Satellite imagery was used to explore if seagrass species in the Mediterranean Sea have differing distributions in relation to annual changes in sea surface temperature.

Methodology

Seagrass Distribution

Seagrass distributions were obtained from the large Metadata set compiled by the The UN Environment World Conservation Monitoring Centre (UNEP-WCMC, 2017). The dataset was subset to include only occurances within the MEOWS mediterranean locations. Rupia species were ommited as they are not taxonomically true seagrasses. This resulted in data from five species of seagrass.

Sea Surface Temperature

Satellite sea surface temperatures files were obtained from the Ocean Biology Processing Group (NASA, 2018). Images used were from the MODIS-Aqua monthly climitology sea surface temperature at a 4km resolution from 2002-2019. The temperature of each raster was adjusted to degrees celcius based on the metadata of the files. A raster of the annual maximum and minimal temperatures was then generated. Using GIS software the minimum annual temperature was subtracted from the maximum annual temperature to genetrate a raster of the temperature fluctuations experienced in each raster cell annually. The vector points of the seagrass distribution dataset were used to extract the data from the temperature fluctuation raster. All GIS work was performed in QGIS 3.4.4.

Statistics

The temperature fluctuation data displayed an leptokurtic distribution with a positive skew when graphed. Using the formal Shapiro-Wilks test for normalilty we confirm that the data is not normally distributed. Only the annual temperature fluctuation of Halophila stipulacea comes from a normally distributed population. No transformation performed on the data fulfilled the assumption of normality and the assumption of homoscedastic residuals was also violated.

ANOVA testing is fairly robust to deviations from normality due to the central limit theorem and are also robust towards deviations of homoscedasticity as long as that ratio between the largest and smallest variance is less than 4X. Due to our large sample size (n = 1542), and the ratio of variances (2.60x) an ANOVA test was chosen to analyse the temperature data.

Results

An ANOVA was performed to examine the differences in annual temperature fluctuation between areas inhabited by the different seagrasses (Figure 1, Table 1). It was found that the seagrass species occupy areas with differing annual temperature fluctuations (F(4, 1537) = 29.36, p < 2.2 x 10-16). A post-hoc test revealed that the invasive species Halophila stipulacea is found in geographic areas with less annual temperature variance than the other four species (Figure 2). Zostera marina occurs in areas with a larger range of annual temperature variance than the rest. Cymodocea nodosa, Posidonia oceanica, and Zostera noltii are found in areas with similar annual temperature variations.

Figure 1. Seagrass species distribution and the annual temperature fluctuations exhibited in the Mediterranean Sea
Table 1. The annual temperature fluctuations (°C) experienced by each seagrass species.
n Mean +/- SE
Zostera marina 60 14.72 0.42
Zostera noltii 526 13.07 0.06
Cymodocea nodosa 754 12.94 0.07
Posidonia oceanica 114 12.82 0.14
Halophila stipulacea 88 11.47 0.14

Figure 2. The annual temperature fluctuation (°C) experienced by each seagrass species.

Discussion

This study analysed the sea surface temperature of the Mediterranean Sea and its relation to seagrass species distribution. It was found that H. stipulacea predominately occupies the eastern basin of the Mediterranean. This location consistently stays warmer than the western basin. The invasion of H. stipulacea into the Mediterranean has created competition with the local species. It could be that the isothermic seperation between these two water masses is controlling the spread of H. stipulacea, or it could be that this species cannot survive in cooler water. It is hypothesized that the Mediterranean sea will exhibit the effects of climate change at a faster rate than other water masses. If the sea water in this location continues to warm it could mean the spread of H. stipulacea further west.

References

NASA Goddard Space Flight Center, Ocean Ecology Laboratory, Ocean Biology Processing Group. Moderate-resolution Imaging Spectroradiometer (MODIS) Aqua Sea Surface Temperature Data; 2018 Reprocessing. NASA OB.DAAC, Greenbelt, MD, USA. doi: 10.5067/AQUA/MODIS/L3M/SST/2014.

UNEP-WCMC, Short FT (2017). Global distribution of seagrasses (version 6.0). Sixth update to the data layer used in Green and Short (2003). Cambridge (UK): UN Environment World Conservation Monitoring Centre. URL: http://data.unepwcmc.org/datasets/7