Climate fluctuation impacts in Astronium urundeuva (M. Allemão) Engl. silvicultural characters in the Brazilian Cerrado

Evidence related to climate change still tends not to be accepted politically, which makes it difficult to create prevention strategies (Martins et al 2010). It is the first time a biotic agent has caused so many changes at the same time, when considering a problem in Earth's history. Even when we take a look at the geological time scale, humans are a small interference point in our planet's existence, it has sacrificed the whole life existence (Steege 2010). Predictions about climate change are usually catastrophic, and the evidence is even more catastrophic, but have an important role in reducing impacts and development strategies, that also fails due to methodology multiplicity and high results variability, which makes it difficult to see a better scenario about our future (Pereira et al 2011, Bellard et al 2012, Santos et al 2020). These predictions are made based on models, used to infer possible scenarios, with data from different sources (paleontological, recent, experiments, observations and meta-analyses) (Lepetz et al 2009, Alvares et al 2021). These models also focus only on one response kind: spatial changes, without considering other aspects. As a science in its early stages, there is still a lot to be improved (Bellard et al 2012).

Their impacts are also measured on ecological processes (Scheffers et al 2016). Climate effects are evaluated, particularly in geographic occurrence areas, changes terms and population dynamics, and even allow for losses quantification in species genetic diversity (Lima et al 2017, Abreu-Jardim 2018). Changes in native forest species simulated distribution in future climate scenarios has been reported (Siqueira and Peterson 2003, Wrege et al 2009, Diaz et al 2011, Simon et al 2013, Guitérrez and Trejo 2014, Diaz et al 2007), including species decline or extinction in natural regions (Guitérrez and Trejo 2014).

Brazil ranks highly amongst regions most vulnerable to climate change, and there is no security, as our economy is heavily dependent on renewable natural resources. Furthermore, great biodiversity present in tropics, with poor adaptive elasticity, intolerant to abrupt changes, shows how vulnerable our heritage is (Nobre 2008). In Cerrado, up to 90% of species will suffer spatial distribution reduction (Siqueira and Peterson 2003), with major impacts on economy, due to viability loss of use (Nabout et al 2011). Losses at the Cerrado species geographic distribution can reach 78%, on average (Simon et al 2013).

The Cerrado occupation is mainly focused on agricultural production, and these are areas which suffer greatly from the pressures of deforestation (Silva et al 2012, Aubertin 2013). Remembering the Cerrado is formed by complex ecosystems , agricultural expansion causes habitat fragmentation, which reduces species viability and reproduction, intensifies native areas replacement, compromises biodiversity and biome's ecosystem functions (Carvalho et al 2009, Silva et al 2012). Furthermore, competition with agricultural species and their frontiers means that forest areas are installed in more restricted conditions, compared to productive areas (Alvares et al 2021). In Brazil, according to IPCC projections (IPCC 2021), there will be a tropical to arid climate expansion, with greater occurrences of extreme events and ecological imbalance (Souza Jr et al 2020, Alvares et al 2021).

Among Cerrado species, 'Aroeira' (Astronium urundeuva (M. Allemão) Engl., Anacardiaceae) is a tropical, dioecious and native species (Silva-Luz et al 2020), with high wood density (1.19 g cm⁻³) (Paes et al 2009), considered as late secondary or anthropic pioneer (Kageyama et al 1994, Ferretti et al 1995), bee-pollinated and wind-dispersed, with occasional dispersal by psittaciformes (Lorenzi et al 1992, Carvalho 1994). Although it is not listed in Brazil's Environment Ministry threatened species list (Silva-Luz et al 2020), it was and still is one that has been legally logged, but mainly illegally (Moraes and Freitas 1997), due to its wood (known as the hardest wood in Brazil) and non-wood multiple uses (widely for medicinal purposes) (Chaves et al 1998, Viana et al 2003, Monteiro et al 2006, Lucena et al 2008, Oliveira et al 2010). Despite efforts to manage and conserve the species, its ability to survive is compromised (Freitas et al 2006).

