The mess of sampling forest biodiversity

Make sure you don’t miss our new paper ‘One taxon does not fit all: Herb-layer diversity and stand structural complexity are weak predictors of biodiversity in Fagus sylvatica forests’ just published Open-Access on the journal ‘Ecological Indicators’ (2016).


Forest ecosystems are extremely complex. They host a wide spectrum of organisms (biologists call them taxa) that together compose their biodiversity assets. Plants, mammals, lichens, insects, fungi, birds, anura (frogs), oligochaetes (worms), spiders, mites, cyanobacteria, there is a lot of life out there and biologists struggle to get a complete picture of the whole set of species that could be find in a given forest stand. It is a fact, sampling biodiversity requires time, money and expertise which are not always available. Nevertheless, if we want to correctly prioritize our conservation and restoration efforts, we require a thorough understanding of the spatial distribution of biodiversity.

How can we deal with this problem?
Over the years, a wealth of surrogates (or indicators) have been developed and used to simplify, represent, and help the management of forest ecosystems. The assumption is that surrogates may readily reflect other biotic or abiotic components of an ecosystem, or can be used to represent the impact of an environmental change on a community or ecosystem. Other surrogates were assumed to be indicative of the abundance of a particular species or of the diversity of some taxa or of the wholesale biodiversity within an area. But what is the rationale behind the use of a surrogate and how well a given surrogate works in a specific context?

In the paper ‘One taxon does not fit all: Herb-layer diversity and stand structural complexity are weak predictors of biodiversity in Fagus sylvatica forests’ (Open-Access on the journal ‘Ecological Indicators’, 2016), we explored the effectiveness of two commonly used surrogates as indicators of the overall biodiversity in six European beech forests in the Apennines. The paper stemmed from the first monitoring phase of the LIFE+ project ‘FAGUS’ and is the result of a three year work by an interdisciplinary team of researchers from three universities and two national parks.


Vascular plants and forest structural complexity as Biodiversity indicators

In temperate forests both vascular plants and the forest stand structural characteristics were often considered as good surrogate candidates of forest overall biodiversity. Vascular plants are a potentially effective taxon-based surrogate, since they are relatively easy to sample and their taxonomy is sufficiently well described.  Plants represent the bulk of both ecosystems’ biomass and net primary productivity, and play a fundamental role in the trophic networks of forest ecosystems.

Forest structural complexity, which synthesizes the variety of structural components,  occurring in a stand, such as large trees, standing dead trees, tree stumps or logs, is also a possible candidate as a habitat-based surrogate in temperate forests. Structural complexity is often considerd as a biodiversity proxy, at least when considering those taxa that utilize specific forest structures. A typical example relates the occurrence of partially decayed deadwood to the presence of saproxylic insects, i.e. insects that depend on deadwood at least in a part of their life cycle for either feeding or shelter. Quantifying the structural complexity of a stand is difficult, since forests are different from one another in unique ways. Fortunately, we have already worked on this issue and, in a previous work, we developed a Structural Heterogeneity Index (SHI) (Sabatini et al. 2015, iForest) that was explicitly designed to account for the main sources of structural complexity occurring in the beech forests of southern Europe.

But how well these two surrogates could help to predict the species richness of five groups of forest-dwelling organisms, including beetles, saproxylic and epigeous fungi, birds and epiphytic lichens? And are the patterns of species complementarity and composition congruent between herb-layer plants and the target taxa? Collecting this data required a lot of people with different expertise, so a particular thank goes to our field crews!

Data were collected in six European beech (Fagus sylvatica) forests located in the Apennines. We focused on two habitats of European priority interest according to the EU Habitats Directive (92/43/EEC) i.e., the habitat 9210* – Apennine beech forests with Taxus and Ilex, and the habitat 9220* – Apennine beech forests with Abies alba and beech forests with Abies nebrodensis. These forests are included within two Italian national parks the “Gran Sasso and Monti della Laga”, in Central Italy, and “Cilento, Vallo di Diano and Alburni”, in southern Italy. In case you are not familiar with these places, they are really worth a visit!

Upper left: Study area. Upper right: Mt. Gran Sasso (2912 m a.s.l.), the highest peak of the Apennines, gives the name to Gran Sasso and Monti della Laga National Park (Photo: S.Burrascano). Down: the Alburni range in Cilento and Vallo di Diano National Park (Photo: M. Azzella).



