The Samaan Grove Wetland System - Part 1

Description of the Project

by Kevin Kenny

Click images to enlarge
Over the years, residential communities that developed in the areas of Mount Pleasant, Buccoo, Golden Grove, Canaan and Bon Accord did so with little care or understanding of the effect that surface water runoff had on the marine environment. In the areas surrounding the new development, residential communities were built on a solid coral substrate where individual septic tanks were constructed without the assistance of mechanical equipment. This resulted in tanks that were undersized and often very shallow. These tanks functioned reasonably well during the dry season. However, in the rainy season when there was a substantial increase in the water flow and when the water table was at its highest, they were too small to handle the extra volume. As a result, the excess overflowed into nearby drains, eventually making its way into the marine environment and on to the reef, where it fertilized marine algae which smothered the coral.

Drains overflowing after a
heavy rainfall

A drain being dug in coral with a mechanical hammer 

The sea was further overloaded with the construction of two low-income housing communities which utilized package sewer plants that always seemed to malfunction. Sewage was efficiently captured but allowed to flow down the drains into the sea every time the waste treatment plants did not function as designed. Both of these developments were located adjacent to Buccoo Reef. This reef, now a national marine reserve, is polluted and stressed to the point that it may no longer be the main tourist attraction of the island.  

In the first stage of the development of the resort community of Golden Grove, the company decided to incorporate some innovative techniques to show how well they would work. Angostura Resorts, the developer, sought low-tech solutions to deal with surface water runoff. In the first phase called Samaan Grove (a 12 hectare [30 acre] residential development), we decided to construct 2.6 hectares (6.4 acres) of fresh water wetlands, ponds and lakes, part of a proposed overall plan to develop 32 hectares (80 acres) of man-made wetlands that would trap sediment.  

We would use bog plants to remove the contaminants and nutrients that flowed down the various drains through the estate on the way to the sea. Silt and garbage traps were developed and installed at the entrance of the development. Deep and shallow lakes were constructed that would slow the speed of the sediment-laden run off and catch some of the heaver particles. An inventory of the local bog plants was compiled, later collected and planted in the new man-made wetlands. Where possible local bog plants were used, though we felt that we needed to have many varieties of plants to improve the effectiveness of the natural filter.

One of the key requirements of the project was to use these water features to visually improve the vista for residents so that they would become connected to the wetlands and the wildlife they attract.

Flowering bog plants were chosen wherever possible to give the impression of a garden, in the hope that residents would think of these green spaces as recreational and not wastelands. We hoped that these landscaping improvements would lift land values to offset for the very high initial cost of developing the wetland system. This would be the ultimate test of the term "sustainable development" where the cost was fully recovered by the additional premium charged to live in this unique community.
Setting the tone for Samaan Grove, were Ed Stone & Associates (EDSA), our landscape architects, in particular Bob Dugan. They developed a magnificent entrance which includes a large copper vessel that holds our signature waterlily, the Nariva red night bloomer. They provided the layout of the wetlands, ponds, lakes, roads and bridges which make the community a very attractive place to live. The beautiful bridges and winding roads combine to make every feature of design a thing of wonder. The same was done for the utility buildings, with the guard hut, sewer lift stations and well combining both form and function, showcasing the design work of John Otway, our design consultant.   

Together with John Otway Design, we conceived four model homes, built and landscaped to set the standard for design, construction and landscaping. Two features of these houses were the use of recycled PET plastic bottles as a roofing material and the elimination of hard surface driveways. The company also developed covenants as part of the lease which defines how owners operate within the development.   

Samanea saman 

Great care was taken to preserve the existing trees on the development. All the samaan trees (Samanea saman) that were in the path of the road were carefully removed and placed in a holding area. Most of them have since been replanted in the development.

The engineering and project management was provided by Alpha Engineering and in particular Fazir Khan, who spent many hours researching the internet and reading countless journals and papers on man-made wetland construction. Alpha also provided the engineering expertise to develop the sewage solution. 

Results and Images After Completion


Although it is uncommon to start a report by giving the end results, we felt that this was the best place to insert the data gathered to date, hopefully to encourage others to follow our lead.   

