoplankton populations. Phytoplankton enumeration will be used to estimate productivity within the lakes. We collected representative taxa from the aquatic macrophyte communities. We examined the fish population using catch and release methods, and collected representative taxa for identification. Samples were collected both from the lake shore and from throughout the lake accessed via an inflatable boat. We walked (Douat and Laborde) or drove (Lachaux) to the top of the largest gullies of each lake to evaluate watershed land use and land cover, and collected macroinvertebrates, fish, and water samples from wells, springs, and stream pools we encountered. To reconstruct a 50-year history of watershed land use, we concluded work at each lake with a community meeting of residents who live near the lakes. A workshop was held at ORE with a group of stakeholders chosen by CIAT and ORE to learn more about their understanding and goals for this watershed and share initial results of the study.

Methods This study spanned 9 – 19 March 2015. On the first day of the study at each lake, we evaluated land use practices and vegetative cover, and familiarized ourselves with lake conditions by walking around the lake to collect macroinvertebrates and water chemistry from 3 shore locations and plant specimens to characterize the macrophyte communities. On the second day we took an inflatable boat on the lake to collect 3 water chemistry samples in conjunction with 3 phytoplankton and zooplankton samples. We also measured lake depth, Secchi depth, and in situ water chemistry from these 3 sites and an additional 8 to 12 sites distributed throughout the lake. On the third day we meet with residents to inquire of the history of lake and surrounding land use. See Appendix B for sample locations.

Etang Douat had no water, affording us an extra day to explore and document conditions in the larger watershed of Etang Laborde. We also spent a morning examining the location, number and condition of culverts along National Highway 7 where it drains to the watershed of Etang Lachaux on the west and north. Qualitative collections of macroinvertebrates and fish were collected from a stream in the Douat watershed, while in situ water measurements and water samples were collected from springs and wells in each watershed. The geographic coordinates of all sampling localities were recorded with a handheld GPS. Import permits for biological samples were obtained from APHIS and USFWS. Export permits for these same samples were obtained from the Haitian government with the help of CIAT after a quarantine period.

Diagnostic study of three lakes in southern Haiti 16 Land use and land cover (LULC) of watersheds We delineated the watershed of each lake from ASTER GDEM source data (see Appendix A for details). Additionally, we obtained LULC maps of the watersheds produced in 2011 by the Earth Institute at Columbia University for UNOPS and UNEP (Appendix F). Their methods are described in their September 2012 report (Earth Institute 2012). The maps are in jpg form thus not compatible for analysis in ArcMap. We have made a request to UNEP for the raw shapefiles of the LULC so that we can overlay the LULC with topography and determine areas where agricultural or bare ground overlap steep (>50%) slopes. These areas of high erosion are the most likely candidates for zones of protection.

LULC was verified while walking the watershed and observational notes were made. We also asked residents of the lake about the history of their use of the lake and land, and changes they would like to see, if any. Questions included flooding and drought history, fish occurrences, past tree planting efforts, and preferred use of the lake land. The population of each watershed was estimated by counting the number of houses that show in Google Earth aerial photos and photographs that we took, and multiplying this number by 5 (Cayemittes et al. 2007 report average rural household size at 4.7 people). The thesis by Enold (2007) contains extensive information about the population demographics and land ownership and practices in the Laborde watershed.

Water level, Secchi depth, in situ water chemistry At 11 – 15 locations distributed throughout each lake we determined water depth, Secchi depth and measured in situ water chemistry with a Horiba U-22 water quality probe that was 2-point calibrated prior to each sampling day. In situ measurements included pH, turbidity, conductivity, salinity, oxidation reduction potential (ORP), temperature, and dissolved oxygen. Time of each measurement was recorded.

Nutrients and bacteria To determine nitrogen, phosphorus, and the presence of Escherichia coli and other coliform bacteria, grab samples were collected at 3 of the on-lake sites, 3 shoreline sites, and at all wells and springs that we encountered. Within 9 hours of collection, samples were returned to our Haiti residence and 1 ml of each was placed on a 3M Petrifilm plate and incubated for 12-15 hours against the body (Metcalf 2010). E. coli appear as blue colonies, while other coliform appear as red colonies with gas bubbles.

Also with 9 hours of collection water samples were analyzed with a HACH DR700 portable colorimeter for ammonia, total phosphorus, and nitrate. The low-range nitrate method has a detection range of 0 to 0.5 and limit of detection of 0.024 mg/L as NO3-N. Ammonia nitrogen test is based on the modified salicylate method (Reardon et. al. 1966) with a detection range of

0 - 1 ml/L and a detection limit of 0.029 mg/L NH3 as N. Total phosphorus concentrations were

Diagnostic study of three lakes in southern Haiti 17 determined using Hach’s Ascorbic Acid method (range 0 – 2.5 mg/L PO4) to determine reactive phosphorus after acid persulfate digestion. This procedure is equivalent to USEPA method 365.2 and Standard Method 4500-P-E for wastewater with a calculated detection limit of about

0.020 mg/L PO4.

However, no nitrate samples could be analyzed using the Hach colorimeter as the nitrate module for the colorimeter was found to be faulty. In addition, before we sampled Laborde, the HACH colorimeter LED display failed so all samples were acid preserved (pH  2) and returned to the US for analysis by the Johnson County Wastewater Water Quality Laboratory (Olathe, KS) (JOCO). Samples were analyzed on April 1-2, 2015 (within the 28 day holding time) for Kjeldahl nitrogen as N (0.5 mg/l detection limit), nitrate + nitrite as N (0.02 mg/l detection limit), and total phosphorus as P (0.05 mg/l detection limit) by JOCO. This water quality testing laboratory is nationally certified under the National Environmental Laboratory Accreditation Conference (NELAC). In addition, some Lachaux and Douat water samples were brought back and analyzed by the Johnson County Laboratory as a check against onsite determinations.

Macroinvertebrates Macroinvertebrates were collected with a 500 m mesh D net using a 2 m sweep from a variety of habitats at each of 3 sites along the lake shore. Samples were placed in 25ml screw-top scintillation vials, preserved with 5% formalin, and transferred to approximately 80% alcohol 2 – 3 weeks later. Macroinvertebrates were returned to the Kansas Biological Survey at the University of Kansas for identification to the lowest taxonomic level feasible. Proportion estimates of abundance will also be calculated.

Plankton At the same 3 on-lake water sample sites, zooplankton samples were collected using an 8 inch (20. 3 cm) diameter by 20 inch (50.8cm) long standard plankton net with a 153 m mesh opening. Samples of each lake were compiled into 250ml vials and preserved with alcohol. Zooplankton tow lengths were recorded so that community metrics can be reported as units per volume. Zooplankton samples were preserved in 80% ethyl alcohol and returned to the Kansas Biological Survey lab for identification and enumeration.

At these same 3 sites, 80 ml of surface water was collected and compiled into a 250ml vial for phytoplankton. Samples were preserved with Lugol’s iodine solution. Lugol's solution was added to achieve an approximate 1% preservative concentration. These phytoplankton samples were also returned to the Kansas Biological Survey lab for identification and enumeration. Again, phytoplankton measures will be reported as units per volume, and will be used as a measure of productivity within the lakes.

Diagnostic study of three lakes in southern Haiti 18 Fish Fish community composition was assessed using a variety of collecting techniques: shoreline seining, setting baited crab nets (2) out for approximately 4 hours, and a 4-hour set using an experimental gill net. Seining was done using a polyethylene minnow seine net (1/4-inch (0.635 cm) mesh, 4 foot (1.22m) deep by 8 foot (2.44m) long) using repeated seine hauls along a 100 to 500 meter strip of the shoreline. Crab nets were baited with oat meal and anchored in about 0.5 meter of water depth while the experimental gill net was set in > one meter of water whenever possible. The gill net was 60 foot (18.3 m) long and consisted of five 12 foot panels of 0.5-inch, 1.0-inch, 1.5-inch, 2-inch and 3-inch mesh openings (made by Miller Net Company). The lakes were actively used by people and livestock, so both crab nets and the gill net could not be left un-attended which accounted for the 4-hour set period.

DNA samples were collected from about 10 fish at each site, placed in 100% ethyl alcohol, and returned to the KU Biodiversity Institute for processing and identifications. Each fish was photographed and retained as a voucher. We also retained 20-30 fish representing all taxa encountered. These fish were preserved on ± 5% formalin and returned to the University of Kansas for identification.

Macrophytes The more dominate and commonly occurring aquatic and semiaquatic plants were collected and general abundances noted so that some general statements could be made regarding their importance to the lake community. Plant specimens were hand collected, dried in plant presses and returned to Kansas University for final determination. Photos of collected plants were taken at the collecting sites as a record of their occurrence and to provide images of live material to aid identification. Plant identifications were made by staff at McGregor Herbarium at the University of Kansas.

Results Bacteria The presence of Escherichia coli has more to do with the possible human health issues associated with primary and secondary contact with lake water than it does with ecological health of the lake. A simple test for recent fecal contamination of water is to determine the level of E. coli bacteria colonies present in the water. E. coli is present in the feces of humans and other mammals. It slowly dies without multiplying once it leaves the body, but it survives in water as long as the bacteria that cause typhoid fever, cholera, and dysentery. Therefore, the presence of E. coli indicates recent fecal contamination and possible presence of these other disease causing bacteria. On the Petrifilm plates, E. coli colonies are blue, non E. coli coliform colonies (not associated with fecal contamination) are red with gas bubbles, and non-

Diagnostic study of three lakes in southern Haiti 19 coliform Gram negative bacteria form red colonies without gas bubbles. (Metcalf and Stordal 2010).

We tested for E. coli using 1 ml of water on a 3M Petrifilm plate. Most of the water sources tested contained E. coli bacteria. Only a pump well at Laborde installed by an Indian NGO produced no bacteria colonies on the Petrifilm plate (this well was retested to confirm that there was no bacteria).

One approach to the interpretation of our Petrifilm test results is the comparison between E. coli counts and relative disease risks such as those offered by the World Health Organization (WHO) (Table 3). The E. coli counts at or above 10 colonies per milliliter were confined to the small pool that was all that remained of water in Etang Douat, and in Laborde watershed a spring run and a hand dug well along the lake shore (Tables 4 and 5). However, the USEPA Puerto Rico Water Quality Standards Regulation (August 2014) list 2 E. coli colonies per 1 ml as the maximum limit. Most of the non-lake samples exceeded this limit, while only 2 of the 12 lake samples exceeded the limit. These results were not surprising in light of the amount of free ranging livestock observed in and around these water bodies.

Table 3. Correlation of Escherichia coli levels with WHO disease risk categories (Metcalf 2010). Level of E. coli WHO disease risk levela WHO action priority MSF actionb 1-10 in 1 ml High Urgent Must be treated >10 in 1 lm Very High Urgent Reject or thoroughly treat aWHO/UNICEF: A Toolkit for Monitoring and Evaluating Household Water Treatment and Safe Storage Programmes (2012), Figure A-1, p.62. bMédecins Sans Frontières (1994) Public Health Engineering in Emergency Situation. Médecins Sans Frontières: Paris.

