I had a look at the source of this table, Ukrainian Research Institute, Harvard University, and found at first that there was no per 1000 map of the deaths of only the rural population although the overall percentage is known (which was a wrong finding, I found that map later in the same source, see at the bottom):
4 percent of the urban and 16 percent of the rural population in 1933.
Full quote of the source
Map of Total Direct Famine Losses in Ukraine by Region, 1932-1934
This map shows absolute numbers of excess deaths in Ukraine's seven
oblasts and the Autonomous Moldavian Socialist Soviet Republic
(Moldavia). The shading of each region represents the total number of
excess deaths during 1932-1934, while the bar charts in each region
represent annual excess deaths. Numbers of excess deaths for 1932-1934
are divided into three intervals, depicted by three different
color-scale shades, and the number for each region is shown on the
The total number of 1932-1934 famine losses in Soviet Ukraine is
estimated at 4.5 million, with 3.9 million direct losses (excess
deaths) and 0.6 million indirect losses (lost births). The 1933
midyear population of Ukraine is estimated at 29.6 million. The 3.9
million direct losses are approximately equivalent to 13 percent of
the total population in 1933. The 3.9 million is composed of 0.3
million in urban areas and 3.6 million in rural areas; these numbers
are approximately equivalent to 4 percent of the urban and 16 percent
of the rural population in 1933. It is important to note that the
Holodomor is commonly understood as occurring in the period 1932-1933.
However, the analysis here shows that significant famine-related
losses also occurred in 1934. This is the first time that regional
estimates of direct losses have been made for Ukraine.
Full source information, again quoted from the source
This map visualizes data based on estimates of deaths during the
Holodomor period derived under the "Estimation of Regional Losses of
the 1932-1934 Famine in Ukraine" project conducted by Oleh Wolowyna
(University of North Carolina at Chapel Hill) and Omelian Rudnytskyi,
Nataliia Levchuk, Pavlo Shevchuk, and Alla Kovbasiuk (Institute of
Demography and Social Studies in Kyiv).
The estimation here of Holodomor losses is based on separate yearly
urban and rural population reconstructions by age and sex, using data
from three Soviet censuses (1926, 1937, and 1939), as well as annual
vital statistics and migration data for the 1927-1939 period. 1926 and
1937 census data were adjusted for undercounts and other problems,
while 1939 census data were adjusted to eliminate falsifications
implemented to cover up large population losses understood to be due
to the famine and other repressive policies. Yearly registered births
and deaths were adjusted for undercounts, especially during famine
years when the undercount reached extreme levels. Based on these data
and detailed data on migration, taking into account urban-rural
reclassification, yearly population figures were reconstructed
according to age and sex for Soviet Ukraine's eight regions, as well
as for urban-rural areas.
One could calculate this for the whole rural areas:
The 3.9 million is composed of 0.3 million in urban areas and 3.6 million in rural areas; these numbers are approximately equivalent to 4 percent of the urban and 16 percent of the rural population in 1933.
With the also said 29.6 Mio population, with the share of 3.6/3.9=92.3% in rural deaths, the sum of "Total Deaths (1932–1934 in thousands)" is 3942.5. Then (own calculation):
and rural population:
and town population:
So that on the whole, the Holodomor excess deaths per 1000 were:
- 161.7 (rural)
- 40.4 (town)
There are maps on rural and urban splits of Holodomor excess deaths per 1000 in the same source:
You can find more maps about 1932, 1933 and 1934 region levels there as well and other maps, but the main thing is that the rural population had to suffer much more than the urban population, rural range from roughly 100 to 500 while urban from 10 to 100 (very roughly, just looking at the colours). I do not have enough reputation to post more than the eight links here, have a look at the absolute number maps on the website.