Millipede taxonomy after 250 years: classification and taxonomic practices in a mega-diverse yet understudied arthropod group

Abstract

Background: The arthropod class Diplopoda is a mega-diverse group comprising >12,000 described millipede species. The history of taxonomic research within the group is tumultuous and, consequently, has yielded a questionable higher-level classification. Few higher-taxa are defined using synapomorphies, and the practice of single taxon descriptions lacking a revisionary framework has produced many monotypic taxa. Additionally, taxonomic and geographic biases render global species diversity estimations unreliable. We test whether the ordinal taxa of the Diplopoda are consistent with regards to underlying taxonomic diversity, attempt to provide estimates for global species diversity, and examine millipede taxonomic effort at a global geographic scale.

Methodology/principal findings: A taxonomic distinctness metric was employed to assess uniformity of millipede ordinal taxa. We found that ordinal-level taxa are not uniform and are likely overinflated with higher-taxa when compared to related groups. Several methods of estimating global species richness were employed (Bayesian, variation in taxonomic productivity, extrapolation from nearly fully described taxa). Two of the three methods provided estimates ranging from 13,413-16,760 species. Variations in geographic diversity show biases to North America and Europe and a paucity of works on tropical taxa.

Conclusions/significance: Before taxa can be used in an extensible way, they must be definable with respect to the diversity they contain and the diagnostic characters used to delineate them. The higher classification for millipedes is shown to be problematic from a number of perspectives. Namely, the ordinal taxa are not uniform in their underlying diversity, and millipedes appear to have a disproportionate number of higher-taxa. Species diversity estimates are unreliable due to inconsistent taxonomic effort at temporal, geographic, and phylogenetic scales. Lack of knowledge concerning many millipede groups compounds these issues. Diplopods are likely not unique in this regard as these issues may persist in many other diverse yet poorly studied groups.

Figure 1. Taxonomic distinctness plots for the ordinal taxa of the Diplopoda.

Plot A shows taxonomic distinctness in relation to the species diversity of the ordinal groups. Plot B shows the same relationship as the previous plot with species diversity log transformed and a best fit line obtained via linear regression (r² = 0.5705, P<0.01; does not include the two millipede orders with only two species). Points representing the orders of the Chilopoda (shown in red; does not include the single species order Craterostigmomorpha) and the Pseudoscorpiones (shown in green) were added after the regression analyses and have no bearing on the best fit line.

Figure 2. Plots showing millipede taxonomic effort over time in terms of species descriptions.

Plot A shows the data as used to calculate global species diversity following the methods of Bebber et al. . The y-axis corresponds to the number of species descriptions in a given year (t) while the x-axis shows the number of species accumulated at time “t-1”. A local regression line is shown fitted to the data. Plot B shows a species accumulation curve fitted with a local regression line with no trend toward asymptote. Plot C shows the taxonomic productivity over time in terms of species descriptions published yearly and is fitted with a local regression line.

Figure 3. Number of millipede species descriptions in countries as a correlate of land area and human population.

Plots A, C, and E show all countries from which millipedes have been described. Plots B, D, and F have the countries designated as tropical (≥50% of total land area in the tropics) or nontropical (<50% of total land area in the tropics) and were analyzed separately. The outliers for human population, India and China, are excluded from plots E and F.