Diagnostic study of three lakes in southern Haiti 20 Table 4. Lake water chemistry, depth, and bacteria. Shore samples (S) were collected about 1 m out from shore. Remainder (L) were collected from a boat. 2013 was collected in May with depths from 10 sites and in situ measurement from 19 near shore sites. HACH - HACH DR700 portable colorimeter. JOCO - Johnson County Kansas Wastewater Water Quality Laboratory. EPA standards (stds) are for Puerto Rico surface waters (EPA 2014). lake Lachaux Laborde EPA site 2013 x6 x8 x10 x14 x18 2015 B1 B5 B6 B9 B10 B15 B19 2015 stds sample location ave. S S L L L ave. S S S S L L L ave. date 9-Mar 9-Mar 10-Mar 10-Mar 10-Mar 14-Mar 18-Mar 18-Mar 18-Mar 19-Mar 19-Mar 19-Mar water temp C 29.47 29.71 30.19 26.87 27.03 27.02 28.16 33.94 26.03 25.44 26.65 25.99 25.18 27.22 27.21 32.20 pH 9.07 7.10 7.77 7.60 7.53 7.52 7.50 8.23 7.86 8.30 8.24 7.94 8.24 8.40 8.17 6 to 9 ORP mV 162 276 226 219 232 242 239 106 126 163 93 167 197 212 152 conductivity mS/cm 0.193 0.382 0.279 0.257 0.296 0.297 0.302 0.416 0.400 0.402 0.404 0.398 0.399 0.399 0.403 turbidity NTU 89.9 30.6 6.1 0.0 0.0 9.18 386.0 193.0 173.0 197.0 170.0 193.0 185.0 213.86 50.0

DO mg/l 9.78 8.64 9.92 4.90 3.09 1.67 5.64 10.48 5.35 7.90 7.04 7.96 6.31 8.03 7.58 5.00 TDS g/l 0.125 0.247 0.182 0.670 0.192 0.193 0.297 0.271 0.260 0.261 0.263 0.258 0.259 0.259 0.262 0.500 % salt 0.01 0.02 0.01 0.01 0.01 0.01 0.01 0.02 0.02 0.02 0.02 0.02 0.02 0.02 0.02 secchi depth cm 5.4 -- -- 60 90 96 82.00 ------10 10 9 9.67 lake depth cm 18.6 ------100 96 98.00 ------90 45 39 58.00 total ammonia (NH3-N ) mg/l HACH 0.04 0.03 0.04 0.03 0.02 0.03 0.3 ------0.16 0.16 0.26 0.22 1 Kjeldahl nitrogen as N mg/l JOCO 2 1.3 1.6 1.2 1.8 1.58 7.1 4.3 5.2 -- 4 4.3 4.5 4.90 nitrate + nitrite as N mg/l JOCO 0.17 0.15 0.16 0.18 0.15 0.16 0.15 0.21 0.22 -- 0.18 0.22 0.17 0.19 10 TN as N mg/l* calc. 2.17 1.45 1.76 1.38 1.95 1.74 7.25 4.51 5.42 -- 4.18 4.52 4.67 5.09

TP as P mg/l** JOCO 0.14 0.06 0.06 0.025 0.07 0.07 0.68 0.69 0.32 -- 0.25 0.27 0.26 0.41 1 N:P calc. 15.50 24.17 29.33 55.20 27.86 30.41 10.66 6.54 16.94 -- 16.72 16.74 17.96 14.26 total coliforms per 1 ml 7 47 18 19 13 20.8 11 10 8 11 5 9 6 8.6 100

E. coli per 1 ml 2 4 7 2 1 3.2 1 0 0 1 1 0 0 0.4 2

*Sum of [Kjeldahl nitrogen as N mg/l] and [nitrate + nitrite as N mg/l] **0.025 represents half of detection limit.

Diagnostic study of three lakes in southern Haiti 21 Table 5. Spring, well, and stream chemistry, depth, and bacteria. HACH - HACH DR700 portable colorimeter. JOCO - Johnson County Kansas Wastewater Water Quality Laboratory. EPA standards (stds) are for Puerto Rico surface waters (EPA 2014). watershed Douat Laborde EPA

B spring B handdug B handdug B NGO B NGO site lab D_spring D_hole*** B pool#1 B8 stds (n. hill) well#1 well#2 well well description spring pool stream spring well by B1 well by road well well dug well date 13-Mar 13-Mar 14-Mar 14-Mar 14-Mar 14-Mar 14-Mar 18-Mar 18-Mar water temp C 27.00 34.53 25.51 26.42 ------25.77 32.20 pH 3.51 8.62 7.00 7.06 ------7.51 6 to 9

ORP mV -38 123 -124 102 ------205 conductivity mS/cm 0.520 0.202 0.540 0.521 ------0.606 turbidity NTU 4.0 451.0 0.0 13.2 ------0.1 50.0

DO mg/l 3.51 14.41 0.77 4.45 ------0.00 5.00

TDS g/l 0.334 0.117 0.345 0.334 ------0.880 0.500

% salt 0.03 0.01 0.03 0.03 ------0.03 secchi depth cm ------lake depth cm -- -- 83 ------total ammonia (NH3-N HACH 0.03 1.60 0.09 0.01 0.02 0.01 0.01 0.00 -- 1 ) mg/l Kjeldahl nitrogen as N JOCO <0.5 11.1 <0.5 <0.5 0.6 <0.5 <0.5 <0.5 0.9 mg/l nitrate + nitrite as N JOCO 0.18 0.12 0.15 0.30 3.76 2.10 2.39 2.44 1.32 10 mg/l TN as N mg/l* 0.43 11.22 0.40 0.55 4.36 2.35 2.64 2.69 2.22

TP as P mg/l** JOCO <0.05 1.44 <0.05 <0.05 <0.05 <0.05 <0.05 <0.05 0.05 1 N:P** 17.2 7.8 16.0 22.0 174.4 94.0 105.6 107.6 44.4 total coliforms per 1 ml 68 105 49 19 44 10 0 0 >105 100

E. coli per 1 ml 8 70 20 3 7 0 0 0 40 2

*Sum of [Kjeldahl nitrogen as N mg/l] and [nitrate + nitrite as N mg/l] **If TP <0.05, half this detection limit was used in calculations. ***Not a spring or well, wet area.

Diagnostic study of three lakes in southern Haiti 22 Nutrients USEPA Puerto Rico Water Quality Standards Regulation (August 2014) are provided for comparison to the nutrients values obtained in the study (Table 4 and 5). The Puerto Rico standards are for class SD surface waters intended as a source of public water supply, propagation and preservation of desirable species, including threatened or endangered species, as well as primary and secondary contact recreation. All lakes, wells, and springs sampled had nutrient (nitrogen and phosphorus) levels below the Puerto Rico standards. The only site that exceeded all nutrient standards, as well as fecal coliform standards, was the residual water pool at Etang Douat, where livestock were being watered.

All of the water samples collected from Etang Laborde had higher TN and TP values than the samples collected from Etang Lachaux, and the values from Etang Laborde were higher than the wells and springs sampled in its watershed.

While the Etang Laborde north shore sample (B1) had the highest TN (7.25 mg/l) of all lake, well, and spring samples collected, the nitrate level there (0.15 mg/L nitrate + nitrite as N) was below the USEPA standard of 10 mg/l nitrate levels for babies less than 6 months of age. All lakes, wells, and springs sampled had nitrate levels below the USEPA standard.

In nearly all samples the N:P ratios were > 10 suggesting that phosphorus was the limiting nutrient in these aquatic ecosystems. University of Florida researchers tested 1500 soil samples from 5 major watersheds in Haiti for nutrient concentrations. What they found was that nitrogen was the limiting factor in plant production followed by phosphorus (Hylkema 2011). Why these aquatic ecosystems are phosphorus limited instead of nitrogen limited as are watersheds is no total understood at this time. It may be that the higher disproportionate levels of total nitrogen are skewing the N:P ratio to where phosphorus appears to be limiting. The low soil nitrogen levels of Hylkema’s study and our high total nitrogen concentrations for the lakes suggest that the more soluble forms of nitrogen are being transported to the lakes via runoff and subsurface transportation while total phosphorus levels remain low.

In situ water chemistry Some dissolved oxygen (DO) values fell near or below the minimum 5.0 mg/L criteria recommended for fish survivability. Etang Lachaux had the lowest average DO value at 5.64 mg/L, which was below its 2013 average value of 9.78 mg/L. Turbidity at Etang Lachuax 9.18 NTU was also much lower than its 2013 value 90 NTU or Etang Laborde 214 NTU. Both DO and turbidity at 2015 Etang Lachaux might be explained by the vast coverage of macrophytes, which would respire at night and lower DO levels, but which also decrease turbidity as roots hold the sediment. The clarity of the water at Etang Lachaux was also indicated by the deeper secchi

Diagnostic study of three lakes in southern Haiti 23 depth (82 cm) than Etang Laborde (10 cm). Etang Lachaux (98 cm) was nearly twice as deep as Etang Laborde (58 cm). pH at both lakes fell within acceptable range for fish (6 to 9), while water temperatures (25 – 30C) remained below the maximum standard of 32C.

Macroinvertebrates Macroinvertebrates are currently being identified to lowest practical taxon level. Orders found at each lake are presented in Table 6. Macroinvertebrates respond to long-term waterbody conditions and thus can be used to evaluate over waterbody health, while a water chemistry samples reflects the conditions at one moment in time. At the generic level macroinvertebrates can be correlated to pristine or impaired ecosystem condition.

Table 6. Macroinvertebrates orders collected at each lake. Orders Lachaux Laborde Amphipooda x x Coleoptera x x Diptera x x Ephemeroptera x x Haplotaxida x Hemiptera x x Odonata x x Trombidiforme x

Plankton Phytoplankton taxa found in the lakes are presented in Table 7. Two taxa are cyanobacteria (blue-green algae), one of which, Oscillatoria, produces a toxin (see http://www.cyanodb.cz/ for cyanobacteria genera). The phytoplankton are currently being enumerated and measured. These values will serve as measures of productivity of the lakes as high numbers can indicate nutrient enrichment. Another method to measure productivity is to measure chlorophyll levels in the lakes. However, this method requires immediately refrigerating water samples, filtering within 24 hours, and freezing the filters until they are processed, which was not feasible during this study due to lack of a freezer or refrigeration.