Since distribution patterns in tropical species are not so well understood (Kanashiro et al 2002), knowing their landscape becomes important not only for natural resource conservation, but also for climate change adaptation. As the Cerrado is considered as a distinct and complex ecoregion, and A. urundeuva, as a species, is widely distributed throughout the biome, this research aimed to investigate whether the species will be able to survive due to climatic fluctuations, through economic silvicultural character evaluation.

2.1. Study site

Test installations took place at different periods, with different size plots (table 1), installed at the Teaching, Research and Extension Farm (Fazenda de Ensino, Pesquisa e Extensão: FEPE), Ilha Solteira Engineering Faculty (Faculdade de Engenharia de Ilha Solteira: FEIS), São Paulo State University (Universidade Estadual Paulista: UNESP), located to Paraná River's right bank, in Selvíria municipality, Mato Grosso do Sul state, at 22° 22' 02' S and 51° 25' 08' W and 335 m above sea level (figure 1).

Zoom In Zoom Out Reset image size

Table 1. General characterization of Astronium urundeuva populations from Cerrado in Selvíria-MS.

PT	Prov.	P/P	Prog.	N	Spc. (m)	Rep.	Installation	Age (2019)
CUIA	Cuiabá/MT	1	29	615	4 × 3	24	07/01/2010	9
GO	Itarumã/GO	3	30	1080	2 × 6	12	28/06/2004	15
PF	Paulo de Faria/SP	6	30	540	3 × 6	3	12/05/1997	22

PT = progeny tests; Prov. = provenance; P/P = number of plants per plot; Prog. = number of progenies; N = number of individuals; Spc. = spacing; Rep. = Repetitions.

Local climate is tropical humid, with rainy season in summer and dry in winter (Aw) (Köppen 1948), 24.5 °C average annual temperature, 1350 mm average annual precipitation and 66% average annual relative humidity (Santos and Hernandez 2013). Soils are moderately flat to wavy relief, composed by a dystrophic typical Red Latosols (LVd), compacted, very deep and moderately acidic (Santos et al 2018).

2.2. Progenies tests populations

2.3. Silvicultural and climate data collection

Silvicultural characters evaluated, in each progeny test, were height (HT, m) and diameter at breast height (DBH, m), for all individuals. General climatic characteristics, such as average annual temperature (${T_m}$, °C), average precipitation (${P_m}$, mm), accumulated precipitation (${P_a}$, mm) (figure 2), average reference evapotranspiration (${\text{Et}}{{\text{o}}_m}$, mm), accumulated reference evapotranspiration (${\text{Et}}{{\text{o}}_a}$, mm) and water available in soil amount (given by ${H_2}{O_d}\,$ = ${P_a}$—${\text{Et}}{{\text{o}}_a}$, mm), from 1996 to 2018, were obtained from UNESP Ilha Solteira Weather Channel (UNESP 2020).

Zoom In Zoom Out Reset image size

2.4. Data analysis

From 1996 to 2018, average annual temperature ranged between 24.4 °C and 25.6 °C, while throughout the 23 years there was always a water deficit (179 mm yr⁻¹ average), although average annual precipitation (1363 mm) complies with region standards (table 2, figure 3). This implies that, over time, there was a lack of water in plantations (${{\text{H}}_{\text{2}}}{{\text{O}}_d}$ = −179.8 mm, figure 4).

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

Table 2. Basic statistical analysis for descriptive water balance in Selvíria-MS between 1996 and 2018.

Parameters	${T_m}$ (°C)	${P_m}$ (mm)	${P_a}$ (mm)	${\text{Et}}{{\text{o}}_m}$ (mm)	${\text{Et}}{{\text{o}}_a}$ (mm)	${{\text{H}}_{\text{2}}}{{\text{O}}_d}$ (mm)
Mean	24.94	3.75	1363.89	4.28	−1543.67	−179.78
Variance	0.16	0.39	48 315.22	0.08	10 872.05	71 441.88
Deviation	0.40	0.63	219.81	0.29	104.27	267.29
CV%	1.59	16.66	16.12	6.77	6.75	148.68
Maximum	25.6	4.7	1727.6	4.7	−1296.0	230.7
Minimum	24.4	2.5	929.8	3.6	−1712.0	−742.4

${T_m}$: annual mean temperature, ${P_m}$: annual mean precipitation, ${P_a}$: annual accumulated precipitation, ${\text{Et}}{{\text{o}}_m}$: annual mean reference evapotranspiration, ${\text{Et}}{{\text{o}}_a}$: annual accumulated reference evapotranspiration, ${{\text{H}}_{\text{2}}}{{\text{O}}_d}$: annual average available water.

