<![CDATA[This community service program aimed to enhance senior high school mathematics teachers’ competence in constructing numeracy-oriented mathematical problems using a deep learning approach integrated with wetland environmental contexts. The program involved 39 mathematics teachers from the Mathematics Teachers’ Working Group (MGMP) of Banjarmasin and was implemented through a Participatory Action Research framework consisting of planning, implementation, observation, and reflection and sustainability with follow-up mentoring. Teachers’ understanding was evaluated using a one-group pretest–posttest design, while participant satisfaction was measured through a four-point Likert-scale questionnaire. Data were analyzed using the Wilcoxon signed-rank test, normalized gain (N-gain), effect size, and descriptive statistics. The average score increased from 85.38 on the pre-test to 88.21 on the post-test. However, the Wilcoxon signed-rank test indicated that the improvement was not statistically significant, which may be explained by participants’ relatively high baseline competence and their participation in a Ministry of Primary and Secondary Education training program on deep learning and numeracy before the mentoring activities. The N-gain was classified as low, and the effect size indicated a small effect, suggesting a small improvement in teachers’ understanding. Nevertheless, participant satisfaction was very high, and qualitative evidence from discussions and teachers’ problem-design assignments demonstrated increased confidence and competence in developing contextual numeracy problems based on wetland environments. These findings indicate that mentoring integrated with authentic local contexts provides a practical and sustainable approach to strengthening teachers’ pedagogical competence and supporting meaningful numeracy instruction.]]>
