**Table 2.**

*List of terrestrial and Martian images used in the analysis of putative microalgae on Mars. (Abbreviations: Up = Upper, Lw = Lower, R = Right, L = Left, F = Fossil).*

distribution of orientations, (5) overall shape of the profiles, i.e., whether fusiform or not fusiform and whether a degree of curvature was present. A Poisson distribution was fitted to the objects from all images. If the objects were distributed at random, then, the probability (P) that the fields contain various numbers of profiles is given by the Poisson distribution [45]. The variance (V) of a Poisson distribution is equal to its mean (M) and hence, the V/M ratio is unity for a random distribution. The V/M ratio can, therefore, be used as an index of "dispersion", uniform distributions having a V/M ratio less than unity and clustered distributions greater than unity. In addition, profile diameters and orientations were used to construct size class-frequency distributions. Two statistical models were fitted to these distributions: (1) a log-normal distribution [46], often used to describe the size distributions of plant populations [47, 48] and (2) a rectangular distribution, to test whether the objects exhibited any preference with regard to orientation. Goodnessof-fit to the various models was tested using the Kolmogorov–Smirnov (KS) test. To study the similarities among the ten images, the data were analyzed using Principal Components Analysis (PCA). The analyses were carried out using the images as variables and the various metrics as defining features. The result of PCA is a scatter plot of the images in relation to the extracted Principal Components (PC) in which the distance between the images reflects their relative similarity or dissimilarity based on the defining metrics. To correlate the location of an image on a PC axis with the specific metrics, correlations (Pearson's 'r') were calculated between the values of each metric from each image and the factor loadings of the case relative to the PC1 and PC2. For example, a significant correlation between a specific metric and PC1 would identify that feature as particularly important in determining the separation of cases along PC1.