There were few years in which there was rainfall higher than evapotranspiration at test regions, for the whole period (1996–2018) (table 2, figure 4). The pattern consists of water deficit, regarding water availability. In 2018, it is noted that evapotranspiration is relatively constant, while average monthly precipitation is poorly distributed, that mainly occurs in spring and summer (from October to January), typical of Selvíria region climate (table 3, figure 5).

Zoom In Zoom Out Reset image size

Table 3. Descriptive statistical analysis of sequential water balance in Selvíria-MS, in 2018.

Parameters	${P_m}$ (mm)	${P_a}$ (mm)	${\text{Et}}{{\text{o}}_m}$ (mm)	${\text{Et}}{{\text{o}}_a}$ (mm)	${\text{BH}}{{\text{S}}_m}$ (mm)	${\text{BH}}{{\text{S}}_a}$ (mm)	${{\text{H}}_{\text{2}}}{{\text{O}}_d}$(mm)
Mean	3.14	95.33	3.55	−108.00	−0.41	−12.68	−12.68
Variance	8.88	8277.26	0.31	297.79	7.20	6755.63	6755.63
Standard deviation	2.98	90.98	0.56	17.26	2.68	82.19	82.19
CV%	94.80	95.44	15.77	−15.98	−656.44	−648.46	−648.46
Maximum	8.4	259.8	4.43	−74.7	4.5	137.1	137.1
Minimum	0.0	0.0	2.5	−137.2	−3.0	−93.4	−93.4

${P_m}$: monthly average precipitation, ${P_a}$: monthly accumulated precipitation, ${\text{Et}}{{\text{o}}_m}$: monthly average reference evapotranspiration, ${\text{Et}}{{\text{o}}_a}$: monthly accumulated reference evapotranspiration, ${\text{BH}}{{\text{S}}_m}$: monthly average sequential hydric balance, ${\text{BH}}{{\text{S}}_a}$: monthly accumulated sequential hydric balance, ${{\text{H}}_{\text{2}}}{{\text{O}}_d}$: monthly average available water.

Average HT ranged between 8.02 m (CUIA) and 10.69 m (PF), while DBH ranged between 0.10 m (CUIA) and 0.13 m (PF), with CVs between 21.99% (ALT-PF) and 34.77% (DAP-GO), that indicated variability (table 4). When we standardized these results with Z statistic, values varied between 2.5629 (HT-CUIA) and 2.6887 (HT-GO) and 2.6347 (DBH-CUIA) and 2.7981 (DBH-GO), with CVs between 37.26% (HT-GO) and 38.73% (HT-CUIA) and 35.05% (DBH-GO) and 37.64% (DAP-CUIA) (table 4).

These characters were collected due to their easy field measurement, their economic importance and their significance related to other characters measured at field. For correlation between silvicultural and climatic characters, with data already standardized, there was a significant correlation in CUIA between HT-${\text{Et}}{{\text{o}}_m}$ and DBH-${\text{Et}}{{\text{o}}_m}$ (−0,870, 1%), HT-${\text{Et}}{{\text{o}}_a}$ and DBH-${\text{Et}}{{\text{o}}_a}$ (0.884, 1%); in GO, between HT-${T_m}\,$ and DBH-${T_m}$ (0.535, 5%), HT-${\text{Et}}{{\text{o}}_m}$ and DBH-${\text{Et}}{{\text{o}}_m}$ (−0.666, 1%), HT-${\text{Et}}{{\text{o}}_a}$and DBH-${\text{Et}}{{\text{o}}_a}$ (0.696, 1%); in PF, HT-$Et{o_a}\,$and DBH-${\text{Et}}{{\text{o}}_a}$ (0.637, 1%) (figure 6). The higher the evapotranspiration mean is, the lower the individuals' growth will be. In practice, considering that A. urundeuva is a long-living species, its growth time would increase even more, which would impact into several physiological processes.

