DIGITALNA ARHIVA ŠUMARSKOG LISTA
prilagođeno pretraživanje po punom tekstu



ŠUMARSKI LIST 7-8/1973 str. 15     <-- 15 -->        PDF

Summary


CORRELATION ANALYSES OF HEIGHT INCREMENT, HEIGHT, DIAMETER


B. H. AND DIAMETER INCREMENT IN CERTAIN STANDS OF SILVER FIR
In stands of Silver Fir belonging to the forest communities of Blechno-Abietum
Horv. (Forest Districts Sušica and Tuški Laz) and Fagetum abietetosum Horv. (Forest
District of Toričak) in the Gorski Kotar Reg on (Croatia) there were taken
measurements of felled Fir trees. In each tree were measured: diameter b. h. (di,a =
1), total height (h = 2), diameter increment of the last five years (ia5 = 3), diameter
increment of the penultimate five years (ijdo-s) = 4), height increment of the last
five years (ihs = 5) and height increment of the penultimate five years (ih(10 5) = 6).
There vere measured a total of 755 Fir trees, of which 231 in the Forest District of
Sušica, 268 in Tuški Laz and 256 in Toričak. The diameter increment was established
by using the Pressler borer, while the height increment with a tape by measuring
distances between whorls. On the basis of measured magnitudes the author wished
to establish the degree of the connection of the mentioned variables in various
different Fir stands and to find out a favourable combination of available variables
on the basis of which it would be possible to perform an assessment of height increment.
In addition to the mentioned six variables there was also taken into computation
the average annual diameter increment of a 10-year period (i,jio = 7) and
the average annual height increment of a 10-year period (ihio = 8).


Through the correlation analysis the author computed the linear partial and
multiple correlation coefficients under the assumption that the studied population
possesses a mult.variable normal distribution. Data processing was performed on
an electric computer of the CAE 90—40 type. Because of the volume of the material
in Tables the author presents — according to his opinion — the interesting partial
(Tabs. 1—4) and multiple (Tab. 5) correlation coefficients. Testing of significance was
performed by means of the null hypothes s and the Z-transformation by R. A.
Fisher. Although in the correlation analysis it is not possible to determine which
is a dependent variable and which not, the author considered the height increment
primarily a dependent variable because he was interested in the assessment of
height increment. Correlation between the height increment and diameter b. h., as
between the height increment and height was very weak or almost absent in all
the three forest districts irrespective of which order of correlation coefficients was
in question. Only in the Forest District of Sušica the partial coefficients of the first
order — under the exclusion of the influence of diameter increment — were of medium
strength (Tab. 1). The height increment of the last and penultimate five years
exhibited a strong connection in the Forest District of Sušica and Tuški Laz, while
in the Forest D strict of Toričak this connection was very weak (Tab. 2). The correlation
between the height increment and diameter increment in all three forest
districts was positive: in the Forest District of Sušica medium, in Tuški Laz weak,
in the Forest District of Toričak a very weak correlation only between the height
increment and diameter increment of the last five years; other partial correlation
coefficients in the Forest Distr ct of Toričak were not signnificantly different from
zero (Tabs. 3 and 2). The diameter increment and diameter b. h., as the diameter
increment and height were in a significant correlat on in all three forest districts
(Tab. 4). However, under the exclusion of the influence of diameter b. h., there was
no correlation between the diameter increment and height in the Forest D´stricts
of Tuški Laz and Toričak, while in the Forest District of Sušica this connection
existed between the 95- and 99-percent limit of significance. There was also no
correlation between the diameter increment and diameter b. h. under the exclus:on
of height in the Forest District of Sušica, while in the Forest Districts of Tuški Laz
und Toričak it did exist. Multiple correlation coefficients between the height increment,
diameter b. h. and heights were very small in the Forest Districts of Sušica
and Toričak, while in the Forest D strict of Tuški Laz they were insignificant (Tab.
5). The height increment exhibited the best correlation with the diameter increment
and with other factors associated with the diameter increment. Multiple correlation
coeff cients between the height increment, diameter b. h. and diameter increment
respectively, or between the height increment, diameter increment and