feat: add predict-carbon module with auto selection and genus fallback

[aitorvv]

Description

This PR implements the carbon prediction module (PR 4 in the development roadmap). It adds functions to estimate carbon content from predicted biomass using species- and component-specific percentages from the internal carbon_models database.

Key Changes

• New functions:
  • silv_predict_carbon(): Base function that calculates carbon (biomass * carbon_percentage / 100) for a specified model.
  • silv_predict_carbon_auto(): Vectorized function with automatic model selection and fallback logic.
• Fallback logic:
  • Automatically falls back to genus-level (e.g. "Pinus spp.") when the exact species matches are not found, informing the user of the shift.
  • If no match or fallback is found, returns NA and warns the user, suggesting they try manually using general groups like "Otras coníferas", "Otras frondosas", or "Otras laurisilvas".
• Integrations:
  • Added new file R/predict-carbon.R to the package.
  • Registered the functions in DESCRIPTION (Collate field) and _pkgdown.yml.
  • Exported functions and rebuilt documentation using devtools::document().
• Tests:
  • Added a new unit test suite in tests/testthat/test-predict-carbon.R covering base, auto-priority, fallback logic, and warnings.

Verification

devtools::load_all()

# 1. Coincidencia exacta
res_exact <- silv_predict_carbon(
  biomass = 100, 
  species = "Pinus pinaster", 
  component = "stem", 
  model = "montero-2005"
)
print(res_exact) # Esperado: 51.1 kg

# 2. Uso automático con fallback a "Eucalyptus spp."
res_auto <- silv_predict_carbon_auto(
  biomass = 100,
  species = "Eucalyptus inventatus",
  component = "tree",
  priority = "montero-2005"
)
print(res_auto)

# 3. Especie totalmente desconocida (Sugiere grupos genéricos)
res_fail <- silv_predict_carbon_auto(
  biomass = 100,
  species = "Roble magico",
  component = "stem"
)
print(res_fail)