Gobius, Linnaeus, 1758

Cooper, David, Yamaguchi, Nobuyuki, Macdonald, David, Nanova, Olga, Yudin, Viktor, Dugmore, Andrew & Kitchener, Andrew, 2020, Identification of past and present gobies: distinguishing Gobius and Pomatoschistus (Teleostei: Gobioidei) species using characters of otoliths, meristics and body morphometry, Contributions to Zoology 89 (1), pp. 282-323 : 290-295

Species separation in Gobius

Otolith morphometry

Using the complete dataset (N, 14 species), ten otolith variables (out of 23) were found to be useful for species separation. These variables were significant for a total of eight species (ANCOVA/ANOVA, table 2a 1). The highest separation success was achieved with the variables SuL/SuH (77% for G. geniporus ) and SuP/SuEndV (77% for G. paganellus ) ( table 2a 1). The five most successful variables were OL/OH, SuL/OP, SuL/SuEndV, SuH/ SuEndV and SuP/SuEndV, but each of these separated no more than three of the 14 included species ( table 2a 1). ANOVA based on PC1–4 calculated from the otolith variables successfully discriminated between five species( table 2b 1). Gobius vittatus was separated from 69%, the others from 54% of the congeners. Notably, G. cruentatus , G. gasteveni and G. roulei , which had not been separated when the individual otolith variables were used, could now be distinguished from each other (compare table 2b 1 vs. 2a1).

When the reduced dataset (N, 10 species) was used, 17 otolith variables (out of 23) were successful in species separation (ANCOVA/ ANOVA, table 2a2 ). The same two variables as in the complete dataset (SuL/SuH, SuP/ SuEndV) had maximum separation success and also discriminated the same species ( G. geniporus , G. paganellus ), each with increased success (89% vs. 77%). Six further variables were also taxonomically indicative, as each discriminated one or two species from all but two of the congeners (78% success). These variables were OL/OH ( G. incognitus ), OP/OL ( G. cobitis ), SuP/OP ( G. vittatus ), SuL/ OP ( G. auratus , G. vittatus ), SuL/OL ( G. auratus ), SuL/OH ( G. incognitus ), and SuL/SuEndV ( G. auratus ) ( table 2a2 ). With respect to the power of a given variable to discriminate between many species, the two most efficient variables were SuL/SuEndV and SuH/SuEndV; each of these could discriminate between six species (vs. three species when the complete dataset was used). In total, all species could be discriminated in the reduced dataset from ≥50% of the congeners ( table 2a2 ). Four species were separated by eight to ten variables ( G. auratus , G. geniporus , G. paganellus , G. vittatus ), four more species were discriminated by four to six variables ( G. cobitis , G. incognitus , G. niger , G. roulei ), and two species ( G. bucchichi , G. cruentatus ) by two variables ( table 2a2 ). Furthermore, the results of ANOVA using PC1–4 calculated from all otolith variables enabled the discrimination of eight species ( table 2b2 ), vs. five when the complete dataset was analysed in the same way ( table 2b 1). Four of these ( G. auratus , G. bucchichi , G. niger , G. paganellus ) had not been separated when the complete dataset was tested. On the other hand, G. cruentatus was not separated when PC1-4 were used, although it had been discriminated when the entire dataset was used. Overall, the approach based on PC1–4 was less effective than the analysis of individual otolith variables because it was unable to distinguish between G. cobitis , G. cruentatus and G. incognitus (compare table 2b2 and 2a2 ).

Fourier shape analysis of the otoliths

Based on PC1 and PC2 of the Fourier descriptors, species separation success was 77–100% and 80–100%, respectively, in the complete and reduced datasets (MANOVA; table 2c 1, c 2 ). Gobius paganellus was the only species that could be discriminated from all others with 100% separation success in both datasets. Gobius cobitis , G. cruentatus and G. geniporus were differentiated with 100% success when the complete dataset was used (80% in the reduced dataset), whereas G. bucchichi and G. incognitus were distinguished with 100% success when the reduced data was analysed (92% and 77% in the complete dataset) ( table 2c 1, c 2 ).

