Impacts of climatic changes on the vegetative development of olive cultivars

: The aim of this study was to simulate the vegetative development of the olive cultivars Arbequina and MGS ASC315 cultivated in Maria da Fé, MG, Brazil (22° 18’ 29” S, 45° 22’ 31” W, 1.276 m of altitude). The development of these cultivars was simulated in nine transplanting dates, considering three distinct scenarios for three periods (2011-2040, 2041-2070 and 2071-2100). Climate changes influence the olive crop management. The increase in air temperature modifies the development rate and the duration of the seedling phase of the cultivars Arbequina and MGSASC315. In colder months, there is accelerated vegetative development and shorter seedling phase duration. Conversely, in warmer months, there is delayed development with longer seedling phase duration.


Introduction
The Intergovernmental Panel on Climate Change (IPCC, 2013) has projected increases in atmospheric CO 2 and air temperature up to 5.0 °C in the state of Minas Gerais, Brazil, in the 21 st century (Santos et al., 2017b). In theory, C3 plants, which represent most terrestrial species, should be favored by the increase of CO 2 since they are highly dependent on atmospheric CO 2 to have the enzyme Rubisco, which is solely responsible for the metabolic process that governs photosynthesis in C3 plants (Fagundes et al., 2010;Taiz et al., 2018). Conversely, the Warburg effect occurs under high temperature conditions, in which O 2 inhibits the fixation of CO 2 by the Rubisco enzyme, thus reducing photosynthesis and increasing photorespiration (Marenco & Lopes, 2009). This impact hinders development because it shortens the cycle duration and reduces yield, due to the reduction of photosynthesis and the accumulation of photoassimilates (Fagundes et al., 2010). Thus, studies examining the potential impacts of climatic changes on the development of annual and perennial species are important in scenarios with promising economic potential (IPCC, 2013;Walter et al., 2014), such as olive trees that are cultivated in the southern and southeastern regions of Brazil Wrege et al., 2015;Santos et al., 2017b).
The impact of climate change on olive crops can be analyzed through the development of simulation models which are commonly used for annual crops (Streck et al., 2011) and scarce for perennial crops (Costa & Streck, 2018). These models simulate the development through leaf appearance rate (LAR), using air temperature response functions, differing basically between linear or additive and non-linear or multiplicative categories . The aim of the present study was to simulate the vegetative development of the cultivars Arbequina and MGS ASC315 in nine transplanting dates, considering different climate projections.

Material and Methods
Due to the expansion of olive crops cultivation in Minas Gerais, Brazil (Garcia et al., 2018), a numerical study was performed in Maria da Fé, MG, Brazil (22° 18' 29" S, 45° 22' 31" W, 1.276 m of altitude). According to Köppen classification, the region presents Cwb climate, characterized by dry winters and rainy summers (Souza & Martins, 2014). The cultivars Arbequina and MGS ASC315 were chosen because both have been used with the objective of producing olive oil in the region .
The vegetative development models (VD) used in this study were differentiated for each cultivar, with the Wang and Engel model (WE) for Arbequina and the Phyllochron model (PHYL) for MGS ASC315, based on the results obtained by Martins et al. (2014). In both models, VD is given by the number of accumulated leaves in the main stem (NL) which integrated in time provides the duration of the seedling phase (DSS), which refers to the number of days between the transplanting date and the threshold of NL = 20 .
Tm data from three General Circulation Models (GCMs) -MIROC5 (M1), IPSL-CM5 (M2) and MRI-CGC (M3) (MCGs) members of the Coupled Models Intercomparison Project Phase 5 (CMIP5) were used in the air temperature functions of the VD models (f(T) of the WE and STac models in the PHYL model). GCMs were chosen due to the unavailability of up-to-date results from regional models, which are highly computationally costly. Besides, the new generation of GCMs (1) represents the state of the art in climate system modeling (Santos et al., 2017b).
Future climate simulations of CMIP5 use the new generation of radiative forcing scenarios called Representative Concentration Pathways (RCPs) -2.6, 4.5 and 8.5, which correspond to approximate radiative forcing of 2.6, 4.5 and 8.5 W m -2 , associated with CO 2 equivalent concentrations of 490, 650 and 1.370 ppm, respectively at the end of the 21 st century (Moss et al., 2010;Vuuren et al., 2011).