Diagnostic study of three lakes in southern Haiti 24 Table 7. Phytoplankton genera collected at each lake. Cyano- Labord Cyano- Taxon bacteria Lachaux e Taxon bacteria Lachaux Laborde Ankistrodesmus sp. x x Pediastrum sp. x Aphanocapsa sp. x x x Peridinium sp. x x Cosmarium sp.1 x x Phacus sp. x Cosmarium sp.2 x Polyhedriopsis sp. x Crucigenia sp.1 x x Scenedesmus sp.1 x x Crucigenia sp.2 round x Scenedesmus sp.2 x Cryptomonas sp. x x Scenedesmus sp.3 x x Cyclotella sp. x x Selanastrum sp. x x Elakatothrix sp. x Staurastrum sp. x x Euglena sp. x x Synedra sp. x x Golenkinia sp. x Tetraedron sp.1 x x Kirchneriella sp. x Tetraedron sp.2 x x Oscillatoria sp.1 sm sq x x

Fish Both native and non-native fish are present in Etang Lachaux and Laborde (Table 8). At both lakes we collected tilapia (2 species, referred to as white tilapia and black tilapia), common carp (Cyprinus carpio), and the native Limia. At Lachaux we captured equal numbers of Limia and small tilapia. Etang Douat, which was dry during this study, has a fissure at the lower southern extent of the lake basin. Residents say fish reside within this subterranean feature when the lake is dry, and come out of this fissure when the lake refills with water.

Table 8. Fish collected in Etang Lachaux and Etang Laborde, with endemism to Haiti indicated. See Appendix C for photos. Family name Genus Species Creole English Lachaux Laborde Status Cichlidae Tilapia Tilapia x x Introduced Cyprinidae Cyprinus C. carpio Kap Common carp x x Introduced Cyprinidae Hypophthalmichthys H. molitrix Boutlang Silver carp x Introduced Poeciliidae Limia mosquito fish x x Endemic

The presence of the tilapia fry in Etang Lachaux and Laborde indicates high reproduction. Lachaux residents said that the tilapia were introduced into the lake by a Haitian in 1962. Residents also reported that in 1975 large numbers of tilapia died when the lake dried, so many that they fed them to the pigs. The Lachaux residents said that in 2001 a variety of carp was introduced that ate small fish, smelled bad, and was not good to eat. It disappeared and a second variety appeared (they do not remember when this first showed up). Older residents of Lachaux reported seeing a large fish called the ‘boutlang’ when they were children (30-40 years ago), but do not remember the last time they saw boutlang at this lake.

Diagnostic study of three lakes in southern Haiti 25

Near Etang Laborde we saw a boutlang caught by a resident; this fish is actually the silver carp Hypophthalmichthys molitrix. Residents reported that foreigners released this species into Laborde in the 1970’s. The silver carp are a preferred species of the fishermen for the large size and money they bring in. Residents said they like the taste. They believe that crab and shrimp populations began to decrease due to the boutlang. Laborde residents also reported that carp was introduced around 1995 but perhaps before 2000. They mentioned two other fish that no longer exist at the lake, teta and mile. The mile was larger than tilapia. The teta was the size of the common carp or boutlang, and they do not know why it disappeared. They saw it until 14 July 2000 when the boutlang was introduced. It may be that both the teta and mile are actually introduced Tilapia species since Tilapia have been known to be introduced throughout Central America and the Caribbean as early as the 1950s and 1960s (Popma and Lovshin 1995). However, there is no way of knowing for sure what fish species local watershed people are referring to when speaking about “teta” and “mile.”

Fish from the family Poeciliidae were found in both Etang Lachaux and Laborde. Identification of fish in this family is still in progress as this requires a careful examination of the diagnostic feature, the gonopodium of male specimens, under a microscope. The gonopodium is a modified anal fin used in reproduction (Figure 8). This family, commonly referred to as mosquito fish, contains the genera Gambusia and Limia that consist of many species endemic to Haiti (Lee et al. 1993).

From both lakes we collected Limia that appear to be a single species with the males displaying two somewhat distinct color patterns. This species appears to be most similar to Limia trident and L. dominicensis, but among other traits these specimens possess some gonopodium differences. These specimens probably represent a new species, but this must be verified by other researchers at the University of Kansas and Louisiana State University. Twelve species of Limia occur in Haiti, with ten of them endemic to the Republic of Haiti and two endemic to Hispaniola (Appendix D). In fact, of the 45 fish species that occur in Haiti, 27% of are endemic to Haiti, occurring nowhere else in the world. A similar high rate of endemism has been noted in the Caribbean with over 40% of the freshwater fishes in the islands being endemic to one or more of the larger islands. When considering both plants and animals, the Caribbean Islands are one of the seven areas with the highest concentration of biodiversity and endemism in the world with Haiti being exceptionally rich and diverse (Myer et al. 2000, Kier et al. 2009, Huber et al. 2010).

Diagnostic study of three lakes in southern Haiti 26 gonopdium at end of anal fin anal fin

hook

1 mm fleshy palp

Figure 8. Photograph of the gonopodium on the anal fin of a male Limia sp. caught in a stream pool (site B_pool#1) located above Etang Laborde. The gonopodium is the diagnostic feature of species in the Poecillidae family.

In addition to the high endemism in fish, about 24% of Haiti’s fish fauna consists of introduced species which may endanger both the native and endemic species. Nearly all of Haiti’s endemic fish species are known only from one or two specific localities, suggesting that these populations have very limited distribution and therefore are at great risk of being lost through introductions and further habitat degradation and loss. One of our fish fauna concerns at this time is related to the introduction of the two invasive species of Gambusia, which are closely related to Haiti’s endemic Limia species (Appendix D). By the late 1990’s Gambusia affinis and G. holbrooki had successfully spread to 40 countries, and both have been introduced to Haiti for mosquito-control purposes (Welcomme 1992, Lever 1996, Rehage 2003). Not only are these species no better at mosquito control than native insectivorous fishes, but their negative impacts on native aquatic communities (see Pyke 2008) earned them a place on the list of 100 worst invasive species (ISSG 2000).

Walshe and co-reviewers (2013) found no reliable studies that reported the effects of introducing larvivorous fishes on malaria infection in nearby communities, on entomological inoculation rate, or on the density of adults in the mosquito genus Anopheles. Their review

Diagnostic study of three lakes in southern Haiti 27 included studies of G. affinis, G. holbrooki, Poecilia reticulata, Cyprinus carpio and Oreochromis niloticus, all of which have been introduced to Haiti and fail to control rates of malaria infection. Rehage (2003) examined endemism and invasiveness of Gambusia throughout the world, noting that G. affinis and G. holbrooki have not only failed to control mosquito populations, but they reduce and/or eliminate native fishes, amphibians, and invertebrates, primarily through predation of the eggs, fry, and larvae of the native biota.

In general, the fish fauna of these lakes is limited and dominated by introduced fish species. However, the native fishes are comprised mainly of endemic species that are both important to Haiti and the world’s biodiversity.

Macrophytes All plant specimens were deposited and identified at the McGregor Herbarium at the University of Kansas using the Flora of Jamaica, other floras for Central American and/or the Caribbean, and several monographs. We attempted to collect specimens of each macrophyte taxon located at each lake. Only one macrophyte was collected at Douat, at the residual pool of water on the south end (Table 9), while 14 taxa were collected or photographed at Lachaux and Laborde (Table 10).

None of the taxa collected are endemic to Haiti. In the Global Invasive Species Database (http://www.issg.org/database/species/search.asp?st=sss&sn=&rn=Haiti&ri=18996&hci=- 1&ei=-1&fr=1&sts=&lang=EN), only two macrophyte species are listed as invasive to Haiti. They are Eichhornia crassipes (water hyacinth) and Ludwigia peruviana (water primrose). We did not observe water hyacinth in any of the lakes but it has been introduced to other aquatic environments in Haiti. Water hyacinth is one of the worst aquatic weeds in the world. Among its many negative impacts, its shading and crowding of native aquatic plants can dramatically reduce biological diversity in aquatic ecosystems. We did observe and collect a species of Lugwigia but could not tell what species our specimens were since we had neither flowers nor fruiting bodies associated with our material. Ludwigia peruviana is a very attractive wetland species that has been introduced as an ornamental but it is very invasive in nature. Ludwigia peruviana populations can clog waterways, damage structures and dominate native vegetation.

Table 9. Plants collected at Etang Douat. See Appendix E for photos. Family Scientific name Creole name English name Alismataceae Echinodorus berteroi (Spreng.) Fassett (sterile) Upright burrhead Poaceae Oryza sativa L. Diri Rice Malvaceae Hibiscus trilobus Aubl. Threelobe rosemallow

Diagnostic study of three lakes in southern Haiti 28 Table 10. Macrophytes collected during this study with Etang Lachaux or Laborde indicated. None of these taxa are endemic to Haiti. See Appendix E for photos. Family Scientific name Creole name English name Lachaux Laborde Bulltongue Alismataceae Sagittaria lancifolia L. arrowhead x Araceae Pistia stratiotes L. Water lettuce x Ceratophyllaceae Ceratophyllum demersum L. Limon femèl Coontail x x Cyperaceae Cyperus odoratus L. Fragrant flatsedge x Cyperaceae Cyperus sp. Sedge x Eleocharis interstincta (Vahl) Cyperaceae Roem. & Schult. Knotted spikerush x Cyperaceae Fimbristylis miliacea (L.) Vahl Fimbry x Schoenoplectus tabernaemontani Cyperaceae (C.C. Gmel.) Palla Softstem bullrush x Menyanthaceae Nymphoides indica (L.) Kuntze Water snowflake x x Najadaceae Najas marina L. Limon mal Spiny water nymph x Nymphaeaceae Nymphaea rudgeana G. Mey. Rudge's waterlily x Onagraceae Ludwigia sp. (sterile) Primrose-willow x x Polygonaceae Persicaria punctata (Elliott) Small Piman dlo Dotted smartweed x x Typhaceae Typha* Cattail x x *Not collected, photos only.

While we did not attempt to collect non-macrophytes, we did collect two plants of interest at Douat, a rice plant and a hibiscus. The hibiscus was of concern because of it brambly nature and our concern that it could be invasive, or at least difficult to control (Figure 9).

Diagnostic study of three lakes in southern Haiti 29

Figure 9. Thicket of Hibiscus trilobus growing along the basin of Etang Douat.

Landuse/landcover While we were not able to obtain shapefiles of the Earth Institute’s 2011 landuse/landcover watershed maps so that we could perform more detailed GIS analyses (Appendix F), we did examine the maps while we walked the watershed, and found that the LU/LC generally agreed with the coverage indicated in the maps. However, the high mix of agriculture, agroforestry, and pasture seemed to render these coverages indistinct from each other, as most agriculture plots on the hill slopes were sparse with crops, pasture was sparse with grass, and agroforestry had scattered shrubs at best (Fig. 10). At the lake edges agriculture was more intense, as water from the lakes could be used to irrigate the crops (Fig. 10 – 13).

The more densely wooded areas in the watersheds corresponded with the forestry cover in the EI maps that defined forest as >40% canopy cover. However, these areas consisted of narrow bands of cultured trees such as kasya (Senna siamea), sèd (Cedrela odorata), trompette (Schefflera morototoni), and castor (Ricinus communis) and were not native tropical forest trees (Fig. 10).

