The same F-move read from top to bottom: d at the top splits into a, b, and c below, with the same coefficient [F^{abc}_d]_{ef}.
Rendering
$\begin{tikzpicture}[thick, baseline=(current bounding box.center)]
\draw (0,0) node[below] {$a$} -- (0.5,0.7) -- (1,0) node[below] {$b$};
\draw (0.5,0.7) -- node[left] {$e$} (1,1.4) -- (2,0) node[below] {$c$};
\draw (1,1.4) -- (1,2) node[above] {$d$};
\end{tikzpicture}
\;=\; \sum_f \bigl[F^{abc}_d\bigr]_{ef}
\begin{tikzpicture}[thick, baseline=(current bounding box.center)]
\draw (0,0) node[below] {$a$} -- (1,1.4);
\draw (1,0) node[below] {$b$} -- (1.5,0.7) -- (2,0) node[below] {$c$};
\draw (1.5,0.7) -- node[right] {$f$} (1,1.4);
\draw (1,1.4) -- (1,2) node[above] {$d$};
\end{tikzpicture}$