Zoom In Zoom Out Reset image size

Among climatic characters, a significant negative correlation between ${\text{Et}}{{\text{o}}_m}$-${{\text{H}}_{\text{2}}}{{\text{O}}_d}$ (CUIA = −0.692, 5%; GO = −0.590, 5%; PF = −0.617, 1%) and a significant positive correlation between ${\text{Et}}{{\text{o}}_a}$-${{\text{H}}_{\text{2}}}{{\text{O}}_d}$ (CUIA = 0.674. 5%; GO = 0.576, 5%; PF = 0.612, 1%) was found. This highlights the environmental role of water availability, which is more important than temperature, for example. For collinearity assessment, it is suggested to apply mean character values. Therefore, ${P_a}$, ${{\text{H}}_{\text{2}}}{{\text{O}}_d}$ and ${\text{Et}}{{\text{o}}_a}$, as accumulated, were discarded before testing. There was general collinearity between climatic and silvicultural characters.

Therefore, average characters (${T_m}$, ${P_m}\,$ and ${\text{Et}}{{\text{o}}_m}$), were tested individually. In CUIA and PF populations, ${T_m}\,$and ${P_{m\,}}$, were discarded due to collinearity, while in GO only ${P_m}$ was discarded. That is, despite ${T_m}\,$ influences silvicultural characters in GO, there is a higher ${\text{Et}}{{\text{o}}_m}$influence on silvicultural characters in populations.

The stepwise analysis (table 5) showed the most important character is ${\text{Et}}{{\text{o}}_m}$, as the only one which appeared in the top of the three models. However, in CUIA and GO, ${T_m}\,$could also be considered important.

Table 5. Models analysis after discarding collinear characters, for Astronium urundeuva, applying Akaike criterion (AIC).

PT	Character	Model	AIC
CUIA	HT	HT ∼ $\boldsymbol{T_m}$ + $\boldsymbol{P_m}$ + $\boldsymbol{Et{o_m}}$	−13.4
		HT ∼ ${{{\boldsymbol{T}}}_{{\boldsymbol{m}}}}$ + ${{\boldsymbol{Et}}}{{{\boldsymbol{o}}}_{{\boldsymbol{m}}}}$	−15.4
	DBH	DBH ∼ $\boldsymbol{T_m}$ + $\boldsymbol{P_m}$ + $\boldsymbol{Et{o_m}}$	−12.91
		DBH ∼ ${{{\boldsymbol{T}}}_{{\boldsymbol{m}}}}$ + ${{\boldsymbol{Et}}}{{{\boldsymbol{o}}}_{{\boldsymbol{m}}}}$	−14.9
GO	HT	HT ∼ $\boldsymbol{T_m}$ + $\boldsymbol{P_m}$ + $\boldsymbol{Et{o_m}}$	−12.87
		HT ∼ ${{{\boldsymbol{T}}}_{{\boldsymbol{m}}}}$ + ${{\boldsymbol{Et}}}{{{\boldsymbol{o}}}_{{\boldsymbol{m}}}}$	−14.17
	DBH	DBH ∼ $\boldsymbol{T_m}$ + $\boldsymbol{P_m}$ + $\boldsymbol{Et{o_m}}$	−11.67
		DBH ∼ ${{{\boldsymbol{T}}}_{{\boldsymbol{m}}}}$ + ${{\boldsymbol{Et}}}{{{\boldsymbol{o}}}_{{\boldsymbol{m}}}}$	−12.98
PF	HT	HT ∼ $\boldsymbol{T_m}$ + $\boldsymbol{P_m}$ + $\boldsymbol{Et{o_m}}$	−18.51
		HT ∼ $\boldsymbol{T_m}$ + $\boldsymbol{Et{o_m}}$	−20.38
		HT ∼ ${{\boldsymbol{Et}}}{{{\boldsymbol{o}}}_{{\boldsymbol{m}}}}$	−20.78
	DBH	DBH ∼ $\boldsymbol{T_m}$ + $\boldsymbol{P_m}$ + $\boldsymbol{Et{o_m}}$	−17.63
		DBH ∼ $\boldsymbol{T_m}$ + $\boldsymbol{Et{o_m}}$	−19.5
		DBH ∼ ${{\boldsymbol{Et}}}{{{\boldsymbol{o}}}_{{\boldsymbol{m}}}}$	−19.89

PT = progeny test, AIC = Akaike's informative criterion.Data in bold show best model for each character, within each progeny test and lowest AIC value.