Diagnostic study of three lakes in southern Haiti 30

Figure 10. Photograph facing east and overlooking the northwest corner Etang Lachaux. A strip of forest cover can be seen in the lower left quadrant of this photo, while intensive agriculture can be seen along the northwest shore of the lake at the center of the photo. The remainder of the landscape was classified as a mix of agriculture, agroforestry, and pasture. See Appendix F for the land use/land cover map of this lake.

Diagnostic study of three lakes in southern Haiti 31

Figure 11. A mix of squash, okra, and corn near the south edge of Etang Laborde.

Diagnostic study of three lakes in southern Haiti 32

Figure 12. Crops irrigated along the south edge of Etang Laborde. Photograph taken facing north.

Diagnostic study of three lakes in southern Haiti 33

Figure 13. The dry basin of Etang Douat, facing northwest. A residual pool of water can be seen left of center.

Slope Using ArcMap 10.2 the KBS KARS program calculated slope of the land within each watershed (see Appendix A for methods) and created watershed maps with slope categorized as 30-40%, 40-50%, and greater than 50% (Fig. 14 – 16). Areas of high slope (>30%) should be priority in protection or restoration since more runoff and erosion will come from disturbed areas of greater slope. The relatively coarse resolution (30x30m) of the elevation dataset (ASTER GDEM) used to calculate slopes results in an apparent flattening of the terrain, and thus underestimates slopes. However, the resulting maps still provide guidance in selecting areas of protection and restoration.

Diagnostic study of three lakes in southern Haiti 34

Figure 14. Slope within the watershed (black outline) of Etang Lachaux. Yellow = 30-40% slope, orange = 40-50% slope, and red = > 50% slope. Light gray indicates the upper watersheds that drain to the alluvial plains located above the lake.

Diagnostic study of three lakes in southern Haiti 35

Figure 15. Slope within the watershed (black outline) of Etang Douat. Yellow = 30-40% slope, orange = 40-50% slope, and red = > 50% slope. Light gray indicates the upper watershed that drains to the alluvial plain located above the lake.

Diagnostic study of three lakes in southern Haiti 36

Figure 16. Slope within the watershed (black outline) of Etang Laborde. Yellow = 30-40% slope, orange = 40-50% slope, and red = > 50% slope. Light gray indicates the upper watershed that drains to the alluvial plain located above the lake.

Diagnostic study of three lakes in southern Haiti 37 Population Only one house was observed in the Lachaux watershed, in the northwest area of the watershed, between the lake and the new highway. Most of the houses near this lake are south of the lake, downhill from the lake. In the Douat watershed, approximately 7 houses are located along the gullies that drain from the west and north sides of the watershed. Many more houses are in the Laborde watershed. Enold (2007) reports the population of the Laborde watershed at 1,680 (280 households x 6 people), and his estimated watershed size of 7.112 km2 was slightly smaller than the 7.129 km2 we calculated in ArcMap . We adjusted this population for consistency with the Cayemittes et al. 2007 estimate of 5 people per rural household (Table 11).

Table 11. Estimated population in the lake watersheds. Lachaux Douat Laborde Population 5 35 1400

Community meetings After walking the watershed and collecting water and biological samples, we met with the residents of each watershed, with the goal of learning more about the 50-year history of change within the watersheds including changes in the land use, reforestation projects, land management projects and lake changes. The number of attendees varied between watershed communities, with a low of 9 at Etang Laborde on a day that it was raining, to a high of 89 at Etang Douat, where school children attended the meeting. A number of people over 50 years of age attended the meetings (Table 12). Discussion was translated by Mr. Ralph Gustave. Our inquiries focused on a variety of topic including but not limited to extreme flood and drought events, introduction of fish, tree planting efforts, and how residents wish the lake area to be used. Opportunities were provided for community attendees to ask project personal questions regarding our studies of the lakes and watersheds.

Table 12. Demographics of attendees at community meetings (excluding CIAT, ORE, and university students who helped). Lake Etang Lachaux Etang Douat Etang Laborde Date 11 March 16 March 20 March # participants 26 89* 9** Ages 19 to 80 9 to 82 28 to 65 # > 50 years old 13 15 4 # < 20 years old 1 42 0 Communities represented Martinere 2 1 0

Diagnostic study of three lakes in southern Haiti 38 Lake Etang Lachaux Etang Douat Etang Laborde Lachaux 23 0 0 Levy 1 4 0 Douat 0 47 0 Sovo/Savau 0 3 0 Deglalis 0 1 0 Guichard 0 20 0 Rolin 0 8 0 Mersant 0 1 0 Kordin 0 1 0 Cliforde 0 1 0 Jouny 0 1 0 Macenot 0 1 0 Laborde 0 0 9 *About 1/2 the attendees were from a nearby school that attended the meeting. **Raining, few people came.

All three lakes have had extreme flooding in the past 50 years, with residents describing the floodwaters flowing into the adjacent watershed, in effect linking all the lakes. In 1986 Lachaux flooded up to the houses closest to its south end, and flowed into the Douat watershed. Douat residents said their lake will have water for up to 6 years at a time, claiming that 4 – 5 years ago the lake had even more fish than Lachaux. It did have water in 2014. One man blamed erosion, not less rain, for making the lake go dry. Laborde residents reported that in 2012 the lake flooded up to the houses near the south end. In 2015 Laborde experience the lowest water levels since 1975, when it was dry 4 – 5 months. It then filled after 1 night of rain and the animals tied up in the lake bed drowned. Lachaux was also reported to be dry in 1975.

The residents of all three watersheds seem to want more trees in the watershed and desire financial incentives to plant trees. However, they spoke of failed tree planting programs. Douat residents said trees were cut 1985 – 1995 after which the lake went dry. Laborde residents remember forests in the watershed until 1975- 1980, and mentioned egrets roosting in kompesh trees. There were also many mango and abriko trees at the time. They spoke of a program called Development Community Cretienne Haitian (DCCH) which planted both fruiting and non-fruiting trees. In 1995/1996 people began to cut these trees. Some trees remaining from that program (cassia, sed, mango, etc.) still exist and can be cut and sold.

All three communities seem to be aware of the impact of erosion on lake levels, and indicated some willingness to give up land to protect against erosion. They also recognize that trees help protect against erosion. The residents of Lachaux blame current low water levels on the new

Diagnostic study of three lakes in southern Haiti 39 National Route 7, saying that rocks from the recent construction are causing the lake to dry up. During our survey of the watershed, we did note that the culverts that drain water from the road are creating gullies in the hillside above the lake, and washing large rocks down to the shore (Fig. 17 – 19).

Figure 17. Top of National Route 7 culvert above Etang Lachaux.

Diagnostic study of three lakes in southern Haiti 40

Figure 18. Bottom of culvert in Figure 9, showing erosion.

Diagnostic study of three lakes in southern Haiti 41

Figure 19. Two meter deep gully created by runoff from road culverts along hillside above Etang Lachaux.

Springs and wells We encountered two types of “springs” in the watersheds: true ground water springs and shallow, hand-dug wells that expose shallow ground water (vadose water). All but one hand- dug well at Laborde, along with a hand-pump well at Laborde, had coliform bacteria, including E. coli. The primary source of contamination of these springs and wells are unrestricted livestock use in or up-slope of springs. Secondary sources include human uses such as bathing

Diagnostic study of three lakes in southern Haiti 42 and washing laundry in or around these springs. Human activities, if not controlled, can introduce soaps and detergents, as well as human transmissible diseases or infections into the water.

Suggestions for protection are:  Limit area of activities around and up-slope of springs and wells,  Post signs with contamination advisory/safe practices for spring uses,  Avoid any construction activities in or near natural springs.  Provide work areas away from the springs for clothes washing and other activities

Watershed findings and conclusions Soil erosion and land use All watershed soils are basically Udepts soils which are typically nutrient poor and form a thin, easily erodible soil profile. All lake watersheds are dominated by mountainous terrain with extensive areas of high slope hillsides (> 40% slope), nearly all of which is under some form of farming practices, either intercropping, cultivation or pasturing of livestock. There were very limited areas within each watershed where structural erosion control practices such as rock terraces and vetiver grass (Chrysopogon zizanioides) hedges have been put in place and maintained. Agroforestry practices were evident in all watersheds and represented about 24 to 28% of the land use found in these watersheds. While such agroforestry areas sometimes show less signs of erosion they represent less than 25% of all the land use in these watersheds while the majority of remaining land use shows high rates of erosion and continues to contribute to lake sedimentation and water quality degradation. It should also be noted that in most areas where agroforestry was being practiced vegetated ground cover remains limited and soil erosion rates remain high. Both our interviews with local residents and agricultural experts as well as our own observations within each watershed suggest that reforestation projects have had limited success. While the actual number of tree plantings that have been done over time in each watershed was not available, in general most interviews indicated that the number of trees planted was very high but few trees remain from such efforts. Those that do remain are mostly fruit trees. No official records of the survival rates of planted trees were found but it appears that most surviving trees that reached harvestable size were cut and used in charcoal production within each watershed. Thus tree planting efforts appear to have had limited long- term success and this may remain so unless watershed residences and landowners support such planting and charcoal production is reduced or eliminated. It should be noted that exploitation and cutting of fruit trees is more limited in these and other watersheds because of their value as a crop for consumption and sale in the market place (personal comm., Dr. M. Pierre, ORE, March 2015).

Diagnostic study of three lakes in southern Haiti 43 Precipitation and runoff Several natural and human-induced factors have collectively created a prominent wet/dry water cycle that now characterizes these lakes. First the thin Udepts soils once associated with the historic forest have been mostly lost from hill slopes so that most rainfall now runs directly off, causing more soil erosion with rapid infiltration of water not contributing to runoff. Historically, vegetative cover and accumulated organic matter would have slowed and limited runoff while also slowing infiltration rates so that subsurface water movement to the lakes would help extend the wet periods of the lakes. Extreme losses of permanent vegetation cover (e.g. forest), cone karst topography and a bi-model rainfall pattern now contribute to a stronger wet/dry cycle for the lakes as well as promote more severe floods and droughts within these watersheds and the region in general. Little can be done regarding the natural bi-model nature of precipitation and the extreme rainfall events associated with this climate. However, increased runoff due to bare soil conditions and severe soil erosion associated with poor farming practices and failure to apply meaningful soil conservation measures can be reduced and groundwater filtration increased by implementing changes in land management practices.

