Odd behavior with optical emission lines

I noticed a strange behavior where after a logzsol value of 0.2 all points in a standard BPT diagram ([NII]/Hα, [OIII]/Hβ) are stuck to a certain value. I am using default stellar libraries: "mist" and "miles" to get the BPT emission-line ratios ([OIII]/Hβ and [NII]/Hα) for an SSP (sfh=0) observed at tage=1Myr (0.001Gyr) with a constant ionization parameter of logU=-2.5 but for different metallicities (logzsol=[-0.6,0.6]), also I set zcontinuous=1. The models are such that logzsol = gas_logz. I attach the resulted BPT diagram bellow. In an attempt to understand what is happening I have calculated the SEDs in two cases; 1) sfh=0 (SSP), logz=0, dust1=0, dust2=0 for different ages; 2) sfh=0 (SSP), dust1=0, dust2=0 for different metallicities (both with default stellar libraries). The UV produced by the two SSPs (I think that) look normal (attached bellow). Finally I am giving you the result of the [OIII]/Hβ as a function of metallicity (from the calculations of BPT).

Is this expected behavior?

Here is my code for the BPT calculations (the two SED tests are calculated from an other script)

import os
import numpy as np
import fsps
from tqdm import tqdm
import matplotlib.pyplot as plt
import time
import json
import pandas as pd
import warnings
import random as random
warnings.filterwarnings('ignore')

sp = fsps.StellarPopulation(zcontinuous=1)


sp.params['imf_type'] = 2
sp.params['sfh'] = 0 
sp.params['sf_start'] = 0.0 
sp.params['sf_trunc'] = 0.0
sp.params['tburst'] = 0.0
sp.params['fburst'] = 0.0
sp.params['tau'] = 10.0
sp.params['const'] = 0.0


sp.params['add_neb_emission'] = True
sp.params['nebemlineinspec'] = True
sp.params['add_neb_continuum'] = True

ilib, slib, dlib = sp.libraries
print(ilib, slib, dlib)

sample_size = 1000
logz_arr = np.linspace(-0.6,0.6,sample_size)

logu_arr = -2.5*np.ones(sample_size) 

comb_arr = np.stack([logz_arr, logu_arr], axis=1)

uid = time.strftime("%Y%m%d_%H%M%S")

file_name1 = 'sfg_syn_sed' #  uid
file_name2 = 'sfg_syn_emlines'

tage = 0.001

w, _ = sp.get_spectrum(tage=tage, peraa=True)

emwv = sp.emline_wavelengths
with open(file_name2 + '-' + uid + '.txt', 'w') as file:
    np.savetxt(file, emwv, newline=' ')
    file.write('\n')
file.close()

dict3 = vars(sp.params)['_params']

keys = []
for key, value in dict3.items():
    k = key
    keys.append(k)

df_params = pd.DataFrame(columns=keys)
 
for i in tqdm(range(len(comb_arr))):
    sp.params['logzsol'] = comb_arr[i][0]
    sp.params['gas_logz'] = comb_arr[i][0]
    sp.params['gas_logu'] = comb_arr[i][1]

    _, spec = sp.get_spectrum(tage=tage, peraa=True)

    # Write emission lines to a file
    with open(file_name2 + '-' + uid + '.txt', 'a') as file:
        np.savetxt(file, sp.emline_luminosity, newline=' ')
        file.write('\n')
    file.close()

    dict3 = vars(sp.params)['_params']

    values = []
    for key, value in dict3.items():
        v = value
        values.append(v)
        
    df1 = pd.DataFrame([values], columns=keys)
    df_params = pd.concat([df_params, df1], axis=0)


df_params.to_csv('params_' + uid + '.csv', index=False)

Start by running the supplied Python script with the stated SSP, stellar-library, metallicity, and ionization-parameter settings, then inspect the generated emission-line values around logzsol=0.2. Compare the resulting [OIII]/Hβ and [NII]/Hα behavior with the attached plots; done means establishing whether the plateau is expected or identifying a reproducible issue to report or fix.