We conducted our first water sampling on the 30 November 2006, still very early in the maturation of the wetland system. We only looked at four indicators: zinc, copper nitrogen and fecal coliform bacteria. These tests were carried out at the Angostura Environmental Laboratory and at the Water and Sewage Authority (WASA). The results were very encouraging. Samples were taken at the start of the main drain upstream of the development and at the storm water outlet at the very end of the wetland system. The findings were:


Head water 

 Wetland outfall

 0.005 ppm

 0.003 ppm

0.004 ppm 

0.001 ppm 

0.327 ppm 

0.109 ppm 
Fecal Coliform 

340 no. 

62 no. 
Five Day Fecal Coliform 

 1.0 ml

0.4 ml 

The level of zinc was reduced by 40%, the level of copper reduced by 75%, the level of nitrogen reduced by 67%, fecal coliform reduced by 91%, and the five day fecal coliform reduced by 62%.

The significant thing about these results is that the wetland is not yet fully established. Further tests will measure exactly where the reductions are being made. By the end of 2007 we will also be testing for levels of phosphates, lead, BOD, dissolved solids and much more.

The other key and perhaps equally important result is that the additional cost was fully recovered in higher real estate values, well in excess of the forecast.

Images from December 2006

Lake 0, the spring-fed lake

Wetland SWW#5

Lake 1A

Lake 1B

Wetland SWW#6A

Wetland SWW#6B

Wetland SWW#6B

Lake 2B

Lake 2A

the Wetlands


The large version of this diagram opens in
a new window, allowing you to follow
the water flow as you read. >

The New Drainage System

From the southeastern corner of the estate, waste water makes its way down a newly constructed concrete drain (D1). Surface water runoff from the estate's northern lots flows into the concrete drain at the back of each property.


This drain ends at the silt trap. All of these silt traps were designed to be easily cleaned using a backhoe. Although not installed at the time of this report, there is a plan to place a litter barrier just above the eastern trap to collect any floating debris that washes down this drain.

Because most of the runoff currently comes from a pasture upstream there is little trash in the drain at this point in time. From here the water flows into the first of the constructed wetlands (SWW#5). This wetland is only 1 metre (3 feet) deep and was planted with rushes taken from the immediate area. It is the first natural sediment trap for suspended solids and as such it is expected to fill in, requiring cleaning and de-silting on a regular basis. The initial design worked so well that, in the first high flow situation, the entire bog filled up and had to be dug out and replanted.

From here the water then flows into Lake 1A (L1A) which is 4 metres (12 feet) deep and is meant to dramatically slow the flow of water and allow some sediment to fall to the bottom. The water next flows through a culvert located under the bridge (B1) or through a self-leveling pipe buried deep under the road at a depth of about 3.5 metres (10 feet). The self-leveling pipes were installed under both bridges to ensure that, when the level of the lake falls below the culvert, both sides of the Lakes 1A-1B and Lakes 2A-2B are always at the same level.  

< Lake 1A soon after completion

and a year later, after it had settled and the waterlilies had started to grow. >

The land beside Lake 1B hosts one of the signature samaan trees from which the development takes its name. From here the water passes into the second wetland (SWW#6) flowing through a 15 centimetre (6 inch) PVC pipe. The height of the pipe is set to keep water flowing into the wetland throughout the year. In times of flood the water will flow over the spillway completely bypassing the wetland system. The photograph below left shows soap suds at the end of the spillway indicating the level of nitrates and phosphates contained in the water.  

Samanea saman 


The volume of water flowing into SWW#6 can be adjusted by changing the height of the inlet in the supply pond L1B, above right. We continue to experiment with the water flow as the objective is to always have water flowing through the wetland even during the driest months of the year.   