Lake quality and uses This study is the first step in describing the structure and function of these depression lakes (i.e. sinkhole lakes) with regard to their ecological services and values. Most noteworthy is the rarity of natural freshwater lakes in the Department of the Sud or elsewhere in Haiti. This is, in part, due to the widespread occurrence of karst geology and topography which tends to limit the development of surface water features such as lakes and streams. This rarity and the habitat opportunities they afford the often unique and endemic aquatic biota make them outstanding natural resources even in their impaired condition. It is estimated that these lakes have suffered from massive sedimentation as a result of the extensive and poor farming practices that have developed in their watersheds over the last century or so. The extent of the development of the alluvial fans or plains at the primary inflow areas of the lakes attest to the nature of the erosion and sedimentation processes which are the result of human activities. Despite the degraded nature of these ecosystems these lakes still provide many direct human services that include livestock and human drinking water, plant foods, building resources, fishing and fishes and grazing and cultivation areas when water levels are low. However, even these basic human services appear in jeopardy due to increased water level fluctuations, poor water quality and bacterial pollution. All of these services, both ecological and human, are in jeopardy primarily due to poverty and poor land use management within each watershed.

While our studies were all done during unusually dry years with rainfalls well below the 20-year average annual amount, much can still be said regarding creation of protected and management areas within these watersheds to both protect and enhance these lake ecosystems and their long-term natural and human resource values. However, because this

Diagnostic study of three lakes in southern Haiti 44 study was limited to one comprehensive sampling event and was done under less than average water conditions, we must cannot make definitive statements regarding overall ecological condition of these lakes. We can offer general recommendations and comments concerning the establishment of protected areas within the watersheds (and management areas) necessary to enhance and protect lake and watershed quality since land uses and practices are better known and observed even within dry years. What cannot be accounted for is high water and runoff conditions and events that are most contributing to lake sedimentation and land erosion. It is clear that some regions within each watershed are greater contributors to soil erosion and soil movement to these lakes but these events are better understood during major rainfall events which never happen during our study periods.

Identification of protection and restoration zones and areas 1. Identify lake buffer zone by land use based assessments or by using identified legislative requirements. Buffer area would be for protection and restoration activities. One of the most commonly used approaches to the protection of aquatic resources such as lakes, rivers and wetlands is the identification and establishment of a “buffer zone” around the ecosystem (Richardson et.al. 2012, Klemas 2014, Sweeney and Newbold 2012). The immediate areas around these lakes should be considered as a buffer zone (protected area) for these lakes in which certain human activities are either controlled or eliminated to reduce shore erosion, pollutants and habitat destruction. The buffer should encircle the lake margin and its width determined by assessment of its potential function width or by legislation. Lacking an assessment to determine functional protective width of a lake buffer we recommend using the existing federal Haitian laws and regulations to protect areas found to be of value and safety for the general community. The following are excerpts (translated by Google Translate) for Haiti articles and regulations pertinent to protected areas around waterbodies and other sensitive area of importance to the community:

Decree-Law of 23 June 1937 on the regulation of forests. Monitor No. 51 Thursday, June 24, 1937 DECREES Article 1. It is forbidden to do, without prior authorization, special written a qualified agent SNPA & E. R., no clearing, damage, cut, uproot or burn any tree; a) on land with a slope equal to or greater than 30o from the horizontal; b) around sources within a radius of 100m; c) on the banks of rivers, rivers, streams, over a 50m width on each side; d) area around lakes, ponds and natural water reservoirs, a distance of 50m. Article 2. It is prohibited without prior authorization, special written a qualified agent SNPA & ER undertake annual crops say: a) on land with a slope equal to or greater than 45 degrees with the horizontal; b) around the springs on a 100m rapyon; c) on the banks of rivers, streams over a 50m width on each side. Article 3. It is prohibited unless authorized by the SNPA & ER burning savannahs, throughout the territory of the Republic, and land designated and extended to article 1 above, to "new wood" to burn the waste to crops, sarclures or other organic debris.

Diagnostic study of three lakes in southern Haiti 45 AREAS UNDER PROTECTION ZONES AND BOOKED - August 1955- Paul Magloire E Article 15. It may be designated under the name "Areas under protection" tracts of land owned or to the State or to individuals whose protection is necessary and urgent for the wellbeing of the community. Article 16. The "Areas under protection" is any area of land and use will be regulated by the Department of Agriculture, in order to combat erosion, protecting the child's Public Health to ensure sound recreation or tourism promotion. Article 17. The boundaries of these areas will be determined by the Department of Agriculture, in conjunction with the Finance and Public Works. Article 18. Are declared "Areas under protection." 1e) any area of land owned by the state or individuals over an area of at least 5 hectares, around cascades, waterfalls and around the springs supplying drinking water to urban and rural areas. 2) any amount of land owned by the state or to private or around hot sulfur springs, usually around any water tanks, over an area of at least 5 hectares. 3) any amount of land owned by the state or individuals forming the watershed sources and rivers. 4) any area of and owned by the state or individuals to be designated by the Department of Agriculture in accordance with Article 16. Article 19. Within the limits under protection areas, areas of land will be according to the law of February 3, 1926, declared "reserved areas" and withdrawn from exploitation.

Based on our interpretation of these articles we suggest that a 50-meter buffer or 5 hectare area (whichever is larger) be established around the most prevalent lake shoreline and used as a protection zone and management area. Establishing exactly where the lake margin should be in interpreting the buffer area and width is somewhat problematic since lake surfaces thus margins expand and contract both on an annual and an inter-annual basis. One approach would be to designate the “official” lake margin as the margin most clearly defined by the long- term water margin and vegetation pattern observed in repeated satellite and photography imagery, such as Google Earth imagery. Within these lake buffers better land use practices should be required and livestock access and use controlled.

2. Identify buffers to protect springs and spring runs (i.e. spring-fed streams) within these watershed areas. Springs and spring runs within these watersheds need to be protected for several reasons. Springs are a source of clean water to all of the lakes at various times of the year and probably contribute to subsurface water flows to the lakes year round. Springs are common sources of drinking water within all watersheds and thus need to be protected from bacteria, disease and environmental contamination. As already noted some springs have dangerous levels of E. coli levels that indicate users have a high risk of obtaining enteric diseases and infections. In addition, it appears that the springs and spring runs are habitats for native fish and invertebrate fauna that repopulate the lakes after dry periods. Therefore, springs are important in maintaining natural and human environmental services and goods provided by the lake and other aquatic resources within these watersheds. Typically these springs and the areas around them are used for washing clothes and bathing, and for watering livestock and grazing. These activities increase soil erosion and disturbances upstream and in the near vicinity of springs that may endanger spring flows and contaminate spring water. Aquatic habitat is also degraded and

Diagnostic study of three lakes in southern Haiti 46 destroyed by intense human and livestock uses if left unsupervised. Therefore springs and spring areas are also identified as protected areas and minimally should have buffer zones established upstream and around them to limit and discourage livestock grazing and certain human activities. The same buffer approach listed above for lakes can be used for springs.

3. Identify ravines and gullies that drain directly to the lake margins for protection and management. Eroded soils not only originate from gullies and ravines but sediment transportation is greatest in these channels due to open channel flows and increased stream energy associated with open channel flows. Also these channels drain directly into the lake thus increasing their potential delivery capacity. While sheet and rill erosion are also a large problem in these watersheds, control of these types of erosion is more difficult due to the large areas involved and thus the need for total community involvement to achieve meaningful reductions in soil loss and transportation. Gullies and ravines, however, are more limited in extent thus transportation of sediments to the lakes is more easily controlled through structural erosion control measures.

4. Identify areas of mass wasting (i.e. landslides) and contain and control downslope soil loss. Mass wasting, which is sometimes called mass movement or slope movement, is defined as the large movement of rock, soil and debris downward due to the force of gravity. More commonly we refer to these events as landslides. Heavy rains, widespread deforestation and high-sloped terrain are causes of shallow landslides and high erosion rates in the mountains, which have steep, fault-riddled topography, weak rocks, and erosion-prone soils. The combination of deforestation and limited or poor soil conservation practices has led to numerous small landslides within these watersheds. Slide areas are not used by the watershed communities but can still contribute to soil loss and sedimentation. Thus these sites should be identified and when possible steps taken to control soil loss from these discrete and easily identified sites.

5. Severely eroded, low fertility, high slope farm plots immediately adjacent to the lake basins should be restored and protected. Many examples of field plots showing the cumulative results of extreme soil erosion and over farming can be found through all watersheds. These high-slope, thin soil plots are common everywhere within these watersheds and restoring and protecting all of these farm plots would require a massive physical and financial effort. Therefore, lake protection efforts should initially focus only on those severely eroded field plots that drain directly to the lake margins. These severely eroded plots are possibly of more concern to land owners and farmers which might promote better cooperation in adopting soil conservation practices, especially practices that may require removing some land from production. Higher potential recruitment of these types of field plots along with the fact that they directly drain to the lake ecosystem give these sites high protection and restoration value.

Diagnostic study of three lakes in southern Haiti 47

6. Identify watershed infrastructure erosion (animal/human trails, vehicle roads, drainage ditches) and restore and protect from further erosion. Many basic human and farming infrastructure features and practices can and do contribute to soil erosion and sedimentation in these watersheds and lakes. Livestock and human trails, paths, roads and drainage ditches are all susceptible to soil loss and erosion. Their direct contribution to lake sedimentation and contamination varies considerably due to their location within the watershed, and hill slope positon and orientation. The likelihood of getting either livestock or people to change their movement habitats is relatively small but some erosion control measures could be given along paths, roads and ditches that both contribute and directly transport eroded soils to the lake environments.

7. Small alluvial fan or plain regions formed along the lake margins need to be protected, restored as necessary, and managed. Over time eroded materials from these watersheds has been transported into the lake basins through channel flows (i.e. stream flows). This sedimentation process has resulted in the development of alluvial fans or plains in the lower portions of the drainage valleys that enter the lake basins. These alluvial fans of fine sediment and soils are now intensively farmed both at the upper and lower ends of the plains when lake levels allow. The small alluvial plains should be protected by erosion control measures put in place where needed. In general, these alluvial plains consist mostly of fine sediment and soils and are highly valued farm plots. These plains are bisected by few gullies or drainage channels as most of the water flows entering from upstream either spread out over this relatively flat plain or infiltrate into the highly porous alluvial fill that underlies the surface materials.

8. Large alluvial plains regions should also be protected and managed to control soil erosion and transport. As with the case of small alluvial plains, these areas (Fig. 20) are highly valued farm land and have direct contact with the lake ecosystems. Therefore, these regions as with the small alluvial plains should be regarded as protected areas in the sense that they can be both sources and sinks for eroded soils and contaminants. In addition, these large alluvial plains now act as effective barriers to the transport of large amounts of sediment that originate in these upstream sub-watersheds. Except where existing stream channels and gullies flow across these plains to the lakes, these large flat and highly permeable alluvial fans intercept and filter most rain-fed runoff waters. The sub-watersheds that drain into these larger alluvial plains are shown in Fig. 21-23 and can be considered of lesser importance in lake protection efforts.