The multiple linear regression analysis was performed considering models with the lowest AIC, for each silvicultural character (table 6). Results showed $Et{o_m}\,$ was the significant climatic character for the three tests.

Table 6. Multiple linear regression for silvicultural and climatic data in three Astronium urundeuva tests installed in Selvíria-MS, for 2019.

PT	Model	Coef	P	${{{R}}}_{{{m}}}^2$	${{{R}}}_{{{{aj}}}}^2$
CUIA	HT ∼ ${\text{Et}}{{\text{o}}_m}$+ ${T_m}$	${{{Et}}}{{{{o}}}_{{{m}}}}$	0.00168 **	0.8644	0.8192
		${T_m}$	0.07126
	DBH ∼ $\,Et{o_m}$+ ${T_m}$	${{{Et}}}{{{{o}}}_{{{m}}}}$	0.00168 **	0.8644	0.8193
		${T_m}$	0.07124
GO	HT ∼ $Et{o_m}$+ ${T_m}$	${{{Et}}}{{{{o}}}_{{{m}}}}$	0.00683 **	0.6215	0.5584
		${T_m}$	0.03520 *	—	—
	DBH ∼ $Et{o_m}$+ ${T_m}$	${{{Et}}}{{{{o}}}_{{{m}}}}$	0.00683 **	0.6215	0.5584
		${T_m}$	0.03520 *	—	—
PF	HT ∼ $Et{o_m}$	${\text{Et}}{{\text{o}}_m}$	0.000299 ***	0.4879	0.4623
	DBH ∼ $Et{o_m}$	${{{Et}}}{{{{o}}}_{{{m}}}}$	0.000299 ***	0.4879	0.4623

PT = progeny test, coef = character coefficient evaluated, p = significance test, $R_m^2$ = multiple r², $R_{aj}^2$ = adjusted r².Data in bold are equivalent to highlighted characters.*** significant at 0.1%, ** significant at 1%, * significant at 5%.

Climatological data found in this research agrees with Damião et al (2010), Schutze et al (2013), Silva et al (2017) for the region. In Eucalyptus spp., temperature mainly influences plant metabolism regulation, favoring cell division (Cunha et al 2009). Studying forest species from Mixed Ombrophilous Forest, Kanieski et al (2012) estimated positive and significant correlations for average temperature, related to height and DBH, with temperature being considered one of the main evaluation factors (Cardoso 1991).

When there is a correlation of any character with temperature, means that this species is more sensitive to temperature changes, linked to its growth. Other factors may also be related, such as phenological behavior, for example, reflecting on cambium growth (Zanon and Finger 2010, Kanieski et al 2012). Maximum day temperature and rainfall may be responsible for influencing diameter growth (Schippers et al 2015). These are characters probably linked to increased respiration, decreased photosynthesis and stomatal conductance, in response to increased water demand. Adaptation and acclimatization are strategies adopted to deal with climate change, based on genotypic and phenotypic adaptations, so that plants can grow and reproduce, despite stress. Therefore, it is important to select the most suitable genotypes and provenances, adopt appropriate techniques in nurseries and monitor plantations (Higa and Pellegrino 2015).

Pereira et al (2011) reported region rainfall distribution is poor and concentrated (80% in summer), with high evapotranspiration rates, summery subjected (Hernandez et al 2003).

