Diameter growth, ingrowth and mortality models for uneven-aged forests in Slovenia

Trifković, Vasilije (author), Ficko, Andrej (mentor), Bončina, Andrej (comentor)

We developed diameter growth, recruitment (ingrowth), and mortality models for the most common tree species in Slovenian uneven-aged forests using tree, stand, site, and climatic predictors. Utilizing permanent sample plot data of the Slovenia Forest Service (1990–2014), we employed cubic spline regression for diameter growth models, Tobit censored regression for recruitment models, and quantile regression and stochastic frontier analysis for Reineke’s density-dependent mortality models. Binomial generalized linear mixed models were used for individual tree mortality probability models. R2 values in the diameter growth models ranged from 5% to 51%, with deciduous species generally exhibiting a lower R2 than conifers. The effect of climate change on diameter growth was studied for European beech. European beech responds asymmetrically to precipitation anomalies: growth reductions in dry periods are up to three times greater than growth increases in wet periods. Tree recruitment is highly stochastic and primarily influenced by stand conditions. The models enable the prediction of optimal and critical ranges of stand, site, and climatic factors for recruitment of the main tree species. Annual tree mortality averaged 3.26 trees ha- 1, i.e., a basal area and a growing stock reduction of ⠒0.32 m2 ha-1 and ⠒0.83 m3 ha-1, respectively. Mono-species stands had a higher maximum stand density than multi-species stands. The proportion of European beech positively correlated with an increased intercept and slope of the maximum stand density line. The models can be used for the development of a growth simulator for uneven-aged forests and forest management optimization.

• Language: English
• Keywords: diameter growth, tree mortality, ingrowth, precipitation anomalies, self-thinning, uneven-aged forests
• Work type: Doctoral dissertation
• Organization: BF - Biotechnical Faculty
• Year: 2024
• PID: 20.500.12556/RUL-156088
• Publication date in RUL: 08.05.2024

Secondary language

• Language: Slovenian
• Title: Modeli debelinske rasti, vrasti in mortalitete drevja v raznomernih gozdovih Slovenije

Abstract

Za najpogostejše drevesne vrste v slovenskih raznomernih gozdovih smo razvili modele debelinske rasti, vrasti in mortalitete, pri čemer smo uporabili podatke stalnih vzorčnih ploskev Zavoda za gozdove Slovenije (1990 2014) in podatke o rastiščnih in podnebnih dejavnikih. Za modeliranje debelinske rasti smo uporabili regresijo naravnih kubičnih zlepkov, za vrast Tobit krnjeno regresijo, za parametrizacijo modelov mortalitete, odvisnih od gostote sestoja, pa kvantilno regresijo in metodo stohastične meje. Za oceno verjetnosti mortalitete drevja smo uporabili posplošene mešane linearne modele. Vrednosti R2 za modele debelinske rasti so bile med 5 % in 51 %, pri čemer so modeli listavcev običajno izkazovali nižje vrednosti R2 v primerjavi z iglavci. Učinke podnebnih sprememb na debelinsko rast smo preučili na primeru bukve; bukev se asimetrično odziva na padavinske anomalije. Zmanjšanje rasti bukve v sušnih obdobjih je v povprečju trikrat večje od povečanja rasti v mokrih obdobjih. Vraščanje dreves v raznomernih gozdovih je precej stohastično, največji vpliv na vrast imajo sestojni dejavniki. Z modeli vrasti lahko napovedujemo optimalne in kritične intervale sestojnih, rastiščnih in podnebnih dejavnikov. Povprečna letna mortaliteta dreves je 3,26 drevesa ha-1, oziroma 0,32 m2 ha-1 sestojne temeljnice in 0,83 m3 ha-1 lesne zaloge. Čisti sestoji dosegajo večje maksimalne gostote kot mešani sestoji. Z naraščanjem deleža bukve se povečujeta maksimalna gostota sestoja in naklon Reinekejeve krivulje. Modele lahko uporabimo za izdelavo simulatorjev razvoja raznomernih sestojev in optimizacijo upravljanja gozdov.

• Keywords: debelinska rast, mortaliteta dreves, vrast, padavinske anomalije, sestojne gostote, raznomerni gozdovi