Diagnostic study of three lakes in southern Haiti 48 ALLUVIAL FANS

Figure 20. Large alluvial fans or plains of Etang Laborde showing field plots in upper ends of the plains. (Taken from Google Earth, April 4 2013).

Diagnostic study of three lakes in southern Haiti 49

Figure 21. Etang Lachaux showing sub-watershed area (blue and yellow) where sediment and surface flow contaminants are intercepted and reduced across the NW and NE alluvial plains into which these watersheds drain.

Diagnostic study of three lakes in southern Haiti 50

Figure 22. Etang Douat showing sub-watershed (blue area) where sediment and surface flow contaminants are intercepted and reduced across the northwestern alluvial plain into which this watershed drains.

Diagnostic study of three lakes in southern Haiti 51

Figure 23. Etang Laborde showing sub-watersheds (blue and yellow areas) where sediment and surface flow contaminants are intercepted and reduced across the two northern alluvial plains into which these watersheds drain.

9. Control fish stocking activities and other practices that could endanger native aquatic species. All current fishing activities within these lakes are directed toward fish consumption or fish as a market item for profit. The only fish of food or commercial value are the introduced species such as tilapia, common carp and silver carp. While we did not collect any other introduced species of fish in these lakes we are aware of several other introduce species that may find their way into these lakes. While understanding the human value and need placed on these introduced fish species, it is also important to understand the native Haitian fish species may be threatened by the introduction of non-native fishes and other aquatic organisms. Introduced species often have no predators in their new environments and can out compete native organisms for food, habitat and other resources that sustain native populations. Many native Haitian fish species are endemic to Haiti and found no other place on the globe. As part of the natural Haitian environment and culture these native species need some level of protection against introduced species. Therefore we suggest that introduction and restocking of non- native fishes be done under the supervision and management of the Ministry of the Environment. A fish management and stocking program should be developed by the

Diagnostic study of three lakes in southern Haiti 52 government to both protect native species and ensure that fish for consumption and market are available to local fishermen.

Protection area summary The identification of protected areas as well as areas in need of restoration is a critical part of any environmental protection and management program. This report has identified a number of areas in need of protection and in most cases concurrent restoration needs. We have attempted to prioritize areas based on our knowledge of the potential magnitude of their impacts and the likelihood of the ability to actually achieve some level of success. Our concern for prioritization of identified areas and possible actions to incorporate our suggestions is based on our assessment of the magnitude of the environmental and soil management problems identified in all of these watersheds which were both wide spread and extreme in nature. Without some sort of prioritization and focus on areas that are big contributors but still offer affordable fixes, the amount of resources needed to address all areas would likely exceed federal and local capabilities.

Specifically, in order of priority, the zones that should be protected or restored within each watershed are (Fig. 25-27): 1. All areas within a 50-meter buffer or 5 hectare area (whichever is larger) around the most prevalent lake shoreline AND all areas within 50 m of stream banks AND all areas within 100 m of springs. 2. High slope (>30%) areas outside of the upper watersheds. 3. High slope (>30%) areas within the upper watersheds.

Diagnostic study of three lakes in southern Haiti 53 Figure 24. Etang Lachaux watershed (blue boundary) showing zones of protection. Red represents the highest priority for protection within a 50m buffer around the lake or a 100m buffer around each spring. Within the watershed boundary, dark blue represents high (>30%) slope areas that are of 2nd highest priority for protection while light blue represents high slopes that are of 3rd highest priority.

Diagnostic study of three lakes in southern Haiti 54

Figure 25. Etang Douat watershed, showing zones of protection. See Fig. 24 for explanation.

Diagnostic study of three lakes in southern Haiti 55

Figure 25. Etang Laborde watershed, showing zones of protection. See Fig. 24 for explanation.

Diagnostic study of three lakes in southern Haiti 56 Institutional management options and actions Along with identifying a series of options regarding areas in need of protection and restoration that would benefit the lakes themselves, we have developed a short-list of management suggestions that might be used to develop and maintain selected protection and management areas. Many more management options exist than we have listed below and many have been tried in various parts of Haiti and the world with greater or lesser successes. We view this list of management suggestions as a “tool box” of options to choose from based on acceptance by watershed residents, appropriateness to the local conditions, potential for success, and affordability in terms of available resources.

Institutional approaches to establishing protection areas, and restoring and managing areas to protect natural resources such as the study lakes often focus on using existing governmental and educational tools to inform and direct people and activities toward an established goal. Perhaps the most important consideration to the establishment of protected areas and restoration projects (that can lead to more protected areas) is that of providing financial incentives to install and maintain erosion control. All lake watersheds support a high proportion of farms and people that are very poor and lack many basic human resources. Many understand the need to conserve the soil and the natural environment but because of the level of poverty are unable to voluntarily provide the time or resources (e.g. remove land from production, build a rock terrace) necessary to make significant changes in their habits and practices. Project incentives need not be costly considering the limited earning power of most watershed farmers and tenants. Incentives should not be provided without obligations and should be based on the long-term establishment and maintenance of erosion management actions. It is critical that local people are made a part of process of creating protection areas so that they take ownership in the successes of the projects. Some of these institutional approaches are listed below: 1. Use existing legislated lake, stream and spring buffer directives to develop mandatory erosion/contamination reduction measures within protected areas and areas in need of restoration to protect a given resource. 2. Encourage/support local watershed restoration and protection organizations. 3. Arrange agricultural training/education programs using local universities and technical schools. 4. Sponsor conservation and environment discussions. 5. Promote local lake protection and restoration activities. 6. Identify and report local watershed problems and needs to cooperating NGOs and government agencies. 7. Provide financial incentives to install and maintain erosion control techniques and structures. Incentives could be cash, tax reductions or other material resources to promote personal investments in projects and project longevity.

Diagnostic study of three lakes in southern Haiti 57 8. Identify funding to jump start protection and restoration programs (e.g. World Bank, Ministry of Agriculture, Ministry of Environment). 9. Develop demonstration projects for farming techniques and soil erosion control. 10. Consider creating demonstration lake watersheds to collectively share, investigate and develop new and evaluate current approaches to better watershed and protected area management. Demonstration watersheds could facilitate the sharing of many agricultural, soil erosion and environmental techniques and programs that are in practice in various parts of Haiti but are not collectively available because of logistical and communication limitations. 11. Create a fish management and stocking program for introduced fish. Uncontrolled fish stocking by citizens and NGOs represents a real danger from introduction of exotic plants and animals that may have long-term detrimental impacts on the native biota 12. Identify sources of fish for lake stocking or develop a local fish hatchery. Governmental oversight on the acquisition and distribution of fish for stocking would reduce potential introduction of unwanted species and expedite 13. Employ youth and others within the watershed communities for project tasks. Development of youth conservation corps or other conservation organizations would create more opportunities to obtain local resident cooperation and ownership thus increasing potential long-term success and maintenance of erosion control projects

Structural erosion and protection options and actions Structural approaches to developing, restoring, or managing protected areas have a long history in soil conservation and erosion control. This often involves a hard engineering approach that is typically costly and sometimes requires a high level of maintenance to maintain the effectiveness. However, bio-engineering techniques are also available that can achieve similar results and levels of protection but are less costly and more self-maintaining. We highly recommend the use of vetiver grass hedges in bio-engineering projects for a vast number of reasons that have been documented from projects throughout the tropics. Vetiver grasses are found throughout all three watersheds, but only one functional vetiver hedge of only 200 meters was observed in our evaluation walks through these watersheds. More probably exist but from our observations the use of vetiver in soil erosion control is minimal and often not applied correctly. We saw in all watersheds that some soil erosion control projects and structures have been put in place and tried in various locations by individuals and organizations. For the most part, pass structural projects were limited in scope, poorly developed, and poorly maintained or abandon. Even reforestation efforts have meet with limited long-term successes as planted trees are often cut for fuel. Fruit tree plants are an exception to this statement. The following are a few projects and activities that are worth considering: 1. Restore prior erosion control structures.

Diagnostic study of three lakes in southern Haiti 58 2. Require or recommend use of rocks removed from field plots to be used to create rows horizontal to the slope, not vertical and perpendicular to the slope. 3. Promote vetiver hedges and support local vetiver nurseries. 4. Install vetiver hedges or rock-wall terraces on all down-slope farm plots margins. 5. Create close set, vetiver hedges and/or rock terraces on high-slope, waste lands (farmed out, low productivity land areas). 6. Continue fruit tree plantings and reforestation efforts. One approach to reforestation is planting cash crop trees in a successional manner, integrating cassava and bananas into the current crop plots, and progressing to cocoa and other valuable trees (Fig. 27).

Figure 27. Sequence of crops that mimic succession. Figure 6-9 from Kricher 2011.

Non-structural erosion and protection options and actions A number of non-structural recommendations are offered in the following list: 1. Prohibit crop residue burning. 2. Apply micro-fertilizer. 3. Reduce or eliminate tree cuttings for charcoal 4. Trim vetiver hedges to promote new growth and produce mulch or fodder. 5. Install and use latrines where needed. 6. Protect springs and spring sources. 7. Limit livestock access to lakes (provide off site constructed water pool adjacent to lake margins).

Some future considerations Our work on these watersheds and lakes was somewhat limited in scope and time, thus some areas and protection/restoration approaches were probably missed in this initial work. Various

Diagnostic study of three lakes in southern Haiti 59 elements of this report should and could be enhanced with the availability of additional resources and time. Many aspects of this study need to be further explored to focus on conditions and issues that did not surface in our original work. For example, our water quality and biological studies are limited temporally and offer only a small window in the dynamic of all these environmental variables. The native aquatic fauna and the springs associated with the lakes needs to be more toughly surveyed and studied as this region of Haiti is a global biodiversity hotspot. Species new to science and unique to Haiti await future discovery and will require actions to protect them for new generations of Haitians and others.

Nutrient enrichment may or may not be a water quality and biological problem but our single samples suggest that the lakes are very eutrophic. The seasonal and diurnal dynamics of nutrients, dissolved oxygen, pH, turbidity and other physio-chemical parameters remains unknown yet changes in their extent, frequency and duration may be controlling factors in the structure and function of the aquatic ecosystems and landscapes. Additionally, the extent and rate of sedimentation within each lake is not understood or documented, so we cannot predict what these lakes were or could be. Modifying the lakes to control water levels or dredging them to deepen them requires more knowledge than we now possess. We encourage our Haitian partners to continue their efforts in improving and managing these wonderful lakes and exploring the health and condition of the many other aquatic systems that are found in and along the southern peninsula.