< This shows wetland SWW#6A when it was first constructed --

-- and one year later. This pond was planted with Vallisnaria which now covers 60% of its bottom. >

Next the water flows down an earthen drain into the second section of wetland SWW#6B. This bog has a wide range of plants growing on a shelf that was built on the northern side of the pond. The shelf was sealed with clay and covered with 15 centimetres (6 inches) of topsoil to speed up the growth of the plants. During the rainy season this shelf is covered by a foot of water. It was designed to grow a host of different bog plants meant to create a diverse habitat. Although only a year and a half old, black-belly tree ducks and the common gallinules have been observed with young in this area. The bog is one of the most beautiful as it winds its way down to the second lake system L2. 

SWW#6B in its first year, just before it was filled
with water

Because there was not enough water coming into the system we had to pump water from Lake 2B the first year.
Water flows into Lake 2B, then down the spillway into an earthen drain which goes through the estate.

Spillway from SWW#6 to Lake 2B

^ Spillway from
Lake 2B >

Construction of the Lakes
One of the main challenges in constructing ponds on the estate is the hard coral surface that exists in many places. When EDSA, the landscape architects, laid out the lakes they did not know the exact boundaries or structure of the coral formations. As the project progressed we found that coral formations would dictate the placement, dimensions, depth and size of many of the ponds. It became far too expensive to consider using mechanical equipment to place them in the exact spot where the landscape architect wanted them to go.   

Pond carved from hard coral
A good example is the pond shown above which was dug out of an area containing hard coral. Had we known the soil condition before we sold the land we would not have placed this pond in this location. However, because of the commitment made to the customer we ended up digging the pond in the coral base as agreed. We also experienced problems with water levels in L1 and SWW#6 because the excavated coral sections were not properly lined with clay and leaked.  

The Wells

Another challenge we faced in the early design of the project was how to deal with periods of severe drought, which would result in reduced water flow through the wetland system. During the dry season wild fires burn many areas of the estate. Since the main purpose of the developed wetland was to improve the quality of surface runoff flowing all year round we had to find a way to keep the wetland productive.

We decided to develop a series of wells located at strategic points which would be used to supply water to the drains feeding the wetlands. It was a bonus that we subsequently found a spring at Samaan II.

This photograph illustrates the problem at SWW#6. The waterlilies were high and dry in May 2005 after the first dry season. Most of them and most of the bog plants survived this condition and were flourishing in July when the water level returned to normal.  

Canaan Village drain

Canaan Drain (D2)

A main drain that flowed through the estate came off of the Cannan Feeder Road south of the development. This drain carried the bulk of the liquid waste from Canaan village. We built a silt trap and trash barrier which are checked and cleaned frequently. We fenced the silt trap to protect children from the village, who were seen playing and catching fish in the trap area. This water flows through another new concrete drain at the back of the southern boundary wall of Samaan II, down through a series of steps and into Lake 2A. 

Silt trap and drain

Silt trap T2 

Step drain

Trash barrier

As the water enters Lake 2A it flows
through some coral rubble placed to
prevent erosion.

Lake 2A
These two photographs were taken one year apart.
The beautiful Samaan tree at the right is one of the
largest found growing on this part of the estate and
every effort was made to preserve it.

Next the water flows through a culvert under the second bridge (B2). As with B1 we installed a self-leveling pipe. From there water flows into Lake 2B, then down a spillway into an earthen drain. There was no coral encountered in constructing these two lakes and all the sides and bottom were sealed with clay found on the site.

Lake 2B when first constructed >

Samaan II

To our delight we found a natural spring when we started clearing the 2.4 hectare (6.6 acre) site known as Samaan II. We had the water tested and were surprised at the high quality. This allowed us to create another very different lake (L0). The spring produces 160,000 liters (40,000 gallons) per day during the rainy season but falls off by the end of the dry season to around 41,000 liters or 10,000 gallons per day. There is no silt entering this lake, making it one of the clearest on the development.

Spring on Samaan II

The first thing we had to determine was the dimensions of the lake, which were defined by a layer of hard coral on the southern boundary. A number of trial holes were dug and we found that there was a wall of hard coral directly under the proposed road. All the other soil had a high component of clay.    

