GROWTH AND YIELD OF EVEN-AGED Cupressus lusitanica PLANTATIONS IN SOUTHERN BRAZIL

Plantation forestry in southern Brazil demands additional timber species to a higher market differentiation by providing high quality timber and exploitation of market niches. Cupressus lusitanica has long been recognized for this purpose but, until now, it was not properly region-wide quantified in terms of growth and yield. The present study delivers the lacking quantitative approach, which may encourage the commercial use of the species. With this study it was aimed at collecting and processing quantitative data from all known C. lusitanica stands in southern Brazil. Inventories were carried out (60 ha, 6-39 years of age) in order to model the development of dominant height (h100), basal area, volume and dominant diameter (d100). Dominant height was the basis for site quality evaluation, delivering site index curves, which, together with the commercial volume of the stands, allowed yield modelling. A wide amplitude of dominant height growth was detected (10-30 m at 20 years), indicating a great site quality variation. At age of 20 years, commercial volumes of 110 and 620 m3 ha were observed for site indexes of 14 and 26, respectively, equivalent to a maximum of 6-31 m3 ha year at ages between 16-18 years. Results demonstrated in a robust manner that C. lusitanica has a high potential for cultivation in southern Brazil. Thus, offering the opportunity of market differentiation by promoting market niches whose demands timber for special solid end-uses.

In southern Brazil, C. lusitanica has long been indicated for cultivation in the highlands (SHIMIZU et al., 1995). This region has an established planted-forest-based sector, mainly supplied by Pinus taeda L (1,5 million ha) (EISFELD; NASCIMENTO, 2015;ACR, 2019;IBÁ, 2019). The predominance of P. taeda is due to growth rates (34 m 3 ha -1 year -1 at age of ~15 years in average), adaptability to the edaphoclimatic conditions and versatility to a wide range of end-uses (ACR, 2019). Although P. taeda is responsible for a successful forest sector, there is an unquestionable need for increasing diversity. Diversity in this context means additional timber species to a higher market differentiation by providing high quality timber with complementary wood properties and thus, allowing exploitation of market niches. Market differentiation might be also faced as risks reduction imposed by a largely based single-specie sector.
According to Dobner Jr. et al. (2018), the species is not commercially used in southern Brazil due to lack of reliable genetic material. The authors reached this conclusion after visiting several stands where a great variability of growth and quality was observed. Additionally, it is hypothesized that a further reason for not commercially regarding the species is the absence of robust quantitative information of growth and yield. Indeed, growth and yield knowledge about C. lusitanica grown in southern Brasil was restricted to a few small experimental plots. With this study it was aimed at collecting and processing quantitative data from all known C. lusitanica stands in southern Brazil and thus, providing a robust quantitative approach within the existing variability. Site index analysis as well as yield modelling of commercial stands are presented.

Study site
The study was conducted in the highlands of southern Brazil (800-1,000 m a.s.l), where the climate is humid, subtropical, oceanic, without dry season and with temperate summer (Cfb, after Köppen classification). The region has an average annual precipitation of 1,600-1,900 mm and a mean annual temperature of 12-16 °C (ALVARES et al., 2013). Severe frosts occur.
Data was collected in seven different municipalities, across the highland of the Santa Catarina state (Tab. 1 and Fig. 1). Extreme sampling places were 250 km distant from each another. In total ~60 ha of C. lusitanica stands were considered in this study, with ages varying between 6-39 years. Stands were established with 1,600-2,000 trees ha -1 , reaching different densities at sampling ages due to natural mortality or light thinnings from below (Tab. 1).

