(
A) Time evolution of the acircularity index
α of the biofilm contour. Data are shown in magenta, blue, and black for three biofilms grown on 0.4% (solid curves) and 1.0% (dashed curves) agar plates. (
B) Profiles of transmitted light intensity
I (black), local curvature
κ (red), and local expansion velocity
Vf (blue) along the periphery of the biofilm shown in
Figure 6B but at 32 hr. At 32 hr, the wrinkling/delamination pattern had not yet reached the boundary, so the biofilm remained approximately circular, and thus both
I and
κ remain constant over the biofilm periphery.
Vf fluctuated modestly, probably because of noise at the growing front and errors in the edge tracking process. (
C) Image of a WT
V. cholerae biofilm grown on 0.7% agar taken 30 hr after inoculation.
Rf and
Rp denote the radius of the entire biofilm (outlined in red) and the distal radius to which the morphological pattern extends (outlined in blue), respectively. Scale bar: 2 mm. (
D) Time evolution of
Rf (solid curves) and
Rp (dashed curves). Data are shown from three different WT
V. cholerae biofilms (denoted by different colors) grown on 0.4% agar. Initially,
Rp lags behind
Rf, indicating a peripheral zone in which the biofilm remains flat and lacks identifiable features.
Rf increases linearly with time, which defines an expansion velocity <
Vf> . At a critical time
t’c (35 hr (black), 42.5 hr (blue), and 45 hr (magenta) for the three cases shown), the region harboring wrinkles or blisters rapidly propagates to the edge of the biofilm (
Rp =
Rf). Concurrently, the global biofilm expansion velocity slows, with a crossover time point at
t’c. Inset: close-up view of
Rf and
Rp versus time for one set of data (blue) around
t’c. The critical time that is defined in this manner matches that defined from the
α – t plot in
Figure 6A (i.e.
tc =
t’c), suggesting that mechanical instability at the biofilm edge triggers the development of contour undulations. (
E) Averaged biofilm expansion velocity
<Vf> before (white) and after (gray) morphological features appear at the rim.
<Vf> was calculated by linear fitting of
Rf versus time. Data are represented as mean ± std with
n = 3.