Adaxial and abaxial epidermal surfaces of johnsongrass leaves were studied to determine which cells contribute to leaf microroughness. Cork-silica cell (CSC) pairs, three types of prickles, macrohairs, bicellular trichomes, stomata, and ordinary short and long epidermal cells were found and described. CSC pairs made up about 22% of all cells and probably contribute more to microroughness than any other single type because each cork cell produces 11 ± 3 wax filaments that are up to 100 μm long. Bicellular trichomes represented 4 to 5% of the total cells but decreased leaf roughness by secreting a type of mucilage that covers microscopic wax crystals. Stomatal complexes comprised 15 to 18% of all cells and contributed to leaf roughness because they are slightly recessed below the leaf surface. Long prickles occur primarily over veins and represent less than 1% of all cells. Small prickles were present primarily on adaxial surfaces and represent only 3% of all cells. Macrohairs were the largest appendages, 237 ± 104 μm, but they represent far less than 1% of all cells and occur primarily over veins. Ordinary short cells comprised 6 to 13% of all cells. Long cells were most common (41%) of all cells. Short and long cells contribute to leaf roughness because the surface is often convex. A typical johnsongrass leaf may contain more than 25 million appendages on each surface that increase the roughness already caused by epicuticular wax crystals.

Johnsongrass leaves were covered with epicuticular wax that varied from 16 to 25 μg/cm2 on leaf blades and 56 to 206 μg/cm2 on leaf sheaths. At emergence, leaves were covered with a layer of smooth amorphous wax, but crystalline wax (wax plates) began to form on the amorphous wax within 1 or 2 days. This continued until all leaf surfaces were covered with wax plates. At 3 to 4 weeks of age, a smooth layer of coalescence wax was deposited over the wax plates. Formation of coalescence wax continued until nearly all leaf surfaces were covered with a smooth wax layer. Production of wax filaments began when plants were 3 to 4 weeks old and these tubular structures extended 100 to 200 μm above all other wax formations. Deposition of amorphous wax continued after stomata closed in the darkness, sealing over stomata, but the wax layer was broken when stomata opened again in the light. A capillary method was devised that was used to evaporate chloroform containing leaf waxes through 0.1- to 1.2-μm pores in inert filters to recrystallize amorphous wax and wax plates similar to that produced on johnsongrass leaves. Recrystallization of wax from wax filaments dissolved in chloroform produced the same structures of amorphous wax and wax plates as when only wax from leaves with amorphous wax and wax plates was used. Wax washed from leaves also produced wax plates and a variety of crystalline structures on the walls of glass vials after chloroform solutions were evaporated. This result indicated that the morphology of epicuticular waxes is influenced more by their inherent chemical and physical properties than by underlying cells or the cuticular membrane.