Data collection and analysis
Inventories were carried out in 2014, 2016 and 2019, thus plots (500 m²) were measured up to three times. Some of them, however, had two or only one measurement since stands were more recently discovered. Measured variables were: (1) diameter at breast height over bark (dbh, 1,3 m), measured with a diametric tape to the closest millimeter; (2) tree height (h), randomly measured over the diametric range, including the dominant trees (the 100 trees with the biggest dbh per ha) and the smallest one, at least 10 trees per plot, with the hypsometer Vertex IV (Haglöf, Sweden); (3) diameter (di) at different heights (hi = 0.1, 0.5, 1.0, 1.5 and 2.0 m and, from this height onwards in 2 m intervals) with caliper to the closest millimeter (scaling). The scaling measurements were conducted in three stands with ages of 8 (34 trees, dbh 5-21 cm), 12 (26 trees, dbh 7-33 cm) and 16 years (28 trees, dbh 21-50 cm) in the regions of CBS and SJC (Tab. 1, Fig. 1). Unfortunately, it was not possible to proceed destructive scaling in older stands. Inventory and scaling data were used to fit hypsometric models, per stand and age, and taper models (Kozak and the 5 th degree polynomial), for the scaled ages, both by linear regression techniques. Additionally, the artificial form factor (f1.3) was obtained dividing the tree volume by the volume of a hypothetical cylinder formed by the dbh and h. For commercial volume estimations, the fitted taper equation with age closer to the stand was regarded.
Dominant height (h100) and dominant diameter (d100) development were assessed by using a time series of paired values to fit the three parametric Richards growth model (Eq. 1), with non-linear least square technique, where a, b and c are the model parameters and A is age: (1) In both cases, models were fit based on the dataset with ages varying between 6-39 years. Dominant height was the base for site quality evaluation, delivering site index curves with help of the guide curve method. Dominant diameter development was than fitted regarding only average site classes. Statistical analysis was performed using RStudio version 1.1.383 (R Core Team, 2017). Statistical significance was assessed at α = 0.05.
In order to quantify log assortment volumes, logs with 2.5 m in length were classified based on their small-end diameter (SED) over bark in 10-cm-wide classes. This quantification was carried out with help of Florexel (Optimber®). Then, the total commercial volume (SED ≥ 10 cm) of unthinned stands was employed for growth and yield modelling with help of the compatible model of Sullivan and Clutter (1972): (2) ln 2 = 0 + 1 + 2 2 −1 + 3 ln 2 A 2 = projection age in years B1 = initial basal area in m² ha -1 B2 = the predicted basal area at age A2 in m² ha -1 S = site index V2 = predicted volume at age A2 in m³ ha -1 αis and is are the model parameters

Taper equations
The taper equations fitted for C. lusitanica with 8, 12 and 16 years of age grown in southern Brazil are given as follow. The best fit was obtained for the 5 th degree polynomial model for all ages, delivering the best coefficients of determination (R 2 ), 0.95, 0.96, 0.94, respectively.
where: dbh = diameter at breast height in cm di = diameter in cm measured at height hi hi= height along bole in m ht = total height in m These are probably the first taper models for C. lusitanica grown in Brazil, since no previous studies with this approach could be found. A graphic representation of the taper equations is presented in Fig. 2, where the expected trend of decreasing tapering with age can be observed.

Site index
Dominant height (h100) paired with age values was the base of the site quality evaluation. In total, 151 pairs were employed to fit the guide curve shown as follow (R² = 0.73, Syx = 2.8 m or 18.3%) and in Figure 3  From Figure 3-a it can be observed the wide amplitude of dominant height growth of C. lusitanica in southern Brazil (10-30 m), which indicate a great variation in site quality. Values of h100 >25 m were observed at two locations, CBS and RN (Tab. 1 and Fig. 1).
Fitted site index curves present three site classes named 14, 20 and 26, the dominant height at age of 20 years, respectively, used as reference for further analyses.

Basal area growth and volume yield
With help of the site index previously shown, a yield model was fitted, firstly to predict future basal area as a function of site quality, age and initial basal area (R² = 0.96, Syx = 1.8 m² ha -1 or 7.7%). Secondly, with the predicted basal area, the predicted commercial volume (SED ≥ 10 cm) was also estimated (R² = 0.96, Syx = 21.4 m³ ha -1 or 12.1 %). Fitted equations are given as follow. Predicted compared to observed values, as an indicator of fitting quality, are shown in Fig. 4.  The fitted yield models can then be used for a better understanding of the real yield potential of C. lusitanica in southern Brazil. The average basal area growth and commercial volume yield for different site indexes are given in Fig. 5. The great site quality differences already verified in terms of dominant height is even more evident when basal area and volume development are regarded. At age of 20 years, a C. lusitanica stand grown at SI=14 will reach a basal area of only 21 m² ha -1 , while one at SI=26 could deliver 57 m² ha -1 . Commercial volume yield curves reinforce this since differences between site indexes are even more disproportionate. A stand at SI=14 delivered only 110 m³ ha -1 at age of 20 years, while SI=20 delivered 270 m 3 ha -1 and SI=26 as much as 620 m³ ha -1 , additional yield of 160 and 350 m³ ha -1 , respectively. In terms of mean annual increment, these three site indexes were able to deliver a maximum of 5.6, 13.8 and 31.3 m³ ha -1 year -1 at ages between 16-18 years. Current annual increment in volume peaked for all site indexes at age of 9 years, with values varying between 8-46 m³ ha -1 year -1 , showing the pioneer behavior of the species.

