Sustainable forest management (SFM) considers the social and environmental implications of forestry. Criteria and indicators (C&I) spread an understanding of the concept ( Wijewardana, 2008 ), whose principles adapt to local circumstances ( Castañeda, 2000 ; Barbati et al ., 2007 ). The consideration of local conditions becomes especially relevant under Mediterranean conditions ( Osem et al ., 2008 ).

Valls-Donderis et al . (2015) identified C&I adapted to ecosystem services of Mediterranean forests and applicable at the forest management unit (FMU) scale. Those C&I are verified with stakeholders, however, their indicators are too technical for not professionals and these are arranged into aspects (general issues covered by a criterion).

Multiple criteria analysis techniques (MCA) help making choices when several criteria apply. Mendoza & Prabhu (2000) conclude that MCA methods are useful to prioritise and evaluate C&I, and that, for indicators, pairwise comparison methods, like the analytic hierarchy process (AHP), are more accurate. The aim of this research was to verify the indicators from Valls-Donderis et al . (2015) with experts using AHP.

The set of indicators identified by Valls-Donderis et al . (2015) consists of 133 indicators. Even though only the ones of the same criterion were meant to be compared, it still meant many indicator pairs to value. To shorten the AHP questionnaire, the method was adapted: one indicator was compared against the others of the same criterion. Results should not have differed from the pure method: given A-B-C, if a respondent said that A was two times more important than B and A was five times C, if that respondent was consistent, she/he would have said B was three times C. Although Saaty (1980 , 2006 ) states that consistency rarely happens, it was assumed in this study. Further considerations on the limitations and advantages of this simplified application of the AHP method are provided in the results and discussion section.

Saaty (1980, 2006) establishes that more than nine items cannot be compared. Most criteria had more indicators; some were deleted or joined and 103 indicators were finally presented to experts. The indicator of each criterion to be compared against the others were selected at random since researchers did not show special preference for any of the indicators (the indicators evaluated, and the criteria they belong to, are in Table S1 [suppl]).

The questionnaire was made of 15 questions (one for every criterion). In each question, n-1 pairs of indicators were shown (being “n” the number of indicators contained within the criterion). Next to each pair of indicators Saaty’s valuation scale (Saaty, 1980, 2006) was provided as shown at the end of Table S1 [suppl].

Potential respondents were selected by means of purposive sampling. As described by Bernard (2000) , the method consisted of deciding the profile of the individuals that would suit the study and going out to find them. A selection of 343 experts from the groups considered suitable (groups are referred in the results section) whose contact details were available in the internet, or by means of colleagues, were approached via e-mail.

Respondents could also make comments on the indicators or their valuations. Answers were aggregated according to the AHP process (quantitative analysis) described in Saaty (1980, 2006). Comments were put together (qualitative analysis); these comments served to select and rephrase indicators.

A group of 44 experts completed the questionnaire: from central and local governments (8), universities (11), research centres (10), private and public enterprises (12), freelance (4), forest owners (9) and forestry associations (1). Concerning comments, 14 respondents made some.

A short version of the indicators appears in Figs. 1 , 2 and 3 , which show the weights of the indicators (see Table S1 [suppl] to check the whole version). Unlike most criteria, for two of them indicators got similar weights: “diversified exploitation of forests” ( Fig. 1 ) and “education” ( Fig. 2 ).