In agriculture, water balance information, whether descriptive or sequential, is important, as they demonstrate an interested area water regime: soil preparation appropriate time, sowing, planting, implementing drainage systems feasibility or irrigation and water shortage times definition (Schutze et al 2013). Due to its large extension, Brazil is divided into three climatic areas (Binkley et al 2017): these categories are mainly related to air and soil water amount, although there is greater evapotranspiration related to precipitation, which causes species to seek water in deeper places (Huang et al 2009, Rocha et al 2020). This information becomes important for this research, since its purpose evaluates three progenies tests behavior, from different origins (and characteristics) at the same biome, in another region with an environment relatively similar to the origins. Average water deficit is 173 mm yr⁻¹ in Selvíria-MS, which is a critical character for forest productivity (Golfari 1978), since there is a direct relationship between water availability and forest productivity, given its direct and indirect effects on tree growth, such as photosynthetic rate decreases and stomatal resistance increases (Vose and Swank 1994), a tree input decreases (by mass flow and diffusion), organic material mineralization rate decreases, even cellular level collapse (Sands and Mulligan 1990). In warmer, drier climates, plants are subject to drought conditions, as a consequence of low rainfall and high temperature, which causes high water deficit levels (Yuan et al 2019, Firrincieli et al 2020).

Meanwhile, soil water availability is mainly influenced by temperature and rainfall spatiotemporal distribution (Souza et al 2004). In dry seasons, plantations undergo seasonal changes throughout the year, as water becomes a limiting factor (Almeida et al 2007). In Eucalyptus grandis W.Hill ex Maiden (Myrtaceae) plantations aged seven, stomatal control was efficient under low soil water availability conditions (Almeida and Soares 2003). In a typical Cerrado species, 'pequi' (Caryocar brasiliense Cambess., Caryocaraceae), growth remained slow, even in dry periods and with soil water conditions limiting most species, due to drought tolerance mechanisms (Alves-Júnior et al 2013).

Doorenbos and Kassam (1994) reported there is a direct relationship between evapotranspiration and growth. Plants have mechanisms and processes that allow them to deal with periodic and severe water deficit, with growth losses, in order to enable their survival (Gonçalves et al 2017). Cerrado species tend to reallocate carbohydrates to enable rapid growth of the root system and reserve organ development, to ensure individuals survive in dry seasons (Alves-Junior et al 2013).

Several plant physiological processes are affected by water deficit. As growth is controlled by cell division and expansion, with less water available to keep flabby growth zone cells (considered to be a cells normal state), vegetative growth declines (Taiz and Zeiger 2004). The fact there are three progenies tests, from different origins/regions, with different anthropization levels, harvesting sampling and tests installation can also alter the results . Other reasons for changes in forest productivity rates are linked to the progenies groups and area identification (Scolforo et al 2019), and genotype effect, genotype and environment interaction (Binkley 2018), but also as a management system function: genetic selection, silvicultural treatments (area preparation, fertilization, spacing, weed competition and pests and diseases protection) (Binkley et al 2020). Dominance between individuals can also be an influential characteristic in plantations involving genetic diversity or between species, when presenting partitioned niches, reducing competition or promoting facilitation (Forrester 2014), a different pattern from monoclonal stands. Furthermore, spacing water availability is related, since with densification, available soil water uptake is higher, in order to maintain the plants physiological needs (Forrester 2015, Hakamada et al 2020), differing from natural populations (Whitehead et al 1984).

The main factors which alter a trees development are related to increased CO₂ concentration, increased temperature, above-normal precipitation and drought (Chmura et al 2011), Cunha et al (2009), for Eucalyptus sp., we found low correlations between ministumps clonal productivity and climatic characters (light, temperature and humidity). Drought effect interferes, mainly in stomata opening regulation, which directly affects photosynthesis (Urban et al 2017). Environmental characters such as radiation, precipitation, temperature and humidity have direct effects on stomata opening and closing, regulating the photosynthesis CO₂ (Firrincieli et al 2020).

Since collinearity is defined by linear relationship between two characters, multicollinearity occurs when two or more independent variables are strongly linearly related, which makes it impossible to obtain parameter coefficients, their estimates become insignificant, in other words, when an independent variable changes, the correlated one changes too (Maia 2017). In forests, evapotranspiration is influenced by several factors: air humidity, atmospheric conditions, temperature, rainfall and soil water availability (Whitehead and Beadle 2004). Characters related to water issues, especially evapotranspiration and water availability (or deficit), are usually in evidence in climate tests, because indirectly other climate characters integrated them, but, in this research, they were not proved to be more accurate than using characters together, as in Binkley et al (2020).