Test hole

Starting the lake

Construction drain for spring water

Cutting the coral drain

Building the drain in the coral

Shelves on the eastern side

Pumping the lake during construction

Cutting the shelf in the clay on the west side

The north side near the drain

< Lining the lake
with coral >

We started the excavation by draining the existing pond. The coral shelf on the western or right side of the pond was left quite steep while on the eastern and northern sides the slope was more gradual as it consisted of clay. We did not line the coral with clay since we wanted the spring water to flow unimpeded into the pond. It was interesting to see that most of the water flowed into the new lake from the south coral shelf just below the point where we constructed the drain under the road.

Lake 0 finished 
The height of the water table was just 60 centimetres (24 inches) below the level of the road. One of the challenges during construction was the heavy ground water flow which required continuous pumping. We built a number of shelves on both sides of the pond which in time will be planted with submersed grasses meant to attract wildlife both above and below the water. 


We had to develop maintenance practices that would keep the ponds, lakes and wetlands looking beautiful all year long. Already there are some interesting lessons learnt.

The first is that certain plants are more aggressive than others and will quickly crowd out the less aggressive species. There are some terrestrial plants that will thrive in wet areas and smother the bog plants. One of these is a water grass (above right) which grows on the sides of drains and ditches. The second is Ipomoea aquatica, a vine locally called "wild potato" (perhaps because it looks similar to the local sweet potato vine) which spreads quickly and will cover the entire pond if left unchecked.  

< ^ Ipomoea aquatica

The direction of the wind with respect to the surface water is something that needs to be considered in lake and pond design. All floating debris is pushed towards the leeward side of the lake. In an ideal world the outfall of this lake should also be at the leeward corner making the collection of floating matter easy. Wind will also destroy water lily pads. Victoria amazonica was planted down wind in lake 2A and we often found the pads flipped over and covered with floating debris which had been pushed on to the top of the leaves, eventually causing them to rot.

Trash barriers on all drains feeding the ponds and lakes are essential for the easy maintenance of the lakes. We now have someone constantly working on the maintenance of the entire wetland system.

< Victoria surrounded by trash | ^ Worker cleans it out

When new ponds are lined with clay there is little nutrient in the soil and the plants need to be fertilized during their first year. In the early days, fertilizing with pond tabs was essential in making the water garden attractive. As the pond matures and collects silt it will eventually sustain itself.

During the initial planting period we use no fertilizer for about one month to allow the roots to grow. Once the plants have become established, we fertilize once a month during the first year to produce good results. The size of the leaf and the flowering rate of the plant is usually a good indicator of when to fertilize. In our case, one of the main reasons we built the wetlands was to remove nutrients from the water, so the use of fertilizer is controlled and limited. There are all sorts of application rates for pond tabs. For small plants we use two tabs per plant placed about 10 centimetres (4 inches) from the root base. With more mature plants we use up to 10 tablets placed about 1.5 metres (5 feet) from the base of the main stems. The root base can become quite large with older plants and we have difficulty finding soil in which to place the tablets. We simply push a finger between the roots placing the tablets among the roots and then cover the hole with clay. Check Victoria-Adventure for additional advice on fertilizing.


About Tobago
and Maps of the Developments

Text by Indar W. Ramnarine
Maps provided by Kevin Kenny

The Republic of Trinidad and Tobago is located between 10°2' and 11°2' North Latitude and 60°30' and 61°50' West Longitude, just off the northeast coast of Venezuela, South America. Its climate is tropical, with an average temperature ranging from 21°C to 34°C (70°F to 93°F). There is a major dry season from late December to mid-May followed by a rainy season extending from late May to early December. The rainy season is interrupted in November by a short dry spell of mean duration two weeks and termed the petit careme. The rainfall will range from 1500mm to 3000mm (59 inches to 118 inches). Tobago occupies an area of 308 square kilometers (119 square miles) and there is a single mountain range called the Main Ridge which is the oldest forest reserve in the western world.

Map of southern Tobago showing the
Angostura properties >

Development map of the proposed Golden Grove and Buccoo Estates

How water flows through the constructed wetlands

Next - Wetlands Inhabitants - Plants (550K)
Introduction & Index

WGI ONLINE Journal Table of Contents

Water Gardeners International
Home | Join WGI | Members' Exclusive | Gateway to Water Gardening