The Effect of Land Development on Coral Health and Ecosystem Services in Belize

by: Shiela Walsh

Collaborating researcher: Jessica Carilli (SIO-Geosciences)
Research Assistant:  Jason Murray (PhD Economics)

Objective
The objectives of this project are to use biological and geological measures of coral health in the Mesoamerican Reef in order to estimate changes in ecosystem services from the reef. The first objective was to measure the change in coral health over time using coral growth rates, mortality and bleaching frequency under conditions of high and low influence of land development. The second is to use coral growth rates over time and coral percent cover, size structure, and diversity across space as a proxy for the change in ecosystems services (such as shoreline protection and dive tourism). Completing the second objective will allow us to estimate the forgone value in terms of coral reef ecosystem services for Belize due to land development.

Motivation
Agricultural development in the steep mountain watersheds of Honduras and Guatemala over the last century are expected to have led to increased sedimentation and decreases in the health of corals in Belize over 200 km north. SeaWiFs images of major sediment plumes and evidence of stress tolerant zooxanthellae (symbiotic algae) in the corals motivated us to test for the effect of land development in Honduras and Guatemala over space and time in the Belize Barrier Reef.

             (adapted from a. Garren et al. in press, b. Andrefouet et al. 2002)

Sites
We chose eight sites in two regions representing a gradient of land-based impact to study in 2005. Four of the sites in the low impact region (denoted by the red circle) were surveyed in 2003 by Sheila Walsh. We hypothesize that the southern region is more heavily impacted by runoff due to its proximity to the watersheds of Honduras and Guatemala as well as the average coastal current direction.  We conducted surveys and took coral core samples at the eight sites on two trips in October 2005 and February 2006. We plan to include benthic data from an additional atoll and barrier reef region (four sites in each, denoted by black circles and collected by Marah Newman and Gustavo Paredes in 2004) in order to control for differences in habitat between the levels of impact.  

Methods
We collected the data that we need to complete our two research objectives by completing one survey trip in which we collected benthic data and scouted for coral core collection sites and then we returned on a second trip to make coral core collections. During the first trip we collected data on percent cover of benthic species (corals, algae, sponges, gorgonians, and zooanthids), coral bleaching and disease frequency, and size structure of the corals.

Species Percent Cover
We measured percent cover of benthic species at each site using the point-intercept method. The point-intercept method measures percent cover by recording the occurrence of the benthic organisms every 25 cm along a transect of 30 m. For example, if a species of coral was observed 12 times out of the 120 possible observations along the transect, the species of coral would have an estimated percent cover of 10% for that transect. Percent cover estimates were made along five transects at each of the four sites per region. Transects were laid perpendicular to the reef crest within a 100 m^2 area at a minimum distance of 5 m apart.
 

Coral Bleaching, Disease, and Size Structure-  
We recorded the percentage of bleaching (rated as pale, partly bleached, bleached), occurrence of disease, and the length/width/height of all corals along each of the five transects per site.

The survey trip gave us important data to test for changes in the coral community over a spatial gradient of land development. It also confirmed the availability of a limited number of large coral colonies that we could use to obtain coral cores to test for changes in coral health over time. On our second trip, we collected coral cores and conducted more surveys of the size structure of the coral community around the areas where the cores were taken.

Coral Health (growth rates)-Cores of coral were collected to test for changes in growth rates as a proxy for coral health as well as various geochemical proxies representing different stressors on the reef over time. Stressors include sedimentation, freshwater plumes, and agrochemical pollution.  

Results
We collected data on the percent cover of benthic species, frequency of coral bleaching and disease, and the size structure of the coral community from twelve sites (four in the low impact area and eight in the high impact area). Our major results showed a decrease in coral cover from high to low impact regions from 11% to 8% on average. At the same time, we observed an increase in algal percent cover 50% to 80% from high to low impact regions. Sponges, potential indicators of land run-off, did not show any pattern. In general, there was greater variability in the reef community in the high impact region. Considering the entire community at once, sites generally sorted from the high impact region (sites 1-4) and the low impact region (5-12) based on their relative percent cover of coral and gorgonians (octocorals) versus algae. Corals were also affected by bleaching at a higher frequency in the low impact site than the high impact site.

We also collected seven coral cores from the low impact area and 34 from the high impact area. These cores allow Jessica Carilli to construct a 50 to 100 year time series of coral health for each region by stacking the cores of from different colonies. She analyzed the cores with x-rays to determine the coral growth rate by identifying high and low density bands that indicate yearly growth. She found a consistent negative growth rate for the high impact region and no change or an increase in the growth rate from the low impact regions.

Discussion
Once the analyses from the coral cores are completed, we plan to look for explanatory variables to understand what is driving the pattern both spatially and temporally. Currently, we expect that the amount of agricultural land developed in each country over time, temperature, and hurricane impact will be important variables. The patterns that we have found, in which reefs in the low impact sites are in worse condition than in the high impact sites, are unexpected. These results suggest that either the disturbances in these areas, not including land run-off disturbance, are extremely different. An alternate explanation could be that coral reefs closer to land are pre-adapted to land run-off and are dealing better with an increase in this stressor than reefs further away. This hypothesis is supported by the general trend in coral cover loss in the Caribbean in the last twenty years, in which reefs near land decreased more slowly. If pre-adaptation proves to be the best explanation for these results, it suggests that policies to deal with the costs of land run-off on ecosystem services will be complex.