*4.2. Source of Joint-bound Blocks*

Equations (1) and (2) from Nott (2007) and Pepe et al. (2018) [18,20] are specific to wave energy applied at the source against joint-bound blocks exposed in rocky shorelines. All materials subsequently transferred to the 450 m long spit at San Luis Gonzaga originated due to wave erosion against andesite cliffs exposed on the north side of the island. The rocky shore extends EW for almost a kilometer, rising topographically to 60 m, or more (Figure 1c). Sea cliffs in the area nearest the spit exhibit andesite flows with bedding planes that dip at a high angle to the west with irregular joints perpendicular to bedding planes (Figure 3a,b). Large blocks of andesite only crudely rounded by abrasion occupy the intertidal zone on a wave-cut platform at the side of an uplifted marine terrace. Fresh material in the supratidal zone lacks the darker tone of blocks colored by organic growth.

#### *4.3. Andesite Specific Gravity*

Five samples of andesite from the north shore of Isla San Luis Gonzaga yielded a range of values for specific gravity between 2.26 and 2.34 gr/cm3. The samples ranged in weight between 215 and 620 gm and were displaced between 151 and 271 mL of water. The mean value calculated from the samples amounts to 2.3 gr/cm3 and this value was uniformly applied to Equations (1) and (2) in estimation of wave heights provided in Tables A1–A7.

**Figure 3.** Rocky shore and intertidal zone at cliffs adjacent to the marine terrace and near the start of the 450-m long spit on Isla San Luis Gonzaga; (**a**) Oblique view with housing of meter tape for scale; (**b**) Head-on view showing tilted andesite flows in the background.

#### *4.4. Comparative Variation in Clast Shapes*

Raw data on boulder size in three dimensions collected from each of the seven transects are available in Appendix A (Tables A1–A7). Due to the wealth of data collected in the field, the size of Tables A1 and A2 is limited to a representative sample based on 50% of the boulders. Table A3 is the smallest, because few boulders were encountered, all of which are included. Likewise, all boulders from transects 4 to 7 are enumerated in Tables A4–A7. With regard to shape, points representing individual clasts (including smaller cobbles) are shown grouped by transect and plotted on a set of Sneed–Folk triangular diagrams (Figure 4a–g). The spread of points across all seven of the plots is remarkably consistent, showing a strong similarity in the variation of shapes from one transect to another. It is seldom that points fall into the upper-most triangle, which represents an origin from a cube-shaped endpoint. The majority of points from all seven plots falls within the middle part of the two tiers below the top triangle. Those points clustered at the core of any given triangular plot are representative of clasts for which two dimensions are closer in value than the third measured along the shortest axis. However, a significant portion of points falls into the middle-right and lower-right domains of the field, which signifies a tendency for development of elongated shapes eroded from bar-shaped blocks of source rock. The composite slope of points across all plots from the few in the topmost triangle to those in the lower right corner of the field demonstrates a tendency for development of moderately oblong shapes. Rarity of points in the middle bottom tier and complete absence of points in the lower-left corner of triangular plots indicates that the wave-eroded material from the parent sea cliffs excludes plate-shaped blocks.

#### *4.5. Comparative Variation in Clast Size*

Clast size is conveniently plotted on bar graphs as a function of maximum length based on the original data (see Tables A1–A7 for boulders). Transect 1 (Figure 5) is the station physically closest to the bedrock source and, in principal, is expected to reflect the highest proportion of boulders.

**Figure 4.** Set of triangular Sneed–Folk diagrams used to appraise variations in cobble and boulder shapes; (**a**) Trend from transect 1 closest to sea cliffs at the source of the clasts; (**b**) Trend from transect 2; (**c**) Trend from transect 4; (**d**) Isolated trend from transect 3 with a markedly different orientation from all other transects; (**e**) Trend from transect 5; (**f**) Trend from transect 6; (**g**) Trend from transect 7 most distant from sea cliffs at the common source of all clasts in the system.

Conversely, transect 7 is most distal from the bedrock source and is expected to show a higher proportion of cobbles. Groupings separated in bins at intervals of 5 cm are arrayed in histograms stacked to show differences in size range between 6 cm and 90 cm among transects 1, 2, and 4 (Figure 6a–c). Data for transect 1 (Figure 6a) registers the highest concentration of small boulders. Transect 3 projects as a side spur on the southeast side of the spit (Figure 7). It is excluded from this analysis due to the relative scarcity of boulders. Each of the three graphs in Figure 6 delineates the boundary between cobbles and boulders with a dashed line. They are consistently skewed with the highest percentage of clasts at or around the border between the largest cobbles and smallest boulders. The ratio between boulders and cobbles remains steady at 2:1 in transects 1 and 2 but is closer to parity in transect 4. Compared to the proximal transects, the more distal transects 5 to 7 (Figure 8a–c) exhibit a marked shift in skewness to a numerical domination by cobbles. In transects 5 and 6, the relationship between boulders to cobbles is roughly consistent dropping to a ratio of 1:3. However, data from transect 8 at the distal end of the spit (Figure 8c) records a ratio at parity. In all cases among the six transects represented by bar graphs, the extreme outlier of large boulders occurs at measured lengths between 76 cm and 85 cm.

**Figure 5.** Transect 1 (dashed line) features the highest concentration of small and intermediate size boulders. The survey line passes through the spit about 40 m beyond the marine terrace at the NE corner of Isla San Luis Gonzaga. Superposition of white arrows defines the outer lip of the marine terrace elevated 8.5 m above sea level.

**Figure 6.** Set of bar graphs used to appraise variations in maximum boulder length from the three transects with similar orientations closest to the source of coastal erosion; (**a**) Bar graphs from transect 1; (**b**) Bar graphs from transect 2; (**c**) Bar graphs from transect 4. Transect 3 is excluded from this treatment on account of its deviant orientation.

**Figure 7.** Transect 3 (dashed line) follows a side spur off the main spit with a compass orientation 90◦ out of phase with other transects. Boulders are uncommon, but the largest in the foreground exceeds 50 cm in diameter. The spur is notably isolated from the opposite side of the spit by a dense thicket of Sweet Mangrove (*Maytenus phyllanthoides*) and is largely submerged during high tide.