The most common reports indicated responses to increased productivity are related to increased temperature, which did not occur here, and that is not always the case with forest species, as their adaptations have limits (Binkley et al 2020). In more arid regions, plants often access groundwater reserves (Yin et al 2015). This environmental stress not only reduces individual growth, but influences the entire stand structure, which generates (more) heterogeneity, at the A. urundeuva tests, which dominated the trees higher proportion, and thus, reduced overall growth (Soares et al 2020).

A. urundeuva leaves loss, in the dry season, observed in the field, may be related to a lack of soil available water. Standing in drier areas does not mean a plant is more efficient at using available soil water (Hubbard et al 2020). Soil water use is directly linked to transpiration: in dry environments, plants tend less to transpiration. Therefore, above-ground growth ends up lower in these areas, as more drought-resistant species tend to allocate more resources below ground line, according to demand, with different water shortage resistance mechanisms: leaf loss, stomatal conductance regulation, roots resource allocation (Campoe et al 2020, Hubbard et al 2020).

Water deficit is the main risk factor for forest production in Brazil, mainly due to the fact that most of the forest plantation areas are located in water restriction regions (Assad et al 2021). It is important to remember that, in the Cerrado, where three of the largest hydrographic basin sources in South America and three large aquifers are located, is a region with a water reserve and biodiversity conservation strategic role (Assad et al 2020).

Furthermore, changes in the Cerrado climate will accentuate genetic diversity and effective population size losses (Colevatti et al 2013), genetic and morphological divergence and environmental isolation (Ribeiro et al 2016). Diversity losses related to climate change happen with occurrence areas reduction (Barbosa et al 2019). These changes can even generate population homogenization (Braga 2019).

Important biological processes may be related to adaptation and, consequently, to species phenotypic plasticity (Scheiner and Callahan 1999), which generates variability and phenotypes selection acts (Cardoso and Lomônaco 2003). A character can only be considered adaptive when it results from a natural selection process, in order to perform a particular biological function, by promoting a reproductive success increase, compared to other individuals (Gould and Vrba 1982). Adaptations in morpho-anatomical characteristics can guarantee species survival and diversification in long dry and fire exposed areas (Rando and Pirani 2011).

As a new science, there is little literature on the impact of climate change on planted forests (Higa and Pellegrino 2013). In Brazil, some of the changes are the result of natural interannual and interdecadal variability , sometimes confused with climate change effects. Historical temperature series detected continuous increase trends over the years, in country regions north, east and south (Higa and Pellegrino 2013, Batista et al 2014). In the coming decades, the Brazilian climate is expected to be warmer, with a gradual and variable increase in average temperature and a decrease in rainfall in central, north and northeast regions, with the opposite in south and southeast (Higa and Pellegrino 2013).

Current forest use and conservation policies are based on how forests were developed under past climatic conditions. Today, it is necessary to accept that climate change is happening and create strategies in order to reduce possible vulnerabilities (Spittlehouse 2005). Since droughts drastically influence productivity, climate information can help to recommend appropriate locations for planting (Scolforo et al 2016, 2017, Marcatti et al 2017).

Unlike most experiments in forests, which usually show temperature most influences growth, in this research, it was evapotranspiration and, indirectly, the amount of water available in the environment, although indirectly linked to the effects of evapotranspiration . The higher the evapotranspiration rates, the lower the individuals growth, in the three progenies tests. The species will be able to survive climatic fluctuations, although this will affect its growth through time. Therefore, strategies to increase productivity, considering progenies more resistant to drought, but not necessarily fast-growing, can be an alternative to keep a species genetic diversity along with climate and temporal changes apace in our country.

To the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the scholarship granted (Postgraduate Program in Planning and Use of Renewable Resources-2018/2020/Postgraduate Support Program), to technician Alonso Angelo for field support. To Professor. Daniela Canuto (UNESP), Professor Alan Panosso (UNESP), Professor Marcela Moraes (UNESP), Professor Celso Marino (UNESP) and Professor Fernando Braz Tangerino Hernandez (UNESP) for their contributions on the subject discussed.

All data that support the findings of this study are included within the article (and any supplementary information files).

To the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) for the scholarship granted (Postgraduate Program in Planning and Use of Renewable Resources-2018/2020/Postgraduate Support Program)

There is no conflict of interest in this research.