Acknowledgements We are grateful to Michele Oriol and Wilcken Destravil of CIAT and Mousson Pierre of ORE who made this study possible. We also thank the Minister of the Environment for the Southern Peninsula, Jean Dunes Gustave, and his employees; and the employees of ORE and CIAT, including Gérald Buno, the drivers, and people who made us coffee and gave us fruit. We thank Ralph Gustave for translating and Phanel Petro for guiding us around the lake. For helping us to understand plants and agricultural practices, we thank the agronomy students at the American University of the Caribbean: Jackson Coudo, Ducson Visne, Dieunau Taveus, Zachary Rochelin, Wikenson Derival, Lampy Franckel, Viviane Celestin, Jean-Jacques Patrick, Etzer Dumont, Civil Pierre Alix; Notre Dame University (Torbek) student Sinvil Claudimir and graduate Agr. Dimy Orgella; and Institut Professionnel et de Formation Technique en Agronomie (IPFTA) teacher Decossard Guemey and his students Neus Roseline, Gedeon RoseNadine, Gedeon Aliette, Rock Renel Annosleus, Chery Jn Troius, Louis-Jean Dieunaissance, Philippe Daney, Antione Micelene, Sanien Antoniel, Logiste Jemsien, Charles Armelle, Cetoute Bernadine. Thanks to our friend Sonny Henry for driving and translating, and fourteen year old Farcelyn Mèns who helped row the boat and carry equipment. We are especially appreciative of the residents of Etang Lachaux, Douat, and Laborde who showed us springs and helped us to understand the history of and their hopes for the lakes. Thanks to those who assisted with the 2013 Lachaux study:

Diagnostic study of three lakes in southern Haiti 60 Gary Welker, and Nathan and Peter Rudenberg. And finally thanks to Rhoda Beutler for helping us navigate Port au Prince.

Contributors at the Kansas Biological Survey include macroinvertebrate taxonomist LeeAnn Bennett, plant taxonomist Dr. Craig Freeman, and GIS specialist Jerry Whistler of the Kansas Applied Remote Sensing Program.

References Bayard, B., C.M. Jolly, and D.A. Shannon. 2006. The adoption and management of soil conservation practices in Haiti: The case of rock walls. Ag. Econ. Review 7: 28-39.

Cayemittes, M., M.F. Placide, So. Mariko, B. Barrère, B. Sévère, and C. Alexandre. 2007. Enquête Mortalité, Morbidité et Utilisation des Services, Haïti, 2005--‐2006. Calverton, Maryland, USA : Ministère de la Santé Publique et de la Population, Institut Haïtien de l’Enfance et Macro International Inc. 516pp.

Day, M.J. 1993. Human impacts on Caribbean and Central American karst. In: Williams PW (ed.) Karst terrains: environmental changes and human impact, Catena suppl 25. Cremlingen- Destedt, Germany, pp. 109–125.

Day, M..J. 2007. Karst landscapes. In: Bundschuh J, Alvarado GE (eds.) Central America: geology, resources, hazards. Taylor and Francis, London, pp. 155–170.

Day, M.J. 2009. Protected karst landscapes: lessons from Central America and the Caribbean. In: Proceedings international cave conference. Korean Speleological Society, Danyang, pp. 13– 25.

Earth Institute. 2012. Land Use and Land Cover Baseline Report: Data and analysis of land use and land cover practices, South Department, Haiti. Earth Institute, Columbia Univ., New York, NY. 34pp.

Enold, L.J. 2007. Analyse-diagnostic des versants qui surplombent l’étang Laborde. Universite Notre Dame D’Haiti. Mémoire de din d’études agronomiques. 103pp.

EPA. 2014. Puerto Rico Water Quality Standards Regulation August 2014.

Huber, B.A., N. Fischer and J.J. Astrin. 2010. High level of endemism in Haiti's last remaining forests: a revision of Modisimus (Araneae: Pholcidae) on Hispaniola, using morphology and molecules. Zool.l J. Linnean Soc. 158:244-299.

Diagnostic study of three lakes in southern Haiti 61

Hylkema, A.L. 2011. Haitian soil fertility analysis and crop interpretations for principal crops in the five WINNER watershed zones of intervention. Univ. Florida, Inst. Food Ag. Sci. thesis. 38pp.

International, C. 2011. Biological diversity in the Caribbean Islands. Retrieved from http://www.eoearth.org/view/article/150636 on 28 June 2015.

ISSG. 2000. 100 of the World’s Worst Invasive Alien Species: A Selection from the Global Invasive Species Database. Invasive Species Specialist Group (IUCN). Univ. Auckland, Auckland, New Zealand. 11 pp.

Kier, G., H. Kreft, T.M. Lee, W. Jetz, P.L. Ibisch, C. Nowicki, J. Multe and W. Barthlott. 2009. A global assessment of endemism and species richness across island and mainland regions. Proc. National Acad. Sci. 106(23): 9322-9327.

Klemas, V. 2014. Remote Sensing of Riparian and Wetland Buffers: An Overview. J. Coastal Research, 30: 869-880.

Kricher, J. 2011. Tropical Ecology. Princeton Univ. Press. 704pp.

Kueny, J.A. and M.J. Day. 1998. An assessment of protected karst landscapes in the Caribbean. Caribb. Geogr. 9:87–100.

Lee, D.S., S.P. Platania, and G.H. Burgess. 1983. Atlas of North American Freshwater Fishes 1983 Supplement. NC Biol. Survey No. 1983-6. 66pp.

Lever, C. 1996. Naturalized Fishes of the World. Academic Press, New York, 408 pp.

MacFadden, B.J. 1986. Geological setting of Macaya and La Visite national parks southern peninsula of Haiti. Prepared for USAID/Haiti under Contract #521-0169-C-00-3083-00. January 1986.

Metcalf, R.H. and L.O. Stordal. 2010. A Practical Method for Rapid Assessment of the Bacterial Quality of Water: A Field - Based Guide. United Nations Human Settlements Programme, HS/179/10E. 16pp.

Myers, N., R.A. Mittermeier, C.G. Mittermeier, G.A. Da Fonseca, and J. Kent. 2000. Biodiversity hotspots for conservation priorities. Nature 403(6772): 853-858.

Diagnostic study of three lakes in southern Haiti 62 Popma, T.J. and L.L. Lovshin. 1995. Worldwide Prospects for Commercial Production of Tilapia. Intern. Center Aquaculture and Aq. Environ., Dept. Fish. and Allied Aquacultures, Auburn Univ. AL. 42pp.

Pyke, G.H. 2008. Plague minnow or mosquito fish? A review of the biology and impacts of introduced Gambusia species. An. Review Ecol, Evol. and Syst. 39:171-191.

Reardon, J., J.A. Foreman and R.L. Searcy. 1966. New reactants for the colorimetric determination of ammonia. Clinical Chimica Acta 14: 403-405.

Rehage, J.S. 2003. Traits Underlying Invasiveness: A Comparison of Widespread and Endemic Species in the Genus Gambusia (Poeciliidae). University of Kentucky Doctoral Dissertations. Paper 262. 143pp. http://uknowledge.uky.edu/gradschool_diss/262

Richardson, J.S., R.J. Naiman and P.A. Bisson. 2012. How did fixed-width buffers become standard practice for protecting freshwaters and their riparian areas from forest harvest practices? Freshwater Science 31: 232-238.

Santiago-Valentin, E. and R.G. Olmstead. 2004. Historical biogeography of Caribbean Plants: Introduction to current knowledge and possibilities from a phylogenetic perspective. Taxon 53 (2):299-319.

Sweeney, B.W. and J. D. Newbold. 2014. Streamside forest buffer width needed to protect stream water quality, habitat, and organisms: a literature review. JAWRA J. Am. Water Res. Assoc. 50: 560-584.

Walshe, D.P., P. Garner, A.A. Abdel‐Hameed Adeel, G.H. Pyke, and T. Burkot. 2013. Larvivorous fish for preventing malaria transmission. The Cochrane Library (Cochrane Database of Systematic Reviews 2013, Issue 12).

Welcomme, R.L. 1992. A history of international introductions of inland aquatic species. ICES Marine Science Symposium 194:3–14.

Diagnostic study of three lakes in southern Haiti 63 Appendix A. Files and methods used by the Kansas Applied Remote Sensing Program to create Haiti watersheds and calculate slope. Written by Jerry Whistler.

Watersheds delineation Grids are listed in order of processing. Intermediate grids and shapefiles used to fine-tune the western-most watershed’s western edge are not listed. A. Grids derived from 2011 Advanced Spaceborne Thermal Emission and Reflection Radiometer (ASTER) Global Digital Elevation Model Version 2 (GDEM V2). 1) haiti_dem – stitched tiles for Hispaniola 2) haiti_demr – data values 0 and greater than 65,000 were eliminated from haiti_dem. Different color scheme employed. 3) haiti_demf – filled sinks in the haiti_demr. Required step in processing Haiti_dem for watershed extraction. 4) haiti_demFD – flow direction derived from haiti_demf. 5) haiti_demFA – flow accumulation derived from haiti_demFD. Used to create stream traces. 6) f_watersheds – the finalized watersheds.

B. Points (shapefile) used to define pour points for the three watersheds. 1) PourPoints – used haiti_demFA to place four pour points on the stream traces. Needed two pour points to define western-most watershed’s western boundary due to somewhat coarse vertical resolution of the dem.

C. Vectors (shapefiles) derived from grids to define watersheds. 1) watersh_hait1 – first cut at defining the three watershed boundaries using pour points 4, 2, and 3. After consultation with Don, discovered that western-most watershed (pour point 4) was poorly defined on its west side. 2) watersh_hait2 – created new pour point (#2) to better define west side of western-most watershed. 3) f_watersheds – the finalized watersheds.

Slope calculation Geographic spatial data are models of the physical world. Some models are better than others, and often that difference is related to the type of data a model uses. Spatial data can be represented (modelled) with points, lines, and polygons (vector data), or it can be represented as contiguous gridded cells (raster data). Surface elevation is commonly represented and analyzed using raster data in the form of a digital elevation model (DEM).1

Diagnostic study of three lakes in southern Haiti 64 It important to realize that the grid cell used in raster data will generalize spatial data. Coarse resolution data will generalize the data more than fine resolution data. The finest resolution available elevation dataset covering the entire country of Haiti is the ASTER Global Digital Elevation Model (ASTER GDEM). The ASTER GDEM is in geographic lat/long coordinates with 1 arc-second (approximately 30m) grid of elevation postings. This means that one grid cell represents the average elevation within a 30m x 30m area.

A consequence of this averaging is that abrupt changes in elevation that occur within relatively short distances are masked out. This has the tendency to “flatten” rough terrain. We see this reflected in underestimated slope values calculated from the ASTER DEM in the three watersheds studied.

