Facing the challenges related to deciding which clone to plant in each stand, preparing the nursery to meet future demands and adapting to the changes inherent in the forestry process is a complex task. It is necessary to take into account climatic variables, clone-site interaction, susceptibility to diseases, in addition to seeking a clonal allocation that looks for the global and not the local optimum.

To achieve this goal, it is essential to use predictive analysis tools, implement well-structured processes and have a team that is engaged and committed to the results and sustainability of the project. business. This article presents how Veracel Celulose successfully developed an integrated clonal allocation and nursery planning tool and process to address these challenges.

In recent years, Veracel has dedicated itself to addressing the challenges of clonal recommendation, nursery planning and clone-site interaction in a parallel and individual manner. The recommendation of clones was made based on a specific planting sequence, with the result being presented as a percentage of each clone by administrative region. There were difficulties in dealing with risk sizing and answering the question: "What is the impact, in terms of risk, when increasing the planting of one clone over another?"

This was more of a qualitative, argument-based exercise than a quantitative analysis. To solve these problems, a clonal recommendation optimizer was developed , whose objective was to maximize forest productivity. With it, risk measurement criteria were established both for the interaction between clone and macro-region and for the overall risk involved in recommending planting concentration in certain clones. In this way, much of the subjectivity associated with the process was eliminated.

Another challenge the company faced was the planning of the forest nursery, which needed to quickly adapt to variations in demand, climate and performance of clones. A mapping of the entire nursery process was then carried out, from the formation of the clonal garden to the dispatch of seedlings, and a nursery planning optimizer was elaborated.

This resulted in recommendations on what actions should be taken, including which clones to implant or remove, which seedlings to send to which projects, which seedlings to buy in the market or even discarding seedlings of a certain genetic material.

As a result, the nursery gained speed, becoming able to adapt more quickly to changes such as adjustments in the planting sequence, in the pace of development of seedlings of different genetic materials and in the availability of seedlings in third-party nurseries. It was also possible to anticipate probable difficulties in complying with the annual clonal recommendation and to involve the necessary people for decision-making and action.

Another fundamental aspect for an optimized clonal allocation is to understand how each clone behaves in different environments (sites). However, before exploring the interaction between clone and site, it was necessary to establish a definition of "site" for the company. One point that was much discussed in this process is that the more specific the site, the greater the number of sites, the more fragmented the forest base becomes and the smaller the amount of data available on the behavior of each clone in each of the environments. Thus, the climate and soil variables that in the company's history had the most impact on forest growth were defined , thus defining 8 sites.

The next step was to provide the productive potential for each clone and company site interaction. For this, a database was used that included historical information on forest inventory, wood quality and experiment results. Statistical modeling and artificial intelligence tools were applied in order to obtain an estimate of the expected productivity in terms of tons of pulp per hectare per year for each interaction.

Once the company had an optimized annual clonal recommendation process, an optimized nursery planning process forestry and an understanding of the interactions between clones and sites, the next step was to bring all these elements together in an integrated process. Verótima was then developed , a prescriptive analysis tool that, each month, determines the most suitable clone to be sent to each plot and indicates the necessary adjustments to be made in the nursery to adapt to the new scenario.

By establishing an integrated process, several advantages were achieved, such as greater interaction between the parties involved, considering different perspectives in decision making. In addition, information on health, disease history and susceptibility of each clone was included. Key indicators were defined, simplifying the decision-making process.

It was also possible to quickly generate comparable scenarios, providing a view of forest production expectations and risks associated with each proposed allocation.

In order to absorb a little of the dynamics of the nursery, in addition to the clone recommended for each plot, the tool provides a classification, so that, if it is not possible for some reason to issue the best clone, the person in charge can quickly decide which would be the second or third best option; this brings some flexibility to the process, without jeopardizing forestry productivity as a whole.

With the new tool, the focus is no longer on monitoring the percentage of each clone by administrative region, in order to look more carefully at the key indicators of the clonal recommendation: productivity (tons of pulp per hectare per year) and two measures of Risk: Confidence and Garcia Variability Index . Confidence indicates how well known the clone and site combination is. The Garcia Variability Index, named after Carla Garcia, Veracel 's genetic improvement coordinator , evaluates the global genetic diversity of the clonal recommendation, taking into account the area planted per clone and the genetic distance between them.

The clonal allocation process with Verótima has been operational since August 2021, with monthly rounds of the tool. It is during these rounds that new lessons and opportunities emerge, allowing further improvement in clonal allocation and nursery planning. It is notable to see that Veracel achieved a 96% adherence between what was carried out and what was planned by the tool throughout 2022. This demonstrates the commitment of the entire team involved in making the algorithm's recommendations a reality.

In conclusion, the successful implementation of Verótima highlights the importance of well-established processes, the power of technological resources and, above all, the active participation of committed and motivated teams to challenge themselves and do their best to generate significant changes.

The combination of knowledge acquired through constant learning and openness to discussions has been fundamental to sustaining the gains obtained and maintaining the company's forestry productivity. By continuously improving data-based management, prioritizing the search for excellence and promoting collaboration among all those involved, Veracel is prepared to face challenges and guarantee a prosperous and sustainable future for the business.