Dominant diameter growth
For the dominant-diameter-growth modelling, only data from plot classified as site indexes between 18-22 were regarded and thus, representing an average site quality (SI=20, n=318). The fitted equation is given below (R² = 0.78, Syx = 4.4 m or 15.9%). All parameters were obtained with high significance (p<0.001). Data and fitted curve are given in Fig. 6.   Fig. 6 it can be observed that, considering the current genetic basis and management standard, 30-35years of production cycle will deliver, in average site conditions, dominant diameters of ~50 cm. However, higher growth levels as a result of genetic breeding and better management practices could deliver the same dominant diameter in production cycles of ~25 years.

Taper
Taper models were successfully fitted, which, in terms of artificial form factor (f1.3) indicated and increasing tapering with age, i.e. decreasing f1.3 values with increasing age. Noteworthy is that the taper analysis regarded relatively young trees, up to 16 years of age. It is expected that f1.3 would increase if older trees were regarded. A similar pattern was observed for Araucaria angustifolia by Martins et al. (2017), when young trees showed a higher f1.3 because of the proximity of dbh to total height which, for C. lusitanica was particularly important because of the strong tapering at tree base, as observed in Fig. 2.

Site index
The range of dominant height across southern Brazil showed that height growth, i.e. site quality, varied widely between stands (10-30 m at age 20 years). The best sites were observed both at the southernmost and northernmost regions, CBS and RN (Tab. 1, Fig. 1), indicating that the whole studied region has favorable edaphoclimatic conditions for cultivating C. lusitanica and that microsite variations together with genetic materials are responsible for the great amplitude of growth potential. Both stands with higher h100 were established with seeds from Colombia, indicating that this is a provenance of high growth potential in southern Brazil, as already reported by Shimizu et al. (1995).
Slightly lower values at age of 20 years were reported by Berril (2004) for C. lusitanica grown in New Zealand. However, when the age of 30 years is regarded, values are identical to the ones observed in the present study, 15 m for poor sites and 35 m for very good ones. According to Miller and Knowles (1996), on good sites in New Zealand, C. lusitanica at 40 years may reach heights of 30 m, which is in line with the performance observed in the upper level of site index classification presented for southern Brazil. Altogether, it can be concluded that growth potential of the species in New Zealand is comparable to the one in southern Brazil.
A similar range of dominant height was also reported by Teshome and Petty (2000) for C. lusitanica in Ethiopia, where site classes with values between 13-30 m at age of 20 years were described. Ngugi et al. (2000) reported for C. lusitanica grown in Kenya, dominant heights at age of 20 years between 20-30 m. Similar to the ones described by Mamo and Sterba (2006) in Ethiopia for the same age, where dominant height amplitudes of 20-35 m were observed. These values match with the upper level described in the present study, been also slightly higher in Kenya, and both with less variation than the ones measured in southern Brazil. However, Pukalla and Phojonen (1993), also for Ethiopia, found a dominant height amplitude of 15-20 m at age of 20 year, which, in comparison to the ones observed in Brazil, are at a middle low level.
One important aspect in this regard is that, at least in CBS and CA (southern Brazil, Tab. 1 and Fig. 1), where the majority of C. lusitanica stands were sampled for the present study, many of them were established on shallow soils, based on the reported ability of the species to grow in poor sites. As a result, growth is obviously low. Yet, one of the best site qualities for southern Brazil was also observed in CBS, as already mentionedmost circles at a disproportional higher level in Fig. 3-a. Indeed, shallow soils are the most frequent and obvious limitation to growth of plantation forestry in the highlands of southern Brazil, as already reported for P. taeda (Horst et al., 2018). Besides this general idea, further studies should invest in understanding which site variables are the driving factor for C. lusitanica growth and yield in order to avoid the poorest ones that deliver as low as 10 m of dominant height at 20 years of age. Nevertheless, being these sites with shallow soils marginal to the majority of soil uses, it should also be verified if, indeed, C. lusitanica is able to comparatively use it in an effective way.
For comparison purpose, the dominant height amplitude for P. taeda grown in southern Brazil is 16-31 m at age of 20 years (SCHUCHOVSKI et al., 2019). The average value of h100 for P. taeda at age of 15 years is 22 m (default in Sispinus, the most used growth and production simulator), which, compared to the curve of SI 26, (at age of 15 years, h100 = 22 m) indicate that, in terms of dominant height, C. lusitanica cultivated in high quality sites and with appropriate genetic material presents comparable growth potential to the average one obtained with P. taeda.

