**1. Introduction**

The Forest Stewardship Council (FSC) forest certification was developed in the early 1990s to reduce the deforestation and forest degradation caused by unsustainable forest managemen<sup>t</sup> [1]. The FSC forest certification assures buyers/consumers that certified forest products have been produced from forests that have been responsibly/sustainably managed in terms of environment, economy and society. Certified forest products are expected to gain better consumer appeal over uncertified products in the market and eventually drive out uncertified products. Through this market mechanism, the deforestation and forest degradation caused by unsustainable forest managemen<sup>t</sup> will eventually be mitigated.

The FSC forest certification is issued when managemen<sup>t</sup> entities comply with a set of criteria for ten principles in their forests and managements. Principle 5 stipulates criteria for enhancing benefits from the forests, including ecosystem services. Based on the FSC forest certification, one assumes that ecosystem services are responsibly managed in certified forests. However, FSC forest certification is by no means a system to quantitatively verify positive managemen<sup>t</sup> impacts or the enhancement of ecosystem services [1,2]. Based on these considerations, the FSC has increasingly received demands to reliably certify important forest ecosystem services [3–9].

The FSC, therefore, pilot-tested a new system to directly certify the maintenance/enhancement of ecosystem services to promote ecosystem services payments and to further incentivize certificate holders [1,2,5,10–12]. The new system is expected to quantitatively demonstrate for businesses and investors that ecosystem services are maintained/enhanced in certified forests [11]. Ecosystem services include biodiversity conservation, carbon sequestration and storage, watershed services, soil conservation and recreational services [2,12,13]. The FSC suggested generic indicators for each of the five ecosystem services (Annex C, FSC-STD-60-004 V1-0-EN, International Generic Indicators; [13]). However, practical methodologies to evaluate the impacts of forest managemen<sup>t</sup> on these ecosystem services are still undergoing improvement. Developing robust methodologies to quantitatively evaluate the impacts of forest managemen<sup>t</sup> on these ecosystem services is urgently needed to support a reliable ecosystem-services certification system [2]. Moreover, many managemen<sup>t</sup> entities, who wish to apply for the ecosystem-services certification (or other paymen<sup>t</sup> for ecosystem services and REDD+ schemes), may lack the technical capacity to measure the maintenance/provision of ecosystem services [1,14–16].

Thus, methodologies to measure the maintenance/provisioning of ecosystem services must be simple and practical enough to be undertaken by managemen<sup>t</sup> entities on the ground, while scientifically robust enough to accurately measure spatiotemporal changes of each ecosystem service [2]. In view of these challenges, the FSC encourages the development of "a small number of powerful and easy-to-measure proxy impact indicators" [17]. In addition to these technical requirements, the cost of measurements is another issue. Many certification/eco-labelling schemes are already associated with high transaction costs [1,18–20], and further adding a high cost of a ground survey for measuring ecosystem services will discourage stakeholders from applying for the ecosystem-services certification [1].

Jaung et al. [1] suggested bundling ecosystem services to reduce the certification cost per ecosystem service. The certification of multiple ecosystem services also has several other benefits: increasing incomes for managemen<sup>t</sup> entities, allowing managemen<sup>t</sup> entities to access diverse ecosystem-services markets, and encouraging managemen<sup>t</sup> entities to adopt a holistic forest managemen<sup>t</sup> approach [1]. Certification of bundled ecosystem services can also reduce the potential trade-offs between carbon sequestration and biodiversity conservation [5], which is also an issue in Reducing Emissions from Deforestation and Forest Degradation-plus (REDD+) as a biodiversity safeguard, because enhancing carbon sequestration by industrial plantation of exotic species will sacrifice biodiversity. However, again, the lack of methodologies to measure bundled ecosystem services presents a major challenge [1].

In view of these technical challenges, we report here the results of testing our method to simultaneously measure carbon stock and biodiversity over a large area of tropical production forests. These two ecosystem services, among others, may be relatively easily incorporated into ecosystem-services certification because stakeholder adaptability is high for these services [10]. Our method "biodiversity observation for land and ecosystem health (BOLEH)" was developed primarily for evaluating forest intactness using a metric of tree community composition in logged-over tropical rain forests [21–23]: either species or genus composition could be used in this method. Earlier studies indicated that the use of genus composition was statistically equally reliable as species composition, and ye<sup>t</sup> could reduce identification cost [21–23]. This approach is directly relevant to the FSC ecosystem-services certification because the following generic indicator for biodiversity conservation can be evaluated with our BOLEH:

• Management activities maintain, enhance or restore natural landscape-level characteristics, including forest diversity, composition and structure (stipulated in Principle 5, Annex C; FSC-STD-60-004 V1-0-EN International Generic Indicators; [13]).

Furthermore, the recently published "Ecosystem Services Procedure" [12] suggests "conservation of intact forest landscape" as an important managemen<sup>t</sup> impact in biodiversity conservation. Our BOLEH method can directly monitor spatiotemporal changes of forest intactness in a landscape context.

In addition, BOLEH can derive carbon stock (density) values using the same dataset as tree community composition. We applied BOLEH to evaluate the spatiotemporal changes of carbon stock and forest intactness (tree community composition) in two contrasting forest managemen<sup>t</sup> units in 2009 and 2014. Our test sites were Deramakot and Tangkulap in Sabah, Malaysia, which are now certified by the FSC ye<sup>t</sup> having contrasting logging histories [24,25]. Deramakot has been harvested continuously with reduced-impact logging (RIL) since 1995, while Tangkulap was highly degraded by high-impact conventional logging which was operated until 2001. Logging is now fully suspended in Tangkulap, but the recovery of ecosystem services may be slow due to ecological aftereffects of the past high logging impacts and associated collateral damages. We report here the accuracy of estimating carbon stock and forest intactness (or tree community composition as an indicator of biodiversity conservation) using BOLEH and the patterns of spatiotemporal changes of these two ecosystem services between the two contrasting forest managemen<sup>t</sup> units.
