Supplementary high-resolution images and videos for "Plant water stress, not termite herbivory, causes Namibia’s fairy circles" (PPEES - doi:10.1016/j.ppees.2022.125698)

These visual data contain the same images and videos as the supplemental Word document (Appendix A), but the images are in high resolution.

Figure captions:

Fig. A.1.
Grass excavations at the Kam-1 plot undertaken in March 2022. Grasses were dead within the FCs and
these wilted plants occurred primarily at the inner edges of the FCs (a,b). Various sizes of still vital and greenish
matrix grasses (upper plants in c-f) have been compared to the dead grasses from within the FCs
(lower plants in c-f). The root morphology of the dead FC grasses is indistinguishable from the roots
of the vital matrix grasses and signs of termite herbivory are completely absent.

Fig. A.2.
Data loggers at the Jag-1 plot. Twelve soil-moisture sensors were installed at 20 cm depth along a transect
connecting two FCs (a,b). No grass cover existed during the drought in 2020, making this setup ideal to study
the effects of active grass growth on the soilwater distribution before, during, and at the end of the rainy seasons (c,d).

Fig. A.3.
Examples of infiltration measurements inside fairy circles in various regions of Namibia such as at Jag-1 in the
NamibRand Nature Reserve (a), at Mar-1 in the Marienfluss Valley (b), and at Tso-1 in the Tsondab Valley (c).

Fig. A.4.
Additional examples of grass excavations undertaken in 2021 at Garub (a,b) and Brandberg
(c-f). Grass death inside FCs at Gar-1 occurred despite the fact that most roots were undamaged and
were covered by an intact rhizosheath (a,b). Also at Brandberg in the Bra-3 plot, the grasses died
inside the FCs (upper plants in c,d) although the roots were undamaged. In the Bra-4 plot, the dead
grasses from within the FCs (upper plants in e,f) had on average longer roots than the vital green
grasses of the matrix (lower plants in e,f).

Fig. A.5.
Continuous soil-moisture measurements from dry to rainy season at Jagkop. This image
shows the download of the data on 25th February 2021, when the grasses were still vital and at full
bloom.

Fig. A.6.
Examples of two FCs where grass excavations have been done in the Jag-1 plot. The drone
images in (a) and (b) have been taken on 23rd February 2021, hence two days prior to the grass
excavations. Comparisons with drone images of the same locations from 15th February 2020 (c,d) and
with Google satellite imagery from 1st August 2016 (e,f) were used to specifically identify new
emerging FCs that are different from the long-lived mature FCs in the surrounding. The arrows in (a)
and (b) point to such exemplary FCs, where grass excavations have been undertaken.

Video A.1.
Short movie of the excavation of grass at the Kam-1 plot near the Kamberg. The movie
was recorded about a week after the grass-triggering rainfall event that occurred on 4th March 2020.

Video A.2.
Short movie of the excavation of grasses at the Bra-4 plot at Brandberg, about 35 days
after grass-triggering rainfall.