- The performance of a constructed wetland (CW) with Salicornia persica ( Chenopodiaceae ) as a biofilter for effluent water drained from a semi-open recirculating mariculture system was studied in southern Israel. The results demonstrate the effectiveness of N, P and total suspended solid (TSS) removal from mariculture effluent by a CW operating with this plant. The CW was exposed to high (3.3 ± 0.26 g N m − 2 d − 1 ) and low (0.13 ± 0.02 g N m − 2 d − 1 ) nutrient loads (NL) in two hydraulic regimes, surface flow (SF) and subsurface flow (SSF). In addition to the CW being studied as marine fishpond effluent water purifier, the Salicornia yield was evaluated as a marketable agricultural by-product. Eight CW ponds were covered under a 750 m 2 greenhouse. The surface area of each CW pond was 24.3 m 2 (13.5 m length, 1.8 m width and 0.30 m average depth) and volume of 7.3 m 3 . The fishponds' effluent drained in a continuous mode into one end of each CW pond and out from the opposite end (plug flow) at a hydraulic load (HL) of 0.5 m 3 h − 1 , a hydraulic loading rate (HLR) of 0.49 m d − 1 , and a hydraulic retention time (HRT) of 1.51 d − 1 . Dissimilation processes such as ammonification, nitrification and denitrification with assimilation by the plants performed as in natural wetlands. Salicornia wet yield was 23 ± 1.6 and 26 ± 4.6 kg m − 2 y − 1 in the SF and SSF flow regimes, respectively, at the low nitrogen load, and 20.1 ± 2.4 and 17.4 ± 3.1 kg m − 2 y − 1 in the SF and SSF flow regimes, respectively, at the high nutrient load. Total nitrogen (TN) uptake by Salicornia tissue itself excluding the wetland activity (including roots) was 29.1 ± 3.3 and 26.9 ± 3.1 g N m − 2 y − 1 in the SF and SSF flows, respectively, at low nitrogen load, and 18.5 ± 3.5 and 13.3 ± 3.3 g N m − 2 y − 1 at high N loads in the SF and SSF flow regimes, respectively. The contribution of Salicornia as a nitrogen biofilter at high NL was negligible (0.5%–0.9% of the total dissolved nitrogen [TDN]) compared to the low NL (56%–61.4% of the TDN) in both SF and SSF regimes respectively. Our results show that the SF regime with Salicornia would likely be more efficient in facilities with low NL (e.g., fish hatcheries), whereas the SSF regimes would be more efficient with high NL (e.g., super-intensive fish farms). Using CW systems for effluent treatment requires a relatively extensive area. According to our results, about 10,000 m 2 of CW with Salicornia are required to remove nitrogen and TSS produced from 900 kg of 45% crude protein fish feed (11 m 2 kg − 1 of feed) during one year. The income generated from selling the Salicornia as an agricultural crop, together with savings on water treatment and potential fines, contributes to the system's economical viability.