Body morphometry

When the complete dataset was analysed, four (out of eight) morphometric variables were taxonomically indicative, i.e., SN/D1, SN/ D2, D2b, Ab (ANOVA; table 2d 1). The highest success rates (77%) were achieved with SN/D1 (for G. gasteveni ) and Ab (for G. fallax ). The most powerful single variable (with respect to the number of species that can be separated) was Ab, which resolved five species, followed by D2b (four species) ( table 2d 1). In total, nine species could be separated based on one, two or three morphometric variables, while five species could not be separated ( G. auratus , G. couchi , G. cruentatus , G. geniporus , G. paganellus ( table 2d 1). Using PC1–4 of the variables derived from body morphometry in the ANOVA analysis was less efficient than when the individual variables were used ( table 2e 1). Only five species were separated in this case and separation success for a given species varied from 54 to 62% ( table 2e 1). Gobius cruentatus and G. geniporus , which could not be distinguished in the preceding analysis, were now separated. The other three species that could be discriminated were G. bucchichi , G. incognitus and G. vittatus .

The results based on the reduced dataset ( table 2d2, e2 ) were very similar to those obtained from the complete dataset. The same four morphometric variables could separate a species (SN/D1, SN/D2, D2b, Ab), and the individual species were separated with the same success (56–78%; see table 2d2 ). As before, most powerful in relation to the number of species that could be separated were Ab (four species) and D2b (five species). With the single exception of G. roulei (not separated), the same species as in the complete data were separated (five in total), each based on one to four variables ( table 2d2 ). When PC1–4 of the morphometric variables were used in the ANOVA, a total of seven species could be separated ( table 2e2 ). Three of those ( G. roulei , G. cruentatus , G. geniporus ) had not been separated based on the approach using single morphometric variables ( table 2d2 ). In contrast, G. niger could not be discriminated based on PC1–4 of the morphometric variables ( table 2e2 ), whereas it was distinguished on the basis of D2b alone ( table 2d2 ).

Meristic counts

None of the individual meristic characters met our criteria for species separation when we used the complete dataset (Kruskal-Wallis; table 2f 1). On the other hand, ANOVA on the basis of PC1–4 of the meristic values distinguished G. incognitus (with 92% success), G. fallax (77%) and G. geniporus (62%; see table 2g 1). A similar outcome was obtained based on the reduced dataset. A single character (ventral procurrent rays) separated with 56% success G. geniporus and G. incognitus from the others ( table 2f2 ). ANOVA based on PC1–4 of the meristic variables also separated G. geniporus and G. incognitus from all others, now with 89% and 100% success, respectively; in addition, G. cobitis was now separated ( table 2g 2 ).

Discriminant analyses

Each LDA was based on the reduced dataset and the variables used were (i) PC1–4 of the otolith morphometric variables, (ii) PC1–23 of the otolith Fourier descriptors, and (iii) PC1–4 of the body morphometric variables (fig. 2, table 3).

The first two functions of the LDA based on PC1–4 of the otolith morphometric variables captured 50.5% and 33.5% of the variation, respectively (fig. 2a). Overall classification success (jack-knifed) was 45.4%; the highest classification success was 60% for both G. bucchichi and G. geniporus , and the lowest was 18% for G. cruentatus , 30% for G. roulei and 33% for G. incognitus (table 3a). The scatter plot revealed overlap of all species (fig. 2a).Three “pairs” of species are tentatively recognizable: G. vittatus and G. auratus , G. geniporus and G. paganellus , as well as G. cruentatus and G. cobitis (fig. 2a). The members of each pair show some overlap, but only relatively moderate overlap with the other species considered.

The first two functions of the LDA based on PC1–23 of the otolith Fourier descriptors captured 39.1% and 19.1% of the variation, respectively (fig. 2b). Overall classification success (jack-knifed) of this LDA was 85% (table 3b). 100% separation success was achieved for G. cruentatus and G. geniporus (vs. 18% and 60% in the previous LDA), and 95% was obtained for G. cobitis (vs. 50% in the previous LDA) ( Table 3b ). The scatter plot delineates three groups: Gobius cobitis , which had broadly overlapped with G. cruentatus in the preceding LDA, is now distinct (fig. 2b). The second group includes G. geniporus , G. niger and G. roulei , and the third one comprises the remaining species (fig. 2b).

