<![CDATA[Environmental education is most consequential when students can connect scientific ideas with visible socio-ecological problems in their own communities. This study developed and conducted a preliminary evaluation of project-based learning (PjBL) student worksheets grounded in the mangrove ecosystem surrounding Lembar, Lombok, Indonesia. The study combined the four-stage 4D instructional-development model—define, design, develop, and disseminate—with a non-equivalent pretest–posttest control-group evaluation. Three experts reviewed the learning objectives, teaching module, worksheet package, and two literacy assessments. A limited classroom trial examined implementation and user practicality, followed by a wide-scale trial involving two intact Grade 7 classes (experimental n = 29; control n = 29). Content validity was acceptable across all products (overall Aiken’s V = 0.71–0.73), while inter-rater agreement ranged from 97.12% to 99.45%. Across four meetings, mean implementation fidelity reached 97.73%; teacher and student practicality ratings were 97.81% and 92.64%, respectively. The experimental class achieved moderate normalized gains in mangrove ecosystem literacy (g = 0.65; 65.16%) and environmental literacy (g = 0.64; 64.33%). Because at least one distribution in each outcome was non-normal, between-group comparisons used the Mann–Whitney U test. The experimental condition was favoured for both mangrove ecosystem literacy (U = 3.50, Z = −6.527, p < .001, r = .86) and environmental literacy (U = 24.50, Z = −6.178, p < .001, r = .81). The findings indicate that locally grounded PjBL worksheets are valid, feasible, and promising for connecting ecosystem knowledge, evidence-based reasoning, and conservation-oriented action. However, the single-school, non-randomized design and the absence of long-term follow-up require the effectiveness findings to be interpreted as preliminary.]]>