Results and Discussion
The air temperature projections for each RCP demonstrate a gradual increase in mean, maximum and minimum air temperatures throughout the 21 st century, especially for RCP 8.5. This is expected because it presents the greatest radiative forcing (8.5 W m -2 ) and concentration of emitted CO 2 (1.370 ppm)   (Figure 1).
This increase can reach 1.5 °C in the short term (PR1), 3.0 °C in the medium term (PR2) and up to 4.0 °C in the long term (PR3) for Maria da Fé, MG, Brazil (Table 2). This corroborates the values recorded by Santos et al. (2017b) for the state of Minas Gerais, Brazil.
Such increases in air temperature ( Figure 1) will result in direct damage to the physiological processes of plants, influencing the duration of the vegetative and reproductive stages in olive crops (Tanasijevic et al., 2014) and, consequently, in the DSS. This occurs because the reactions of photosynthesis and photorespiration, enzymatic activities, rate of transport and translocation of solutes (Freitas et al., 2017) are intensified or reduced based on the temperature (Taiz et al., 2018).
It was necessary to make Tm corrections in the three GCMs. Therefore, the methodology proposed by Lenderink et al. (2007) was used. It is based on the monthly bias between the climatic means (NP) obtained with the meteorological station data and the climatic means of the historical period of the GCMs, both for 1971 to 2000. Subsequently, the bias was applied to the daily values of Tm for the month in question (January: 3.4; February: 2.8; March: 3.1; April: 3.2; May: 4.7; Due to their values, the axes of the air temperature variables have different scales and TT3 (20/07) in both cultivars. Except for TT1, the increase in DSS occurred in simulations performed in periods with the highest air temperatures (Table 2). On the contrary, the reduction of DSS (about 20 days for Arbequina and 50 for   Table 1 MGS ASC315) occurred in periods with the lowest values of air temperature, especially in TT3 and in RCP 8.5 (Figure 2).
The increase in DSS occurred mainly in the spring and summer (MGS ASC 315) (TT6, TT7, TT8 and TT9) (Figure 2), in which the highest values of Tm were projected; this is similar to what was observed for the VD of red rice (Streck et al., 2013) and potato (Streck et al., 2006). This occurs because at higher temperatures, especially above Tot (Arbequina = 16.1 °C and MGS ASC315 = 14.7 °C) (Lisboa et al., 2012), there is a reduction in the development rate, thus increasing the DSS. Reduced development occurs due to higher breath maintenance rate, energy-consuming process (ATP and NADPH) or carbon skeletons from photosynthesis (Taiz et al., 2018), which raise maintenance costs by increasing protein turnover and maintaining ionic balance (Marenco & Lopes, 2009;Ruelland & Zachowski, 2010).
Significant effect of triple interaction was verified among transplanting dates, radiative forcing and projections for both cultivars (p ≤ 0.05) on the DSS. Among the main effects, the highest values of the f-test were for the transplanting date  (Table 3), similar to what was observed by Streck et al. (2013) in different rice genotypes.
There is a tendency of increase in DSS, except for TT2 and TT5 for Arbequina and TT2 and TT3 for MGSASC315, intensified toward PR3, especially in RCP 8.5 in both cultivars. More specifically, TT7 and TT8 of the Arbequina cultivar presented higher DSS in the three radiative forcing and in the three periods, around 86 days (± 0.311) for PR1 and RCP 2.6 and 97 days (± 0.919) for PR3 in RCP 8.5. Conversely, TT3 and TT4 will have the lowest DSS in radiative forcing in the short and medium term (PR1 and PR2, respectively). In addition, long-term TT1 and TT2 (PR3), especially in RCP 8.5, simulated in May and June, will cause a reduction in DSS. In the cultivar MGS ASC 315 in TT2 there was also a reduction of DSS (around 47 days) in the three RCPs throughout the projections. From TT5 to TT9, especially those simulated in September (TT5), October (TT6) and November (TT7), presented an increase in DSS throughout the projections. For both olive cultivars, the VD through DSS was shortened or elongated depending on the simulated transplanting date, like that observed by Streck et al. (2006) with potato cv. Asterix, Streck et al. (2012) with corn, Liu & Tao (2013) with wheat, and Streck et al. (2013) with different rice genotypes.