Basal area growth and volume yield
Additional to dominant height development, basal area is an important feature in forest management, delivering even more robust insights in relation to yield and scheduling silvicultural practices. The fact that basal area at age 20 years varies between 20-55 m² ha -1 shows the impressive effect of site quality on the growth of C. lusitanica grown in southern Brazil. In Kenya, basal area between 40-50 m 2 ha -1 at age of 20 years are reported for unthinned C. lusitanica stands (NGUGI et al., 2000) showing that, in terms of basal area, the best sites of southern Brazil are comparable to the ones in Kenya.
Volume yield, however, is the key variable. Maximum mean annual increments observed for C. lusitanica in southern Brazil varied between 6-31 m³ ha -1 year -1 at ages of 16-18 years. Yield between 10-36 m 3 ha -1 year -1 have already been reported by Shimizu et al. (1995) for Brazil, the best one obtained by progenies from Colombia and the lowest ones by local seed source (Camanducaia, MG). Thus, indicating that genetic material may also play an important role in the wide range of observed yield. Unfortunately, only few studied stands have some information about their genetic provenance and, therefore, no further discussion can be supported in this regard. It indicates, however, that more attention to the seed source should be regarded. Pukalla and Pohjonen (1993) reported yields between 6-17 m 3 ha -1 year -1 for C. lusitanica in Ethiopia in rotation length ranging from 25 (best sites) to 34 years (poorest sites). The authors pointed out that proper management of plantations at that time were neglected. Berril (2004) mentioned the same values (6-17 m 3 ha -1 year -1 ) at ages of 15-30 years in New Zealand, however, with one single record of 36 m 3 ha -1 year -1 .
Standing volumes between 100-600 m³ ha -1 at 20 years of age for full stoked stands were observed as a result of site quality in southern Brazil. Noteworthy is that stands with only 100 m³ ha -1 are substantially and negatively influenced by exposed bedrocks, a physical impediment which reduces soil and water availability, the number of cultivated trees and obviously yield. Stands of C. lusitanica in New Zealand around 20 years of age showed stockings ranging from 240-1,000 trees ha -1 , with total standing volumes between 160-650 m 3 ha -1 and current annual increments of 17-58 m 3 ha -1 year -1 . By comparison, mature (>40 years) and unthinned stands of C. macrocarpa usually maintain very high stocking levels (1,200-1,300 trees ha -1 ), carrying total standing volumes as high as 1,000-1,600 m 3 ha -1 , still growing with exceptional current annual increment of 30-70 m 3 ha -1 year -1 . On poor sites, by contrast, stands are dismal in performance with total volumes at similar ages of no more than 200-450 m 3 ha -1 (MILLER; KNOWLES, 1996).
Average mean annual increments for P. taeda cultivated in southern Brazil are 34 m 3 ha -1 year -1 at age of ~15 years (ACR, 2019), once again, the values of ~30 m 3 ha -1 year -1 at age of 16-18 years observed for C. lusitanica in this study reinforces that its productivity is comparable to P. taeda, even if slightly lower. It is important to note that C. lusitanica, if regarded for timber production, would not compete for maximum volume production. Instead, it would be an option for timber production aiming at special solid end-uses with complementary wood properties (VIVIAN et al., 2020).

Dominant diameter growth
The cultivation of C. lusitanica is only meaningful in the context of market niches of high-quality timber, which implies big-sized logs. Therefore, individual growth assumes a greater importance than volume yield per hectare. By analyzing the dominant diameter of the stands, it is possible to evaluate individual growth of the potential crop trees. Dominant trees with dbh close to 50 cm can be obtained in production periods of 30-35 years regarding current genetic material and management standard. Previous study reported a substantial improvement potential from genetic material since a great variability was observed and already rescued from stands in southern Brazil (DOBNER JR. et al., 2018). Furthermore, it is noteworthy that there is no management standard for C. lusitanica in southern Brazil. Measured stands for the present study, probably all existing stands of this species in the region, were not consciously managed. The great majority had no thinning at all or only thinnings from below of low intensity. There was only one exception, the stand presented in Fig. 7-b, where intense thinnings from above were conducted since early ages and then, the diameter growth under appropriate management regimes can be evaluate. Fortunately, this was carried out at a very good site, expressing therefore the probable species upper threshold for the studied region.
Huss and Dobner (2020) reported for P. taeda grown in southern Brazil dominant diameter between 38-54 cm at 20 years of age, as a result of different planting densities and thinning intensities. The comparison provides the perception that C. lusitanica can only reach the bottom level verified for P. taeda, obtained with the highest initial density (2,500 trees ha -1 ) and without any thinning. However, at age of 40 years, C. lusitanica reach a d100 value >50 cm, similar to the one obtained from P. taeda stands with medium initial density (1,250 trees ha -1 ) subjected to heavy thinnings. Thus, C. lusitanica in southern Brazil, even without a conscious management strategy was already able to deliver dominant diameters similar to those obtained in well managed P. taeda stands, reinforcing the growth potential of the species.