Quantifying spawning biomass of commercially relevant fish species is important to generate fishing quotas. This will mostly rely on the annual or daily production of fish eggs. However, these have to be identified precisely to species level to obtain a reliable estimate of offspring production of the different species. Because morphological identification can be very difficult, recent developments are heading towards application of molecular tools. Methods such as COI barcoding have long handling times and cause high costs for single specimen identifications. In order to test MALDI-TOF MS, a rapid and cost-effective alternative for species identification, we identified fish eggs using COI barcoding and used the same specimens to set up a MALDI-TOF MS reference library. This library, constructed from two different MALDI-TOF MS instruments, was then used to identify unknown eggs from a different sampling occasion. By using a line of evidence from hierarchical clustering and different supervised identification approaches we obtained concordant species identifications for 97.5% of the unknown fish eggs, proving MALDI-TOF MS a good tool for rapid species level identification of fish eggs. At the same time we point out the necessity of adjusting identification scores of supervised methods for identification to optimize identification success.

Of the 40 µl mixture containing water and the crushed eggs, for all samples 5 µl were used for extraction of peptides and proteins (Fig. 2). To this mixture, 5 µl of 5% trifluoroacetic acid were added and incubated for 1h at room temperature (RT). After 1h, 1 µl of this solution was transferred to a MALDI-TOF target plate and dried at RT. The dried extract was covered with a layer of 1 µl α-Cyano-4-hydroxycinnamic acid (HCCA) as a saturated solution in 50% acetonitrile, 47.5% molecular grade water and 2.5% trifluoroacetic acid. For each egg, the solution was applied to one spot and air dried for co-crystallization of proteins and peptides with the matrix.

Measurements were carried out on a Microflex LT/SH System (Bruker Daltonics) using method MBTAuto with laser intensity between 50% and 60%. Peak evaluation during measurement was carried out in a mass peak range between 2k – 10k Dalton (Da) using a centroid peak detection algorithm, a signal to noise threshold of 2 and a minimum intensity threshold of 600, with a peak resolution higher than 400. Proteins/Oligonucleotide method was employed for fuzzy control with a maximal resolution ten times above the threshold. To create a sum spectrum, 240 satisfactory shots were summed up. Each spot was measured once. Before measurements, system calibration was carried out according to the Bruker default procedure using Bacterial standard (BTS).

Alongside raw data an Excel file containing meta data refering to sampling time and place is enclosed. In total raw data from 511 fish egg specimens is included. However, some of these were found to be of inferior quality and are not usefull for specimen identification. These are accordingly marked in the supplied meta data file.