Late growing seasons in autumn and early winter will increase VD and reduce DSS; and those cultivated in spring and summer will delay VD and increase DSS. This is mainly due to the fact that Tm is higher than the Tot of the cultivars equal to 14.7 ºC for MGSASC315 and 16.1 ºC for Arbequina (Lisboa et al., 2012) and close to a cardinal temperature (TB) of 37.8 ºC (Denney et al., 1985). The development rate increased when Tm was between Tb and Tot but decreased when it moved away from Tot and approached TB (Streck et al., 2006;Ruelland & Zachowski, 2010;Martins et al., 2014). This occurs because in olive crops, Rubisco is the sole enzyme controlling photosynthesis in the metabolic process (Taiz et al., 2018). In conditions with high Tm, the Warburg effect occurs and O 2 inhibits CO 2 fixation by Rubisco, thereby reducing photosynthesis, increasing photorespiration (Marenco & Lopes, 2009) and DSS. In practice, it is necessary to reduce the negative impact of increasing Tm in warmer seasons in nurseries where olive seedlings are cultivated. For this, two options can be used namely: shading screens and sprinkler irrigation (Costa & Streck, 2018).
Considering climate change projections, studies by García-Mozo et al. (2010), Tanasijevic et al. (2014) and Moriondo   Table 3. Comparison of means for the duration of seedling phase (DSS, days) for both olive cultivars (Arbequina and MGS ASC315) considering three radiative forcing (RCP 2.6, RCP 4.5 and RCP 8.5) and three projections (PR1 = 2011-2040, PR2 = 2041-2070, PR3 = 2071-2100 (García-Mozo et al., 2010;Tanasijevic et al., 2014), while during pollination and flowering affects stigma receptivity, ovule longevity and pollen tube growth, impairing fertilization and fruit formation (Garcia et al., 2018). Concurrently, the duration of flowering and fruiting is shortened (Tanasijevic et al., 2014), indicating that olive crops are vulnerable to climatic changes. In addition, the increase in temperature will reduce the accumulation of chilling hours, which is required by the olive tree to interrupt dormancy, start differentiation of flower buds and have normal flowering (Santos et al., 2017a;Garcia et al., 2018). The insufficiency of chilling accumulation causes a series of physiological anomalies, generating irregularities in sprouting and flowering, which adversely affect the development (Santos et al., 2017a) of the olive tree.
Thus, both the vegetative development, represented by the DSS, and the reproductive development of the olive tree are vulnerable to the projected climatic changes throughout the 21 st century. This corroborates the results presented by Santos et al. (2017b) for climatic zoning in MG, Brazil. Such information will affect the choice of the best dates for transplanting and the techniques for seedling production, development of the seedlings in projections of climatic changes, affecting the olive tree management and the success of seedlings establishment in the field.

Conclusions
1. The increase in air temperature, reaching 3 ºC at the end of the century, will probably modify the development rate and duration of the seedling phase in the cultivars Arbequina and MGSASC315 cultivated in Maria da Fé, MG, Brazil.
2. In colder months, there is increased vegetative development in olive crops and reduced seedling phase duration. Conversely, in warmer months, there is delayed development and longer seedling phase duration.