How well do these indicators work?

Now, the bad news is that our data provided only a limited support to the hypothesis that the herb-layer plants or the forest structural complexity can be used as surrogates of the overall multi-taxon species richness of a site, at least in our case study. Across the three facets of biodiversity we considered (i.e. species richness, complementarity and composition), we observed only a few significant congruent patterns between these two surrogates and the target taxa.

Nevertheless, when considering species richness, the herb-layer plants performed better than stand structural complexity at predicting multi-taxon species richness, and was consistently selected as the best predictor even if the overall magnitude of this relationship was weak. This is probably due to the fact that the response of each group of organisms to increasing levels of herb-layer diversity were taxon-specific. For instance, for increasing levels of herb-layer richness, the richness of lichens showed a marked increase, while the richness of saproxylic fungi decreased. We also found significantly similar complementarity patterns between the herb-layer plants and beetles, as well as a significant congruence in species composition between herb-layer plants and saproxylic fungi.


With regards to the structural heterogeneity of a given forest stand we did find that for increasing levels of SHI the species richness of the target taxa increases, although this effect was extremely weak. Nevertheless, SHI performed worse than the herb-layer and we only found a moderate evidence that SHI may be used to complement the information provided by the herb-layer.

How can we explain these results and how to proceed from here?

Surrogates are often used, but for many taxa, the assumptions behind cross-taxon congruence have not yet been sufficiently tested, nor were the identity of the factors driving the huge variation observed across systems and scales. For instance, in our study data were collected at a very fine-scale, and recent evidences showed that cross-taxon congruence is usually pretty low at this scale. When the spatial scale is broad, a higher environmental variability is usually sampled, and thus the diversity patterns of different taxa are more likely to co-vary, also in relation to the common biogeographical and evolutionary history shared by these taxa. Furthermore, although the spatial domain of our study was relatively broad (the farthest plots were about 300 km apart), the dataset was designed not to encompass any major environmental gradients. Indeed, we limited our analysis to relatively well-preserved, low intensity managed beech forests. In this context, we cannot exclude that sampling stands along a broader environmental gradient, or stands characterized by different conservation statuses, management regimes or disturbance histories would result in stronger pattern of cross-taxon congruence.

We conclude that when a strong cross-taxon congruence cannot be assumed, like in our case, than it is better to carry out an extensive assessment of multi-taxonomical diversity rather than using weak or ineffective surrogates. It is indeed a lot of work, but only in this way we can be sure that we are minimizing the likelihood of taking poor conservation decisions based on wrong or incomplete information. Not to mention that conservation priorities of different taxa may be contrasting. Better then acknowledging these contrasts clearly before taking any decisions, rather than relying on over-optimistic win-win scenarios based on incomplete data.

In short, in this world of information-technology, fancy electronic devices and complex mathematical model we still badly need well-trained field biologists to conduct extensive field-surveys!

Ah…and economical resources for going out in the field and collect data, of course!



Relevant Literature:

Burrascano, S., Keeton, W.S., Sabatini, F.M. & Blasi, C. (2013) Commonality and variability in the structural attributes of moist temperate old-growth forests: A global review. Forest Ecology and Management, 291, 458-479.

Lindenmayer, D.B., Barton, P.S., Lane, P.W., Westgate, M.J., McBurney, L., Blair, D., Gibbons, P. & Likens, G.E. (2014) An Empirical Assessment and Comparison of Species-Based and Habitat-Based Surrogates: A Case Study of Forest Vertebrates and Large Old Trees. PLoS ONE, 9, e89807.

Sabatini, F., Burrascano, S., Lombardi, F., Chirici, G. & Blasi, C. (2015) An index of structural complexity for Apennine beech forests. iForest – Biogeosciences and Forestry, 8, 314-323.

Sabatini, F.M., Burrascano, S., Azzella, M.M., Barbati, A., De Paulis, S., Di Santo, D., Facioni, L., Giuliarelli, D., Lombardi, F., Maggi, O., Mattioli, W., Parisi, F., Persiani, A., Ravera, S. & Blasi, C. (2016) One taxon does not fit all: Herb-layer diversity and stand structural complexity are weak predictors of biodiversity in Fagus sylvatica forests. Ecological Indicators, 69, 126-137.