Slope represents the rate of change of elevation for each DEM cell, and the inclination of slope can be output as either a value in degrees or percent rise. We chose to represent slope as percentage values. The values range from 0 to essentially infinity. A flat surface is 0 percent and a 45-degree surface is 100 percent, and as the surface becomes more vertical, the percent rise becomes increasingly larger.2

1 Other surface elevation representations are triangulated irregular networks (TINs) and LiDAR point clouds. 2 ArcGIS 10.3 Online Help. URL http://desktop.arcgis.com/en/desktop/latest/manage- data/raster-and-images/slope-function.htm

Diagnostic study of three lakes in southern Haiti 65 Appendix B. Study locations with approximate altitude. Samples collected from shore were taken approximately 1 m out into the water. Watershed Site code date location latitude longitude approx alt (ft) x1 9/Mar/2015 from shore 18.30727 -73.84896 x2 9/Mar/2015 from shore 18.30944 -73.84888 x3 9/Mar/2015 from shore 18.3113 -73.84903 x4 9/Mar/2015 from shore 18.31627 -73.85034 x5 9/Mar/2015 from shore 18.31718 -73.84669 x6 9/Mar/2015 from shore 18.31362 -73.84379 x7 9/Mar/2015 from shore x8 9/Mar/2015 from shore 18.30876 -73.84465 x9 9/Mar/2015 from shore x10 10/Mar/2015 on lake 18.30832 -73.84627 x11 10/Mar/2015 on lake 18.30906 -73.84585 x12 10/Mar/2015 on lake 18.30927 -73.84522 Lachaux x13 10/Mar/2015 on lake 18.30989 -73.84584 x14 10/Mar/2015 on lake 18.31123 -73.84633 x15 10/Mar/2015 on lake 18.31283 -73.84572 x16 10/Mar/2015 on lake 18.31361 -73.84656 x17 10/Mar/2015 on lake 18.31441 -73.84798 x18 10/Mar/2015 on lake 18.31282 -73.84882 500 x19 10/Mar/2015 on lake 18.31152 -73.84864 x20 10/Mar/2015 on lake x21 10/Mar/2015 on lake 18.31001 -73.84889 x22 10/Mar/2015 on lake 18.30925 -73.84845 x23 10/Mar/2015 on lake 18.30832 -73.84791 x24 10/Mar/2015 on lake 18.3081 -73.84751 D_spring 13/Mar/2015 spring 18.32137 -73.82909 648 Douat D_hole 13/Mar/2015 pool 18.30667 -73.828 400 B pool#1 14/Mar/2015 stream 18.31799 -73.80691 500 B spring (n. hill) 14/Mar/2015 spring 18.31943 -73.80662 B handdugwell#1 14/Mar/2015 well by B1 18.30601 -73.80027 B handdugwell#2 14/Mar/2015 well by parking 18.30464 -73.80334 B NGOwell 14/Mar/2015 well 18.30437 -73.80412 321 B8 18/Mar/2015 dug well 18.30308 -73.79644 B1 14/Mar/2015 from shore 18.30592 -73.80012 B3 18/Mar/2015 from shore 18.30362 -73.80267 B5 18/Mar/2015 from shore 18.3019 -73.80228 B6 18/Mar/2015 from shore 18.30014 -73.79888 B9 18/Mar/2015 from shore 18.3034 -73.79791 Laborde B10 19/Mar/2015 lake 18.30161 -73.80129 B11 19/Mar/2015 lake 18.30097 -73.80109 250 B12 19/Mar/2015 lake 18.3008 -73.80067 B13 19/Mar/2015 lake 18.30112 -73.79973 B14 19/Mar/2015 lake 18.30151 -73.79864 B15 19/Mar/2015 lake 18.3022 -73.79805 B16 19/Mar/2015 lake 18.30299 -73.79822 B17 19/Mar/2015 lake 18.30485 -73.79919 B18 19/Mar/2015 lake 18.30501 -73.80106 B19 19/Mar/2015 lake 18.30441 -73.80187 B20 19/Mar/2015 lake 18.30358 -73.80156 B21 19/Mar/2015 lake 18.3027 -73.80182

Diagnostic study of three lakes in southern Haiti 66 Appendix C. Photographs of fish collected during this study at Etang Lachaux and Laborde.

Family name Genus Species Creole English Lachaux Laborde Cichlidae Tilapia Tilapia x x Cyprinidae Cyprinus C. carpio Kap Common carp x x Cyprinidae Hypophthalmichthys H. molitrix Boutlang Silver carp x Poeciliidae Gambusia, Limia mosquito fish x x

Photo Taxon

Tilapia

Cyprinus carpio (mirror form on bottom)

Diagnostic study of three lakes in southern Haiti 67 Photo Taxon

Hypophthalmichthys molitrix

Poeciliidae

Diagnostic study of three lakes in southern Haiti 68 Appendix D. Fishes of Haiti with occurrence classes. Endemic species are species known to occur only in the Republic of Haiti. Except for Poecilia reticulata all introduced species were introduced as mosquito control or food for humans. Modified from FishBase (http://www.fishbase.org/Country/CountryChecklist.php?c_code=332&vhabitat=fresh&csub_c ode=, output generated 26 May 2015). Additional introduced species listings are from the Database on Introductions of Aquatic Species (http://www.fao.org/fishery/dias/en) and the Global Invasive Species Database (http://www.issg.org/database/welcome).

Order Family Species Occurrence FishBase name Anguilliformes Anguillidae Anguilla rostrata native American eel Carcharhiniformes Carcharhinidae Carcharhinus leucas native Bull shark Clupeiformes Engraulidae Anchoa parva native Little anchovy Cypriniformes Cyprinidae Ctenopharyngodon idella introduced Grass Carp Cypriniformes Cyprinidae Cyprinus carpio introduced Common carp Hypophthalmichthy Cypriniformes Cyprinidae molitrix introduced Silver carp Hypophthalmichthy Cypriniformes Cyprinidae nobilis introduced Bighead carp Cyprinodontiformes Cyprinodontidae Cyprinodon bondi native Hispaniola pupfish Cyprinodontiformes Poeciliidae Gambusia affinis introduced Western Mosquitofish Cyprinodontiformes Poeciliidae Gambusia beebei endemic Miragoane gambusia Cyprinodontiformes Poeciliidae Gambusia dominicensis native Dominican gambusia Cyprinodontiformes Poeciliidae Gambusia hispaniolae native Hispaniolan gambusia Cyprinodontiformes Poeciliidae Gambusia holbrooki introduced Eastern Mosquitofish Gambusia Tiburon Peninsula Cyprinodontiformes Poeciliidae pseudopunctata endemic gambusia Cyprinodontiformes Poeciliidae Limia dominicensis endemic Tiburon Peninsula limia Cyprinodontiformes Poeciliidae Limia fuscomaculata endemic Blotched limia Cyprinodontiformes Poeciliidae Limia garnieri endemic Garnier's limia Cyprinodontiformes Poeciliidae Limia grossidens endemic Largetooth limia Cyprinodontiformes Poeciliidae Limia immaculata endemic Plain limia Cyprinodontiformes Poeciliidae Limia melanonotata native Blackbanded limia Cyprinodontiformes Poeciliidae Limia miragoanensis endemic Miragoane limia Cyprinodontiformes Poeciliidae Limia nigrofasciata endemic Blackbarred limia Cyprinodontiformes Poeciliidae Limia ornata endemic Ornate limia Cyprinodontiformes Poeciliidae Limia pauciradiata endemic Few-rayed limia Cyprinodontiformes Poeciliidae Limia rivasi endemic Rivas's limia Cyprinodontiformes Poeciliidae Limia tridens native Trident limia1 Cyprinodontiformes Poeciliidae Poecilia hispaniolana native Hispaniola molly Cyprinodontiformes Poeciliidae Poecilia reticulata introduced Guppy Elopiformes Megalopidae Megalops atlanticus native Tarpon Mugiliformes Mugilidae Agonostomus monticola native Mountain mullet Mugiliformes Mugilidae Joturus pichardi native Bobo mullet Mugiliformes Mugilidae Mugil liza native Lebranche mullet Perciformes Centropomidae Centropomus ensiferus native Swordspine snook Perciformes Centropomidae Centropomus parallelus native Fat snook Perciformes Centropomidae Centropomus pectinatus native Tarpon snook Perciformes Cichlidae Nandopsis haitiensis native Haitian cichlid

Diagnostic study of three lakes in southern Haiti 69 Order Family Species Occurrence FishBase name Perciformes Cichlidae Oreochromis aureus introduced Blue tilapia Oreochromis Perciformes Cichlidae mossambicus introduced Mozambique tilapia Perciformes Cichlidae Oreochromis niloticus introduced Nile tilapia Perciformes Gerreidae Eugerres plumieri native Striped mojarra Perciformes Gerreidae Gerres cinereus native Yellow fin mojarra Perciformes Gobiidae Awaous banana native River goby Perciformes Haemulidae Pomadasys crocro native Burro grunt Siluriformes Ictaluridae Ictalurus punctatus introduced Channel catfish Syngnathiformes Syngnathidae Microphis lineatus native Opossum pipefish

1 Common name “Trident limia” was taken from Lee et al. 1983.

Diagnostic study of three lakes in southern Haiti 70 Appendix E. Photographs of macrophytes collected during this study at Etang Lachaux and Laborde. Creole English Photo Family Scientific name Lachaux Laborde name name

Sagittaria lancifolia Bulltongue Alismataceae x L. arrowhead

Water Araceae Pistia stratiotes L. x lettuce

Ceratophyllum Limon Ceratophyllaceae Coontail x x demersum L. femèl

Cyperus odoratus Fragrant Cyperaceae x L. flatsedge

Cyperaceae Cyperus sp. Sedge x

Eleocharis Knotted Cyperaceae interstincta (Vahl) x spikerush Roem. & Schult.

Diagnostic study of three lakes in southern Haiti 71 Creole English Photo Family Scientific name Lachaux Laborde name name

Fimbristylis Cyperaceae Fimbry x miliacea (L.) Vahl

Schoenoplectus Softstem Cyperaceae tabernaemontani x bullrush (C.C. Gmel.) Palla

Nymphoides indica Water Menyanthaceae x x (L.) Kuntze snowflake

Limon Spiny water Najadaceae Najas marina L. x mal nymph

Nymphaea Rudge's Nymphaeaceae x rudgeana G. Mey. waterlily

Ludwigia sp. Primrose- Onagraceae x x (sterile) willow

Diagnostic study of three lakes in southern Haiti 72 Creole English Photo Family Scientific name Lachaux Laborde name name

Persicaria punctata Dotted Polygonaceae x x (Elliott) Small smartweed

Typhaceae Typha* Cattail x x

*Not collected, photos only.

Plants collected at Douat. Creole Photo Family Scientific name English name name Echinodorus berteroi (Spreng.) Alismataceae Upright burrhead Fassett (sterile) Poaceae Oryza sativa L. Diri Rice

Threelobe Malvaceae Hibiscus trilobus Aubl. rosemallow

Diagnostic study of three lakes in southern Haiti 73 Appendix F. 2011 landuse/landcover maps of each lake watershed. Maps were generated by the Earth Institute for this study, using watershed boundaries created by the Kansas Applied Remote Sensing Program.

Diagnostic study of three lakes in southern Haiti 74

Diagnostic study of three lakes in southern Haiti 75

Diagnostic study of three lakes in southern Haiti 76