The first two functions of the LDA based on the body morphometric variables accounted for 66.3% and 18.6% of the variation, respectively. Overall classification success (jack-knifed) was 54.6% (table 3c). Classification success was highest for Gobius vittatus and G.bucchichi , with 81.8% and 70%, respectively, while the success rates were lowest for G.auratus , G.cobitis and G.roulei (27–30%) (table 3c). The scatter plot revealed overlap of almost all species, except G. vittatus , which is comparatively distinct, and G. bucchichi and G. incognitus , which overlap with each other, but not or not very much with the remainder (fig. 2c).

Tables

TABLE 2 Results of the statistical analyses based on the Gobius datasets and the indicated variables.The success of a certain variable in separating a species is indicated in % (100% means that the species is separated from all congeners by this variable).Light / dark green color denotes that at least two species were separated by the given variable, and each of those two species from ≥ 50% / ≥ 80% of the congeners (in the respective dataset)

Classification success based on otolith variables (AN(C)OVA, P <0.05)
a1 complete dataset															a2 reduced dataset
Species aur buc			cob	cou	cru	fal	gas	gen	inc	kol	nig	pag	rou	vit	aur	buc	cob	cru	gen	inc	nig	pag	rou	vit
N	10	10	11	4	11	2	1	10	6	2	7	9	10	10	10	10	11	11	10	6	7	9	10	10
OL/OH	31	31	38	54	23	15	8	54	62	46	38	46	38	46	33	33	44	33	67	78	44	56	33	56
OP/OL	15	15	62	23	15	0	15	31	38	31	0	15	15	54	22	22	78	44	44	44	11	22	22	67
OP/OH	31	31	31	62	38	15	0	38	23	23	23	46	31	38	33	44	33	44	56	33	33	56	56	56
SuP/OP	38	15	15	8	8	0	15	23	62	0	8	46	23	62	44	22	22	11	22	67	11	22	33	78
SuL/OP	62	15	15	8	8	0	23	38	62	0	15	23	23	62	78	22	33	11	56	67	22	22	33	78
SuH/OP	23	23	0	0	15	0	0	15	23	0	0	23	0	0	33	33	0	33	33	33	0	33	0	0
SuTipV/OP	46	15	23	54	38	23	8	31	23	8	23	38	31	54	56	33	22	44	33	22	33	44	44	67
SuEndV/OP	38	38	15	15	23	23	15	46	23	15	46	54	38	23	56	56	33	33	44	33	44	67	67	33
SuA/OA	38	15	15	8	23	0	15	15	54	0	0	0	0	31	44	22	22	22	22	56	0	0	0	33
SuL/OL	54	8	23	31	0	0	0	23	23	0	0	15	8	38	78	11	22	11	22	22	0	22	11	56
SuL/OH	38	8	0	38	0	0	0	0	69	0	0	0	0	38	44	33	33	44	67	78	67	44	67	44
SuL/SuH	15	15	15	31	15	0	8	77	8	15	23	54	15	31	22	22	22	22	89	22	22	78	22	33
SuL/SuP	0	15	8	15	0	0	0	23	0	0	0	31	0	15	0	22	22	0	33	0	0	33	0	22
SuL/SuTipV	38	23	23	62	15	23	23	46	38	15	46	46	46	38	56	22	22	33	56	44	67	56	67	44
SuL/SuEndV	54	23	46	23	23	15	23	54	46	15	23	54	31	38	78	33	56	22	56	56	22	56	44	56
SuH/OL	15	15	23	0	23	0	0	23	0	0	8	23	0	0	22	22	33	33	33	0	11	33	0	0
SuH/OH	31	15	8	0	23	0	0	0	23	0	15	0	0	15	44	33	22	33	0	33	33	0	0	33
SuH/SuP	15	15	15	38	15	0	0	54	15	15	15	46	15	46	22	22	22	22	67	11	11	56	22	56
SuH/SuTipV	38	15	15	8	46	0	8	15	23	0	15	8	8	31	56	22	22	56	22	33	22	11	22	44
SuH/SuEndV 54		31	54	38	38	0	31	31	62	0	46	31	46	8	67	44	56	56	44	67	56	44	56	11
SuP/SuTipV	54	31	23	62	23	23	23	38	38	0	38	38	46	38	56	33	22	33	44	44	67	44	67	44
SuP/SuEndV	31	62	62	15	23	23	0	23	23	15	23	77	38	46	44	67	67	33	33	44	33	89	44	56
SuTipV/ SuEndV	8	8	8	8	8	15	0	0	8	0	54	0	0	8	11	11	11	11	11	11	100	11	11	11