CombineResult.plot_forest(ax=ax) does not pass ax to dot_plot()

[mjrosen]

statsmodels/statsmodels/stats/meta_analysis.py

Line 322 in 6a9ce0a

fig = dot_plot(points=res_df["eff"], intervals=hw,

CombineResults.plot_forest(ax=ax) does not actually pass the mpl axis when invoking dot_plot(). So a new axis is created by dot_plot() whether or not an axis is passed to plot_forest(). I came across this with version 0.13.2 but appears still to be the case.
Removing ax as an explicit argument from plot_forest() would even solve the problem, as optional *kwds are passed to dot_plot() anyway.